nisa1 by 67Gw8304

VIEWS: 57 PAGES: 309

									Institut für Höhere Studien (IHS), Wien
Institute for Advanced Studies, Vienna

The Austrian Innovation System
A New Epigenetic Framework
Final Report
Volume 1

Karl H. Müller
The Austrian Innovation System
A New Epigenetic Framework

Prof. Bernhard Felderer (Director, IHS)
Karl H. Müller (Project-Coordinator, IHS)
Beate Littig, Christoph Hofinger(IHS)

in Collaboration with

Peter Biegelbauer (IHS/MIT)
Prof. Günter Haag (University of Stuttgart)
Richard Költringer (Institute for Panel-Research, IPR)

Final Report
Volume 1

Study for the Federal Ministry of Science, Transport and the Arts and the OECD

June 1996

Institut für Höhere Studien
Stumpergasse 56, A-1060 Wien
Fax: +43/1/597 06 35

Karl H. Müller
Phone: +43/1/599 91-212

Beate Littig
Phone: +43/1/599 91-215

Christoph Hofinger
Phone: +43/1/599 91-219

Institut für Höhere Studien (IHS), Wien
Institute for Advanced Studies, Vienna

Project Report
Not to Be Quoted without Permission by the Author.


The IHS-Group:

Prof. Bernhard Felderer, Director of the IHS

Karl H. Müller: Project-Coordinator
Research Tasks: Theoretical and Methodological Background; Specification of Historical Development Patterns; Integration of NIS-Data
with Complex, Dynamic Models; Recommendations for Science and Technology Policies; Coordination of the Overall Project-Designs,
Author of the Interim Reports and of the Final Reports, etc.

Christoph Hofinger: Research Assistant
Research Tasks: Coordination of the Data; Data Bank Management; Co-worker for the Survey-Design; Data Search etc.

Beate Littig: Research Assistant
Research Tasks: Coordination of the Full Investigation for Intermediary Institutions; Co-worker for the Survey-Design; Expert Interviews,

National Cooperation:

Richard Költringer (Institute for Panel-Research, Vienna): Austrian Survey of Innovation and Transfer
Friedrich Tobil: Internet Research on Innovation and Technology Policies
Thomas Ullrich (Compass, Vienna): Graphical Design

International Cooperation:

Prof. Günter Haag (Department of Physics, University of Stuttgart): Complex Modeling
Peter Biegelbauer (Political Science Department, MIT/IHS): Institutional Analysis

National Project-Management: Federal Ministry for Science, Research and the Arts:

Norbert Rozsenich, Head of Section II (Research and Technology)
Eva Schmitzer, Head of the Department for Technology Policy II/8

OECD-Project Management:

Prof. Dominique Foray (CNRS)
Jean Guinet (OECD)

National Advisory Members:

Karl Messmann, Head of the R&D Department (Central Austrian Statistical Office)
Reinhard Schurawitzki, Department for Basic Research Policy (Federal Ministry for Science, Transport and the Arts)
Prof. Gunter Tichy (Academy of Sciences)

International Advisory Members:

Prof. Michael Gibbons (SPRU, University of Sussex)
Prof. Karin Knorr-Cetina (University of Bielefeld)
Prof. Bruno Latour (CIS, Paris)
Prof. Helga Nowotny (ETH Zürich, University of Vienna)
Prof. Klaus G. Troitzsch (University of Koblenz)
Prof. Björn Wittrock (SCASSS, Uppsala)
Prof. Wolfgang Zapf (Science Center, Berlin)

Institut für Höhere Studien
Institute for Advanced Studies
Stumpergasse 56
A-1060 Vienna

Phone: +43-1-599 91-216
Fax: +43-1-597 06

                                     TABLE OF CONTENTS1

PART 0:        OVERVIEW ON VOLUME I AND VOLUME II                 10

0.1            A COGNITIVE MAP FOR VOLUME I                            18

1.             PROLEGOMENA TO

               AN INTRODUCTION                                         33
               A METATHEORETICAL FRAMEWORK                             42
               AND INNOVATION                                          46
1.5.           MULTI-LEVEL ORGANIZATION                                63
1.6.           REDUCIBILITY - REDUCED                                  67

2.             THE GENOTYPES OF

2.1.           EMBEDDED CODE-SYSTEMS                                   71
2.2.           RECOMBINATIONS IN
               EMBEDDED CODE-SYSTEMS                                   98
               KNOWLEDGE AND SCIENTIFIC PRODUCTION                 120
2.4.           NIS-HELICES AS “KNOWLEDGE BASES”                    135
2.6.           AT THE END - A NEW BEGINNING                        145

     Volume I has been entirely produced by Karl H. Müller.

                  TABLE OF CONTENTS

       IN NATIONAL INNOVATION SYSTEMS                        151

       THE ANALYSIS OF AGENCIES                              154
       AT THE PHENOTYPE-LEVELS                               186
3.3.   THE EMBEDDEDNESS RELATIONS                            191
3.4.   THE EMERGENCE OF HYPER-COMPLEXITY                   209
       COMPLEXITY THEORY                                     211

       NATIONAL INNOVATION SYSTEMS                           211

4.1.   PHENOTYPE NETWORKS                                  213
       NATIONAL INNOVATION SYSTEMS                           217
       NATIONAL INNOVATION SYSTEMS                           221
       AN EXEMPLAR BASED CONCLUSION                        247
       TENTATIVE CONCLUSIONS                                 259

                                                   FINAL REPORTS

Within the OECD-project on National Innovation Systems it has been the intention on part of the IHS to enhance the process of
co-operation and communication by producing a comprehensive set of final reports which highlight the theoretical foundations,
the modeling background as well as the empirical analyses for a major part of the overall Austrian contribution to the NIS
project.2 According to the IHS workplan, the final report has been divided into eight separate volumes.

VOLUME 0:                                                            SCOPE OF THE PROJECT.
                                                                     Central Topics:  Separation into Major Areas and
                                                                                      Content Distribution.

VOLUME I:                                                            A NEW EPIGENETIC FRAMEWORK.
                                                                     Central Topics: Theoretical Background: Evolution,
                                                                                     “Embedded Code Systems”, the
                                                                                     Multiple Constitution of NIS.
VOLUME II:                                                           HISTORY AND METHODOLOGY.
                                                                     Central Topics: A Co-Evolutionary Historical Framework.
                                                                                     Overall Methodology: Systems
                                                                                     and Utilization Contexts.

VOLUME III:                                                          INSTITUTIONAL ANALYSES: INTERMEDIATE
                                                                     NIS-ORGANIZATIONS AND NIS-CASE STUDIES:
                                                                     Central Topics: Selected Components of the
                                                                                     Intermediate Institutional Network
                                                                                     Case Studies of the Austrian IS.

VOLUME IV:                                                           THE DISTRIBUTION POWER OF THE
                                                                     AUSTRIAN INNOVATION SYSTEM: AN EMPIRICAL
                                                                     SUMMARY FROM THE AUSTRIAN SURVEY OF
                                                                     INNOVATION AND TRANSFER (ASIT)
                                                                     Central Topics:  A Comprehensive Picture of the Austrian
                                                                                      System of Knowledge Production, Transfers,
                                                                                      and Innovations.

VOLUME V::                                                           COMPLEX MODELING WITH NIS-DATA.
                                                                     Central Topics: The Multiple Constitution and “Building
                                                                                     Blocks” for NIS;
                                                                                     The Use of NIS-Data for Complex Types
                                                                                     of Innovation and Diffusion Models:
                                                                                     Neural Networks and Master-Equations.

VOLUME VI:                                                           POLICY IMPLICATIONS: MAIN CONSEQUENCES
                                                                     FOR SCIENCE AND TECHNOLOGY POLICIES
                                                                     IN THE FUTURE.
                                                                     Central Topics:  A Comprehensive Set of Policy Measures
                                                                                      for Enhancing the Austrian IS-Distribution
                                                                                      Power at the Micro- and at the Macro-Level

VOLUME VII:                                                          APPENDICES: SUMMARY OF DATA-SETS AND BIBLIOGRAPHY
                                                                     Central Topics: The ASIT-Data Sets and Other NIS-
                                                                                     Relevant Data from Official Sourcs (ÖSTAT,
                                                                                     Chamber of Commerce, etc.)

    It must be noted that the IAS-team is only part of the total Austrian contribution to the NIS-project since a second team from TIP will
concentrate on cluster-analyses as well as on the macro-economic side of the efficiency of innovation patterns so that, for the final report, a
very comprehensive picture of the Austrian NIS should emerge.

                                                    OVERVIEW OF

                                      VOLUME I AND VOLUME II

The first two volumes on National Innovation Systems contain three different perspectives, namely -

                            Volume I as the theoretical background, centered around the
                         “genotypes” and the “phenotypes” of National Innovation Systems3
                                 Part I of Volume II as the historical framework on the
                                co-evolutionary development of modern societal systems
                     Part III of Volume II as methodological “description device” for systems,
                                    complex models and their utilization contexts

These three parts, in conjunction, lay down the conceptual apparatus for the subsequent investigations,
especially for Volume IV on the Austrian Innovation System in the year 1994, for Volume V on complex
modeling and, finally, for Volume VI on science and technology policies. Thus, Volume I and Volume II
have their main focus on the analytical foundations for possible NIS-investigations and are couched,
horribile dictu, in a meta-design or, more appropriate, in a transdisciplinary approach4 far away from the
established “basins of conceptual attraction” which have been established for studying the socio-economic
universes on micro or macro levels. Thus, Volume I and Volume II, though being centered on the
specifications for different families of National Innovation Systems, are couched in a conceptual
configuration of high generality which clearly transcends the boundaries of the traditional social sciences.
In particular, the following three main parts can be distinguished.

    With respect to biological domains, genotypes and phenotypes are defined in the following sense -

       The structures and functions of an organism that can be observed and measured are called its phenotype ... The genotype,
       on the other hand, is defined entirely by the sequence of nucleotides that make up the DNA. For a given genotype, differnet
       phenortypes may be realized, depending on the environment in which the organism finds itself. (FELDMAN 1988:43)

Thus, tt will become the central cognitive task in Volume I to extend the duality genotype - phenotype from biological domains to societal
ensembles ...
    Following Erich Jantsch (1972), one is invited to distinguish between three types of collaboration across disciplines, namely multi-
disciplinarity (common topic, various unrelated disciplinary methods and theories), inter-disciplinarity (common topic, common methods,
separated theories), and trans-disciplinarity (common topic, common methods and common theoretical core).

Diagram 0.1: The Scope of Volume I and Volume II

VOLUME I: THEORY                             VOLUME II: HISTORY                     VOLUME II: METHODOLOGY
                                                        (PART I)                               (PART II)

SECTION I:                                   The Long-Run Development               SECTION I:
Prolegomena for Evolutionary                 of Societal Systems.                   A Systemic Approach under
Systems: Speicification Schemes,             Major Phases in the Evolution          the Heading of “Systemicity”;
Innovation and Denovation Pro-               of the Modern World system;            Basic Definitions for Concepts
cesses.                                      Identification of the Main             like “Systems”, “Structures”,
Main Characteristics of Evolu-               Characteristics of Markets;-        “Processes”, “History”, etc.
tionary Systems: the Dual-Level

SECTION II:                                                                         SECTION II:
The Genotypes of National                                                           Introduction of a Family
Innovation Systems:                                                                 of Complex or Transdiscipli-
Embedded Code Systems                                                               nary Models, Main General
Knowledge, Information and                                                          Features of this Modeling
Scientific Production                                                            Type.

                                                                           
                                          META-THEORETICAL FRAMEWORK
                                                FOR THE STUDY OF
                                             EVOLUTIONARY SYSTEMS
                                             AND THEIR INNOVATIONS
                                            IN BIOLOGY, ECOLOGY AND

                                                                           
SECTION III:                                 PART I (CONTINUED)                    SECTION II (CONTINUED):
The “Embeddedness Relations”              The Co-Evolutionary Loops              Two Main Utilization Con-
between Genotypes and Phenotypes.            between a Market Sphere               texts for Complex or Trans-
Embeddedness-Domains as Pub-                 and Social “Protective Belts”         disciplinary Models;
lic Domains; a Multiplicity of Em-           or, Alternatively, Metabolism         Different Rule Systems and
beddedness Relations within and              Repair-Systems (M, R).                Evaluation Criteria for both
between Genotype- and Phenotype                                                    Types of Modeling Practices.

The Phenotypes of National
Innovation Systems.
The Structure of NIS-Helices;
Families of Single, Dual, Trial ...
and N-Tuple NIS-Ensembles.

       Volume I will develop a new meta-evolutionary perspective, suited not only for the purpose of
       the evolution of National Innovation Systems in particular or of socio-economic systems in
       general but for the evolution of a much richer variety of systems in biological, ecological and
       socio-economic or, alternatively, societal5 fields. In doing so, a heavy emphasis will be placed
       on a particular feature of evolutionary systems which, so far, has been mostly neglected within
       the social science literature, namely, in general terms, the problem of epigenesis and, in
       particular, the duality of (re)production levels as well as the intriguing relationships between
       genotypes and phenotypes.
       Likewise, Part I of Volume II will serve as a meta-historical framework for describing long-
       term processes of macro-evolution in societal ensembles. Phrased differently, Part I will show
       an interesting way of integrating economic and social domains in form of a “mechanism
       protection-” or, alternatively, of a “mechanism-repair-system”. (CASTI 1992, ROSEN 1991)
       Finally, Part II of Volume II will present a meta-systemic summary of the “systemic stance”
       or, more concretely, an transdisciplinary approach for the analysis of a wide array of systems
       like “code systems”, physical systems, biological ones, ecological configurations, socio-
       economic ensembles and the like.
       Moreover, two different utilization contexts with highly different sets of rules and evaluation
       criteria will be identified in which systemic investigations of systems can take place.

In this manner, three important domains have been covered which, following traditional accounts of
National Innovation Systems, are hardly dealt with in more than a marginal and superficial manner. It will
become one of the acid tests for the whole project whether the inclusion of theoretical, of historical as
well as of methodological issues has broadened and enriched the subsequent constructions as well as
interpretations of the empirical results obtained and whether, most importantly, the policy implications
include new options, not encountered in contemporary discussions. According to the parallel network-
character of the overall project design it may even be considered as the most decisive evaluation criterion
whether visible and clear linkages have been established between the theory and modeling parts of the
projects and its empirical and policy counterparts.

    Throughout the subsequent discussions, the terms “societal”, social and “socio-economic” will be used in an equivalent manner, since all
three expressions will refer to any type of human ensembles - interactions with non-humans not withstanding - where practices, actions,
routines, evaluations or, in short particular “aspects of forms of life” play an essential role.


          A NEW





                                              A COGNITIVE MAP

                                                     FOR VOLUME I

The subsequent sections will reveal a new evolutionary perspective on the constitution and on the
development of National Innovation Systems in general, taking its prime focus not, as usual, in human
agencies, institutions or collective actors on the one hand and processes of learning and knowledge
production on the other hand, but in an (almost) natural way of a dual level interplay between socio-
economic genotypes and socio-economic phenotypes.6 Thus, the four sections of the theoretical part
which run under the labels of -

                                                 Prolegomena to
                                 Evolutionary Systems and Innovations (Section 1)
                             The Genotypes of National Innovation Systems (Section II)
                         The Embeddedness Relations between Genotypes and Phenotypes in
                                    National Innovation Systems (Section III)
                            The Phenotypes of National Innovation Systems (Section IV)

will offer a new and as yet relatively unexplored research trajectory in the analysis of innovation and
diffusion processes in general. The basic rationale for departing from the traditional ways lies in the hope
that, as it often happens in the case of fitness landscapes, long jumps and radical deviations may lead to a
largely improved fitness region. After all, the phenomena to be studied in the case of National Innovation
Systems bear strong and deep similarities with the common processes of searching, exploring and
adapting in biological domains.

       Organisms arise from the crafting of natural order and natural selection, artifacts from the crafting of
       Homo Sapiens. Organisms and artifacts so different in scale, complexity and grandeur, so different in

    As a terminological remark, phenotypes and genotypes are to be understood throughout the project reports as generic terms, comprising
any well-defined ensemble or system either on the level of code-systems (genotypes) or on the levels of organisms and artefacts
(phenotypes). Thus, phenotypes and genotypes can be located at many different levels each, including, most prominently, the conventional
micro- or macro-systems for organisms and artefacts. Moreover, the term “socio-economic” or, alternatively, “societal” refers to a sufficiently
general large scale ensemble at the genotype and at the phenotype level. (For a more precise demaractaion of “sufficient generality”, see
chapter 3.3 in Section IV)

      the time scales over which they evolved, yet it is difficult not to see parallels ... Evolution explores its
      landscapes without the benefit of intention. We explore the landscapes of technological opportunity
      with intention, under the selective pressure of market forces. But if the underlying design problems
      result in similar rugged landscapes of conflicting constraints, it would not be astonishing if the same
      laws governed both biological and technological evolution. (KAUFFMAN 1995:191p.)

In this manner, the subsequent four sections will try to establish notions like “rugged fitness landscapes”,
“explorations in fitness landscapes” in a heuristically fruitful way, furnishing thus new conceptual tools in
the analysis of National Innovation Systems in general. As a cognitive map, the theoretical landscapes to
be explored throughout the four sections of the theory-part, exhibit the following characteristics.

Table 0.1: The Basic Chapters in Volume I

                              0/1 conditions for the evolvability of evolution:
                                Two specification requirements (chapter 1)
                  A transdisciplinary concept for innovations and denovations (chapter 2)
         Core-requirements for evolutionary sytems: genotype-phenotype architectures (chapter 3)
                                   Further ramifications (chapters 4 - 6)

                      Code-Systems and Embedded Code-Systems (chapter 1)
                      Recombinations in embedded code systems (chapters 2)
       Embedded code systems: integrating development processes for knowledge, for information
                              and for scientific production (chapter 3)
                           Helices as NIS-”knowledge bases” (chapter 4)
              Recombination and a theory of scientific genotype-creativity (chapter 5)

                           Pseudo-enclosures at the phenotype level (chapter 1)
                        “Agencies” at genotype and phenotype levels (chapter 2)
                       Embeddedness domains and embeddedness relations (chapter 3)
                      Multiple embedded code systems and hypercomplexity (chapter 4)
                         Embeddedness relations and complexity theory (chapter 5)

Table 0.1: The Basic Chapters in Volume I (Continued)

                           Multiple network configurations (chapters 1,2)
                  Multiple ensembles of NIS-helices and NIS-networks (chapter 3)
                    Phenomenology of National Innovation Systems (chapter 4)
                       A theory of scientific phenotype-creativity (chapter 5)

In this manner, a systematic account of the constitution of National Innovation Systems can be provided
which goes well beyond the established knowledge or learning based conceptual foundations which have
been proposed for NIS-analyses so far.

Diagram 0.1: A Cognitive Map for Volume I

                    Phenomenology of National Innovation Systems (SECTION IV)

                                            SECTION IV
                    Phenotype                                        Phenotype

                                                                         
     SECTION III                                                                   SECTION III

                                                                         
                    Genotype                                         Genotype
                                            SECTION II

                       General Requirements for Evolutionary Systems (SECTION I)

As can be seen from Diagram 0.1 (previous page), the new evolutionary or, to be more precise, the new
epigenetic framework will be introduced in four major steps, starting from the most general requirements
for any evolutionary architecture (Section I) and proceeding from here, first, to the genotype or knowledge
bases of National Innovation Systems (Section II) and then to the genotype - phenotype interactions
within a NIS-context (Section III) as well as, finally, to the phenotype organization of National Innovation
Systems and to an extremely important genotype - phenotype NIS phenomenology (Section IV).
Due to this new epigenetic framework, a “Caveat lector” must be made right at the beginning. The new
framework is not restricted to a particular segment of the societal fabric, concentrating on some
connections between science and society only, but is to be understood as a genuine contribution to the
basic architecture of modern societies, past, present and, especially important, future. Thus, notions like
“National Innovation Systems”, “Regional Innovation Systems” (RIS) or “Global Innovation Systems”
(GIS) are, from the way they will be introduced in the subsequent sections, the quintessential “building
blocks” for an evolutionary theory of societies at different levels of spatial distribution. In this sense, the
subsequent explorations, in conjunction with Volume II on History and Methodology, are to be
understood as a new type of approach in modernization theory (For a summary, see e.g. ZAPF 1971,
1994) far away from the established research traditions within this field.
Finally, due to the large number of new concepts and contexts to be introduced, almost no direct
comparisons on the comparative advantages or disadvantages of the present epigenetic framwork with
current modernization programs like “knowledge and information societies”, “modernization I, II or
reflexive” or “postmodernism” will be undertaken. It is hoped, however, that the new epigenetic view or,
to use a core phrase in Section II, the recombination of well-known phenomena and processes within an
epigenetic design will offer sufficiently new cognitive gains even at first sight.
After all, the next four sections develop, step by step, a transdisciplinary approach on evolution,
applicable to biological domains, to ecology, to human societies and, above all, to the interactions
between these three areas ...






                         1. EVOLUTIONARY SYSTEMS -

                        PRELIMINARY REQUIREMENTS

Approaching the domain of National Innovation Systems from a very broad and, above all, from a meta-
theoretical perspective, the introductory chapter will be devoted to two areas:

     The first chapter (1.1) will introduce a series of heuristic devices which should turn out to be
     useful for the specification and the design of socio-economic systems in general and of
     evolutionary socio-economic systems in particular.
     The following chapter (1.2) will be centered around a classification problem with respect to
     systemic states or systemic components and will suggest, under the heading of Evolutionary
     Stable Classifications (ESC), a specification requirement of long-term invariance.

With these two steps, two preliminary research tasks in the creation of a proper design for evolutionary
systems and, consequently, of National Innovation systems, have been completed.

            1.1. Heuristic Rules for Evolutionary Specifications

From a theoretically oriented perspective, the following demarcations and restrictions can be outlined
which offer some convenient guidelines for the range of potential applications for evolutionary models.

     First, the processes under investigation should be characterized by attributes such as
     increasing complexity, critical fluctuations, pattern formations, discontinuities, non-
     linearities, sensitivity for differences in the initial conditions, structural changes, chaotic
     oscillations and the like.
     Second, it should be reasonable, at least prima vista, to assume that these attributes are not the
     consequence of a central steering or control unit but the outcome of the interactions between

       the systemic components.7 Moreover, the systems in question should be composed of a large
       number of distinctive sub-components.
       Third, the prevalent relations of the field under investigation must lie in the internal dynamics
       of the systems components and not, at least not in a predominant way, in the systems-
       environment relations.8
       Fourth, the essential processes and structures of the systemic components - their within-
       organization - should be, again in principle, observable and historically as well as actually
       Fifth, the general requirements presented in Table 1.1 must be, at least in their essential
       segments, specifiable.9

In sum, evolutionary models in their existing variations and offsprings should be, as a heuristic rule of
thumb, utilized and applied if and only if systems components, their mode of reproduction, attributes of
advantageous changes at the level of systems components, relevant segments in the environment, both the
internal and the external systems structures, and, finally, types of internal innovations and external
turbulences can be identified and, to a sufficient degree, couched in quantitative terms.
Thus, very general requirements with respect to the applications of evolutionary models have been
introduced. It must be emphasized that these heuristic rules are at best necessary - and by no means
sufficient for the specification of genuinely evolutionary models. The five rules, presented above,
constitute only a bare sketch for a systemic identification of a wide range of socio-economic models,
some of them being evolutionary, some of them without any evolutionary features.
To present a list of examples, four large scale socio-economic systems, namely an economic system, an
employment system, an education system, and a scientific system, will be specified with all the necessary
ingredients required by Table 1.1. Moreover, these examples should be sufficient to show that some of the
most general requirements for evolutionary systems can be met in the case of typical large-scale socio-
economic systems.

    Especially the second point should make it clear that the self-organization paradigm has, also in the social sciences, a long established
tradition since the insistence on non-intentional outcomes of a comparatively large number of intentional actions can be found both in the
sociology of figurations by Norbert Elias (see e.g., ELIAS 1971, 1988) or in the idea of a spontaneous social order by Friedrich A. Hayek.
(See, e.g. HAYEK 1949, 1967)
    As a more formal corollary to the third requirement one may postulate, following Karl W. Deutsch and Bruno Fritsch, that the number of
internal systems relations must exceed the number of the external ones. (See DEUTSCH/FRITSCH 1980:40)
    The following demarcations offered by Francisco J. Varela use, more or less, a very similar tune:

       (1) Self-organization is a behaviour which is proper to autonomous units;
       (2) autonomous units can be appropriately characterized if we change from an input-type to a closure-type stance;
       (3) specifying the closure of a system leads to an understanding of the internal coherence (eigenbehaviors) such units have;
       (4) if a system has enough structural plasticity the landscape of its eigenbehaviors will be divers and complex, and the
       pathways of change from one to another will be constrained, but not uniquely specified: there is a natural drift ...;
       (5) such self-determined internal coherences and their natural drift, when observed under contingencies of interactions, will
       appear as the making of sense, novelty, and unpredictability, in brief as the 'laying down' of a world. (VARELA 1984:30)

Table 1.1: General Requirements for the Application of Evolutionary Models

              SYSTEM-COMPONENTS                               ENVIRONMENTAL-                   SYSTEM-ENVIRONMENT
                                                              COMPONENTS                       RELATIONS

     Units                       Changes                              Units                 Structure             Changes

Types         Type of        Type of            Types of              Types                 Int.   Ext.           Int.   Ext.
              Repro-         Diversi-           Compa-
              duction        fica-              rative
                             tion               Advantages

│        │      │           │         │           │   │      │    │
│        |      │           │         │           │   │      │    │

                                                          SPECIFICATION OF

│                   │                     │                  │
│                   │                     │                  │

SYSTEMIC-                    COMPONENT-                            ENVIRONMENTAL                   MICRO- AND MACRO-
COMPONENTS                   CHANGES                               COMPONENTS                      ORGANIZATION

What follows next, is a widely ignored intermediate step, viz. some afterthoughts on conceptual
prerequirements which basic classifications for evolutionary socio-economic systems should fulfill.10

  Surprisingly, the point of systems demarcations and orderings has been widely discussed outside the social sciences, for example in
SCHWARZENBACH 1991:341 - 367.

Table 1.2: Basic Systemic Ingredients for Four
           Large Scale Socio-Economic Systems

                             (1) BASIC FEATURES OF AN ECONOMIC SYSTEM

                                                        COMPONENTS              RELATIONS

        Units                     Changes                     Units        Structure          Changes
Type            Repro-      Differen-     Comparative         Type         Int. and Ext.      Int. and Ext.
                duction     tiation       Advantages

Firms           Average     Vertical and    New Products; Foreign          Input-Output-      Economic Inno-
                Growth,     Horizontal      New Production Firms;          Flows (Dome-       vations; Political
                Average     Differentia-    Processes;     et al.          stic and Abroad)   Turbulences; Turbu-
                Profits;    tion            New Forms of                   et al.             lences abroad et al.

                           (2) BASIC FEATURES OF AN EMPLOYMENT SYSTEM

                                                        COMPONENTS              RELATIONS

        Units                     Changes                     Units        Structure             Changes
Type            Repro-      Differen-     Comparative         Type         Int. and Ext.         Int. and Ext.
                duction     tiation       Advantages

Sectors         Average     Vertical        Incen-            Total        Flows of Emp-         Economic
                Growth      and Hori-       tives;            Popula-      loyees between -      Innova-
                            zontal          Attracti-         tion         Sectors; Mi-          tions et
                            Differen-       vity              et al.       grations              al.
                            tiation                                        et al.

                            (3) BASIC FEATURES OF AN EDUCATION SYSTEM

                                                        COMPONENTS              RELATIONS

        Units                     Changes                  Units           Structure             Changes
Type            Repro-      Differen-          Comparative Type            Int. and Ext.         Int. and Ext.
                duction     tiation            Advantages

School            Average     Vertical and     Incentives;    Total        Flows of Pupils       Economic or
Types             Growth      Horizontal       Attractivity   Population   between School-       Socio-Cultural
                              Differentia-                    et al.       Types; Mi-            Innovations et
                              tion                                         grations et al.       al.

                                 (4) BASIC FEATURES OF A SCIENTIFIC SYSTEM

               SYSTEMIC COMPONENTS                               ENVIRONMENTAL                  SYSTEM-ENVIRONMENT
                                                                 COMPONENTS                          RELATIONS

       Units                        Changes                              Units                  Structure                  Changes
Type           Repro-         Differen-     Comparative                  Type                   Int. and Ext.              Int. and Ext.
               duction        tiation       Advantages

Research       Average        Vertical and        Cognitive              Extrascientific        Scientific Ex-             Scientific or
Units          Growth         Horizontal          Attractivity           Ensembles              changes;                   Technological
                              Cognitive                                  et al.                 Migrations                 Innovations et
                              Differentia-                                                      et al.                     al.

                     1.2. Evolutionary Stable Classifications (ESC)

One of the apparently most widely neglected problems in the field of the dynamics of evolutionary social
systems is a conceptual one, namely the problem of partitioning and classifying the areas under
investigation in an evolutionary fitting and conducive perspective. In most cases of applied dynamics, be
it on the level of individual interactions, of so-called social subsystems, on the level of national societies
or the world system, it seems to hold as a general assumption that the selections for a specific type of
system assume a self-evident character in which, via systems ex machina, all the necessary conceptual
demarcations and boundary conditions have been already accomplished. Quite contrary to this popular
belief and, even worse, common procedure11, a distinctly different approach will be chosen in the
subsequent investigation, by taking none of the common sense notions for socio-economic systems as
granted or established. Even more, the following approach to systems modeling starts with a distinctive
requirement stating that any utilization of transdisciplinary or, alternatively, complex approaches has to
meet, from its very outset, a special requirement, viz. the condition of evolutionary stable classifications
(ESC).12 The ESC-postulate implies that any conceptual differentiation for socio-economic systems has to
search for those intrasystemic components and extrasystemic environment elements which exhibit, in the
long run, a sufficient degree of constancy and thus, at least metaphorically, of structural stability. The
rationale for such a demand lies in four different areas:

    If one takes, as one example among too many, George Psacharopoulos' (1987) seminal handbook on the economics of education, one
finds hardly any remark, let alone a single article which would adress to the problem of an adequate conceptualization and demarcation of
the educational system. Apparently, it seems too obvious that young children have to be enrolled worldwide, for a specific period in time,
into a public or private school system.
    Any family resemblance to the notion of evolutionary stable strategies (see e.g. MAYNARD SMITH 1985) is neither accidental nor
undesired, since, in both cases, the attempted solution adresses the same general problem, namely the identification of those strategies or
concepts which, despite marginal or even major long-run changes can be assumed to remain stable and unchanged.

       First, the time horizon for modeling the evolution of large scale systems makes it almost
       imperative that the classifications chosen offer a sufficiently long-term stability or, to be more
       precise, constancy. Since complex models are not only utilized in the explanation context for
       past processes and development patterns, but also for future scenarios, it must be demanded,
       even as a condition for the possibility of applying future scenarios, that all important changes
       within the period under investigation can be accounted for by the basic conceptual framework.
       Second, the ESC-postulate may be seen as an indispensable prerequirement for establishing
       relatively clear-cut subdivisions for large scale socio-economic systems since the ESC-
       requirement implies, almost by necessity, a general focus on the dominant differentiation
       dimensions of the socio-economic system under consideration. Why? Simply because only
       those components become acceptable ESC-candidates which exhibit both a sufficient long-
       term duration and a high diversity to account for the heterogeneity of historical forms and
       patterns. Consequently, any successful solution which fulfills the ESC-requirement must be
       qualified as core dimension in the evolution of the socio-economic systems to be analyzed.
       Third, at the same time the ESC-demand paves the way, as an intended side-effect, for a
       morphological procedure (DUBACH 1977) in the sense that sufficiently closed and
       homogeneous fields with uneven distributions are generated where some regions exhibit a
       strong historical record while other areas have, at least until now, rarely been occupied and
       some fields have not even come into existence as yet.13 This uneven distribution leads, in turn,
       to the formation of simple research heuristics, in which these fields can be taken as primitive
       phase spaces for the socio-economic system under consideration and in which basic types of
       trajectories can be identified and classified.14
       Fourth, a final reason may be mentioned which justifies the aforementioned postulate for
       evolutionary stable classifications: Problems with respect to the origins of qualitatively new
       development patterns, although usually, like the Popperian truth, hard to come by (POPPER
       1965:373), can be dealt with as long as at least one of the following two conditions aare met.
       In the first case, the hitherto unrealized new forms can be accounted for, at least in principle,
       by the classification system already established since the new components, the new patterns or
       the new phase transitions have been integrated into the specified ESC-network15; in the
       second case, the emergence of ex ante unforseeable new forms, be they on the level of
       components, of structures or of processes, will and must remain, by definition, unforseeable
       ex ante and cannot be in any meaningful way anticipated beforehand, at least not in the
       conventional conceptual manner. Here, the ESC-demand imposes one important restriction

    To give just one example from Table 1.3 (C). The combination singular/compulsory segment within the educational system has not been,
at least not to the present point in time, occupied yet and will, with an extremely low and even decreasing probability, ever come into
    It seems even a promising endeavor, to use the basic dimensions of ESC as a primitive phase space - and to identify basic trajectories,
basic patterns, and the like.
    On the possibility to incorporate the emergence of qualitatively new phenomena within the context of neural nets, see, e.g.

       however, for any subsequent ex post analysis, since the resulting conceptual framework has,
       once again, to fulfill the requirement of offering a new, but evolutionary stable classification
       scheme which allows the combination of the ex ante unaccountable elements and the
       historical record in the long run ...

Despite its apparent rationality and justifiability, the demand for evolutionary stable classifications has
rarely been raised, neither in the fields of general methodology for the social sciences, nor in the area of
systems research.16 The consequences, though, for the modeling endeavor in the social sciences are
considerable. In Table 1.3 one finds a condensed summary of the implications of the ESC-requirement in
the modeling of four distinctive large scale socio-economic systems, namely the economic, the
employment the education and the scientific system. Substituting the basic dimensions for each of the four
systems into the general requirements for the application of evolutionary models, one arrives, as one
among many possible forms of conceptualizations, at the demarcations and categories which are then
summarized in Table 3.3.
A few explanatory remarks on the main categories for the various socio-economic systems, as shown in
Table 1.2 and Table 1.3, seem appropriate.

       For the economic system, a Schumpeterian perspective has been adopted in which various
       types of evolutionary stable classifications with respect to the innovation activities of firms,
       namely product, process and organizational innovations, have been combined with two
       diffusion patterns, one labeled core and the other periphery. This differentiation must be
       qualified, from a perspective of une très longue durée, as invariant, following Schumpeter
       (1961) and others (e.g., POLLARD 1980), from the onset of the Industrial Revolution to the
       present day (THERBORN 1995).
       For the employment system, the three basic dimensions selected, viz. type of work location,
       type of products, and type of work, yielded, as one of their compatible solutions, a sectoral
       mix which can be applied to the separation of the labor force of core or, to a lesser extent, of
       semiperipheral areas17 since the middle of the 18 century. Moreover, other sectoral

    Take, once again, the excellent exposition of systems methodology by Mario Bunge, then one is confronted with a simple pyramid,
composed of the set of physical things, the set of chemical systems, the set of bio-systems, the set of sociosystems, and the set of artificial
things. (Ibid., 45f.). Even worse, Bunge goes on to define socio-economic systems like the economic system, the cultural system or the
political system in the following hyper-simplistic manner -

       The economic system of a society is composed of its primary workers (1PL), cultural workers (2PL), and managerial workers
       (3PL) (BUNGE 1979: 208) ... All(!) the members of a cultural system share the same(!) environment and are linked by
       certain relations Sk (Ibid., 211) ... The political system of a society is the subsystem of the latter that controls(!) (to some
       extent) economic and cultural work as well as other types of social behavior (Ibid., 215) -

and arrives at very elementary and, at times, pre-modern conceptualization strategies for socio-economic systems and structures in which the
problem at hand, viz. the quest for evolutionary stable classifications, due to Bunge's extremely simple partitionings of the socio-economic
universes, does not even arise.
    On the notions of core-regions, semiperiphery, periphery, and external areas see especially WALLERSTEIN (1974/1980/1991).

       specifications like agriculture, industry, firm-related services, household-related services18 or
       the government segment do offer both a long term perspective and a sufficient richness in
       variation to account for an adequate partitioning of the labor force ...
       For the education system, the basic two dimensions for differentiating between various
       school-forms follow, on the on hand, along the axis of degrees of compulsion, dividing
       schools according to the attribute compulsory (no legal exit option) and post compulsory (exit
       option), and, on the other hand, along the line of employment relations, dividing schools
       according to the subjects taught and to the transition flows from school to the employment
       sphere, into singular components (a specific school type can be strongly related to one and
       only one of the three main economic sectors, viz. agriculture, industry, and services) and into
       multiple elements (a specific school type can be related to more than one of the three main
       economic sectors).19 Both the compulsion and the employment dimension can be applied from
       the very beginnings of a sufficiently general public school system, since the divisions into
       compulsory and post-compulsory segments and into employment related or university linked
       school forms was present from the 18 century, since the beginnings of a state organized
       system of education, onwards. Thus, a total of nineteen school forms, five from type1
       (multiple-compulsory), five from type2 (multiple-post compulsory), and nine from type4
       (singular-post compulsory) has been selected which could form the set of components for an
       education network. (de SWAAN 1994)
       Finally, the scientific system is decomposed in an evolutionary stable manner into various
       output segments, differentiating, in a Schumpeterian-Kuhnian-Nelsonian (KUHN 1971,
       NELSON/WINTER 1982) manner, between innovative and replicative types of output
       activities and, similar to the firm network, between core and peripheral cognitive domains.
       (On an evolutionary perspective for science, see esp. HULL 1988) In this way, an interesting
       partitioning for the mapping of the cognitive dynamics of research units has been established
       whose empirical trajectories are, until now, largely or, to be more precise, almost totally

The separations along two or more dimensions in the four examples just given fulfill both the criterion of
a sufficiently longue durée and the demand for the requisite variety. Again, alternative ways of
partitionings both on the level of basic dimensions and on the number of components within the
dimensions chosen are not only possible in principle, but are, in both cases, in actual use ...20

    It should be interesting enough to point to the fact that the sectors of agriculture and household related services occupied a prominent
position throughout the nineteenth century and became, by and large, marginalized in the subsequent decades only. In Germany for instance,
one finds, out of a total labor force of 14.8 million people in 1849, roughly 8.3 million people in agriculture and a surprisingly high number
of 1.8 million people in household related services compared to only 0.35 million people in firm related services like banking, insurances
and the like. On these numbers see HOFFMANN (1965:202ff.)
    For an operationalization of the singular-multiple distinction see MÜLLER 1992.
    On alternative ways of conceptualizing the domain of schools even in the case of a small country like Austria, see, e.g. CLEMENT et al.
1980, dell´ MOUR 1985 or HOLZINGER 1991.

Table 1.3: Evolutionary Stable Classifications for Four Large Scale Socio-Economic Systems


Dimension1:        Type of Innovation Activity (Product - Process - Organization)
Dimension2:        Type of Diffusion (Core - Periphery)

                        Product                 Process             Organization

     Core               Type1                   Type2                  Type3


     Periphery          Type4                   Type5                  Type6

Sub-populations:     Firms engaged in           Core product innovations, core process innovations, core-
                                                organization innovations, peripheral product innovations,
                                                peripheral process innovations, peripheral organization inno-


Dimension1:        Type of Work-Location (Household - Outside)
Dimension2:        Type of Products (Goods - Services)
Dimension3:        Type of Work (Paid- Unpaid)

                          Household-Work                               Outside-Work
                        Market   Non-Market                            Market    Non-Market

     Goods              Type1           Type2                             Type3          Type4


     Services           Type5           Type6                             Type7          Type8

Sub-populations:     People employed in         paid household sector (goods and services), unpaid household

sector (goods and services), goods sector, service sector,
unpaid external goods and services.


Dimension1:        Type of Schools (Compulsory - Post-Compulsory)
Dimension2:        Employment-Relation (Multiple - Singular)

                           Compulsory Segment                           Post-Compulsory Segment

     Multiple                    Type1                                          Type2


     Singular                    Type3                                          Type4

Sub-populations:     Pupils enrolled in         Primary School, Special Primary School, General
                                                Secondary School, Academic Secondary School I,
                                                Special Secondary School, Polytechnical School,
                                                Dual Vocational School1-4, Intermediate Vocational
                                                School1-4, Upper Vocational School1-4, Academic
                                                Secondary School II.


Dimension1:        Type of Scientific Output (Innovative-Replicative)
Dimension2:        Type of Cognitive Width (Core, Periphery)

                           Innovative                            Replicative

     Core Areas               Type1                                     Type2


     Peripheral               Type3                                     Type4

Sub-populations:     Research Units in             Innovative in core areas, Replicative in core areas
                                                Innovative in peripheral domains, Replicative in peripheral

                                      2. INNOVATIONS IN SOCIO-


The previous discussions on preliminary specification heuristics for evolutionary systems have helped to
prepare the ground for the introduction of a comprehensive understanding of innovations which,
moreover, can be seen as a second order concept (FOERSTER 1985, 1995), since the analysis of
innovations should proceed in an innovative manner, too.21 The first important differentiation which must
be made in establishing an encompassing notion of innovations, including its self-referential potential,
refers to two distinctive levels, namely the level of components and the systemic level. Table 2.1 gives a
first impression of important basic concepts which will be introduced in the present section in order to
enable a comprehensive analysis of innovation processes.

                             2.1. Innovations at the Component Level

From Table 2.1 it becomes clear that at the component level two fundamentally different types of changes
can be distinguished, namely innovations and denovations. Both concepts can be defined with respect to
the same evolutionary reference frame, namely with respect criteria like average fitness, average
reproduction rate, average degree of competitiveness and the like. Changes, which on the level of
components increase the average fitness (reproduction, competitiveness) of a systemic unit, can be
classified as innovations whereas changes, deteriorating the average fitness (reproduction,
competitiveness) of a component, will be labeled as denovation. Consequently, the trajectory of systemic
components can undergo evolutionary or devolutionary phases, depending on the respective changes from
the average fitness (reproduction, competitiveness).

    This announcement is not to be understood as an immodest self-laudation. Rather, the notion of “second order concepts” refers to the
possibility that concepts, theories or models possess a self-referential realm of applications. Thus, the following considerations are produced
in a way which exemplifies directly the general framework of evolution, of innovations and, as will be seen later, of recombinations ...

Table 2.1. Specification Requirements for Innovations

                                                  COMPONENTS                     RELATIONS

  Units                     Changes                        Units             Structure            Changes

Type of      Type of      Type of        Types of          Types             Int.     Ext.        Int.   Ext.
Unit         Repro-       Diversi-       Compa-
             duction      fica-          rative
                          tion           Advantages
│        │      │           │         │           │   │      │    │
│        |      │           │         │           │   │      │    │

                                              SPECIFICATION STEPS

                                               FOR INNOVATIONS

                  LEVEL OF COMPONENTS                                                 SYSTEMS LEVEL
                    │                                                                             │
┌───────────────────┼─────────────────────┐                                                       │
│                   │                     │                                                       │
│                   │                     │                                                       │

Potential range of          Justification of the           Specification of                  Specification of the
innovations (denova-        comparative advantage          excitatory or inhibi-             resulting systemic
tions) for components:      (disadvantage) for each        tory turbulences at               interaction patterns
output-innovations          of the innovation (denova-        the level of environ-             and the in(de-)novation
input-innovations           tion types                     mental components                 turbulence induced macro-
withinput-innovations                                                                        dynamics
(and the respective de-

Table 2.2:     Changes at the Level of Systemic Components


INNOVATIONS:      Comparative Advantage               DENOVATIONS:     Comparative Disadvantage
                  of Components                                        of Components
                  (Increase in Overall                                 (Decrease in Overall

               Fitness)                                           Fitness)
A second fundamental concept for the creation of novelty comes from the environmental components of a
particular system . Here again, two fundamentally different types of changes can be distinguished,
namely excitatory turbulences and inhibitory turbulences. Both concepts can be defined with respect to
the same evolutionary reference frame, namely with respect to criteria like average fitness, average
reproduction rate, average degree of competitiveness and the like. Outside changes, which on the level of
systemic components increase the average fitness (reproduction, competitiveness) of a systemic unit, can
be classified as excitatory turbulence whereas changes, deteriorating the average fitness (reproduction,
competitiveness) of a systemic component, will be labeled as inhibitory turbulence. Consequently, the
trajectory of systemic components can undergo evolutionary or devolutionary phases, depending on the
effects of outside changes on their average fitness (reproduction, competitiveness).

Table 2.3:   Changes at the Level of Environmental Components


EXCITATORY       Comparative Advantage           INHIBITORY        Comparative Disadvantage
TURBULENCE:      for Systemic Components         TURBULENCE:       for Systemic Components
                 (Increase in Overall                              (Decrease in Overall
                 Fitness of a Component)                           Fitness of a Component)

The innovation concepts at the level of components will turn out to be sufficiently general to be
applicable to a wide range of domains, including large scale socio-economic systems.

                       2.2. Innovations at the Systemic Level

At the systems level, only one concept will be introduced, namely that of systemic fitness which, once
again, can be defined with respect to the same evolutionary reference frame, namely with respect to
criteria like average fitness, average reproduction rate, average degree of competitiveness and the like.
Changes, due to innovations and denovations at the level of components, increasing on the systemic level
the average fitness (reproduction, competitiveness) of a system , can be classified as systemic
innovations whereas changes, due to innovations and denovations at the level of components,

deteriorating the average fitness (reproduction, competitiveness) of a unit , will be labeled as systemic
denovations. To complete the introduction of relevant concepts for the analysis of novelties at the
systemic levels, changes, due to excitatory or inhibitory turbulences at the level of components,
increasing the level of average systemic fitness (reproduction, competitiveness) of a unit , can be
classified as excitatory systemic turbulences whereas changes, due to turbulences at the level of
environmental components, deteriorating the average fitness (reproduction, competitiveness) of a system
, will be labeled as inhibitory systemic turbulences. Consequently, the trajectory of systems can undergo
evolutionary or devolutionary phases, depending on the respective changes from the average fitness
(reproduction, competitiveness).22
To sum up, Table 2.4. and 2.5 introduce systemic selection procedures for innovations, leaving room for
the possibilities of increasing the diffusion and imitation rate of a particular innovation set or cluster as
well as of sub-critical diffusion and imitation processes.

Table 2.4:          Internal Changes at the Level of Systems


SUCCESSFUL              A Specific Interaction Pattern,                  UNSUCCESSFUL A Specific Interaction Pattern
INNOVATIONS             Selecting a Particular Set                       INNOVATIONS  Neglecting a Particular Set
                        of Innovations                                                of Innovations
                        (High or Supra-Critical)                                      (No or Sub-Critical
                        Diffusion Rate)                                                        Diffusion Rate)

Table 2.5:          External Changes at the Level of Systems


EXCITATORY              A Specific Interaction Pattern                   INHIBITORY    A Specific Interaction Pattern
TURBULENCES:            with Excitatory Turbulences                      TURBULENCES : with Inhibitory Turbulences
                        from Outside                                                           from Outside
                        (Increase of the)                                                      (Decrease of the
                        Systemic Fitness)                                                      Systemic Fitness)

     The concept of average fitnessboth for the genotype and the phenotype levels will be precisely defined in Section II.

                  2.3. Types of Innovation- or Denovation-Processes

After having established a rich conceptual vocabulary compatible with the multi-level configuration of
many evolutionary systems, the next step will introduce three systemic types of innovations which may be
classified as elementary and which, following a long historical tradition in the theory of innovations, are
sufficiently general to include the existing spectrum of innovative procsses. The starting point comes, not
surprisingly, from an Austrian summary of an original Austrian contribution to the theory of innovations -

       Of the five ways to innovate stipulated by Schumpeter a long time ago, the first two, innovation
       through new products or processes, have since long become overshadowed and dependent upon the
       other three: finding a market, new energy sources (or financement, one could add) and new social
       institutions. They have become interwoven and make up the conditions without which no primary
       innovation has a chance to develop into a socially recognizable and accepted product. (NOWOTNY

The explicit recognition of Joseph A. Schumpeter´s highly complex conceptual innovation framework as
well as the stress on the importance of the interlinked character of various innovation types makes it
almost imperative to introduce a systemic innovation or denovation framework in the following manner.

       First, the distinction between different levels has to be re-iterated, leading to a distinction
       between innovations across many systemic levels, including genotype and phenotype levels.
       Second, assuming for the level of components an input  withinput  output-perspective,
       the five Schumpeterian innovation types can be condensed into three different areas, leaving
       room for many additional features not covered by the five Schumpeterian domains.23 These
       four innovation areas include, with a strictly systemic focus on various types of different
       components -

           innovations in the output-segment (new types of outputs (products), new environmental
           areas for output linkages, etc.)
           innovations in the input section (reliance on new types of input-connections, new
           environmental input domains, etc.)
           innovations in the input-output related structures and processes within a component,
           including innovations in the input  withinput  output organization (process

    Since the definitions for innovation types have been constructed in a purely formal way, relying only on systemic notions of inputs,
outputs or withinputs, the relation between conventional innovation concepts like the Schumpeterian and the systemic ones must be a
subclass relation .

        innovations and organizational changes in the input relations, within a component, or in
        the output segment)

     Third, assuming for the level of systems, a slightly modified input  withinput  output-
     perspective too, the five Schumpeterian innovation types at the systemic level can be mapped
     into three different areas. These three systemic innovation areas include, again with a special
     emphasis on systems in general -

        innovations in the external relations of a system (new outputs, new environmental areas,
        innovations in the input section (reliance on new elements in the input, new
        environmental input-domains)
        innovations in the input-output related structures and processes within a system,
        including innovations in the input  withinput  output organization (process
        innovations and organizational changes mainly within a system, but also in the input or
        output relations)

Thus, the main innovation types for the duality or the multiplicity of levels in socio-economic ensembles
can be classified as -

                                          output innovations
                                           input innovations
                                input  withinput  output innovations

It goes almost without saying that the same types of classifications and taxonomies can be used for
different types of denovations which follow exactly the same pattern of differentiation and can be
classified, thus, as output denovations, input denovations, product denovations and input  withinput 
output denovations.

                         2.4. Socio-Economic Innovations -

                                   Prototypical Examples

At this stage, a set of examples highly relevant for the analysis of National Innovation Systems will be
introduced in order to demonstrate the fruitfulness of the innovation (denovation) concepts just
Starting with an obvious and almost classical innovation (denovation) candidate, namely with firms, the
following specifications for innovations can be made for the phenotype levels:24

           innovations in the output-segment of firms (product innovations, new markets, new
           types of waste- and emission-flows, etc.)
           innovations in the input section (reliance on new raw materials, on new finished input-
           materials, on new energy flows, on a new mix of suppliers, new supplier markets, etc.)
           innovations in the input  withinput  output related structures and processes within a
           component (process innovations in firms as defined by the Frascati Manual-Family,
           organizational changes in firms with respect to supply organization, new management
           forms, re-organization of labor relations, new types of marketing, etc.)

Second, a research unit - a university institute or a research institute - can and, at times, must become an
innovation element, too, where the following three types can be distinguished:

           innovations in the output-segment of research units (prototypes, patents, innovative
           publications (new theories, new methods, new research topics, etc.)
           innovations in the input section of research units (new patterns of recruitment, new ways
           for outside information retrieval, incorporation of new cognitive domains, etc.)
           innovations in the input  withinput  output related structures and processes within a
           research unit (process innovations like the installment of new information and
           communication technologies, new machinery for research purposes, redesign of research
           laboratories25, organizational changes in a research unit, ranging from new ways of
           distributing incoming information to a re-shuffling of research tasks among the staff
           members or the installment of additional personnel for external relations, etc.)

Third, school types, also an important ingredient for a comprehensive study of National Innovation
Systems, can be classified according to their innovation potential in the following manner:

           innovations in the output-segment of a school-type (new types of degrees, new
           possibilities for job-segments, new linkages for careers abroad, etc.)

    As will become clear later on, especially in the chapter on scientific genotype creativity (Section II), a similar scheme can be developed
for genotype-levels, too.
    It should be stressed that the notion of a laboratory is to be understood in its wider sense, namely in the meaning of any space in time
where scientific practices are actually performed. (On this point, see esp. KNORR-CETINA 1984/1992)

        innovations in the input section of a school-type (new supply sources for pupils, new
        recruitment patterns for teachers, etc.)
        innovations in the input  withinput  output related structures and processes within a
        school-type (process innovations, comprising, above all, the installment of new teaching
        and learning technologies, new teaching and learning methods, organizational changes
        with respect to recruitments, in the distribution of teachers and classes, in the relation
        with parents, etc.)

Fourth, another essential component in a wider conception of National Innovation Systems can be seen in
the area of health care-institutions where, very generally, the following innovation forms can be

        innovations in the output-segment of a health care-unit (patients with new types of
        therapies, with new forms of implantations, with new ways of surgery, etc.)
        innovations in the input section of a health care unit (change in the admission criteria,
        new types of medical supply, new groups of patients, etc.)
        innovations in the input  withinput  output related structures and processes within a
        health care unit (process innovations, consisting of new medical machinery for
        diagnosis, of medical equipment for therapy, of a redesign of medical laboratories, a
        new mix between institutional care and households, changes in the size of institutes,
        reforms toward a decrease or an increase of hierarchies within a health care institution,

Fifth, an additional example will be introduced, namely households which, once again in an
encompassing NIS-analysis, should play a vital role as potential innovators:

        innovations in the output-segment of households (new patterns of household
        composition, new forms of child-rearing, new ways of care for elderly, etc.)
        innovations in the input section of households (reliance on new types of supply,
        including energy supply, shift to new supply markets, etc.)
        innovations in the input  withinput  output related structures and processes within
        households (process innovations, due to new household machinery for reproductive
        household functions, to new information- and communication technologies (“tele-
        shopping”, “tele-banking”), organizational changes with respect to work re-distributions
        among household members, etc.)

It goes without saying that this type of analysis can be generalized to other areas, too: to the political
realm, focusing on political parties (HOFINGER/GRÜTZMANN 1994), to the artistic fields, emphasizing

artistic schools or styles (MARTINDALE 1990), to the life world realm, specializing to the area of life
styles, of religion ... In each of these realms, the systemic innovation concepts - and one may add: the
systemic denovation terms - can be successfully applied.
What follows next however, is a different type of generalization, namely a generalization across
phenotype levels in which the concept of innovation, introduced so far, will be utilized for the macro-
systemic level, too. To shorten the list of examples, only two systemic fields, namely the economic and
the scientific systems, will be chosen. For a national economy, the three innovation types can be specified
in the following manner.

         innovations in the external output relations of a national economy (new product mix in
         exports, new markets, etc.)
         innovations in the external input section (reliance on new elements in the import mix,
         innovations in the input  withinput  output structures and processes within a system
         (process innovations within a national economy, organizational changes mainly within
         an economic system and, partly, in the input or output relations in a national economy

Likewise, a scientific system at the national level can exhibit the following innovation features:

         innovations in the external output relations of a national scientific system (new research
         products (new cognitive elements (theories, methods, topics) as well as patents and
         prototypes) for the international community, etc.)
         innovations in the external input section of a national scientific system (new research
         products from the international community, etc.)
         innovations in the input  withinput  output structures and processes within a
         national scientific system (process innovations within the scientific arena, organizational
         changes mainly within a scientific ensemble and, partly, in the input or output relations
         in a national scientific system)

The main rationale for this wide range of examples was a conscious attempt to show that the conceptual
prerequirements for evolutionary systems and, moreover, the proposed definitions for innovations for
output, input, input  withinput  output as well as for the phenotype level of components and the
systemic one are sufficiently powerful to account for the creation of new phenomena in areas as different
as the arts, the sciences, the economic spheres, the social arena, religion or the political realm.

                         3. EVOLUTIONARY SYSTEMS -


The present chapter is devoted to a series of “Gestalt-Wechsel” with respect to the nature of evolutionary
systems in general. More concretely, the following remarks are aimed at the necessary specification
requirements for any type of evolutionary system, be it biological, ecological, socio-economic or
otherwise in nature. To reach this goal, the present chapter is divided into three major parts:

      First, in the present chapter, the differentia specifica of evolutionary systems will be
      presented, namely the duality of levels and, more importantly, the duality of state spaces and
      forms of (re)production.
      Second, the next chapter (chapter 4), will, due to the duality or, in (almost) all instances, due
      to the multiplicity of levels at the genotype and phenotype niveau, summarize essential
      implications and ramifications for core-concepts in the study of National Innovation Systems
      like emergence or reducibility or the dual arrow of causality, operative in National Innovation
      Third, the final chapters of the meta-theoretical framework offer additional specifications and
      heuristic guidelines for an appropriate level organization. In this manner, further consequences
      of multi level organizations for any NIS-investigation will be discussed. (chapter 5 and
      chapter 6)

With these very brief content descriptions, the meta-theoretical framework will start with an unusual and
unconventional thought-experiment.

                            3.1. The Weizenbaum Experiment

One of the most fundamental approaches to evolutionary systems comes via a thought experiment which,
under the heading of Weizenbaum Experiment, has not become very popular by now although it contains

all the essential ingredients for a deep understanding of evolutionary systems in general. Thus, few
explanatory remarks seem appropriate which help to introduce the basic components for the Weizenbaum-
At first, three basic concepts must be introduced, namely that of a space, of a metric space and, finally, of
a distance.

      A space X is a set. The points of the space are the elements of the set ...
      A metric space (X,d) is a space X together with a real-valued function d: X x X  R, which measures
      the distance between pairs of points x and y in X. We require that d obeys the following axioms:

                   (i)       d(x,y) = d(y,x)                         Vx,y  X
                   (ii)      0  d(x,y)                            Vx,y  X, x y
                   (iii)     d(x,x) = 0                              Vx  X
                   (iv)      d(x,y) d(x,z) + d(z,y)                  Vx,y,z  X (BARNSLEY 1988:6ff.)

With these ingredients, the experimental setting can be summarized in the following manner:

      Biologists realize that there are two spaces in life. There is the DNA and there is the organism
      himself. We say that in both spaces a natural metric exists - the natural metric - and that development
      proceeds in both spaces according to this metric ... In the DNA space, there is a natural probability
      distribution ... Strings that are close to each other here will be likely to change into nearby strings.
      This is random mutation. Now we also suppose that this DNA space is mapped to the other space, the
      organismic space. After all, everything that lives has a certain batch of DNA, and vice versa. We call
      this the natural mapping ... We have also a mirror. In the DNA space we suppose that there is a second
      measure of distance that reflects the distances in the organismic space ... So when we talk of strings of
      DNA or proteins, we can talk of the natural distance between them, and the artificial distance, the
      mirror distance. (BERLINSKI 1986:245)

The experiment itself is conducted in a way that a basin of attraction or a goal area is specified within the
second metric space which, however, must be reached via probabilistic variations in the first one.

      We suppose that we have strings of DNA, and there is some initial probability distribution on them.
      That is to say, at the beginning of the experiment, the strings are most likely to be in a certain
      configuration. We also have ... a probability transition system, that is to say, a system that tells us
      what changes in strings are probable. We also fix a distinguished element in the second space right
      from the first ..., like a distinguished point in topology ... We start with a distinguished element ... in
      the second space and we allow the sequences in the first space to undergo a series of transformations

     according to their probability distribution. The Weizenbaum Experiment is successful if after a time
     the sequences come closer and closer to the original (distinguished element). (IBID.)

The reason why the Weizenbaum Experiment has been introduced in a metatheoretical framework for the
analysis of evolutionary systems should become clear, once the final sentence of the following
conversation has been read:

     'We ask now what we conclude from the fact that a Weizenbaum Experiment is successful?
     I looked up, concluding nothing.
     'We conclude, naturally, that the probability system cannot be arbitrary with respect to its induced
     metric structure.' ...
     'So a successful Weizenbaum Experiment demonstrates that the probability transition system is not
     arbitrary,' I said.
     'Of course,' Schützenberger added, 'no one has ever seen a successful Weizenbaum Experiment.'
     I wrinkled my brows to signify perplexity.
     'With one exception', said Schützenberger.
     `And that is ...?'
     'Life itself.' (IBID:247)

Meanwhile, the differentia specifica of evolutionary systems must have become clear, namely the dualism
of state spaces as well as the dualism of their mode of (re)production. (See also LLOYD 1994) Not
surprisingly, an intense discussion has been conducted and is still going on with respect to the relations
between the two fundamental levels of evolutionary systems, between the genotype and the phenotype,
between the DNA space and the organism and its environment.

     On the one side, the river out of Eden (Richard Dawkins) can and should be studied with a
     single focus, namely with respect to the genotype, its variations and offsprings, and the
     genetic fitness landscapes.

        Genes are pure information - information that can be encoded, recoded and decoded, without
        any degradation or change of meaning. Pure information can be copied and, since it is digital
        information, the fidelity of the copying can be immense. DNA characters are copied with an
        accuracy that rivals anything modern engineers can do. They are copied down the generations,
        with just enough occasional errors to introduce variety. Among this variety, those coded
        combinations that become numerous in the world will obviously and automatically be the ones
        that, when decoded and obeyed inside bodies, make those bodies take active steps to preserve
        and propagate those same DNA messages. We - and that means all living things - are survival
        machines programmed to propagate the digital database that did the programming. Darwinism

         is now seen to be the survival of the survivors at the level of pure, digital code. (DAWKINS

     On the other side, one finds a dualist conception, stressing the importance of the phenotype
     level which offers, contrary to the first version, an interesting field, sometimes called
     morphogenetic field, of investigation. At times, the fundamental dogma of genetics is
     supplemented with a more open version in which causally relevant impacts at the level of
     organisms and the environment act directly on the genetic level, too.

         Instead of ... DNA as the carrier of the specific inherited factors from the parent that influence
         the formation of particular structures in the offspring we have what I call inherited particulars.
         These can be either particular base sequences of DNA that define the genes, or particular
         structures of the parent organism, such as the melon stripe in Paramecium, that get transmitted.
         These act on the generative field, the organization of the egg cell or the organism itself that
         grows and develops to produce a new individual with characteristics inherited from the parents.
         Separating these two aspects of an organism is actually artificial, because in reality the two are
         part of a single system: the inherited particulars are part of the organism, which is a field that is
         organized in space and time. (GOODWIN 1995:39)

While any detailed discussion of this point, though extremely interesting and challenging, would
transcend the boundaries of a study on National Innovation Systems, the following general heuristic rules
can be specified with respect to the evolutionary demarcations and definitions of National Innovation

     First, any evolutionary inspired NIS-analysis must identify a solution in which the dualism of
     levels and, moreover, the dualism of (re)production modes occupies a central position.
     Second, the problem of the interaction between levels must be solved in the case of NIS,
     yielding solutions (n)either close to the first (n)or to the second extreme point outlined above.
     Third, the question of (re)production across and between levels requires specific NIS-
     solutions, too.

Thus, the notion of the Weizenbaum Experiment has led to three additional, clearly meta-theoretical
requirements in the specification of evolutionary systems which will play a vital role within the next three
Sections where the duality of genotype and phenotype domains as well as the multiplicity of levels for
genotypes and phenotypes respectively or alternatively, the multiple units of evolution (ERESHEFSKY
1992) will be described in greater detail.

                                     4. Level-Features: Emergence,

                                       Irreducibility and Innovation

Before entering the domains of NIS-genotype and NIS-phenotype levels as well as the connections
between them, a short “intermezzo”, aimed at major epistemological issues related to the duality of levels
for evolutionary systems, will be performed. Here, core notions for evolutionary systems like emergence,
irreducibility, arrows of causality and others will be discussed which will not only remain consistent with
the overall epigenetic perspective but which will be re-inforcing it.

                                 4.1. The Emergence of Emergence

The next step in the meta-theoretical framework is devoted to two related problems, which are the
necessary consequence of the duality of levels and structures, namely the topics of emergence as well as
of the arrows of cuasality related to dual-level architectures. Moreover, the topic of emergence or, to use
Konrad Lorenz´ term fulguration (LORENZ 1973:48ff.), will be analyzed and will be dealt with in two
different versions where the second one can be shown to be particularly suited for multi-level systems.
Moreover, the notion of innovations, already introduced in the previous chapter, will be related to the two
concepts of emergence in order to avoid an unnecessary, albeit widely distributed confusion.
   Despite its hypertrophic and sometimes even spiritualistic connotations26, the concept of emergence can
be understood in two different versions:
    For the first case, a system  undergoes, within a specific time interval, an emergent1 change in its
attributes or structures iff one of the two following conditions holds27 -

                                                      ,t  ,t (t´  t) t

                                              (t,t´) =   () - (t) t
                                                        t  t'

 designates, as will be introduced in the methodological part on systemicity, the structure of a particular
system,  the set of properties, t and t´ different points in time with the ordering t´  t. Moreover, the
changes in structures or properties should consist, as the differentia specifica for the emergence1 concept,

    See, for example, CORNING 1983 or WILBER 1986.
    For the notation, see especially Section I of Part II in Volume II which contains a detailed exposition of a generalized systemic

of something genuinely new or, alternatively, of a principally unforseeable and unpredictable nature,
expressed in terms of t, the available knowledge stock at time t.

      To introduce a paradigmatic example, the diffusion of factories equipped with steam engines
      must be seen, around the year 1600, as an emergent1 phenomenon since this type of
      production process needed both technological and organizational innovations as well as a
      general consolidation period in the world-economy (WALLERSTEIN 1980). Thus,
      emergence1 must be understood, as has been noted by Carl Gustav Hempel a long time ago
      already (HEMPEL 1965:259-264), as a highly context-dependent phenomenon, since this
      concept requires a precise specification of a theoretic corpus as well as of antecedens
      conditions. Emergence1 an sich is, like a variety of other things an sich, a necessarily
      unrecognizable phenomenon.

Generally, emergent1 attributes or structures may be generated along three different lines -

      first, via unpredictable intrastructural changes of systemic components
      second, via new and not anticipated interstructural changes between components (due, for
      example, to critical increases or decreases in the number of systemic components)
      third, via unforseeable turbulences from outside.

These few remarks are already sufficient, to restrict the concept of emergence1 to a comparatively small
set of property or structural changes only. For example, any context-independent assertions like the one
on the emergence1 of consciousness or the human language and the human brain

      Zu den wichtigsten emergenten Ereignissen nach heutiger kosmologischer Auffassung zählen wohl
      die folgenden:
      - die Emergenz des Bewußtseins
      - die Emergenz der menschlichen Sprache und des menschlichen Gehirns. (ECCLES/POPPER 1982:50) -

should be viewed as highly insufficient and elliptic.
Emergence in its common descriptions like the following -

      they (i.e. Locke´s secondary qualities, K.H.M.) are real, but 'emergent', the essentially novel and
      unpredictable product of a fantastically complex collaboration of the primary qualities of particles
      (DENNETT 1986:142) -

possesses, aside from its genetic meaning, i.e. an un-anticipated property or process of a specific
ensemble as well as an un-foreseeable structural change in time - a time-independent semantic version too

which has been described in a masterly manner by Stephen Jay Gould. Here, complex, multi-level
ensembles are seen as being -

       constructed not as a smooth and seamless continuum, permitting simple extrapolation from the lowest
       level to the highest, but as a series of ascending levels, each bound to the one below it in some ways
       and independent in others. Discontinuities and seams characterize the transitions; 'emergent' features
       not implicit in the operations of processes at lower levels, may control events at higher levels.
       (GOULD 1982:132)

In a more systematic manner, the concept of emergence2 can be applied especially in the context of
complex, stratified systems in which, moreover, processes of emergence1 happen quite frequently and not
seldom in a spontaneous manner. Once again, the knowledge dependency of the emergence2-utilization
must be stressed since the knowledge stock t becomes an indispensable element for the determination of
emergence2, too:

                                                    ,l  ,l´ (l´  l v l  l´) t

                                                 (l,l´) =   () - (l) t
                                                           l  l'

Thus, the range and the potential of emergence2-phenomena is inversely linked to the notion reducibility,
since a high potential for successful reductions across levels implies a small range for emergence2
structures or properties and vice versa. Consequently, emergence2 will emerge1 especially in those areas
which are characterized by a complex stratification pattern and a multitude of mutually irreducible levels
of analysis.28
What remains unresolved at this stage, is the precise meaning of the term reducibility and, alternatively,
irreducibility. Loosely formulated, the position, taken here on the reducibility potential, can be
summarized in the following propositions. First, heterogeneous forms of reductions between different
levels of analysis are, at least in principle, possible. Moreover, any successful form of level reduction
must be considered as an important cognitive gain and is duly qualified as a significant contribution for
the structure and the organization of t, the knowledge stock. However, a strong and eliminative general
direction of reductions from

    A metaphorical summary of the general view, underlying not only the notion of emergence, but the theory part as a whole,, one could
imagine complex pictures of closely interlinked, though mostly irreducible systemic landscapes in a rich manifold of levels, with islands of
highest complexity like the human brain, with a large amount of contingent couplings, with decay and entropic degradation in the
environments - a gallery of Great Chains of Becoming. On similar visions, see e.g. BURKS 1970, CALVIN 1990, CARVALLO 1988,
GOULD 1989 or ZELENY 1980.

                                             (4) Ecology / Sociology
                                                   (3) Biology
                                                  (2) Chemistry

down to the level of elementary wave-particles in

                                                   (1) Physics

must be considered as a misplaced piece of research libido. The cognitive evolution of scientific
disciplines over the last two hundred years follows, in all probability, a highly complex emergent 1,2 mode
and does not exhibit a strong or even a weak reductionist drift from the eco- or the social sciences to the
levels of quantum mechanics. As Peter Medawar, seen through the lenses of Karl Popper, notes, it is
extremely important to emphasize -

     that the true relation of the higher to the lower of these subjects (the four domains from
     ecology/sociology to physics, K.H.M.) is not simply one of logical reducibility, but rather comparable
     to the relation between the subjects

                                       (4) Metrical (Euclidean) geometry
                                              (3) Affine Geometry
                                             (2) Projective Geometry
                                                  (1) Topology

     The fundamental relation between the higher geometrical disciplines ... and the lower ones is not quite
     easy to describe, but it is certainly not one of reducibility. For example, metrical geometry, especially
     in the form of Euclidean geometry, is only very partially reducible to projective geometry, even
     though the results of projective geometry are all valid in a metrical geometry embedded in a language
     rich enough to employ the concepts of projective geometry. Thus we may regard metrical geometry an
     enrichment of projective geometry. Similar relations hold between the other levels ... The enrichment
     is one of concepts, but mainly one of theorems. (POPPER 1982:166f.)

The interesting point in the analogy with geometry lies, so Medawar and Popper, in the easy way of
generalizing the special relationships between any pair of hierarchically ordered domains of scientific
discourse like (sociology/biology), (biology/chemistry) or (chemistry/physics) to different fields of

     Medawar proposes that the relations between consecutive levels of Table 3 (i.e., the listing with four
     different scientific domains, K.H.M.) may be analogous to those of Table 4 (i.e., the enumeration of

         four types of geometry, K.H.M.). Thus chemistry may be regarded as an enrichment of physics; which
         explains why it is partly though not wholly reducible to physics; and similar the higher levels ... Thus
         the subjects ... are clearly not reducible to the ones on the lower levels, even though the lower levels
         remain, in a very clear sense, valid within the higher levels, and even though they are somehow
         contained in the higher levels. Moreover, some of the propositions on the higher levels are reducible
         to the lower levels.(IBID:167)

Consequently, emergence1,2 and very limited forms of reducibility should be considered as the
unavoidable and necessary blind spots with respect to the explanatory or predictive potential for
evolutionary systems which, moreover, differ markedly across time, across culture, and, very generally
speaking, across forms of life.29
Finally, the links between an innovation concept and the notion of emergence1,2 can be established in a
very clear-cut manner:

         First, compared to innovations and denovations, emergence1,2 must be considered as a very
         small class of processes and phenomena since only those elements qualify as emergent1,2
         which could not be anticipated, given a stock of knowledge at a particular time t. Since an
         overwhelming number of innovations at the level of firms, institutes, etc. are the result of
         imitation and learning processes, this type of innovations cannot be described as emergent.
         Second, the relationship between innovation- and emergence1,2-processes cannot be
         conceptualized as a proper subset-relation however, since the occurrence of emergence2-
         phenomena is not necessarily connected with innovations. Emergence2 may be attributed in
         complex, stratified systems without any particular innovation activities.
         Third, the reference points for innovations and emergence1,2 point into very different
         directions, namely to the average fitness of components in the case of innovations and to the
         knowledge stock t at a particular point in time. In this sense, the two concepts, though very
         closely linked at first sight, can be clearly distinguished from one another.

In this manner, some of the vital problems in the areas of emergence, irreducibility and innovations have
been solved, at least at the conceptual level.

     For more details on this particular point, see esp. Section III of the present theoretical part.

              4.2. The Arrows of Causality - Causality from Above

Having completed the section of level differentiation as well as a first round of essential level features,
the present chapter continues with another series of special level features in evolutionary, i.e. dual level
systems. More concretely, within chapter 4.2 the problem of the arrow of causality will be analyzed for a
specific direction, namely for the causality from above, going from higher levels down to the lower ones.
It is hoped that these various metatheoretical elements help to clarify essential features of dual level
(re)production systems, including socio-economic ensembles like National Innovation Systems. The
starting point for the subsequent deliberations comes from Hilary Putnam who has established a dual
direction of justification in science -

     In fact, justification in science proceeds in any direction that may be handy - more observational
     assertions sometimes being justified with the aid of more theoretical ones, and vice versa . (PUTNAM

At first sight, Putnam´s statement seems too far away to have any bearing for the problem of the arrows of
causality. But the subsequent discussion will show that, in the end, the answer to the causality directions
across levels will, in fact, be equivalent to the aforementioned quotation from Putnam. Thus, causality
from above or causality from below - what does it mean from an empirical point of view?
In the present chapter, the first part of the answer will be provided, focusing mainly on those explanatory
practices in the natural or in the social sciences where a configuration of the following type takes place:

Table 4.1: Causality across Levels - Causality from Above

Macro-Level:     Explanatory Macro Variables
                 F1, ..., Fn

Micro-Level:     Explanatory Micro                                    g1, ..., gl as
                 Variables f1, ..., fm                             Explanandum

Arrow of         Explanatory Variables F1, ..., Fn                 g1, ..., go
Causality        of the Macro-Level co-determine,
from Above:      in conjunction with f1, ..., fm

Thus, two questions arise immediately. First, can one find, within the confines of normal science,
configurations of the above type? And second, under which general circumstances do they arise? The
answer will point, generally speaking, to three different domains for causality arrows from above.
For the first domain, an example from the field of economics may be useful. Consider the following
function from the established economic equation stock -

                                                  Xi= i (RP i, WT)       i=(1,...,n)

where X designates the exports of a particular country i, RP the relative prices, WT the world trade and
the subscript i goes over all n countries in the world. In such an equation, the global trade becomes an
explanatory variable for the exports of a national economy. World trade which, qua definition, is
composed of exports and imports of individual countries and nothing else, becomes, thus, under the
Granger conditions for causal relations30, an explanatory and thus causally relevant factor for the transfer
of national goods abroad.31 Another example is provided by a demand equation of a particular
qualification segment Lh/L which is modeled -

       as a function of overall labor productivity, given the availability of capital:

                                                   Lh/L = h ((Y/L)/(K/L))32

These two examples make it very clear that the social sciences with their common distinctions between
macro- and micro-levels exhibit numerous examples of a causality from above, and, consequently, of a
downward causality arrow. It comes into operation in all those cases where an explanation at the level Li
utilizes explanatory factors from levels Li+j (j=1,2,...) - factors from the global economy as relevant for the
national economy, macro-economic factors for micro-economic components like firms, households,
family patterns or small group-configurations for individual behavior, high level brain patterns for neural
groups (CALVIN/OJEMANN 1995, CAUDILL 1992, CHURCHLAND 1995, WILLS 1993) ...
In all these examples, an important piece of reduction of complexity occurs since multi-level systems with
a large number of components are explained with the help of macro-variables, acting as a substitute for
highly dense and complex micro-structures. Thus, the first instance for causality from above can be
identified for the case of micro-ensembles where dense, non-linear, simultaneous processes occur
A second variant of causality from above comes into play in all those ensembles where overall or
systemic context and configurational aspects of socio-economic systems are utilized in an explanatory

   On this point, see e.g. DESAI 1981.
   One can find various attempts to find more specific criteria for causality than just assuming functional relationships (CARTWRIGHT
1989). Nevertheless, in most instances these criteria create more additional new problems than solved old ones
   On this type of function, see PSACHAROPOULOS 1987:341.

scheme. A justification for such a procedure could be given, albeit restricted for economic systems, in the
following manner.

        Whereas in microeconomics, each component 'factor of production' ... is responsible for some specific
        'piece of the action', in macro economics, individual production factors cannot be assigned such clear
        roles. It is the tertiary structure as a whole which determines the mode in which an economy will
        function. There is no way one can say 'This individual production factor's presence means that such
        and such a macro-operation will get performed'. In other words, in the macro world, an individual
        production factor's contribution to the economy's overall function is not 'context free'. However, this
        should not be construed in any way as ammunition for an anti-reductionist argument to the effect that
        'the whole cannot be explained as the sum of its parts.' That would be wholly unjustified. What is
        justified, is rejection of the simpler claim that 'each individual production factor contributes to the
        sum in a manner which is independent of the other production factors present'. In other words, the
        function of an economy cannot be considered to be built up from context free functions of its parts;
        rather, one must consider how the parts interact. (HOFSTADTER 1982:520f.).33

The remarkable feature in the above quotation lies in the fact that it has been slightly modified version of
a text devoted, in its original form, to genetics. Why this analogy from genetics?
The most important part of the answer has to refer to the fact that, once again in the case of highly
complex systems independent of the systemic domain the arrangements, the ordering, the clustering and
the distribution of positions become an important ingredient in explaining the micro-behavior. Two
additional examples directly from the socio-economic realm seem appropriate:

        In the discussion on patterns of development, especially in its dependencia-variants (See, e.g.,
        SENGHAAS 1982, TODARO 1985), one finds an emphasis on configurational and
        distributional aspects as central explanatory elements, differentiating between the socio-
        economic evolution in the First World and the Third World. Factors like structural
        heterogneity should be able, according to the dependencia-school, to account for the

     D.R. Hofstadter´s original version exhibits the following differences:

        Whereas in Typogenetics each component amino acid of an enzyme is responsible for some specific 'piece of the action', in
        real enzymes, individual amino acids cannot be assigned such clear roles. It is the tertiary structure as a whole which
        determines the mode in which an enzyme will function; there is no way one can say 'This amino acid's presence means that
        such-and-such an operation will get performed'. In other words, in real genetics, an individual amino acid's contribution to
        the enzyme's overall function is not 'context-free'. However, this fact should not be construed in any way as ammunition for
        an anti-reductionist argument to the effect that the 'whole cannot be explained as the sum of its parts'. That would be wholly
        unjustified., What is justified is rejection of the simpler claim that 'each amino acid contributes to the sum in a manner
        which is independent of the other amino acids present'. In other words, the function of a protein cannot be considered to be
        built up from context-free functions of its parts: rather, one must consider how the parts interact. (HOFSTADTER

      characteristic differences in the long-term historical trajectories of core areas in peripheral
      In the construction of master equations (HAAG 1989, HAAG/MÜLLER 1992, HAKEN
      1982/1983, MÜLLER/HAAG 1994, WEIDLICH/HAAG 1983/1988) for large scale socio-
      economic systems, it has become a standard practice to incorporate intrasystemic barriers and
      distances which act as an inhibitory factor for micro-processes like changes between large
      towns (migration system), changes between employment sectors (employment system),
      change between political parties in national elections (political system), changes between
      cognitive production states (scientific system) or changes between status-groups (social
      system), etc.

With these examples, a second semantic field for the downward arrow of causality can be identified -
causality from above through the incorporation of systemic context- or distribution-effects.
Third, causality from above arises in those systemic ensembles which can be qualified as hierarchical and
which possess a high systemic unit for control and steering.35 Referring to early dynamic multi level-
models (MESAROVIC/MACKO/TAKAHARA 1970, WHYTE/WILSON/WILSON 1969), a clear line of
connections has been drawn from upper levels to lower ones. Likewise, in control theory (CASTI 1992,
WARFIELD 1976) or in stratified organization models one can identify linkages from higher strata to
lower ones, leaving, thus, ample room for causalities from above ...
The most noteworthy common element in the three applications of causality from above lies in the overall
systemic configuration. Typically, causality from above can be utilized in systems with various levels of
its own internal configuration, the existence of a higher-level way of looking at such a system
(HOFSTADTER/DENNETT 1982:199), in short: of a systemic organization that is stratified.(IBID.) In
other words, causality from above is, almost by necessity, an attribute of complex systems. And within
this context, it becomes quite obvious that a whole can become causally relevant for its constituent parts
So far, the discussion has concentrated on three proto-typical instances for a downward arrow of causality.
What follows next is a clear warning sign that these three cases must not be generalized along two
different lines:

    On this point, see esp. a study for UNIDO (MÜLLER 1984b) in which configurational aspects like economic dualism have been
incorporated into production functions.
    See, for example, a harsh methodological individualist who, nevertheless, supports this view of causality from above :

      Die interessantesten Beispiele für Verursachung nach unten sind in Organismen und ihren ökologishen Systemen und in
      Organismengesellschaften zu finden. Eine Gesellschaft kann ungehindert weiter funktionieren, auch wenn viele ihrer
      Mitglieder sterben; doch ein Streik in einem maßgeblichen Industriezweig, etwa in den elektrischen Kraftwerken, kann für
      viele Menschen großen Schaden anrichten ... Ich glaube, diese Beispiele machen die Tatsache einer Verursachung nach
      unten offenkundig; und sie machen das vollständige Gelingen eines reduktionistischen Programms zumindest problematisch.
      (POPPER/ECCLES 1982:42)

     First, causality from above should not be misunderstood as a justification for a high reference
     level which, then, becomes the most fundamental and most important stratum of analysis. In
     the case of socio-economic systems, one could refer to Immanuel Wallerstein and his
     foundational attempt to characterize the level of world systems-analysis as the most profound
     level of investigation. (WALLERSTEIN 1994). It will belong to the clear conclusions of the
     present section that reference levels at any height, from the string level to the stratum of world
     systems and beyond must be considered as detrimental for the cognitive evolution in science.
     Any attempt to establish a reference level creates, both from a methodological and from a
     theoretical viewpoint, more problems than it allegedly solves ...
     Second, causality from above should not be overstressed by postulating that macro-ensembles
     of factors, in turn, determine and restrict the actions, expectation and the interactions of
     micro-actors. More concretely, in Marxist studies of classes and consciousness (See, for
     example, PROJEKT KLASSENANALYSE 1976) one finds quite frequently an explanation
     scheme which must be qualified as deductive fallacy and which can be summarized by the
     following table.

Table 4.2: Errors across Levels - The Deductive Fallacy

Macro-Level:     Explanatory Variables F1, ..., Fn          Explained variables G1, ..., Gn

Micro-Level:     F1, ..., Fn correspond                     G1, ..., Gn correspond
                 with Micro-Variables                          with Micro-Variables
                 f1, ..., fm                                 g1, ..., gl

Level-Error:     Explanatory Variables F1, ..., Fn          Explanatory Relevant for g1, ..., go
                 must be                                       or for f1, ..., fm

Against a deductive or, alternatively, an ecological fallacy one may identify three different groups of

     First, in many instances it becomes a simple matter of a level fallacy to assume that
     combinations and relations at a higher level have to exert a causally relevant impact on lower
     levels, too. Behind any microstructure one does not detect a generative or an ultimate macro-
     structure, but, following Umberto Eco (1972) a consecutive number of new structures, some
     of them micro-based, some, especially in the three instances of causality from above, macro-
     logical in character ...

        Second, quite frequently the macro-micro transfers cannot be accomplished simply because
        the necessary one to one correspondences between the macro sets and the micro components
        cannot be established. Take, once again, economic classes, then one can easily see that one-
        many relations between a macro-variable and micro-elements have to be excluded. But the
        overwhelming evidence, both theoretically and empirically, indicates that a functional
        distribution say, of incomes, cannot be translated into a corresponding personal distribution
        (BLÜMLE 1975:12). Moreover, the set of macro-factors must be both necessary and
        sufficient for the determination of the micro domains which, as the case of the causality from
        above has amply demonstrated, must be considered as a totally unwarranted assumption.
        Third, due to the highly risky, highly hypothetical and, methodologically speaking, highly
        erroneous nature of postulating macro-connections as micro-configurations too (See esp.
        FLORA 1974:110ff.), any success in a macro-micro transfer can be established only a
        posteriori, i.e. after a large number of empirical tests and confirmations has been performed

With these twofold explorations, on the one hand on the legitimate possibilities of employing a downward
arrow of causality and, on the other hand, on the illegitimate generalization of these instances, the present
chapter has to come to end. And one of its possible conclusions can be seen in a re-normalization of
assertions relating to key concepts like emergence, irreducibility, comprehensible parts and
incomprehensible, qualitatively different wholes. From a normal science point of view, the notion of
irreducible interactions in emergent1,2 systems can be given a precise definition beyond mythology,
beyond science-fiction - and beyond spiritualism ...

                 4.3. The Arrows of Causality - Causality from Below

The next problem in the level-construction of evolutionary systems, social and otherwise, leads directly
into one of the most heated controversies in the architecture of the social sciences in general, viz. into the
question of the exact nature of micro-macro-relations.37

   To present just a single example which highlights the methodologicasl position for level separations just outlined, one may refer to W.W.
Rostow who made a remarkable statement on the micro-foundations of his macro-appraoch to modernization:

        Whatever further empirical research may reveal about the motives which have led men to undertake the constructive
        entrepreneurial acts of the take-off period, this much appears sure: these motives have varied greatly, from one society to
        another; and they have rarely, if ever, been motives of an unmixed material character. (ROSTOW 1971:52)

     On this problem areas, see. e.g.ALEXANDER et al. 1987 or RITZER 1988:366 - 384.

The starting point should be fairly obvious and consists in a schematic representation of a normal
aggregation procedure and, thus, of an upward direction in the arrow of causality.

Table 4.3: Causality across Levels - Causality from Below

Micro-Level:       Explanatory Micro                                   g1, ..., gl as
                   Variables f1, ..., fm                            Explanandum

Micro-Macro        f1, ..., fm can be                               g1, ..., gn can be
Level:             aggregated to the                                   aggregated to the
                   Macro-Variables F1, ..., Fm                     Macro-Variables G1, ..., Gn

Arrow of           Aggregated Variables                               G1, ..., Go
Causality          F1, ..., Fmof the
from Below:        Macro-Level determine

In this sense, the normal science literature is full of instances where successful aggregation procedures
lead from webs of micro-interactions to the corresponding macro-dynamics. Especially in economics and
in the social science literature, a very early search directive and heuristic device has been developed
which, under the heading of methodological individualism, requires a micro-foundation and aggregation
procedures as a generalized construction rule for the analysis of macro-level domains. Thus, since more
than hundred years already, a postulate for a specific reference level L*, at least within the social sciences,
has been formulated. It will become the main task of the present section to dissolve the reference
character of the micro level.
Starting with a typical contemporary example for the justification of a reference level in a particular
branch in science, namely in the neural sciences -

      A description of that activity of a single nerve cell which is transmitted to and influences other nerve
      cells and of a nerve cell's response to such influences from other cells, is a complete enough
      description for functional understanding of the nervous system. There is nothing else 'looking at' or
      controlling this activity, which must therefore provide a basis for understanding how the brain
      controls behaviour. (BARLOW 1972:380, for a historical background of the neural doctrine, see
      BARLOW 1995) -

one may identify a specific configuration which may be labeled as constitutive fallacy and which can be
summarized in the following table.

Table 4.4: Causality across Levels - The Constitutive Fallacy

Micro-Level:           Explanatory Micro                                               g1, ..., gl as
                       Variables f1, ..., fm                                        Explanandum

Micro-Macro            f1, ..., fn constitute                                   g1, ..., gn constitute
Level:                 Macro-Variables                                                Macro-Variables
                       F1, ..., Fm                                               G1, ..., Gl

Level                  Explanatory Variables f1, ..., fm                            G1, ..., Go
Error                  of the Micro-Level must                                         or F1, ..., Fo
        :              determine

The point of reference for the subsequent remarks is given by a meanwhile famous and widely cited
diagram developed by James S. Coleman38 stating, by and large, that any satisfying explanation of social
processes on the macro scale must include at least three distinctive steps, viz. a macro-micro part, a
micro-micro segment, and, by force of logical reasoning, a micro-macro component. Seen from this
perspective, the first question may be phrased in the following way: Which of the explanation elements,
demanded by James S. Coleman, can be rationally accounted for by dual level evolutionary models? And
the answer must state clearly that, at least within the frameworks of the existing model stock, evolutionary
formalisms offer, by and large, only an explanation for the macro-macro sphere, for micro-micro-
relations, for micro-macro-relations in a possible world configuration (MÜLLER 1995) and, at least not
in any essential sense39, neither for the macro-micro processes nor for the micro-macro steps in an actual
world design. Thus, one is invariably confronted with the alleged incompleteness of evolutionary
approaches at hand. Since justifications of the necessity of specific forms of incompleteness have become
however, part and parcel of the scientific climate, at least since the days of Kurt Gödel, it should come as
no surprise that Coleman's completeness postulate, similarly to David Hilbert's dream, will be dismissed.
It is definitely not the case, as James S. Coleman wants us to believe, that two level-crossings are missing
in the established approaches of evolutionary dual level systems.
The justification for the necessity of incomplete explanations where the term incomplete has to be
understood according to the standards of Coleman's completeness perspective, must be given, due to
restrictions in space, in an extremely apodictic way40:

    See, e.g. COLEMAN 1986.
    To argue in a less metaphorical manner, the notion of essentiality can be related to genuine steps in micro investigations of the actual
world type such as micro theory construction, micro empirical data collection, testing, the design of aggregation as well as the test design for
these aggregation procedures. And taking these procedures as points of reference one is forced to concede that none of them are normally
employed in the case of complex models ...
    It should be stresses that the present discussion will be restricted to micro- and macro phenotype levels. A fortiori, a similar argument
can be developed for the genotype-phenotype interconnections.

       First, a genuine macro-evolutionary theory, subsequently called MAC, must pass, aside from
       Coleman's completeness criteria, a large number of test instances, ranging from the question
       whether the MAC approach allows for long term historical explanations, for long term
       predictions, up to a variety of test procedures on the significance of the factors chosen, on the
       correspondence with the historical data or on the sensitivity to shocks. Thus, a large number
       of test criteria, independent on the micro-foundations of the MAC approach, can be put
       forward to evaluate a given macro-macro explanation scheme. And according to this set of
       test criteria, one can effectively arrive at a comprehensive summary, stating that a specific
       MAC framework like the proposed master equation network performs, given other MAC
       alternatives, as superior to ..., as equally satisfactory than .... or as inferior compared to ...
       Second, any micro evolutionary model (MIC) allowing for macro extrapolations has to fulfill,
       in order to obtain an a priori credit on the virtue of its being based in the micro universe, an
       extremely important and restrictive requirement, namely the triviality of its aggregation
       procedures41; that is, no additional assumptions outside the micro-sphere are allowed in the
       aggregation heuristics from the micro to the macro sphere. Given the common aggregation
       heuristics like perfect, consistent or other forms of aggregation (SCHLICHT 1985) however,
       it seems extremely questionable that the triviality condition can be upheld even in the simplest
       possible cases.42 Thus, while micro-macro transfers have become a widely used and also
       successful research strategy in the social sciences, successful judged in the sense of the pure
       accomplishment of the micro-macro transition, the procedures normally utilized turn out to be
       non-trivial in their basic character.
       Third, from a methodological point of view, no a priori credits should be attached therefore,
       to an explanatory framework of the form MIC = {micro-factors & non trivial aggregation
       assumptions} compared to its MAC counterpart which uses, by and large, macro factors and
       macro-assumptions alone. It may very well be the case that basically satisfying micro-models,
       judged by the conventional standards for micro approaches, yield, due to risky and, moreover,
       erroneous aggregation heuristics, an unsatisfying macro model whereas, judged by the
       conventional evaluation standards for the macro realm, a model unrelated with the micro
       spheres fulfills the macro criteria in a highly satisfactory manner.
       Fourth, an extremely important qualification must be added to Coleman's completeness
       requirement, therefore: Given two models, one of the MIC-type and the other of the MAC-
       variety, the MIC approach, being firmly rooted in the allegedly observable social micro

     The term triviality of aggregation procedures can be given the following intuitive interpretation: It must not be the case that a MIC
approach, while true on the micro level, turns out to be false on the macro level. More to the point, the MIC aggregation should turn out as a
nonampliative procedure. (For this, see also SALMON 1975:5ff.)
     Consider, to take the most elementary example, the case of the average height of a given population in a specific region, say R, then a
trivial aggregation procedure would consist in measuring the heights of n individuals, and in calculating the resulting mean value h *. The
feature of non-triviality enters however, even in this simple procedure, if one interprets the results in the sense that the average height in
region R has the average value h*, since, for purely logical reasons, an additional macro-statement of the form Region R consists of a total of
n people becomes necessary to arrive at the desired conclusion.

worlds, offers a comparative advantage, namely the advantage of greater level generality and
applicability, if and only if both the MAC and the MIC-model, judged by the evaluation
results on the macro level, offer roughly similar and results and can be considered of equal
quality. If this equality relation does not hold, one is immediately confronted with the question
why anyone should prefer, from the viewpoint of rational theory selection, a less satisfying
MIC-approach over a comparatively better performing macro model without corresponding
micro foundations.
Fifth, a surprising observation must be added, however: Though it has been readily
acknowledged that satisfying micro-macro models do offer the genuine advantage of greater
generality compared to equally performing macro models without micro foundations, the
micro models normally proposed for the explanation of macro processes, viz. variations of
rational decision models share almost unanimously a striking feature: when evaluated in
terms of their actual world micro foundations they take resort to extremely simplified micro-
assumptions. Consider, for example, a genuine MIC-approach on collective marriage behavior
which fulfills, moreover, Coleman's completeness criteria, then one is confronted with
introductory statements like the following:

   I assume(!) that all male participants(!) and all female participants(!) in the marriage market are
   identical (!). An equilibrium allocation of participants to different mates in an efficient
   marriage market would(!) provide all men and all women with the same expected utility(!). If
   the commodity outputs of households can be combined into a single household commodity (!),
   such as the quantity of children ...., if the output of all marriages is known with certainty(!) ...,
   and if the output is distributed as income to mates(!), the following accounting identity holds
   for all(!) marriages ... (BECKER 1981:40)

Despite the free floating procedures with respect to an actual world micro theory and even
worse, an actual world micro theory testing, James S. Coleman qualifies this type of micro-
macro transition as - successful since -

   the neoclassical economic theory of perfect market exchange systems constitutes a model for
   the micro-to-macro transition, although the model is appropriate only for an idealized(!!!) social
   system with complete communication; ... that using the conceptual framework of a market but
   with certain modifications other micro-to-macro transitions may be successfully made - as, for
   example, in marriage markets, labor markets, and other matching markets; ... that the micro-to-
   macro transition in certain areas, such as escape panics and placement of trust, may be built on
   a model of individual rational behavior but without markets or exchange. (COLEMAN

Consequently, one is even forced to the conclusion that the widely held advantages of micro-
macro approaches, viz. comparatively easy ways of understanding and intelligibility, of
observability and accessibility, and, finally, of testability and confirmation are seldom, if at
all, utilized in the actual furnishing of micro-macro explanation schemes. (See, once again,
Sixth, aside from methodological “wargames” and long standing demarcations with respect to
individualism and holism both in their ontological and methodological variations, a MAC
approach and a conventional MIC framework, based on a version of simple minds (Dan
Lloyd), are confronted, by and large, with identical test procedures with respect to their
plausibility in the micro worlds. It goes without saying that also for the MAC script, the
macro relations of the explanatory scheme must have and do possess a micro corollary at least
in the sense that, on the average, individuals prefer, decide or act in a manner consistent with
the general pattern of the macro-spheres. And this, in turn, implies very clearly that a
statement like the following -

   An explanation based on internal analysis of system behavior in terms of actions and
   orientations of lower-level units is likely to be more stable and general than an explanation
   which remains at the system level. Since the system's behavior is in fact a resultant of the
   actions of its components parts, knowledge of how the actions of these parts combine to
   produce systemic behavior can be expected to give greater predictability than will explanation
   based on statistical relations of surface characteristics of the system (COLEMAN 1990:3) -

becomes, in the final analysis, totally irrelevant to the evaluation scheme both for the MAC or
the conventional, i.e. minimal minds MIC approach under consideration.
Seventh, any MIC framework however, which, at the micro sphere, is able to account for the
highly complex pattern of individual actions and decision making, e.g. the general action
model by Raymond Boudon (See, e.g. BOUDON 1980), is and will be confronted with a
considerable complexity problem, namely the near or even complete impossibility, due to a
highly linked, context sensitive and non-linear interaction network at the micro level to arrive
at any form of aggregation at all. Even worse, an argument to the contrary can be developed
that, for purely formal reasons, a disaggregation procedure starting from a macro framework
should be considered as a more robust strategy than its aggregation counterpart (CASTI 1983)
although no methodological consequence, such as an a priori preference for macro
frameworks, should be attached to this point either.
Eighth, MAC frameworks and MIC approaches, thus, do co-exist, as a matter of fact, side by
side. An argument can be made however, that, from a methodological point of view, they
should do so. The argument stems both from artificial intelligence and from learning

     algorithms and has a single attribute, namely gracefulness. Gracefulness implies, inter alia,
     that theories, be they on the macro or on the micro level, consistently making -

        poor predictions when invoked have their strength steadily decreased to the point that they are
        displaced by newer candidates. The newer candidates must in turn compete, usually doing well
        in 'niches'... The combination of competition and confirmation contributes to the system's
        gracefulness: Large number of new candidates can be injected without disturbing the
        performance in well-practiced domains. (HOLLAND 1986:600)

     Why should an effective learning process within the scientific system be structured basically
     different from learning in non-scientific domains?
     Ninth, it should thus be viewed as a genuinely pragmatic question which approach, either one
     of the MIC-type or one out of the MAC-repertoire, one chooses. The selection problem should
     be considered, consequently, to be largely dependent on the specific research interests of the
     person or the research team in question. Moreover, at least in one instance, one and the same
     process, viz. the movement of pedestrians, has been extensively studied by one and the same
     author both from a micro- and from a macro-perspective. (HELBING 1993) The conclusion
     reached there reproduces and reiterates in a very concrete manner the methodological
     principle of gracefulness offered above: For processes such as the formation of groups, the
     behavior in queues, the avoidance of collisions and the like one should choose the micro-
     model whereas for processes such as the development of walking lanes, the propagation of
     waves or for the purpose of traffic planning one should preferably select the macro version.
     Both approaches to the phenomenon of pedestrian movement, far from being reducible to each
     other, turn out to be as complete in their own right as they are, under different research
     contexts, useful ...

Returning, once again, to Coleman's completeness criteria, one can hardly fail the conclusion that a
genuine macro-macro approach without Coleman's micro-foundations for the evolution of socio-economic
systems must and should be judged as complete. It might well be the case that, from the perspective of
rational theory selection, an equally satisfying micro approach for macro processes must and should be
preferred, especially in the case of research interests in the changing motivation structures, decision
configurations, and action patterns of individuals. But, as the points five, six and seven above have
shown, it must and should be judged as extremely unlikely that, independent of the problems to be
analyzed and of the specific research interests involved, a singular, and at the same time generally
superior micro approach will be reached in the foreseeable future ...
Contrary to the insistence on micro-foundations and to the necessity of successful micro-macro transfers,
the present approach offers a deep-seated principle of tolerance and a convenient and thoroughly
pragmatic way out of the conventional micro-macro and empirical-rational dilemmas by pointing out, as it

has been done over and over already, on the legitimate multiplicity of levels, on the legitimate multiplicity
of selection perspectives, on the irreducibility both of levels and of intra-level perspectives respectively43,
and, finally, on the critical dependence of the process of model selection on primary research interests
which, in the course of the last decades, have, in all probability, increased significantly and will continue
to do so ...44

                                          5. Multi-Level Organization

The discussion so far has established a variety of different results which serve as meta-theoretical
building blocks for the study of evolutionary systems in general and, thus, for the investigations of socio-
economic systems in particular. These building blocks or schemes consist of -

       An (almost) irreducible multiplicity of perspectives and approaches across a duality of
       genotype and phenotype levels LG and LP ...
       A multiplicity of different sub-levels within the genotype and phenotype domains ...
       A multiplicity of approaches and perspectives at the same sub-level Lj ...
       A dual arrow of causality (causality from below and causality from above)
       No reference level L* ...
       Innovations (denovations) as basic concepts for intrasystemic novelty across evolutionary
       systems in general and across different levels ...
       Excitatory or inhibitory turbulences as a basic term for extrasystemic novelty across
       evolutionary systems in general and across different levels ...

    The present alternative may be formulated in contrast to the common insistence on the desirability and rationality of microfoundations
for macroprocesses on behalf of methodological individualists (see, e.g. BOUDON 1979) methodological situationalists (see, e.g. KNORR-
CETINA 1988) and the like. The present approach reiterates a point which has been forcefully developed by Ross W. Ashby already, namely
the legitimate multiplicity of model descriptions especially for complex systems, ranging from the subindividual level to the individual and,
finally, to the macroscopic scales.
    Despite the increasing heterogeneity within the scientific enterprise in general, there seems to exist a deep seated sentiment to the
contrary, where one sees clearly the impediments for any from of unification program -

       It should be quite obvious that microreductions of social theories will involve enormous complexities ... Adequate
       microreductions of social theories cannot be accomplished in the foreseeable future. (CAUSEY 1977:158)

but where, despite the appearances to the contrary, the sentiment for a somewhere hidden singular point for the cognitive trajectory of the
scientific system in general remains unaffected:

       I remain optimistic that scientific progress will continue, and that the unification program will play a very significant role in
       this process ... Unified science is an ideal towards which we can strive; the nearer (!) we approach this ideal, the deeper (!)
       will be our understanding of the world. (IBID: 6)

       Emergence1,2 as knowledge-dependent special feature of evolutionary systems in time or
       across levels ...
       Systemicity as a universal description device across domains and across levels ...45
       Transdisciplinary or, alternatively, complex models as a universal explanation device across
       domains and across levels ...46

With respect to National Innovation Systems, the preceding list of meta-theoretical elements leads to the
following set of general characteristics.

Table 5.1: Multi-Level-Organization and National Innovation Systems

MULTIPLICITY OF                                 ..., alphabets, library of symbols, numbers ...47
GENOTYPE LEVELS                                 ..., words, symbol-configurations, compound numbers, ...
                                                ..., sentences, diagrams, algorithms, ..
                                                ..., articles, books, sets of diagrams, classes of algorithms, ...

MULTIPLICITY OF                                 ..., employees, scientists, ...48
PHENOTYPE LEVELS                                ..., departments of firms, departments of institutes, ...
                                                ..., firms, institutes, bridging institutions, ..
                                                ..., sectors, clusters, disciplines, modes of knowledge production, ...
                                                ..., national economy, national science system, ...
                                                ..., global economy, global scientific system, ...

DUALITY OF (RE)PRODUCTION                      (Re)production at the Level of Code Spaces and at
LEVELS                                         the Level of Socio-Technical Systems, Institutions, etc.

LIMITED CAUSALITY FROM                     Restricted Aggregation Possibilities (due to the Non-linear
BELOW                                        and Context-dependent Relationships at the Micro-Levels,
                                             Co-existence and Complementarity of Descriptions and
                                             Explanations across and within Levels (Weak Consistency
                                             Requirements Only)

LIMITED CAUSALITY FROM                         Restricted Possibilities to Include Macro-Variables into
ABOVE                                          Micro-configurations (due to the Non-linear and Context-
                                               dependent Relationships at the Micro-Levels)

INNOVATIONS AND                                Specifiable for a Multiplicity of Phenotype-Levels and Domains
    On this point, see especially Section I in Part II of Volume II ...
    For a survey on the transdisciplinary or complex model stock, see Section II and III of Part II in Volume II ...
    Consistent with the multiplicity of genotype levels, the level of alphabets or symbol libraries is by no means the most basic one for NIS,
since letters and symbols can be decomposed into a large number of sub-levels, leading all the way down to basic geometric elements, points
or even to the atomic compositions ...
    Consistent with the multiplicity of phenotype levels, the individual level is by no means the most basic one for NIS, since individuals can
be decomposed into a large number of sub-levels, leading all the way down to the cellular, molecular or to atomic compositions ...

DENOVATIONS                         Specifiable for a Multiplicity of Genotype-Levels and Domains

Table 5.1: Multi-Level-Organization and National Innovation Systems (Continued)

EXCITATORY AND IN-                  Specifiable for a Multiplicity of Phenotype-Levels and Domains
HIBITORY TURBULENCES                Specifiable for a Multiplicity of Genotype-Levels and Domains

EMERGENCE                           Relatively Frequent Phenomenon (due to the Focus on Innova-
                                    tions and Denovations and due to the Multiplicity of
                                    Phenotype- and Genotype Levels)

Looking at the long list of components for the meta-theoretical background, one element requires special
and additional attention, namely the assertion of a multiplicity of perspectives for one and the same level.
Take the domain of organizational innovations, one may distinguish, for many interesting problem areas
at a special level, say at the organizational level of research units of a National Innovation System, at least
three principally different utilization contexts. (With modifications from MARR 1981:25)

Table 5.2: Three Perspectives for Investigations at the Organizational Level

ORGANIZATION                        MODELING RELATIONS                             IMPLEMENTATION
THEORY                              AND ALGORITHM                                  OF ORGANIZATION

What are the goals of               How can this organization                      How can the “mo-
the organization,                   theory be implemented?                         deling reltions” and
why are they appropriate,           In particular, what is the                     the algorithms be rea-
and what is the logic                  “modeling relation” for the inputs,            lized physically?
of the strategy by which            the input-withinput-output structures
it can be carried out?              and the output? What are the
                                    algorithms for the transformation?

Despite the almost self-evident differentiation proposed by David Marr, some additional remarks seem
appropriate which highlight the “interpenetrations” and the “interdependencies” between these three

perspectives. In a little variation, or, to be more precise: generalization of David Marr´s view, the
following distinctions can be made:

        Each of the three perspectives of description will have its place in the eventual understanding of the
        organizational processes under investigation, and of course they are logically and causally related. But
        an important point to note is that since the three perspectives are only rather loosely related, some
        phenomena may be explained at only one or two of them ... For some phenomena, the type of
        explanation required is fairly obvious. Reconfigurations of specific organizational units are clearly
        tied principally to the third perspective, the physical realization. ... Theories of the Recombination
        between formal and informal rule-systems, on the other hand, are related more directly to the
        perspective of algorithm and representation. Different algorithms tend to fall in radically different
        ways as they are pushed to the limits of their performance or are deprived of critical information ...
        Although algorithms and mechanisms are empirically more accessible, it is the top perspective, the
        perspective of organization theory, which is critically important from a goal-oriented point of view ...
        To phrase the matter another way, an algorithm is likely to be understood more readily by
        understanding the nature of the problem being solved than by examining the mechanism (and the
        hardware) in which it is embodied (MARR 1981: 25ff.).49

For a better understanding of the above relations, a concrete example will be provided which, once again,
will be generalized for evolutionary processes:

        Trying to understand the dynamics of evolutionary systems by studying only their most elementary
        interactions is like trying to understand bird flight by studying only feathers: It just cannot be done. In
        order to understand bird flight, we have to understand aerodynamics; only then do the structures of
        feathers and the different shapes of birds' wings make sense. More to the point, we cannot understand
        why molecules, cells, neurons or individuals have the structures they do just by studying their
        anatomy and physiology. We can understand how these cells, neurons or individuals behave as they
        do by studying their interactions, but in order to understand why their structures are as they are - we
        have to know a little of the theory of differential operators, band-pass channels, and the mathematics
        of the uncertainty principle, to mention just a few vital theory-elements. (IBID., 27f.).

     Marr´s original version has the following format where, for obvious reasons, the terms level and perspective have been exchanged.

        Each of the three levels of description will have its place in the eventual understanding of the perceptual information
        processing, and of course they are logically and causally related. But an important point to note is that since the three levels
        are only rather loosely related, some phenomena may be explained at only one or two of them ... For some phenomena, the
        type of explanation required is fairly obvious. Neuroanatomy, for example, is clearly tied principally to the third level, the
        physical realization of the computation. The same holds for synaptic mechanisms, action potentials, inhibitory interactions,
        and so forth. Neurophysiology, too, is related mostly to this level ... Psychophysics, on the other hand, is related more
        directly to the level of algorithm and representation. Different algorithms tend to fail in radically different ways as they are
        pushed to the limits of their performance or are deprived of critical information ... Although algorithms and mechanisms are
        empirically more accessible, it is the top level, the level of computational theory, which is critically important from an
        information processing point of view ... To phrase the matter another way, an algorithm is likely to be understood more
        readily by understanding the nature of the problem being solved than by examining the mechanism (and the hardware) in
        which it is embodied.(IBID: 25ff.)

This perspective which can be found in slight variations in MILLIKAN (1984) or PENROSE (1995) too,
makes it very clear that differences in scientific interests and research goals will and must lead to a
multiplicity of perspectives at any given level. Taking, additionally, into account, that each of these
perspectives, within any given level, has, under normal circumstances, a large set of conflicting
descriptive and, even more so, of explanatory devices, one approaches the actual complexities of the
current cognitive landscapes of the scientific systems, global and otherwise. What remains to be done at
this point, is a final look at the precise meaning of the notion of reduction which should help to determine
the possibilities for, to introduce a variation of DAMASIO´s (1994) term, cognitive markers in these
highly differentiated landscapes.

                                    6. Reducibility - Reduced

Thus, the present chapter will effectively conclude the first section of the meta-theoretical framework on
evolutionary systems which has been devoted to core notions like evolution, emergence, innovations,
arrows of causality and the like and which will find its logical conclusion by a discussion of the
possibilities and the limitations of an operation called reducibility across levels. The following quotation
serves as a useful basic orientation:

     We are invited to distinguish between the following kinds of reduction:
     definability of terms (terminological or 'weak' reduction vs. derivability of propositions (deductive or
     'strong' reduction),
     homogeneous reduction (where the concepts of the theory to be reduced are already contained in the
     reducing theory) vs. heterogeneous or inhomogeneous reduction (where this is not the case),
     strict reduction (or reduction by deduction) vs. approximate reduction (where previous theories appear
     as 'limiting cases' of newer ones, possibly under very special or even counter-factual conditions,
     domain-combining (derivational, explanatory or unificational) reduction vs. domain-preserving
     (successional or justificational) reduction,
     explanation of a theory vs. explanation of its success,
     reduction by deduction (or by approximation) vs. replacement (elimination, dislodgement) of theories,
     reduction between theories vs. reduction between fields or branches of science. (VOLLMER

In the subsequent remarks, three different versions of reducibility or, alternatively, of irreducibility will be
differentiated. The first notion, reducibility1, can be qualified as strong and eliminative reduction and can
be circumscribed in the following manner.
Reduction1 is a methodological position according to which the construction of new theories or
explanatory schemes must be generated or constituted ex ante from a generalized reference level, valid
over all scientific disciplines (Chromodynamics, Grand Unified Theories, Symmetry Groups, etc.), or
from a local reference level, applicable to one or several scientific research fields - the neuro-sciences just
from the interactions of neurons, biology written in a genetic code only, social sciences from the practices
of individuals exclusively ... After the elaborated discussion on emergence, innovations or the dual arrow
of causality, it should come as no surprise that reduction1 is seen neither as a viable short-term nor a
sustainable long term objective in science.
In the second version, reduction2 demands, ideally exemplified in Coleman´s version, a criterion of
completeness according to which any macro-account, descriptive or explanatory, needs separate macro-
micro relations and additional micro-macro bridges. Again, the preceding discussion on the causality from
above or from below has furnished very powerful arguments that a completeness criterion in both areas,
for the macro-micro links and for the micro-macro connections, must be regarded as counter-productive
and as too strong since reduction2 pre-supposes a local or a global reference level L*.
Finally, one may find a third version of reducibility which however must be seen primarily as a general
heuristic orientation and a search strategy than as a well-defined program. Reducibility3, contrary to its
strong, eliminative or completeness counterparts, is connected with -

      a more general view of the logical structure of related empirical theories ... Theories may be
      represented as an intricate array, web or net of linked theory elements. Indeed, one may envision the
      whole of empirical science - not just single theories - as a 'net' of linked theory elements ...
      Intertheoretical links provide a unified treatment of particular, well-known intertheoretical relations
      such as reduction, specialization and theoretization and include, as well, other types of relations
      among empirical theories. (SNEED 1984:95f.)

In this version, reducibility3 can be seen as an open quest for the detection and identification of inter-
theoretical linkages by means of a set theoretical (LUDWIG 1990), of a structuralist
(BALZER/MOULINES/SNEED 1987), of a statement view or of other approaches (For a survey, see
BALZER/PEARCE/SCHMIDT 1984:3ff.). Via reduction3, he cognitive networks, the cognitive maps of
and between scientific disciplines become accessible and more clearly recognizable. In this manner,
reducibility3 might contribute to an increased transparency and visibility of the cognitive landscapes of
science which have been established over the last two centuries.
Thus, the plea for a peaceful explanatory co-existence across the duality of micro- and macro-levels and,
more generally, between a multiplicity of levels can be found, at times, in the methodological literature,

too. For the social sciences, Richard Münch and Neil J. Smelser conclude their search for micro-macro
linkages in a similar perspective -

      Both microscopic processes that constitute the web of interactions in society(!) and the macroscopic
      frameworks that result from and condition(!) those processes are essential levels for understanding
      and explaining social life. Moreover, those who have argued polemically that one level is more
      fundamental than the other (in some kind of zero-sum way), or who have argued for the complete
      independence of the two levels, must be regarded as in error. (MÜNCH/SMELSER 1987:385)

At this point, a set of basic ingredients for a meta-theoretical framework of evolutionary systems in
general has been put forward. It will be the task of the subsequent three sections, to develop, within this
specific dual level framework, suitable systemic spaces for the two fundamental levels of National
Innovation Systems, namely for the genotype-analogon of “knowledge” and “knowledge (re)production”
(Section II), for the phenotype-analogon of firms, universities or research institutes (Section IV) and, most
important, for the relations between NIS-phenotypes and NIS-genotypes (Section III).

        SECTION II:



Section II will be concentrated on a central domain for National Innovation Systems, namely, in a formal
mode of description, on the genotype-levels of National Innovation Systems or, in a material mode, on
areas like knowledge, knowledge production, knowledge transfers, information gathering, scientific
production and the like. It will become the main task of Section II to provide a slightly unconventional
framework in which notions of “code-systems” and “embedded code-systems” (ECS)50 at various levels
will occupy the center stage and in which “knowledge”, aside from “information” and “scientific output”,
will turn out to be but one of at least three different genotype-level domains for growth and development
processes within the scientific system. Thus, the three key notions of “knowledge”, “information” and
“scientific production” will be unified under the new heading of “embedded code-systems”. Moreover,
the new evolutionary or, more concretely, epigenetic framework will allow to focus more clearly on those
areas in “knowledge production”, in “information processing” or in “scientific output” which -

       like other products of the economic system, can be manufactured, stored, sold or exchanged ...
       like other services of the economic system, cannot be be manufactured, stored or sold ...

Similarly, the production and manufacturing processes within scientific research laboratories, within
universities, within firm-specific R&D departments or within the confines of small rooms at a research
institute will be summarized under the heading of strings, grammars, programs and recombinations
which at least in the traditional understanding can be classified, to modify a phrase from Stuart Kauffman,
as “programs for free”. (KAUFFMAN 1995) In turn, the comparatively wide distributions of “programs
for free” facilitate in a mode of self-organization the emergence1,2,3 of highly ordered secondary, tertiary,
etc. structures. Likewise, the notion of “(re)productive” as well as of “non (re)productive programs” will
serve as an additional evolutionary fitting entity to account for the traditional domains of knowledge,
information or scientific output-transfers ...51
The introduction, so far, has used a lage number of new concepts which are, at present, far from the point
of intersubjective accessibility and understanding. At the end of Section II however, the same introduction
will be repeated in a slightly altered version, this time, hopefully, near or within the domains of
intersubjective accessibilities and understandings ....
In order to achieve this ambitious journey from the land of marginal to deep understanding, the present
section will highlight four major areas which will become of core-relevance for the introduction of the

    One of the interesting reasons for the introduction of Embedded Code-Systems (ECS) or, alternatively, Eigen-Code-Systems, lies in the
recombinative character of the triplet C-E-S which has appeared already in Section I under the permutation of ESC (Evolutionary Stable
Classification), which will acquire a distinct new meaning in Section III under the heading of Complex Embedded Systems (CES) and which
will, under the title of SCE (Self-Contained Ensembles), lead to a new perspective on the multiple constitution of National Innovation
    In Dretske´s words -

       What is knowledge? A traditional answer is that knowledge is a form of justified true belief. To know that s is F is to be fully
       justified in one´s (true) belief that s is F. Normally, these conditions are interpreted so as to be independent of one another.
       Beliefs can be false, and the truth may not be believed. (DRETSKE 1981:85) -

the common background forknowledge-analyses consists in the trias of belief, truth and justification which, in conjunction, form the central
societal knowledge basis.

concept of embeddedness relations (Section III) as well as for the multiplicity of “phenotypes” for
National Innovation Systems, developed in Section IV:

        First, one will find an elaborated discussion on different “ways of code-making” in which
        code-systems at various levels, instead of knowledge, information or scientific output, will
        become the central focus of analysis.
        Second, the problem of recombinations in embedded code-systems will be analyzed across
        various levels, leading, at times, to new insights into universal modes of changes in areas as
        different as the genetic code in biological systems and the conceptual revulsions and
        revolutions within the scientific arena.
        Third, code-systems will be identified not only in the biological field but also, not
        surprisingly, in the area of languages and, probably surprising, in a domain of which is
        normally considered off-code records, namely in the area of social interactions.
        Fourth, the relations between code-systems, knowledge, information and scientific output will
        be dealt with extensively. The main purpose for this elaborated discussion lies in the
        development of a satisficing argument why code-systems across levels provide a fruitful
        research strategy with a high potential for new insights into processes of changes, adaptations
        or innovations.

With Section II, all the necessary ingredients should be provided which, together with the elements in
Section I on evolutionary systems and innovations, allow for a partially new perspective on the genotypes
(G) of National Innovation Systems, their G-network structures, their G-linkages and the ongoing
processes of recombinations within these systems. After all, the basic question for the whole project on
National Innovation Systems -

        The mystery deepens when we observe the kaleidoscopic nature of National Innovation Systems.
        Firms, research institutes, public administrations, buildings, communication infrastructures, and
        computer hardware are always changing, so that a NIS´s coherence is somehow imposed on a
        perpetual flux of people and structures. Like the standing wave in front of a rock in a fast moving
        stream, a National Innovation System is a pattern in time. No single constituent remains in place, but
        the National Innovation System persists ... What enables NIS to retain their coherence despite
        continual disruptions and a lack of central planning? 52 -

     The above quotation has been slightly modified in substance, though not in spirit from the following original version:

        The mystery deepens when we observe the kaleidoscopic nature of cities. Buyers, sellers, administratioons, streets, bridges,
        and buildings are always changing, so that a city´s coherence is somehow imposed on a perpetual flux of people and
        structures. Like the standing wave in front of a rock in a fast moving stream, a city is a pattern in time. No single constituent
        remains in place, but the city persists ... What enables cities to retain their coherence despite continual disruptions and a
        lack of central planning? (HOLLAND 1995:1)

must be answered with at least some new and counter-intuitive insights, research strategies and
methodological results.

                         1. EMBEDDED CODE-SYSTEMS

Section I has established, due to its emphasis on the dual level character of evolutionary systems, a
general and encompassing basis for the subsequent explorations into “ways of code-making”, into code-
based reproduction modes of varying complexity, and into the subtle relations between embedded code-
systems, information, knowledge and scientific production. In particular, the current section on NIS-
genotypes has been motivated by a current wave of heterogeneous analyses of the basic practices and
procedures in National Innovation Systems where concepts like knowledge, learning by doing, tacit
knowledge, learning by using, knowledge spaces, learning by interacting etc. occupy a set of highly
diverging semantic fields which make it increasingly difficult to relate core notions like that of a
“knowledge based economy” to empirically accessible and “tangible” phenomena and processes of “what
scientists do” (KNORR-CETINA 1995), “what technicians do” or “what firm managers do”. Thus, the
present section will attempt a rigorous integration of existing approaches and will arrive at a
comprehensive genotype-framework which will offer a sufficiently broad basis for the correspinding
phenotype-approaches in Sections III and IV.



One of the surprising features in the discussion of National Innovation Systems lies in the unresolved
status of two central notions, namely that of knowledge and that of learning. The following excerpts from
recent NIS-publications reveal an astonishing discrepancy between a knowledge based version on the one
hand and an approach, based on learning and adaptations on the other hand.

     Learning processes, leading to growth in the stock of knowledge, are basic in the dynamics of a
     modern economy ... In a society knowledge is stored in many ways, and institutions are important for

     determining how this is done. Rules, traditions, customs, norms and even habits help to transfer
     knowledge from one generation to the next. Some of this knowledge will prove conducive to the
     further development and accumulation of knowledge, while other parts of it may retard this process by
     preserving unproductive habits of thought. However, without the support of institutions knowledge
     probably could not accumulate at all. Society would not be able to „remember‟, and would soon
     „forget‟ what it had learnt. A system of production, consisting of more or less interrelated firms, stores
     its knowledge in different ways. Some is stored in „the book of blueprints‟ and some in the heads,
     hands and backbones of the individual producers in the form of both explicit and tacit knowledge.
     (JOHNSON 1992:28)
     Innovation is a ubiquitous phenomenon in the modern economy. In practically all parts of the
     economy, and at all times, we expect to find on-going processes of learning, searching and exploring,
     which result in new products, new techniques, new forms of organization and new markets ... The
     most relevant performance indicators of national systems should reflect the efficiency and
     effectiveness in producing, diffusing and exploiting economically useful knowledge. (LUNDVALL
     Economic agents draw upon stocks of knowledge that exist, either in their own conscious or
     unconscious mind, or are held by others and may be acquired by resort to transfer mechanisms of
     varying degrees of formality. In referring to knowledge products, however, we specifically indicate
     flows, that is knowledge which is being generated within the economic entity under consideration.
     Formal instruction, via the medium of reading a description of a technological process, or a set of
     specifications of the components of a complex piece of machinery, or the chemical constituents of a
     compound listed on a commercial product label, represents the acquisition of new knowledge by the
     learning individual. To say whether the flow observed at that individual level represents an increase in
     the gross stock of knowledge is not possible until we have defined the boundaries of the social entity
     under analysis. If the latter is confined to the individual, there is obviously a correspondence between
     flows and stock changes. Otherwise, if what going on corresponds to a transfer of knowledge between
     agents, we may justly say that the social stock of knowledge has not automatically increased thereby,
     even though the existing knowledge was being more intensively utilized by virtue of becoming more
     widely disseminated. (DAVID/FORAY/OECD-SECRETARIAT 1994:33)

In all these general and, in most instances, highly useful characterizations of the basic architecture of
National Innovation Systems, an inherent tension can be identified between a stock model of innovations,
based on an existing gross knowledge stock, knowledge production, knowledge transfers and knowledge
utilizations, and a non-corresponding flow model, centered around cognitive practices like learning,
searching, exploring and “implicit knowledge”. To make the basic differences more transparent, one may
ask the following set of questions:

        If, on the one hand, the basic process to be analyzed in National Innovation Systems consists
        in the increase and the utilization of a given stock of knowledge, one must clearly distinguish
        between the context of knowledge production and the context of knowledge utilization which,
        by and large, has no or little effect on the existing knowledge stock. Moreover, the core
        problem in the first perspective becomes the determination of the main dimensions of the
        local, national or global knowledge stock, its inflows, its depreciation rates, etc. Consequently,
        a NIS-investigation based on knowledge stocks will become successful if concepts like the
        “generation of new knowledge”, “durability of knowledge stock”, “utilization of a specific
        segment of the knowledge stock” and the like have been defined in an intertemporal and
        interregional comparable and satisficing manner.53
        If, on the other hand, the essential innovation process at the level of firms, research institutes
        or universities lie in the adaptations of new practices and routines which are generated at
        some space-point in the Global Innovation System and which are partially codified - making
        local use of “formal instruction, via the medium of reading a description of a technological
        process, (of) a set of specifications of the components of a complex piece of machinery, (of)
        the chemical constituents of a compound listed on a commercial product label”
        (DAVID/FORAY/OECD-SECRETARIAT 1994:33) -, then the prime interest shifts to the
        problem of the local accessibility of new practices and routines, to the subtle relations
        between codified forms of instructions and necessary environmental changes, to the local
        competencies in generating successful reproductions and adaptations, to organizational and
        institutional problems of learning and thinking (DOUGLAS 1987, 1992) etc. Here, the
        process of successful reproduction, of local adaptations and of generating appropriate “copies”
        (HANAPPI 1994) become the main target domains for analysis.54

It can be argued that both perspectives are not only useful and highly instructive, but also very
encompassing and, moreover, mutually irreducible. Due to integration problems resulting from the
generality of two perspectives, the knowledge stock-model and the adaptation model of scientific
production and transfers are impossible to combine within a “knowledge-learning framework”. For
research design purposes, a decision must be made which of the two main trajectories in the “NIS-space”
should be chosen. It will not be extremely surprising that for Section II, Section III and Section IV a “third
way”-strategy will be pursued, centering on “embedded code systems” (ECS), on “embeddedness
domains” and, finally on “embeddedness relations”, encompassing both approaches under a new

    “Satisficing” means, in the present context, that the concepts listed above should be defined in a comparative manner, fulfilling the
normal requirements for comparative concepts like transitivity, asymmetry and the like and allowing for sentences of the form “The
knowledge generation (knowledge stock) during the interval t1-t2 in regioni was higher than during the following period t2-t3 within the same
    An analogy to the biological sphere might be helpful at this point to clarify the two basic perspectives. Within the first view, the genetic
code for a specific and well-defined biological domain is seen as a stock of information which undergoes, as time goes by, a permanent
change and which determines the distribution of a single or groups of species. For the alternative perspective, however, the emphasis lies in
the local processes of reproduction, the conditions of local environments, the local adaptations, the resulting global distribution patterns, the
emergence of new local recombinations, their impact on global reproduction processes and so on ...

transdisciplinary, code-based “umbrella”. Moreover, the ECS-trajectory for analyzing National Innovation
Systems has the distinctive advantage of being able to locate the two traditional approaches in different
domains of the ECS-framework. The knowledge-strategy has its main focus on the genotype-levels of
National Innovation Systems, whereas the learning approach is situated, as will become clear throughout
Sections III and IV, primarily at the phenotype levels.

Table 1.1: National Innovation Systems - Two Heterogeneous Perspectives

DOMAIN                               KNOWLEDGE-BASED               LEARNING-BASED
                                     APPROACHES                    PERSPECTIVES

                 BASIC UNITS         Knowledge                     Practice- and Routine-
                                     Components                    Components
                 BASIC CHANGES       Changes in the                Changes in Practices and
                 (MICRO)             Stock of Knowledge            Routines (Learning and Adaptation)
PRODUCTION       GROWTH              Accumulative                  Structural
                 PROCESSES           (Changes in the compo-        (Changes in the linkages and in
                 (MACRO)             sition of the know-           the linkage weights of NIS-Actor
                                     ledge stock)                  Networks)

                 CODIFICATION        Explicit/Implicit             Implicit, Focusing on Processes of
                                     (with Implicit Com-           Learning and Adaptations
                                     ponents of „Tacit             (with Explicit Elements of
                                     Knowledge‟)                   Knowledge, Information or
                                                                   Language Instructions)

                 ORGANIZATION        Hierarchic                 Heterarchic

                 UNITS OF            Explicit Know-                Implicit Learning Packages
                 TRANSFER            ledge Components              (Including Explicit Knowledge
TRANSFER                             (and „Tacit Knowledge‟)       Components etc.)
SCIENCE AND      TYPE OF             Transmission                  Imitation

MAIN           Hierarchic        Heterarchic

Table 1.2: National Innovation Systems - A New Paradigm

DOMAIN                           TRADITIONAL                       EMBEDDED CODE-SYSTEMS
                                 APPROACHES                        AS NEW PERSPECTIVE

               BASIC UNITS       Knowledge or Practice             Code-Strings, Programs, Programs
                                 Components                        of Programs ...
               BASIC CHANGES     Changes in the                    Recombinations and Corresponding
               (Genotype)        Stock of Knowledge                Forms of (Re)production
PRODUCTION     GROWTH            Changes in Routines,           Embedded Interplay between Genotype
               PROCESSES         Diffusions of Imitation etc.      and Phenotype Development as well as
               (Phenotype)                                         Genotype-Independent Differentiations
                                                                   in the Diffusion Patterns within the

               CODIFICATION      Explicit/Implicit                 Code-Based at Various Genotype-Levels,
                                                                   Including Social Rule-Codes.

               ORGANIZATION      Hierarchic or                     Heterarchic, Embedded

               UNITS OF          Explicit/implicit Know-           Programs at Various Genotype-Levels,
               TRANSFER          ledge Components or               Including Social Rule-Codes
                                 Learning Packages
BETWEEN        TYPE OF           Transmission or                   (Re)production
SCIENCE AND    TRANSFER          Imitation

               MAIN              Hierarchic or                     Heterarchic, Embedded
               ORGANIZATION      Heterarchic

The main reason for the research choice across the paths of knowledge and adaptation lies, moreover, in
the fact that it will become possible to establish direct links between the previous section on evolutionary
systems and innovations, the subsequent domains on a network-based multiplicity of National Innovation
Systems (Sections III, IV) and the Volume V of the NIS-project on complex, dynamic models for which
processes of knowledge production, adaptation and learning become the necessary and, above all, the
evolutionary fitting foundation.
Thus, Table 1.1. gives a brief summary of the current bifurcation area, by separating clearly between
knowledge stock models of scientific production and an alternative focus which has been labeled,
consequently, as flow models of adaptations and learning. Table 1.2 points, then, to some significant
differences between the traditional approaches to National Innovation Systems and the new evolutionary
or, alternatively, transdisciplinary framework, built up within the four sections of the present volume.
In this manner, the basic direction to be followed within the next chapters has been specified. Following
the definition of evolutionary systems as dual level-reproduction systems, it will become the main
objective to specify, for the hyper-complex NIS-configuration, the genotype level of evolutionary systems
by an elaborate analysis of the, at times, surprising and counter-intuitive relations between embedded
code-systems, knowledge, information and scientific output. 55

                                1.2. EMBEDDED CODE-SYSTEMS

                                                AS BASIC CONCEPT

The main-task for the subsequent chapters is clearly defined. It consists, on the one hand, in an
elaboration of an integrated conceptual apparatus for describing and analyzing processes of knowledge
production, learning, tacit knowledge, local adaptations, knowledge stocks, and, on the other hand, in the
conservation of an overall evolutionary perspective and in the specification of the linkages between
phenotype and genotype levels. Reading the title of the present chapter, it should come as a minimal
surprise that the notion of codes and embedded code systems will be proposed as an unifying perspective
which should secure the necessary bridging functions between linguistic entities, their storage and
retrieval processes on the one hand and operations of learning and exploring on the other hand. Codes
have been selected since, even on an entirely intuitive level, they fulfill four requirements simultaneously:

    In later parts of the theoretical exposition it will become clear that evolutionary systems like a National Innovation System do not exhibit
a dual level, but a multiplicity of levels both for the genotypes and for the phenotypes. Moreover, one will be confronted in the case of NIS
with a comparatively large number of different embedded code systems which, then, render the resulting processes as hyper-complex.

      First, in biology, codes as for example the DNA-code, play literally a vital role in the
      maintenance and in the production of organisms. Thus, embedded biological code-systems
      stand, under configurations specified below, in very close connection to problems of
      reproduction, differentiation or maintenance.
      Second, codes have a long standing in areas like linguistics and semiotics and are able,
      consequently, to capture and to incorporate linguistic entities and structures as well.
      Consequently, code-systems are sufficiently powerful to account for the emergence and for the
      development of languages, human and otherwise.
      Third, code-systems are closely related to fields like information-theory offering, in addition,
      the possibility for formalized investigations with respect to information contents, entropy,
      order, etc.
      Fourth, codes occupy a prominent role, especially in postmodern social sciences, for
      describing and analyzing communication processes, media, films, architectures, etc. Thus,
      code-systems can also be utilized for non-linguistic contexts of social, technical or economic
      environments. (ECO 1972, 1981, 1992, 1993)

Due to the fourfold application areas for processes of ontogeny, for linguistic structures, for information-
theory and, finally, for socio-economic life-worlds, code-systems should offer, by means of analogical
reasoning, an extremely useful tool in the domain of National Innovation Systems, too. Moreover,
processes of “knowledge production”, of local adaptations or of searching in “knowledge space” should
be analyzable, after all, in a comparatively more encompassing manner, since a code-based perspective
allows not only a dual-level investigation, but also an intricate combination of issues like codes, learning,
recombinations, adaptations, reproduction or information. Thus, it will become a highly demanding
research challenge whether a unifying framework for codes, code spaces and, finally, for embedded code-
systems can be identified which fulfill the necessary role of offering a homogeneous analysis for the
duality of levels and their entangled interactions within any type of evolutionary system, biological, social
or otherwise.
The inclusion of the domains of organisms, including humans, and the domain of human NIS-artifacts -
books, problems solved, and enginneering triumphs has bee undertaken -

      by design, not accident, of course. It was to help set the stage for ... a Central Salvo: there is only one
      Design Space, and everything actual in it is united with everything else. (DENNETT 1995:135)

                                   1.2.1. BASIC REQUIREMENTS

                               FOR EMBEDDED CODE-SYSTEMS

The present chapter will summarize in a brief and condensed fashion five essential ingredients which are
necessary for the characterization of embedded code systems at the genotype levels. (See esp.
GOODMAN 1973) Four of the five requirements are located at the level of code-elements, the fifth
condition refers to the so-called “embeddedness-relations”, i.e. to the linkages between code-systems and
their wider environments.

       Indifference of Code-Elements; The first condition refers to the exchangeability of specific
       “marks” of a basic component or “character” in a code-system. Thus, taking the alphabetic
       code as reference point, A, A, A or A are clearly different marks for a specific basic character
       which, nevertheless, can be freely exchanged without distorting the code-message.
       Indifference is, thus, a typical equivalence relation, being reflexive, symmetric and transitive.
       Finite Differentiation: Second, the condition of finite differentiation refers to the decidability,
       in principle, whether a given mark belongs to a specific code-character - or not. Being strictly
       independent from the indifference condition, the requirement of finite differentiation asks for
       a code-system which, under normal conditions, possesses a very low rate of reproduction
       errors. Take, once again, the alphabetic code, then the following sequence of code- characters

                                                                 {a, b, c}

       should be reproduced by any system capable of code-retrieval and processing as

                                                                 {a, b, c}

       Combination of Code-Elements: The most essential requirement for code-systems, their
       differentia specifica, lies, third, in the combinability of code-elements or characters. Code-
       characters are able, via recombinations and a small number of basic operations, to produce
       new composite sequences. Moreover, code-characters can be integrated into larger sequences
       and can be synthesized as programs which serve an indispensable construction- or recipe-
       function56 in the production of organisms, species, socio-technical complexes, socio-
       economic systems, etc. The following chapter will bring a more elaborated summary of the
       three necessary characteristics in the combination of code-characters.
     Thus, programs should be placed, within the “metaphor space”, in close distances to concepts like production, constitution, construction
... and in a long distance from terms like “representations”, imitations and the like ...

       Fitness Functions: Fourth, combinations and recombinations of code elements should become
       evaluable in terms of a fitness function . Stated in a loose fashion, combinations and
       recombinations should be assigned different “degrees of fitness”, ranging from simple
       nominal arrangements (no fitness, small degree, medium degree, high degree, maximum
       fitness) to ordinal and, finally, to cardinal ensembles, defined over the interval 0,1.
       Embeddedness: Understood in a wide and metaphorical sense, the fifth and final requirement
       refers to the necessity for code systems to be dynamically embedded in a larger surrounding
       whose production and reproduction depends, however, crucially on the code-system itself.
       Thus, instead of the semantic postulates to be found in Goodman which, as can be easily
       demonstrated, are impossible to fulfill57, the “embeddedness condition” refers to the dual level
       character of evolutionary systems and, consequently, to the connections between a code-based
       genotype and a code-dependent phenotype which, again very loosely formulated, are
       dynamically interwoven and entangled with each other.

It will become the main task in Section III to analyze the so-called “embeddedness-relations” in greater
detail. It will be shown, in the end, that the embeddedness-relations hold the key for a partially new
understanding of the constitution and the dynamics of socio-economic systems in general and of National
Innovation Systems in particular.

                                  1.2.2. STRINGS, GRAMMARS,

                                           PROGRAMS, GENRES

The next step in the elaboration of an evolutionary, genotype-based approach to National Innovation
Systems consists in a condensed discussion of the third requirement for embedded code-systems, namely
the combinability-condition. The goal of this chapter lies in the definition of three concepts which must be
simultaneously present in order to fulfill the combination and recombination postulate. For a basic
introduction, the following quotation might be very helpful and instructive.

       ENDLESS CHAOTIC SEQUENCES (ECS): If a trillion monkeys were to type ten keys a second at
       random, it would take more than a trillion times as long as the universe has been in existence merely

     In order for Goodman´s semantic conditions for notation systems to hold (GOODMAN 1973), a positive solution must be required to
both Hume´s and Goodman´s induction riddles. Since non-ambiguity belongs to the first semantic demand, it can be easily shown that,
invoking Goodman´s induction paradox against Goodman as “languages of art”author, any proposed notation scheme, by necessity, fails to
fulfill such a demand.

       to produce the sentence 'To be or not to be: that is the question.' ... By applying certain quite simple
       rules of probability ... imaginary monkeys could, in a matter of minutes, turn out passages which
       contain striking resemblance to lines from Shakespeare‟s plays ...
       STEP I: A program, arranging for certain letters to appear more frequently than others ...
       STEP II: Some statistical rules about which letters are likely to begin at the beginning and end of
       words, and which pairs of letters, such as th, he, qu and ex, are used most often ...
       STEP III: A program ... to take into account triplets of letters, in which the probability of one letter is
       affected by the two letters which come before it ...
       STEP IV: At a fourth level of programming, where groups of four letters were considered, only 10
       percent of the words produced were gibberish ...(CAMPBELL 1984:116ff.)

Thus, a set of rules on relative frequencies of code-components, on start- and termination-conditions as
well as on code-triplets or code-quadruplets etc., is able to generate large classes of code-sequences which
resemble those compound code-elements - words - which have been developed in the embedded code-
systems of natural languages. In sum, an ECS needs, aside from a repertoire of code-characters, additional
provisions and rules for code-strings and the continuation of code-strings ...
Following the above quotation and sticking, for the time being, to the domain of natural languages58, the
starting point in the transformation of an endless chaotic sequence into an embedded code-system lies in a
complex rule-system for the elementary combination of code-characters into small sequences of such
characters, i.e. into code-strings and, finally, of strings to well-formed sentences, etc. Since a heterarchic
architecture of such a rule-system would be extremely hard to come by, especially in real-time, a different
evolutionary path has been chosen for the “language lottery”(LIGHTFOOT 1986, 1991) which exhibits a
clear-cut differentiation of several genotype-levels.

       At the basic level of alphabetic characters, rules for cluster-formation transform single letters
       into compounds or syllables which form the elementary two string-, three string-, four string-,
       n-string-components of natural languages.
       At the level of compounds of syllables or, alternatively, of well-formed strings, a rich
       dictionary can be built up, consisting in the case of natural language, of a truly infinite number
       of “words”, since a mode of “recombinations through combinations” can be set up which, like
       Cantor´s diagonalization method, creates, by necessity, new combinations not included in
       existing dictionaries ...59
       At the level of words, an extremely important rule system or grammar can be established for
       the transformation of words into word-sequences or, alternatively, sentences by linking and

    It must be immediately added that the restriction to natural languages occurs simply for reasons of simplicity and of an exemplar-guided
procedure. In later parts, the language-independent requirements for programs, etc. will be laid out.
    To make this point clearer, the following procedure can be set up: Start with a lexicon L of well-formed strings, then it will be possible,
via string-combinations, to find new strings not included in L; add this set to L for the new lexicon L 1; then a new set of words, not
contained in L1, can be constructed in a combinative mode, leading to L2; subsequently, the power of combinations can be used for L3, L4,
and so on ....

     directing the patterns of words, the starting conditions, the end-conditions, etc. (See, e.g.,
     WINOGRAD 1983)
     Furthermore, a variety of code-specific genres has been developed, which combine the set of
     well-formed sentences into special forms ranging from “narratives” to the literary fields (short
     stories, novels, plays ...), to the scientific area (articles, scientific books, ...) .... Here, a set of
     weakly defined and, by necessity, fuzzy production rules have been worked out over the last
     centuries, leading to a repertoire of core-rules and core-requirements for each of these genres
     Finally, at the phenotype levels, a manifold of “language games” has emerged1,2,3, is
     continuing to emerge1,2,3 and will do so in the future while at the same time a large quantity of
     such language games falls into oblivion. As a NIS-relevant example, one may think of the
     enormous number of professions and their systems of practices which have effectively
     disappeared in the course of socio-economic evolution. (For an interesting summary, see
     PALLA 1994)

It seems appropriate, at this point, to recapitulate the previous rule levels with an enlightening quotation
from Daniel C. Dennett:

     Let´s consider ... a pandemonium of word-demons ... First we go into vocal noise-making mode
                                Beeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeep ....
     ... The internal „noise‟ excites various demons in us who begin trying to modulate the horn in all sorts
     of random ways by interfering with its stream. The result is gibberish, but at least it´s English
     gibberish (in English speakers)
                            Yabba-dabba-doo-fiddledy-dee-tiddly-pom-fi-fi-fo-fum ...
     But before any of this embarrassing stuff actually hits the outside world, further demons, sensitive to
     patterns in the chaos, start shaping it up into words, phrases, clichés ....
                        And so, how about that? baseball, don´t you know, in point of fact,
                        strawberries, happenstance, okay? That´s the ticket. Well, then ...
     which incites demons to make further serendipitous discoveries, augmented by opportunistic shaping,
     yielding longer bits of more acceptable verbage, until finally a whole sentence emerges:
                                  I´m going to knock your teeth down your throat!
     Fortunately, however, this gets aside, unspoken, since at the same time (in parallel) other candidates
     have been brewing and are now in the offering, including a few obvious losers, such as
                                                   You big meany!
                                            Read any good books lately?
     and a winner by default, which gets spoken:
                                                Your feet are too big!

       ... We can suppose that all of this happens in swift generations of „wasteful‟ parallel processing, with
       hordes of anonymous demons and their hopeful constructions never seeing the light of the day ...
       (DENNETT 1991:237p.)

Here, the metaphorical notion of a “pandemonium of demons” has been employed where each of the
demons is operating at a specific level, and where their interaction produces, after a period of internal
competition and selection, a specific output which fulfills apparently some optimality criterion.60 Thus, the
notion of a “grammar” G will be understood, in a slightly unusual manner, as the total set of production
rules G = {PRNL} for a natural language, ranging from syllable formation to the level of syntax and,
finally, to the level of language games. In short, a grammar will be conceptualized as the total set of
procedures to set the “pandemonium of demons” in a mode of successful operation.
Having introduced the term “grammar” in a very encompassing manner, the next concept, namely that of
programs, will be used in a non-standard way, too. Contrary to conventional definitions of programs as a
set of rules aimed at the solution of a particular problem, i.e. as an algorithm, which can be run or
executed on a computer (BREUER 1995:13), programs are to be understood in a much wider sense,
comprising any grammatically well-formed code-arrangements and its embedded relations. The following
set of clearly meta-theoretical definitions, applicable to biological as well as to socio-economic domains,
will be proposed:

       Potential programs - any well formed, grammatically valid string-arrangement of an
       embedded code-system ...
       Actual programs - any well formed, grammatically valid string-arrangement of an embedded
       code-system which is used within its embedded environment ...61
       Algorithmic programs - any well formed, grammatically valid string-arrangement of an
       embedded code-system which can be implemented and run on a computer ...
       Reproductive programs - any well formed, grammatically valid string-arrangement of an
       embedded code-system which has an essential and indispensable function in the reproduction
       of its embedded system - the paradigmatic example being the genetic program of organisms ...
       Non-reproductive programs - any well formed, grammatically valid string-arrangement of an
       embedded code-system which has no essential and indispensable function in the reproduction
       of its embedded system.
       Productive programs - any well formed, grammatically valid string-arrangement of an
       embedded code-system which has an essential and indispensable function in the production of
       its embedded system - the paradigmatic example being, once again, the genetic program of
       organisms ...
    It should be added that the sentence “Any spoken output is an optimal one” would be an outright tautology. Here, hoewever, the sentence
should be read as “Any spoken output is the result of an internal bidding process” which makes a substantial and highly non-trivial
contribution to the basic constitution of utterances.
    A library consists, following the above definitions, of a comparatively large amount of potential programs, with a small subset proper of
actual ones ...

     Non-productive programs - any well formed, grammatically valid string-arrangement of an
     embedded code-system which has no essential and indispensable function in the production of
     its embedded system.
     Programs of programs - any well-formed, grammatically valid string-arrangement of a well-
     formed, grammatically valid string-arrangement of an embedded code-system which is used
     within its embedded environment ...
     Programs of programs of programs of ... - any well-formed, grammatically valid string-
     arrangement of a well-formed, grammatically valid string-arrangement of a well-formed,
     grammatically valid string-arrangement of ... an embedded code-system which is used within
     its embedded environment ...

Turning now to the final notion, namely to that of genres, it should be emphasized, first, that “genres”
must be considered as a mixed genotype-phenotype concept, referring to “clusters” or “types” of specific
programs which are utilized in specific contexts. Thus, “scientific genres” are linked to the totality of
scientific utilization contexts in which programs are subject to additional, system-specific constraints.
Subsequently, four important scientific genres will be introduced which occupy the center stage in the
scientific output domain.

     Scientific articles - any program which exhibits the specific features of a scientific articles like
     an intersubjectively accessible transformation procedure from methods and other input
     devices (data, observations, etc.) to specific results, including bibliographic references,
     relevance to current debates and discussions within a scientific domain, consistency, etc.
     Research reports - any program fulfilling the main rule-components of research reports (a
     program, containing a well-defined research problem, an intersubjectively reproducible link
     between input devices (data, observations, ...) and results, recommendations for implementing
     changes, new programs for actions ...)
     Books - any program of sufficient length which offers the qualifications of a scientific
     publication, i.e. essentially the same characteristics as for a scientific article, but with
     additional qualifications like length, more detailed empirical or historical information, more
     background materials, more disgressions, any collection of programs, fulfilling the criteria of
     scientific articles, etc.
     Lectures - any program which is communicated to an audience either on a fully pre-arranged
     codified basis (“reading of a manuscript”), in a partially pre-codified manner (via
     transparencies, etc.) or as a real-time “narrative” without prepared encoded materials.

Having established the different domains for grammars, programs and genres, the final battery of
examples will be devoted to a combination of these three concepts which results, then, in the following

     Potential scientific program genres - any program which exhibits specific rules for scientific
     genres (a scientific article on innovation models, a project report on NIS, a book on the basic
     genres in transfer science ...)
     Actual scientific program genres - any program of an embedded code-system within a
     scientific genre which is used within its embedded scientific environment (a scientific article,
     a project report, a book on transfer science which are read, criticized, copied, used in the
     founding of new institutes, ...)
     Algorithmic scientific program genres - any scientifically relevant program within a well-
     defined scientific genre which can be implemented and run on a computer (a scientific article
     on formal models of evolutionary development, a book on estimation procedures for an
     evolutionary model or a lecture on an econometric test which can all be operated on a
     computer ...)
     Reproductive scientific program genres - any program of an embedded code system within a
     scientific genre which has an essential and indispensable function in the reproduction of its
     embedded scientific system (Here, one may think of a project-report on the constitution of
     National Innovation Systems which, under appropriate circumstances, may acquire an
     indispensable role in the future scientific reproduction processes ...)
     Non-reproductive scientific program genres - any program of an embedded code system
     within a scientific genre which has no essential and indispensable function in the reproduction
     of its embedded system. (A scientific article on economic cycles in the Late Roman Empire
     may serve as a paradigmatic example for a non-reproductive scientific program ...)
     Scientific program genres of scientific program genres - any program of an embedded code
     system within a scientific genre which is based on a specific scientific program (a scientific
     article, summarizing the content of a big research project, a summary of a comprehensive
     scientific book, a widely distributed “catch-phrase” from a single article ...)
     Scientific program genres of scientific program genres of scientific program genres of ... -
     any program of an embedded code system within a scientific genre which is based on a
     program of a program of .... a specific program within a scientific genre (a scientific article,
     summarizing the content of a big research report, focusing on major achievements of a large
     overall research project, a summary of a comprehensive scientific survey of the state of the art
     within a specific discipline, a widely distributed “catch-phrase” from an overview of a series
     of different articles summarizing large research reports respectively ...)

In this manner, an important number of different code-configurations has been introduced, ranging from
simple alphabets and strings to absolutely vital notions like “(re)productive programs”. The distinctive
advantage of the approach, chosen so far, lies primarily in two areas:

      First, the overall evolutionary background has remained in its full power, since programs both
      of the reproductive and the non-reproductive variety occupy the center stage at the genotype-
      Second, a meta-level perspective has been successfully retained which, even after the
      introduction of many code-specific terms like grammars, programs, strings et al., can be
      applied to many different levels, ranging from the genetic code to the world of Morse-codes,
      ASCII-codes, pictorial codes and, as will be shown in later chapters of Section II, social rule

Thus, a scientific article on innovation and diffusion models, a book on evolutionary theory or a research
report on National Innovation Systems - three of the many genres within the code-system of natural
languages - will be understood here as programs which enter into a manifold of relations within scientific
and extra-scientific “life-worlds”. After all,

      conscious human minds are more-or-less serial virtual machines implemented - inefficiently - on the
      parallel hardware that evolution has provided for us (DENNETT 1991:218) -

                                        1.3. A WORLD OF

                             EMBEDDED CODE-SYSTEMS

In a variation to Mario Bunges “A World of Systems” (1978), the “World of Embedded Code-Systems”
refers to a special feature of the biological or the socio-economic-environments since one of the probably
surprising points in the discussion on basic code-requirements lies in the fact that these five conditions are
sufficient to identify a large number of code-systems at various levels, ranging from the biological code,
to the code of alphabets, of musical schemes, of numbers and, finally, to the code of actions, human or
otherwise. In order to avoid a detailed discussion, three simple schematic summaries must be sufficient to
account for the deeper meanings in the heading of the present chapter. Table 1.3 presents, first, a long co-
evolutionary history of code-based (re)productions where three additional comments must be added:

      First, the basic configuration of Table 1.3 follows directly from the dual level-scheme of
      evolutionary systems which have been laid out in Section I.
      Second, the encoding of the genetic code into the codes of natural and scientific languages
      may be considered as a most significant co-evolutionary turning point since a new type of an

     interlinked triangle between the development of the genetic code, of human code-systems and
     human societies has been set in motion.

     Diagram 1.1. The Code Triangle

     Third, with the help of Table 1.4 (next page), the overwhelming importance of code-systems
     for the constitution and the reproduction of human lifeworlds or, pace Habermas, for socio-
     economic systems should become transparent.

Table 1.3: The Long Co-Evolutionary Chains of Becoming

PHENOTYPE LEVEL                       Complexification of the “extended phenotypes” 

                   Eucariotic cells      Cell Assemblies    Organisms     Code producing Organisms

                                                                                        

                   Genetic Codes1        Genetic Codes2     Genetic Codes3         Genetic Codes4

GENOTYPE LEVEL                          Complexification of the genetic codes 

Table 1.4: A World of Embedded Code-Systems

DOMAIN                    CODE-SYSTEM            CODE-ELEMENTS                            ORGANIZATION

BIOLOGICAL                Genetic                   Four Bases: Adenin,                       Double Helix-
                          Code                      Cytosin, Guanin, Thymin                   Configuration

HUMAN                     Natural                   Letters of an                             Grammars
“LIFEWORLDS”              Languages                 Alphabet

HUMAN                     Number                    Sets of Various                           Algorithms
“LIFEWORLDS”              Codes                     Numbers {}, {},

HUMAN                     Pictorial                 Symbols from a                            Picture
“LIFEWORLDS”              Codes                     Symbol-Library                            Programs62

HUMAN                     Musical                   Musical Notes                             Musical
“LIFEWORLDS”              Codes                                                               Schemes

HUMAN                     Rule-                     Rules as Parts of                         Action-
“LIFEWORLDS”              Codes                     “Language Games”                          Programs

                                           1.3.1. CODE - SPACES

The next basic feature of code-systems to be discussed at some length consists in the elaboration of a
basic feature of any code-system, namely the configurations of code-specific spaces. The following
quotation, focusing on the genotype-space of the genetic code serves as an useful introduction both for the
concept of spaces and for one of the peculiar features of code-spaces, namely the striking discrepancy
between the potential space-dimensions and the actually occupied space-regions:

    So far, very few explicit picture schemes are available at the moment, one of the most prominent being ISOTYPE (International System
of Typographic Education) by Otto Neurath, Gerd Arntz et al. in the 1930´s. (MÜLLER 1991a,b) It should be added though that the early
code-systems within human history had been devised as pictorial or symbolic codes (WHITE 1995, CALVIN 1996)

       Suppose we consider a haploid organism with N different genes, each of which comes in two
       versions, or alleles, 1 and 0. With N genes, each with the two alleles, the number of possible
       genotypes is now familiar: 2N. The bacterium E.coli living merrily in the intestines, has about 3000
       genes; hence its genotype space might have 23000 or 10900 possible genotypes. Even PPLO,
       pleuromona, the simplest free-living organisms with only 500 to 800 genes, has between 10150 and
       10240 potential genotypes. Now consider diploid organisms, with two copies of each gene. Plants may
       have 20.000 genes. If each has two alleles, the number of diploid genes is 2 20000 for the maternal
       chromosomes times 220000 for the paternal chromosomes, or about 1012000. Clearly, genotype spaces are
       vast; even for a modestly long genome, there are an astronomical number of states it can assume. Any
       population of a species represents at any time a very, very small fraction of the space of its possible
       genotypes. (KAUFFMAN 1995:163f.)

With this quotation, it becomes relatively straightforward to introduce the concept of “code-spaces” since
it can be defined for all code-systems as the set of all possible configurations of code-characters. Thus,
taking the 26 letters of the alphabet, and strings of four characters only, the resulting code-space is of the
moderate size 264 or 456976. Allowing, however, a number of strings in the range of 106 (like the present
overall report) and, moreover, roughly 74 additional code-components like {, . ? “ - „ ( )   { }     
 ... } and the extremely important character for an empty space, the resulting code-space CS becomes
transcomputable due to its size of CS  1010 000 000. Thus, it is, by (almost) all standards, certain that the
present explorations on code spaces and on the genotypes of National Innovation Systems will remain
highly original since it is impossible to be reproduced in exactly this format, including the following two
typing erors (the second one of them is self-referential!) ...
Thus, the most important point with respect to the construction of code-spaces lies in the specification of
the code-elements E and the number of code-strings N which, then, define the code space CS as -

                                                              CS = NE

It is important to add that code-spaces can be constructed within any type of embedded code system,
including the genetic code, natural language, music, societal rules and the like.
Moreover, codes can be operative on a multiplicity of levels, generating, aside from primary structures,
secondary, tertiary, quartiary ... levels which contain regular code-sequences, too. Take, as an extremely
simple example, the following diagram with an alphabetic code in which, on the primary level, small M´s
produce, at the secondary level, thirteen large M´s which, finally, are combined on the tertiary level to a
very large, though hardly recognizable U.63
   For a definitely more elegant version, consisting of four levels, see D.R. HOFSTADTER (1985b:563) where the following messages
have been coded at the four different levels:

       Die Globalstruktur - bestehend aus den Buchstaben 'M' und 'U' - ist eine Quartärstruktur, dann besitzt jeder dieser Teile
       eine Tertiärstruktur, die aus 'HOLISMUS' und 'REDUKTIONISMUS' besteht; dann existiert das entgegengesetzte Wort auf
       der Sekundärebene, und ganz unten ist die Primärstruktur, wieder das Wort MU, endlos wiederholt. (HOFSTADTER

The distinctive advantages offered via secondary, tertiary, ... n-ary levels need hardly be emphasized. Two
main groups of advantages will be briefly mentioned.

       First, multiple levels allow for a highly differentiated network of control-relations since
       higher level strings and programs can serve as control instances for lower levels. Through
       level differentiation, higher levels can act on lower ones which, in turn, take influence on the
       lowest levels ...
       Second, multiple level organization makes a dense encoding of “information” possible which
       would be impossible to achieve in an accumulative single level-manner.64 Even in the simple
       Diagram 1.2., the secondary and tertiary levels have added an analogical surplus amount of
       M´s and U as well as, depending on the specific grammatical rules, a large amount of
       additional letters like V´s, I´s, H´s, etc.

Diagram 1.2: Primary, Secondary and Tertiary Levels within an Alphabetic Code-System

                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM      M               M   MM                      MM                 M               M   MM
                   MM          M       M       MM                      MM                     M       M       MM
                   MM              M           MM                      MM                         M           MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM      M               M   MM                      MM                 M               M   MM
                   MM          M       M       MM                      MM                     M       M       MM
                   MM              M           MM                      MM                         M           MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM      M               M   MM                      MM                 M               M   MM
                   MM          M       M       MM                      MM                     M       M       MM
                   MM              M           MM                      MM                         M           MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM      M               M   MM                      MM                 M               M   MM
                   MM          M       M       MM                      MM                     M       M       MM
                   MM              M           MM                      MM                         M           MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM                          MM                      MM                                     MM
                   MM      M               M   MM                      MM                 M               M   MM
                   MM          M       M       MM                      MM                     M       M       MM
                   MM              M           MM                      MM                         M           MM
                   MM                          MM                      MM                                     MM
                   MM                          MMMM                  MMMM                                     MM
                   MM                          MMMM                  MMMM                                     MM
                   MM      M               M   MMMM        M       M MMMM                 M               M   MM
                   MM          M       M       MMMM          M   M   MMMM                     M       M       MM
                   MM              M           MMMM            M     MMMM                         M           MM
                   MM                          MMMM                  MMMM                                     MM
                   MM                          MMMM                  MMMM                                     MM

    An instructive example can be provided by early systems of counting large quantities of animals where a multi-level organization
managed to encode even four or five digit numbers of cattle. How? Simply by one person being responsible for counting from 1 to 9, a
second person for the registration of two digit occurences 10, 20, 30, ..., a third person being in charge of three digit numbers, etc. For an
extremely interesting history on number systems from an evolutionary point of view, see esp. IFRAH (1989).

In this manner, hierarchical rules and dense encodings can take place, transcending the one-dimensional
and linear forms which have become typical for natural language encodings.

                             1.3.2. FITNESS LANDSCAPES

Having established a multiplicity of levels within code spaces of a given code-system, the next step
consists simply in the attribution of fitness-measures for various code-spaces. In a metaphorical
introduction, code-spaces can be seen, alternatively, as a specific landscape of an embedded code-system
which should be evaluable in terms of a fitness-function , a basin of attraction S in which (S) = 1 and
an evaluation measure which, in its simplest form, involves only a mapping from strings to 0 and 1.

     At S, where (S) = 1, it (the evaluation measure, K.H.M.) orders the system simply to stop; at all
     other strings, the command is to mush on. (BERLINSKI 1986:321)

In (almost) all cases of embedded code-systems, the evaluation measure must become more diversified,
differentiating not only between 0 and 1 but between small intervals within the 0-1 domain. Thus, fitness
should not be considered -

        an all-or-nothing affair;  thus takes values, let us say, between 0 and 1. Scanning every new
        string, the evaluation measure selects those strings si such that (si)  (si-1 ). These the systems
        retain until it finds a string superior in point of fitness. The result is a sequence of strings that
        ascends in fitness. At S, as before, the system stops.(BERLINSKI 1986:322)

For biological and, a fortiori, for societal domains, the notion of fitness landscapes can be given no
precise and well-defined meaning since -

     „fitness‟ applies principally to an entire organism. It has components of fecundity, fertility and other
     factors leading to reproductive success. These include complex issues such as the frequency of each
     genotype variant of the organism in the popuzlation, the density of each genotype variant in a region,
     and even the entire ecosystem with which each organism interacts. Therefore, in the general context,
     it is difficult to assign a fitness to a gene or even to a genotype, since all these factors depend upon the
     other organisms in the population. (KAUFFMAN 1993:37)

Nevertheless, the search for appropriate fitness landscapes must be considered as a highly desirable
research goal, at least within the contexts of explorations ...
“Coupled” or, alternatively, “rugged fitness landscapes” are built up, not surprisingly as configurations
with code-strings of length N - genes in the genotype, amino acids in a protein, letters in words, words in
a sentence, sentences in stories, musical notes in a tune where -

     each part makes a fitness contribution which depends upon that part and upon K other parts among
     the N. That is, K reflects how richly cross-coupled the system is. In the geneticist´s term, K measures
     the richness of epistatic interactions among the components of the system. (IBID:40)

It would go well beyond the scope of the present conceptual introduction of fitness landscapes to present a
summary of the basic features of fitness landscapes between the extremes of K = 0 with unipeaked
configurations and highly erratic, random landscapes with a multiplicity of small local peaks for K = N -
1. Suffice it to say that highly intriguing and fascinating dynamic features of adaptations and “hill
climbings” can be recorded especially in the case of small K´s. (K = 2, 3, ..) (KAUFFMAN 1993, 1995)
“Multiply coupled fitness landscapes”, the configuiration of highest relevance in the case of National
Innovation Systems, consists of ensembles, integrating fitness landscapes at the genotype as well as on
the phenotype levels. Here, the NK-structure is replaced by at least a multiply coupled NGKG(NPKP)-type
where, in a variation to Kauffman´s NK-decription -

     each part at the genotype level makes a fitness contribution which depends upon that part and upon
     KG other parts among the NG at the genotype level as well as on phenotype parts NP and their co-
     determinative factors KP. That is, KG, KP reflect how richly cross-coupled the systems at the genotype
     and phenotype level are. In reproductive systems, KG and KP measure the richness of epistatic
     interactions among the components of the system.

One point should be added to the above recombination of Kauffman´s original quotation. Especially in
socio-economic systems with their multiplicity of embedded code-systems (ECS) the coupling of fitness
landscapesG and fitness landscapesP may seem an unusual or even unwarranted phenomenon. The next
chapters and especially Section III on different forms of ECS will make it clear that the rapid successes in
the evolution of socio-technical systems is directly linked to the cross-level adaptation processes within
genotype variations and phenotype recombinations - and vice versa ...

                                  1.4. BASIC DYNAMICS

                        FOR EMBEDDED CODE-SYSTEMS

Having established the smooth, the rugged or even the bizarre fitness landscapes for any given code-
system, the next point will elaborate some highly characteristic dynamic features of embedded code-
systems, viz. the emergence of code-specific drifts. The essential question asks simply for the principal
ways of movement, of “hill-climbing” and of “hill-descending” within fitness landscapes at the genotype
level. Very generally speaking, the following four conditions can be identified which become necessary
for any dynamic investigation of embedded code systems.

     First, embedded code systems, especially more complex ones like those constitutive of
     vertrebrates or used by humans, are processed within a recursive, internal state-determined
     organization, in which the output of internal processings becomes the most essential input for
     subsequent operations. With respect to the genetic code, the closed production cycle has been
     identified, under the heading of the “Central Dogma of Molecular Biology” (Francis Crick), as
     the prime organizational feature. But also with respect to human code-processings like
     reading, probably the most astonishing characteristic from the viewpoint of cognitive
     architectures lies in the almost totally internal organization of the relevant processes of de-

        Wir sind weitgehend auf unser eigenes System eingestimmt. Das Verhältnis von internen und
        externen Sensoren beträgt 100.000:1. Das bedeutet, daß auf jedes Stäbchen oder jeden Zapfen
        in der Netzhaut des Auges, die die Funktion haben, externe Stimuli (Photonen) wahrzunehmen,
        100.000 Reizpunkte kommen, die auf interne Stimuli reagieren. (SEGAL 1988:118)

     Second, an important requirement for an ECS lies in the nature of adaptation processes,
     distinguishing very generally between an asymmetric pattern of “early experimentation and
     later standardization”. Following Stephen Jay Gould (1989), an ECS must be seen as a very
     special survivor in the lotteries called “Life” -

        Major lineages seem able to generate remakable disparity of ... design at the outset of their
        history - early experimentation. Few of these designs survive an initial decimation, and later
        diversification occurs only within the restricted ... boundaries of these survivors - later
        standardization ... Maximal initial disperity and later decimation give the broadest possible role
        to contingency, for if the current taxonomic structure of life records the few fortunate survivors

   in a lottery of decimation, rather than the end result of progressive diversification by adaptive
   improvement, then a replay of life´s tape yould yield a substantially different set ... and a later
   history making perfect sense in its own terms but markedly different from the one we know.
   (GOULD 1989:304)

Thus, survivor ECS are definitely not the final result of a long “cone of increasing diversity”,
capitalizing step by step on its partial successes by staying at the levels of fitness reached so
far and, consequently, by exploring and accepting new local solutions with a comparatively
higher degree of fitness.
Third, a successful ECS will proceed in all probability in a de-composite fashion, safe-
guarding the system from too powerful disturbances from outside. For a sustainable tinkering
process, a prime requirement consists, phrased in a post-modernist fashion, in the recursive
de-construction of the overall outside “influences” into small subsets with comparatively
fewer and simpler modes of changes and with relatively short time-intervals for the successful

   Das Erstaunliche ... ist, daß ... ein Stäbchen oder Zapfen auf der Retina des Auges ... nur die
   Sprache 'Klick' sprechen: Die physikalische Ursache der Erregung einer Nervenzelle ist nicht in
   ihrer Aktivität enthalten, sondern ausschließlich die Intensität der Störung, die ihre Aktivität
   verursachte. Die Signale, die dem Gehirn zugeführt werden, sagen also nicht 'blau'...., sondern
   'Klick','Klick','Klick', d.h. sie sprechen nur von der Intensität einer Störung und nicht von 'was',
   nur von 'wieviel' und 'woher'. (FOERSTER 1987:138f.)

The most prominent metaphor for such an evolutionary path of independently distributed
errors and successes has been provided by Herbert A. Simon in his highly illuminating
narrative of the two watch-makers or, in a typical “variation on a thema”, of two

   Two programmers assemble fine programs, each program containing ten thousand lines. Each
   programmer is interrupted frequently to answer the phone. The first has organized his total
   assembly operation into a sequence of subassemblies; each subassembly is a stable arrangement
   of 100 lines, and each program, a stable arrangement of 100 sub-programs. The second
   programmer has developed no such organization. The average interval between phone interrup-
   tions is a time long enough to assemble about 150 lines. An interruption causes any sequence of
   lines that does not yet form a stable system to fall apart completely. By the time he has
   answered about eleven phone calls, the first programmer will usually have finished assembling

            a program. The second programmer will almost never succeed in assembling one - he will
            suffer the fate of Sisyphus: As often as he rolls the rock up the hill, it will roll down again.65

        Finally, a fourth requirement for a successful ECS can be identified by demanding a
        distributed solution for the so-called frame-problems. Since the two central terms in the last
        sentence - “frame problem” and “distributed solution” - require special attention, another
        metaphor will serve as a useful basis for the introduction of the peculiar notion of “frame
        problems”. The story starts with an artificially designed roboter, having to fulfill a specific
        task set, namely to fend for itself.

            One day its designers arranged for it to learn that its spare battery, its precious energy supply,
            was locked in a room with a time bomb to go off soon. R1 located the room, and the key to the
            door, and formulated a plan to rescue its battery. There was a wagon in the room, and the
            battery was on the wagon, and R1 hypothesized that a certain action which it called PULLOUT
            (WAGON, ROOM) would result in the battery being removed from the room. Straightway it
            acted, and did succeed in getting the battery out of the room before the bomb went off.
            Unfortunately, however, the bomb was also on the wagon. R1 knew that the bomb was on the
            wagon in the room, but didn't realize that pulling the wagon would bring the bomb out with the
            battery. Poor R1 had missed that obvious implication of its planned act.
            Back to the drawing board. 'The solution is obvious', said the designers. 'Our next robot must be
            made to recognize not just the intended implications of its acts, but also the implications about
            their side effects, by deducing these implications from the descriptions it uses in formulating its
            plans.' They called their next model, the robot-deducer, R1D1. They placed R1D1 in much the
            same predicament that R1 had succumbed to, and as it too hit upon the idea of PULLOUT
            (WAGON, ROOM) it began, as designed, to consider the implications of such a course of
            action. It had just finished deducing that pulling the wagon out of the room would not change
            the color of the room's walls, and was embarking on a proof of the further implication that
            pulling the wagon out would cause its wheels to turn more revolutions than there were wheels
            on the wagon - when the bomb exploded.

     Simon´s original “seeing watchmaker”-metaphor has, as is well known, the following format -

        Two watchmakers assemble fine watches, each watch containing ten thousand parts. Each watchmaker is interrupted
        frequently to answer the phone. The first has organized his total assembly operation into a sequence of subassemblies; each
        subassembly is a stable arrangement of 100 elements, and each watch, a stable arrangement of 100 subassemblies. The
        second watchmaker has developed no such organization. The average interval between phone interruptions is a time long
        enough to assemble about 150 elements. An interruption causes any set of elements that does not yet form a stable system to
        fall apart completely. By the time he has answered about eleven phone calls, the first watchmaker will usually have finished
        assembling a watch. The second watchmaker will almost never succeed in assembling one - he will suffer the fate of
        Sisyphus: As often as he rolls the rock up the hill, it will roll down again. (SIMON 1977:248)

Thus, for Sisyphus to become an evolutionary “lucky man”, his main task would have consisted in decomposing the stone into a number of
sub-components and in carrying, piece by piece, the stone on the top of the hill ..

             Back to the drawing board. 'We must teach it the difference between relevant implications and
             irrelevant implications', said the designers, 'and teach it to ignore the irrelevant ones.' So they
             developed a method of tagging implications as either relevant or irrelevant to the project at
             hand, and installed the method in their next model, the robot-relevant-deducer, or R2D1 for
             short. When they subjected R2D1 to the test that had so unequivocally selected its ancestors for
             extinction, they were surprised to see it sitting, Hamlet-like, outside the room containing the
             ticking bomb, the native hue of its resolution sicklied o'er with the pale cast of thought, as
             Shakespeare (and more recently Fodor) has aptly put it. 'Do something!' they yelled at it. 'I am',
             it retorted. 'I'm busily ignoring some thousands of implications I have determined to be
             irrelevant. Just as soon as I find an irrelevant implication, I put it on the list of those I must
             ignore, and ...' the bomb went off.
             All these robots suffer from the frame problem. (DENNETT 1986b:129p.)

         While the story, its underlying problems and, most important, the possibility for successful
         solutions seem to be ill-defined or, to be more precise, of a typical NP-structure66, they point
         to a special requirement for embedded code-systems in general, namely to the availability of
         self-organization mechanisms or, alternatively, distributed solutions for code-based
         aggregation or integration problems. While, by necessity, a successful solution to the “frame-
         problem” cannot be provided, it can be demanded, nevertheless, that embedded code-systems
         have, at the level of their “embedded code-carriers” a rich repertoire of evolutionary fitting
         strategies for combining and integrating different code-sequences. Thus, the problems for the
         robot generation R1, R2, etc. lies primarily in the finding of salient selections, depending on
         the distribution of “drafts” or “tasks” as well as on the perceived states of the enviroment. It
         will become one of the targets within the Section III of the theoretical part to identify a
         number of pre-conditions for the possibilities of arriving at salient solutions.

In this manner, four basic conditions have been set up to “switch” a code-system into an embedded
dynamic configuration, generating its highly specific drifts and, moreover, its reproductions and
recombinations at the level of its embedded code-environment.

                                           1.5. THE COMBINATION

                                    OF EMBEDDED CODE-SYSTEMS

     On this point, see especially the final chapter of Section III.

A final characteristic already visible in Table 1.4 lies in the combination of embedded code-systems into
integrated ensembles, consisting of two or three well defined ECS. Consider the following examples
which may seem trivial at first, but which turn out to be highly interesting when viewed from a code-
based perspective.

       Performing a work task in an organization (like supervising a machine-controlled production
       process), making a telephone call and drawing symbols on a piece of paper ...
       Doing scientific routine work (like the writing of the present report), listening to music
       through headphones and thinking of radically new research designs for National Innovation
       Systems ...
       Driving a car from A to B (as part of a work-task), listening to music (background) and
       communicating at the same time ...

These three examples exhibit the possibilities for code-combinations at the phenotype level. At the same
time, these three seemingly trivial instances make it clear that the integration of different code-based
activities has its definite capacity limits and barriers. Take the following variations to the above cases -

       Performing a new work task in an organization (like supervising a new machine-controlled
       production process), neglecting the ring of the telephone and making notes of the observed
       process on a piece of paper ...
       Doing creative scientific work in total silence and refusing to discuss some scientific problem
       with a colleague ...
       Driving a car from A to B (as part of a work-task) under adverse weather-conditions like
       heavy fog and restricting communication processes to the road conditions ...

These three modifications should be sufficient to demonstrate, on the one hand, the inherent limitations in
complex code-processing capacities. On the other hand, the long-term technological development pattern
clearly points into the direction of integrating embedded code-systems into new comprehensive forms.
This point, while extremely interesting, must be confined to a single sentence on the emergence of new
code-systems which run under the headings of “multi-media systems”, “hyper-text” and the like where the
combination of code-systems will be used, almost by necessity, for an evolutionary hill-climbing with one
of the most efficient tools, namely via the help of the multiplicity of level-device, creating, “as we go
along”, secondary, tertiary, ... n-ary plattforms for encodings ...67

   A highly intriguing history could be written by focusing on the phase transitions from single code-based activities to multiple ones.
Radio and television, to mention two prime examples, are well in the process of changing into a mode of easy combinability with different
code-based activities like reading, talking, writing and the like ...

                                               2. RECOMBINATIONS

                                   IN EMBEDDED CODE-SYSTEMS

The next major area of investigation consists in the universal problem of recombinations at various levels
of embedded code-systems (ECS). Sticking to the format of a meta-theoretical framework, the following
pre-requirements must be laid out:

        First, one must be prepared to find a universal mode of recombinations across different code
        levels ...
        Second, the recombination mechanisms can be described in a homogeneous manner across
        different code levels ...
        Third, one may identify, in one of the most fascinating “second-order” research fields (Heinz
        von Foerster), namely in the area of the “evolution of evolution”, an evolutionary pattern, too,
        leading, over the extremely long run, to a more refined and complex recombination repertoire.
        Nevertheless, the clockworksfor changes of the second-order type, are set at a vastly slower
        pace than the evolutionary changes and development patterns in organisms, ecosystems, let
        alone in the code-based areas of human life-worlds.

The starting point lies in a definition of recombinations across the multiplicity of code-system levels.
Accordingly, the following set of requirements must be fulfilled for code-based changes in any type of
evolutionary system:

        Full-scale change potential for an embedded code-system consists in having a rich repertoire for
        recombinations, following them recursively, applying them at the meta-level, and modifying them
        accordingly. 68

According to the definitions of innovations and denovations in Section I, one can immediately add the
following statements.

     The sentence above is a variation on a dfinition which Douglas R. Hofstadter has proposed for “creativity” -

        Full-scale creativity consists in having a keen sense for what is interesting, following it recursively, applying it at the meta-
        level, and modifying it accordingly. (HOFSTADTER 1995:313)

It will become one of the main targets within the present recombination-chapter to demonstrate the very close “family resemblances”
(Ludwig Wittgenstein) between recombinations at different levels of code-systems, including, especially, the phenomenon of scientific
creativity as a particular case in question.

     Full-scale innovation potential for an embedded code-system consists in having a rich repertoire for
     recombinations with comparative advantages, following them recursively, applying them at the meta-
     level, and modifying them accordingly.

Likewise, the opposite direction can be defined as well -

     Full-scale denovation potential for an embedded code-system consists in having a rich repertoire for
     recombinations with comparative disadvantages, following them recursively, applying them at the
     meta-level, and modifying them accordingly.

The six basic requirements for innovations or denovations in embedded code-systems which can be
extracted from the three definitions above need further elaborations and more specific in order to render
them empirically accessible.

                               2.1. CODE-SYSTEMS AND


                                   THE GENERAL CASE

For the general case, the recombinations in embedded code systems across levels, one can identify six
conditions which must be present simultaneously.

     The first set of basic requirements is marked by the “rich repertoire-condition“ which states
     that successful recombinations are dependent on a “requisite variety” (Ross Ashby) of the
     embedded code system. In other words, an embedded code-system with only random
     mutations as sole source of recombinations must be considered as a very poorly equipped
     recombination repertoire, whereas a “pandemonium of recombinative demons” across
     different levels fulfills the first requirement in an optimal way.
     The „rich repertoire-requirement“ needs, second, the availability of code-spaces, which should
     have, in the general case, a single distinctive feature, namely a comparatively large area of
     unrealized code-sequences and, thus, a high potential for new sequences.

        The central area for recombinations resides, however, in the third requirement, namely in the
        availability of recombination operators which are able to generate in a recursive manner,
        starting from an initial scheme, new code-strings or programs. For the general case, one is
        able to distinguish at least ten recursive operators which, following mostly Douglas R.
        Hofstadter (1995:77), can be recombined by using some “adding operations” and which, then,
        can be subsumed under the following headings69-

            Adding, the integration of new building blocks into an existing scheme ...
            Breaking, the differentiation of at least one scheme into two disjunctive building blocks ...
            Crossing-over, the breaking of at least two schemes and their merging into a new ensemble ...
            Deletion, the destruction of a specific building block from a set of schemes ...
            Duplication, the repeated insertion of at least one identical scheme ...
            Inverting, the making of copies with an opposite sequence of elements ...
            Merging, the integration of at least two existing schemes into a new one ...
            Moving, the shifting of code-elements or of established boundaries ...
            Replacing, the substitution of a code-element by another one ...
            Swapping, the movement from a level Li to a different level Lj ...

        The important point which cannot be over-emphasized lies in the universality of these
        recombination operations across various embedded code-systems.

            At the level of the genetic code, operations like inversion, adding, crossing-over or
            replacing occupy, as will be shown later on, an important role in the generation of new
            Within natural languages, recombinations occur in practically all utilization contexts,
            leading to the creation of new words by merging, adding, replacing ... operations, of
            replacing words in very specific contexts with new inverted ones70, of recombining
            phrases into new ensembles via moving, swapping, crossing-over ...
    In the following enumeration, terms like “building blocks”, “scheme” or “code-elements” will be used simultaneously. In order to avoid
a possible misunderstanding, it should be added that these three expressions refer to different degrees of complexity in an embedded code-
systems, from simple code-elements like letters to code-strings up to the level of programs or programs of programs ...
    Take, as but one example from the language of children, the shift from “hot” and “absolutely hot” to “cool” and “ultra-cool” which,
apparently, occupies the central position at the present time. For an interesting cultural inversion, one may refer to the classical goal domains
way above, exemplified, for example, in Goethe´s “Prologue” -

                                                      Denn das ist der Kunst Bestreben,
                                                       Jeden aus sich selbst zu heben,
                                                        Ihn dem Boden zu entführen;
                                                      Link und recht muß er verlieren,
                                                         Ohne zauderndes Entsagen;

           In social rule-systems, a recombinative operation like crossing-over can be identified,
           for example, in a surprising number of ways, yielding two different and, following
           Holland´s “schema-theorem” (1992), probably more useful specialized rules ...
           Within the scientifc system, the recombinations, leading to new research programs
           follow, as will become clear in the final chapter of the present section, operations like
           swapping, merging, adding, replacing ... too, whereby new and potentially more fruitful
           problem solutions, research trajectories, research agenda or research agenda for research
           agenda ... come into view ...
           In the music domain, “variations on a thema”, aside from being a musical genre in itself,
           offer a powerful route for the re-arrangement and re-configuration of music-programs ...
           In socio-technical systems, operations like adding, breaking, replacing or merging play,
           quite naturally, a central role in the construction, adaptation and reconfiguration of
           established programs into new ones ...

       The requirements four and five demand a sufficient degree of flexibility - a capacity to salient
       adaptations (requirement four) - as well as of efficiency in approaching the target domains
       within a relatively small amount of time (requirement five).
       Finally, a control-capacity as well as a sufficiently powerful support system must be present
       which are not only able to secure the partial gains reached so far, but which, furthermore,
       develop at least some “gate-keeping”-functions and safe-guards against detrimental
       trajectories (requirement six) ...

This small set of examples has paved the way, hopefully, for a truly meta-theoretical or, alternatively,
transdisciplinary recombination perspective, relying entirely on code specific strings and programs and on
recombinative operations which can be identified throughout the spaces of embedded code-systems.

                                                    Aufwärts fühlt er sich getragen!
                                                    Und in diesen höhern Sphären
                                                    Kann das Ohr viel feiner hören,
                                                     Kann das Auge weiter tragen,
                                          Können Herzen freier schlagen. (GOETHE 1977:650) -

to romantic target areas “in the deep” like Nietzsche´s romantic summary in his “Also sprach Zarahthustra” -

                                                       Denn alle Lust will Ewigkeit,
                                                        will tiefe, tiefe Ewigkeit ...

                                        2.2. CODE-SYSTEMS

                                  AND RECOMBINATIONS

                                        IN A MULTIPLICITY

                              OF EMBEDDED ENSEMBLES

The following chapters will recapitulate the main points and features of recombinations in embedded
code-systems by applying them to five different domains, namely to the genetic code, to action-patterns in
animals, to the field of natural languages, to the realm of scientific languages and, finally, to the domain
of social rule-systems. Thus, the main target of this chapter will lie in a demonstration of the usefulness of
the conceptual apparatus developed so far, despite the long distances which separate a morphogenetic
field in biology from the present concerns for National Innovation Systems. Nevertheless, it will turn out
as a highly rewarding research task to continue the construction of a meta-evolutionary approach on the
constitution and on the development patterns in complex embedded systems since an astonishing number
of similarities between biological genetics on the one hand and social genetics on the other hand will be
brought to light.
The following table (see Table 2.1, next page) will give an introductory overview of at least one
surprising equivalence which can be identified between so diverse components as socio-technical systems,
molecular biology and mathematics. Following Table 2.1, one will detect in all probability a rather
unexplored feature of socio-economic systems, namely their necessary “blind spots”. (For more details,
see HOFSTADTER 1982:532pp.)
In this manner, a rare type of a Gödelian exploration into the incompletenesses of socio-economic
systems could be set in motion. It might well be that the dense amount of examples on systems failures
(see esp. GALL 1990) has as its common theoretical background an “Incompleteness Theorem” of the
format -

                        “There always exists an unreproducible (genotype) program,
                                 given a particular (phenotype) system” ...

And this “Incompleteness Theorem” can be applied in domains as diverse as institutions, bacteria, firms,
machines or scientific institutes ...

Table 2.1: Socio-Biological Genetics - A Gödelian Survey

MATHEMATICS                MOLECULAR BIOLOGY                       TECHNO-SOCIAL                   SOCIO-TECHNICAL
                                                                   SYSTEM                          SYSTEM

Axiomatic Number           Cells                                   Phonograph                      Organization
Theory-System (NT)
“Perfect” System           “Perfect” Cell                       “Perfect” Phonograph            “Perfect” Organization
String of NT               Strand of DNA                           Record                          Rule
String deducible           Strand of DNA repro-                    Record playable                 Task performable
by a given NT              ducible by a given cell                 by a given phonograph           by a given organization
String not deducible       Strand of DNA not repro-                Record unplayable               Task not performable
by a given NT              ducible by a given cell                 by a given phonograph           by a given organization
Process of interpreting    Process of transcription                Process of converting           Process of converting
NT  Numbers               of DNA onto mRNA                        record grooves into sounds      rule-sequences into work
Arithmetization            Translation of mRNA                     Translation of sounds into      Transformation of rules
N  meta-NT                into proteins                           vibrations of phonograph        into tasks
Gödel-Code                 Genetic Code                            Phonograph Code                 Organization Code
(mapping from triplet      (mapping from mRNA                      (mapping from external       (mapping from rules
of digits onto meta-NT)    triplets onto amino acids)              sounds onto vibrations of       onto organizational tasks)
Inconsistency of NT        Destruction of the cell                 Breaking of phonograph       Dissolution of an
“Imperfect” NT             “Imperfect” Cell (a cell for            “Imperfect” phonograph       “Imperfect” organization
                           which there exists at least one         (a phonograph for which         (an organization for which
                           DNA strand which it cannot           there exists at least one re-   there exists at least one
                           produce)                                cord which it cannot         set of rules which it cannot
                                                                   reproduce)                      reproduce)

                                   2.3. THE GENETIC CODE

The main focus in the section on the genetic code lies in a brief discussion of some of its main
characteristic features, especially its recombinative capacities. In doing so, the close family resemablances
between reconfigurations at the genotype levels of biological and socio-economic systems should become
more transparent and recognizable. Of its central characteristics, some of them are too well known to be
recapitulated here in greater detail. Consequently, they will be compressed into a small listing, starting
with the code-elements.

       The genetic code consists, first, of triplets from four bases, adeninin, cytosin, guanin, uracil,
       which form a codon like UUU, GGG, AUU, CCU and the like ...
       Second, triplets are the necessary basis for producing one out of a total of twenty amino acids
       like glycin, alanin, prolin, valin, isoleucin, leucin, serin, histidin, arginin and the like ...
       Third, the gentic code is a degenerate one, since the codons GGU, GGC, GGA and GGG
       encode the production of glycin, GCU, GCC, GCA, GCG the “genofacturing” of alanin ...71
       Fourth, the sequence of codons follows a continuous pattern without blanks or commas
       between the different triplets. What is present however, is a set of termination triplets, mainly
       starting with uracil bases like UAA, UAG or UGA ...
       Fifth, the genetic code is almost context-free and, thus, almost universal. In recent years, some
       deviations to the universal encoding schemes from triplets to amino acids have been
       identified, for example in mitochondrial DNA of humans where AGA and AGG which
       normally encode the production of arginin act as a stop-symbol.

With respect to the recombinative capacities of the genetic code, one may distinguish, aside from single
random mutaions and random replication errors, between two modes of reconfigurations.

       On the one hand, the genetic program for the production of new phenotypes is the result of a
       standard procedure of crossing over, i.e. the breaking of two strands of chromosomes and
       their merging into a new ensemble.
       On the other hand, one finds at least five additional systematic ways of rearrangements in the
       genetic program.

Sticking to the general recombination operators, introduced above, they can be subsumed under the
following headings (HENNIG 1995:485) -

           Breaking or Translocation, the breaking of at least two schemes or chromosomes into new
           building blocks ...
           Deletion, the destruction of a specific building block from a set of schemes ...
           Duplication, the repeated insertion of at least one identical scheme ...
           Inverting, the making of copies with an opposite sequence of code-elements ...
           Merging or, Fusion, the integration of at least two existing schemes into a new one ...
           Moving or Transposition, the shifting of building blocks into a new place ...

In this manner, a surprisingly rich repertoire has been identified which is responsible for the persistent
genotype changes in plants and organisms, simple, complex and otherwise ...

    It is highly informative to note that even in the case of the genetic code, the semantic requirements for code-systems which have been put
forward by Nelson Goodman, are not fulfilled ...


One of the persistent mysteries in the studies of National Innovation Systems lies in the recognition of a
domain of “implicit knowledge” or of “tacit learning” which has the peculiar property of being non-
codified and of residing only within the “hearts and minds”, or, alternatively, in the hands and operations
of scientists, engineers, technicians, workers, managers ... While the NIS-relevant characterization of
“implicit knowledge” domains will be dealt with in a more elaborated manner in Section III, an important
genotype-phenotype linkage outside the genetic code must be discussed within the present context,
namely the relations between action patterns and their neural counterparts. Take as a characteristic
example the dominance relations, defined “in terms of the direction of approach-retreat interactions
between two individuals” (CHENEY/SEYFARTH 1990:29), then one may observe the following patterns
which have evolved in the case of male vervets -

      If one vervet ... is dominant to another when competing for food, the same animal will also be
      dominant when competing for gropoming partners, mates, or resting sites. If one animal is dominant
      to another in approach-retreat interactions, she will also be dominant in fights. Finally, in most vervet
      relationships the subordinate both grooms the dominant and forms alliances with her at higher rates
      than vice versa. (CHENEY/SEYFARTH 1990:30)

A similar action pattern with a large cluster of homologies cannot be observed, however, in a close
relative to vervets, namely in baboons -

      Among male baboons ..., dominance relations are quite unpredictable from one context to the next.

Likewise, dominance relations between female and male vervets follow a context-dependent relation of
indeterminacy as well -

      Among vervets, dominance relations between males and females are similarly context-dependent.

Here, one is clearly confronted with an emergent1 behavior which, in all probability, must be independent
of the genetic code since it is restricted, despite an (almost) identical genetic configuration, to a subset of
a subset of the species only. What one can observe in this case, is a special relation between a consistent
action pattern, being -

      not only consistent across contexts but also transitive: if A is dominant to B and B is dominant to C, A
      is invariably dominant to C (IBID.) -

and a corresponding neural ensemble which, following Gerald M. Edelman, may be labeled as a “neural
group”, a group of “neural groups” ... (EDELMAN 1987, 1990, 1992) or, following recent expoloration in
Artificial Intelligence, in Cognitive Science or in Artificial Life, as “agents” (Marvin Minsky), as “drafts”
(Daniel C. Dennett), or as “tasks” (Rodney A. Brooks). Moreover, for purposes of a consistent
terminology throughout the epigenetic sections, the term “building blocks” or, sometimes, “autonomous
units” will be reserved for the characterization of ensembles across the genotype and phenotype levels
whereas the notion of “neural groups” (agents, drafts, tasks) will be applied for the genotype level and the
concepts of “action patterns” (action sequences, action programs) will be used for its phenotype
counterparts. While the main characteristics in the “embeddedness relations” in the case of neural groups
 action patterns, especially the sharply increased importance of the phenotype-domains, will be
analyzed within Sections III and IV, the important transdisciplinary point to be discussed within the
present context lies in the applicability of the general recombination operators to the case of action
sequences and neural groups. The following list of genotype examples offers an interesting neural
perspective on recombinations, making use, in a purely formal description of neural groups, of the varying
composition of neural ensembles -

         Adding, the creation of a larger neural group by integrating new neurons or another group ...
         Breaking, the differentiation of an established neural group into two separate ones ...
         Crossing-over, the breaking of at least two neural groups and their merging into two new
         ensembles ...
         Deletion, the destruction of specific neurons within a neural group ...
         Duplication, the (re)production of identical neural groups (especially important, due to the
         tremendous complexity barrier between the human genome and the resulting complexity of the
         human brain)
         Inverting, the complete reversal of the linkage structure between a neural group ...
         Merging, the integration of two neural groups to a new and more complex one ...
         Moving, the process of “horizontal” transfers through lateral recursions and re-entries ...
         Swapping, the process of “vertical” transfers via recursions and re-entries with a control-level
         Replacing, the substitution of a specific neuron or neural group by another one ...

To complete the “power of recombinations”, a small summary will be provided for recombinative changes
on the phenotype level which, surprisingly, follow along the same types of recombinative operations -

        Adding, the combination of a specific action pattern with another one like introducing an
        element of fighting during a game period ...
        Breaking, the differentiation of an established everyday action pattern into two separate ones
        like the search and the consumption of food into two distinct phases ...
        Crossing-over, the breaking of at least two everyday action patterns and their merging into a
        new ensemble, to be found, for example, in bird songs ...
        Deletion, the destruction of a specific sub-routine within an action pattern ...
        Duplication, the utilization of a specific action pattern twice ...
        Inverting, the complete reversal of an established action pattern ...
        Merging, the integration of two distinct action patterns to a new and more complex one,
        exemplified in the increasing complexity of the hunting behavior of animals ...
        Moving, the process of “horizontal” analogy building like the extension of dominance relations
        to other social domains ...
        Swapping, the analogy-making from a level Li to a different level Lj like the emergence of
        group behaviors ...
        Replacing, the substitution of a specific sub-routine in an action pattern by another one ...

In this manner, an important “missing evolutionary link” has been built up, bridging the gaps between
genetic codes on the one hand and human actions and interactions on the other hand. Due to the genotype
chains of -

         Genetic Code (internal)  Neural Programs (internal)  Natural Languages (external)

and due to the phenotype lineages from -

                                      Metazoa  Vertebrates  Humans

it becomes possible to identify a well integrated pattern of different types of embedded code systems at
the genotype level and corresponding species, socio-technical systems and societal formations (both for
animals, animats and humans) on the phenotype levels.
Table 2.2 summarizes the discussion so far - and offers an outlook on the subsequent analyses of “natural
languages” or of “scientific languages” which, for obvious reasons, fall under the heading of external code
systems. Moreover, Table 2.2 could have carried a different heading as well, namely the -

                                        Externalization of Code Systems

whereby, phrased in the rythms of a cosmic clockwork, the emergence1,2,3 of an external code system
(natural languages, scripts ...), after millions of years, has been successful at last in integrating the internal
code-systems which have dominated the evolution of species so far ...

Table 2.2: The Long Co-Evolutionary Complexification of Embedded Code-Systems

PHENOTYPE LEVEL                            Complexification of the “extended phenotypes” 

                         Metazoa                   Vertebrates                 Humans
                                                   Action Patterns             Action Patterns
                                                                               Code-Utilizing Routines

                                                                                    

                                                                               External Code-Systems
                                                   Neural Programs             Neural Programs
                      Genetic Codes                Genetic Codes               Genetic Codes

GENOTYPE LEVEL                                  Increase in embedded code systems         

                              2.5. NATURAL LANGUAGES

It becomes almost a truism to stress the importance of the general recombinative repertoire at the level of
natural languages. The following set of examples should be sufficient to demonstrate the flexibility as
well as the infinity of the recombinative space of natural languages and their utilization contexts.

         Adding, the using of connectives like “and”, “but”, “or” and the like which form a new scheme
         by linking two existing ones ...
         Breaking, the differentiation of a compound sentence into two separate ones ...

           Crossing-over, the breaking of at least two sentences and their merging into a new ensemble
           Deletion, the destruction of a specific element in a verb-phrase ...
           Duplication, the making of expressions like bonbon, beri-beri ...
           Inverting, the negation of a sentence ...
           Merging, the integration of two distinct words like “house” and “hold” to “household” ...
           Moving, the process of “horizontal” analogy or metaphor building like the construction of new
           utilization contexts for concepts or catch-phrases ...
           Swapping, the “vertical” movement from a level Li to a different level Lj like the movement of a
           well-defined sentence up or down the “reduction ladder” ...73
           Replacing, the substitution of an idiom by another one like the current switch (at least in
           Germany) from “o.k.” to “alles klar” ...

Through this list of examples, the flexibility and adaptability of the code system of natural languages
should become clear and visible. According to the ubiquity of the “recombination game”, a conjecture
will be put forward that the development of natural languages (genotype-level) and, especially important,
the complexification of language games (genotype and phenotype levels) can be described by the same
types of recombinations, which have been identified in the general case or in the particular instances so
far, namely by -

           Adding, the combination of a specific language game with another one like the usage of a set of
           service-oriented part of questions and answers (“How did you like our meal?”, etc.) within the
           routine of serving food in restaurants...
           Breaking, the differentiation of an established everyday pattern of practices into two separate
           ones like the change in decision making processes from an autocratic (individual decision
           making) to a democratic mode ( individual decision making + group consensus formation) ...
           Crossing-over, the breaking of at least two everyday routines and their merging into a new
           ensemble, as exemplified in the heavy exchanges of verses and tunes within the pools of
           traditional songs ...

    Aside from being a source to some games, crossing over occurs in a number of different ways, in poetry, for example, by breaking two
lines in a poem and combining them into two new verses ...
    Take the following sentence from the original level of particle physics -

       We begin with the observation that if particles didn't interact with each other, things would be incredibly simple ... Particles
       without interactions are called bare particles, and they are purely hypothetical creations; they don't exist.
       Now when you 'turn on' the interactions, then particles get tangled up together in the way that functions F and M are
       tangled together ... These real particles are said to be renormalized - an ugly but intriguing term. What happens is that no
       particle can even be defined without referring to all other particles, whose definitions in turn depend on the first particles,
       etc. Round and round, in a never-ending loop (HOFSTADTER 1982:127) -

and apply these sentences “up the reduction ladder” to bacteria, animals, human interactions, group exchanges ...

        Deletion, the destruction of a specific sub-routine within an everyday setting like, to use an
        example from an outdated socio-technological system, calling persons directly without the sub-
        routine for dialing an intermediary operator first ...
        Duplication, the utilization of a specific practice twice like reading the same book, listening to
        the same musical piece, watching the same film again and again ...
        Inverting, the complete reversal of an established everyday practice like the children´s play “the
        world turned upside down” ...
        Merging, the integration of two distinct everyday routines to a new and more complex one,
        exemplified in listening to the radio while reading newspapers ...
        Moving, the process of “horizontal” analogy building like the application of an “internal
        (intentional, emotional, etc.) vocabulary” to the domain of techno-social systems (“The
        computer refuses (does not want, is not in the mood ...) to work in proper speed today ...”) ...
        Swapping, the analogy-making from a level Li to a different level Lj like the usage of dialogical
        language components within the context of audio-visual mass media (e.g., rituals like greeting,
        making excuses for disturbances in programs, etc.) ...
        Replacing, the substitution of a specific element by another one, exemplified in the permanent
        rule changes even in well established sports games like, for example, soccer (the replacements
        with respect to the number of exchange players, the shifting limits in penalties for foul play, the
        changing behavior patterns for goal keepers, etc.) ...

In this way, recombinative operators, the most important actors within Dennett´s “pandemonium of
demons”, are constantly at work, leading to new recombinations at practically all genotype or genotype-
phenotype levels from Alphabets to Zeitgeist-induced changes of entire language games ...

                           2.6. SCIENTIFIC LANGUAGES

It comes with minimal surprises that the language of science exhibits the same amount of variations and
the same recombination repertoire which has been identified in the case of natural languages already.
Thus, the following list of examples, this time centered on scientific programs, will be sufficient to
demonstrate the ubiquity of the operation called recombination.

        Adding, the utilization of an additional control routine within an experimental setting ...
        Breaking, the differentiation of a complex program into two separate sets ...
        Crossing-over, the breaking of at least two methods and their merging into a new ensemble ...

        Deletion, the destruction of a specific sub-routine within a well-defined experiment ...
        Duplication, the utilization of a specific method twice ...
        Inverting, the complete reversal in an experimental setting ...
        Merging, the integration of two distinct knowledge domains like “genetics” and “semiotics” to
        “social genetics” ...
        Moving, the process of “horizontal” analogy building like the utilization of the genetic code as
        analogy for other code-systems, especially for language and language instructions ...
        Swapping, the analogy-movement from a level Li to a different level Lj like the building of
        “second order-concepts” (FOERSTER 1995) like “knowing knowledge”, explaining
        explanations”, “understanding understanding”, “observing systems” or, in a self-referential
        mode, the movement of well-defined genotype-domains up to the level National Innovation
        Systems ...
        Replacing, the substitution of a specific element by another one like the change from a linear
        model specification to a non-linear one, from a specific estimation technique to a new one and
        the like ...

Another point must be made with respect to the differentiation and the evolution of scientific languages -

           Recombinative Operators Languages of Everyday Life  Scientific Languages
      Recombinative Operators Language Games of Everyday Life  Scientific Language Games

More concretely, the same set of “recombinative demons” can be identified in which everyday routines
are transformed, via the meanwhile well-known recombinative operators, into proper scientific ones. In
other words, scientific language games are characterized by minor or major variations on the thema of
everyday communication, relying on -

        Adding, the combination of an established scientific control method with an everyday control-
        routine, exemplified in the early works in botany (Linné, Mendel ...)
        Breaking, the differentiation of an everyday pattern of practices into two separate scientific
        ones like the breaking of observing the stars into a telescopic form of observation and a
        specialized form of measuring and reporting ...
        Crossing-over, the breaking of at least two everyday routines and their merging into a new
        ensemble better suited for the purpose of scientific investigations, to be found, for example, in
        the so-called triangulation (GIDDENS 1989:682pp.) of empirical methods in the social
        sciences where standard practices of the quantitative or qualitative type can be re-arranged into
        quantitative-qualitative and into qualitative-quantitative sequences ...
        Deletion, the destruction of a specific everyday sub-routine within a scientific setting ...

           Duplication, the utilization of a specific method twice, securing, thus, higher accuracy and,
           thus, scientific credibility ...
           Inverting, the complete reversal of an established everyday practice into a scientific one like the
           Copernican inversion of a highly confirmed everyday interpretation of centre and periphery ...
           Merging, the integration of two distinct everyday routines to a new and more complex scientific
           one, exemplified in Oevermann´s “objective hermeneutic” where everyday competencies of
           attributions, projection of different contexts and accurate reporting are integrated into a special
           method of qualitative understanding (OEVERMANN 1983) ...
           Moving, the process of “horizontal” analogy building like taking societal relations of an
           evolving capitalist system as an analogy for ecological domains and for the origin of species ...
           Swapping, the analogy-movement from a level Li to a different level Lj like the revolutionary
           move of applying terrestrial mechanics to the vertically different domain of the movement of
           stars and planets ...
           Replacing, the substitution of a specific everyday routine element by a scientific one like the
           introduction of a new instrument for the purpose of observation ...

Seen in this perspective, it should come as minimal surprise that “life and death” inside the sciences,
consensus formation, winning alliances, persuading someone ... exhibit most of the traits and
characteristics of “life and death” outside the sciences, consensus formation, winning alliances,
persuading someone ... Unfortunately, this particular feature has only entered very recently into the realm
of scientific reflexions of the processes of scientific development.74

                                 2.7. CODIFIED RULE - SYSTEMS

One of the decisive steps in the elaboration of the code-perspective especially for complex areas like
socio-economic systems in general and National Innovation Systems in particular comes in the present
chapter since an unusual move will be undertaken, namely the elaboration of embedded code-systems
which do not possess bases, letters, words or set-theoretical entities as their basic components, but -
codified rules. The starting point for the subsequent deliberations lies in the very general observation that
codified rule-systems in general are not an emergent or characteristic feature of human “life-worlds” but
applicable to biological systems at the phenotype level as well. Consider a small set of properties relevant
for the “frog eye to tell the frog brain” like -

    As one of the rare exceptions to this almost universal “blind spot” in the history and philosophy of science, see NEURATH 1981, who
had proposed already in the 1930‟s a behavioristic approach (a, so-called “Gelehrtenbehavioristik”) for the study of scientific practices and
routines ...

      moving, striped, large, near ... (HOLLAND 1995:45)

and a small set of actions relevant for the frog brain to tell the frog´s senso-motoric apparatus like -

      flee, pursue, turn, head, extend tongue ... (IBID.)

then it becomes possible to construct a set of “production-rules” in the If-then format like -

      IF (small flying object to the left) THEN (turn head 15o left)
      IF (small flying object centered) THEN (send @)
      IF @, THEN (extend tongue) (IBID:43pp.)

In this manner, codified rule-systems become the genotype counterparts to a huge class of phenotype
actions and interactions. For reasons of a systematic account of the genotype superstructure in biological
and social domains, the subsequent analysis will be focused primarily on codified rules within societal
ensembles ...

                   2.7.1. WHAT SOCIETAL CODES ARE NOT

The introductory chapter on codified rule systems will be devoted to a total rejection of the ways of code-
making which have been proposed by Niklas Luhmann. (LUHMANN 1984/1987/1988/1989). The main
code-relevant point in Luhmann´s analysis of socio-economic systems lies in his insistence that any
sufficiently differentiated socio-economic subsystem like the economy, the political system, the legal
system, the system of education, the scientific system, the system of sports, the media, etc. possesses a
binary code which acts as a prime criterion for system-specific selection procedures. Thus, the scientific
system selects primarily according to truth/falsity, the legal system according to legality/illegality, the
economic system according to monetary/non-monetary transactions, etc.

      Codes sind Unterscheidungen, mit denen ein System seine eigenen Operationen beobachtet, sie im
      Falle der Wissenschaft zum Beispiel nach wahr und unwahr unterscheidet. Außerdem ist zu beachten,
      daß binäre Codes immer die Selbstbeobachtung und Selbstbeschreibung eines Systems strukturieren,
      also nicht etwa irgendwelche Unterscheidungen sind, die durch einen externen Beobachter
      herangetragen werden. Ein externer Beobachter kann mithin ein solches System nur angemessen

      verstehen, wenn er berücksichtigt, daß es die eigenen Beobachtungen binär codiert und damit sich
      selbst dazu zwingt, sich selbst von der Ebene zweiter Ordnung aus zu beobachten. Oder verkürzt
      gesagt: Über binäre Codierung zwingt ein System sich zum Prozessieren von Selbstreferenz, und ein
      externer Beobachter, der dies nicht sieht, versteht das System nicht. Das ist nur dann ergiebig, wenn
      mit der Beschränkung auf nur zwei Werte ein Ausschließungseffekt verbunden ist ... Die Code-Werte
      öffnen nur einen Kontingenzraum und stellen sicher, daß alle Operationen des Systems auch der
      entgegengesetzten Wertung unterliegen könnten; aber sie geben nicht an, wie zu entscheiden ist.
      (LUHMANN 1990:194pp.)

Apart from the primary “difference” and “differentiation” for societal ensembles, any socio-economic
system creates, in addition, a number of “programs” compatible with the primary code-separations -

      Die Programme sind dagegen vorgegebene Bedingungen für die Richtigkeit der Selektion von
      Operationen. Sie ermöglichen einerseits eine gewisse „Konkretisierung‟ (oder:‟Operationalisierung‟)
      der Anforderungen, die an ein Funktionssystem gestellt werden, und müssen andererseits eben in
      gewissem Umgfange änderbar bleiben. Auf der Ebene der Programme kann ein System, ohne seine
      durch den Code festgelegte Identität zu verlieren, Strukturen auswechseln (!!!). Auf der Ebene der
      Programme kann daher in gewissem Umfange Lernfähigkeit organisiert werden. Durch die
      Differenzierung von Codierung und Programmierung gewinnt ein System also die Möglichkeit, als
      geschlossenes und als offenes System zugleich zu operieren. Deshalb ist diese Differenzierung
      mitsamt der dadurch gewonnen Artikulationsfähigkeit der Schlüssel für das Problem der
      gesellschaftlichen Resonanz auf Gefährdungen durch die Umwelt. (LUHMANN 1986:91)

Thus, the scientific system has developed a large number of discipline-specific “programs” which become
relevant in the day to day operations and routines and which guarantee the “reproduction” of the basic
binary distinction even in remote areas of a specific societal system. Intriguing as it stands, especially due
to the reliance on a similar vocabulary like “codes”, “programs” and the like, one can find a large number
of reasons to abandon Luhmann´s trajectories into code-land entirely. Subsequently, only four basic
counter-arguments will be provided.

      First, the fuzzy notion of Luhmann´s primary code must be seen as one of the central
      weaknesses in the overall approach. In case of scientific systems which will serve as reference
      point for the subsequent critique, a number of alternatives can be named instantly which could
      serve as primary code, too. Starting from Larry Laudan´s observation on the practical
      irrelevance of truth -

         Philosophers and scientists since the time of Parmenides and Plato have been seeking to justify
         science as a truth-seeking enterprise. Without exception, these efforts have floundered ... If

           rationality consists in believing only what we can reasonably presume to be true, and if we
           define „truth‟ in its classical, non-pragmatic sense, then science is (and will forever remain)
           irrational (LAUDAN 1977:125)

       as well as on the insistence on the practical relevance of “problem solving” -

           To make rational choices is, on this view, to make choices which are progressive (i.e., which
           increase the problem solving effectiveness of the theories we accept) By thus linking rationality
           to progressiviness, I am suggesting that we can have a theory of rationality without
           presupposing anything about the veracity or verisinmilitude of the theories we judge to be
           rational or irrational (IBID) -

       one might invoke, as the prime requirement for accepting or rejecting hypotheses, theories,
       models or research programs, criteria like degrees of “problem solving effectiveness”
       (LAUDAN 1977/1981), Thus, the scientific system selects, following Laudan, in its most
       basic distinctions, highly interesting solvable or as yet highly relevant unsolved problems
       from a wider set of problems already solved, of irrelevant problems or of problems which
       cannot be solved in principle. Using a mode of “one case-inductivism”, one is confronted,
       thus, with the serious problem that for any socio-economic system a legitimate as well as
       irreducible multiplicity of primary codes, some of them binary, some of them non-binary, can
       be established.75
       Second, the internal linkages between primary codes and secondary programs have to remain,
       by necessity, highly indirect and ex post at best. Take, once again, a special methodology on
       evaluations or a particular design for consensus building like the Delphi-method, a special
       research program in number theory, the archeological programs for recovering and analyzing
       historical objects, the modeling programs associated with the “homo oeconomicus”,
       therapeutic programs in the area of systemic therapy or family therapy, then one is confronted
       with a highly inconsistent set of programs where any link to a primary code or program has to
       become, by logical necessity, tautological at best.
       Following Feyerabend´s “Anything goes” in its intended applications, one will be able, at any
       point in time, to identify special instances within the diversified history of science where two
       scientific programs at the same level of discourse exhibited a contradictory set of research
       operations. Thus, the Luhmannian distinction between codes and programs has to remain
       without any content since any type of program has to be compatible, a priori, with a specific
       primary code.

     A counter-objective in the spirit of Niklas Luhmann might be that the openness and vagueness of the primary code is an unpleasant but,
at the same time, unavoidable feature of complex socio-economic systems which operate in a strictly recursively closed manner, irrespective
of the “indeterminacy” of observer-dependent attributions. Despite this rescue argument, the “fuzziness” with respect to primary codes
prevails ...

     Third, Luhmann´s code-program considerations are subject to a severe constraint, namely to a
     focus on communications, and communications alone. Due to the overall focus on code-
     systems and their embeddedness relations, one has, within the present transdisciplinary
     framework, a far more comprehensive systemic view of processes at the genotype level, at
     phenotype processes as well as on the multiplicities of embeddedness relations between both
     levels. Idiosyncratic restrictions on psychic systems, consciousness and the like are clearly not
     necessary within a framework which, like the late BCL program (FOERSTER 1995), starts
     from a general systems perspective, goes on on to adaptation processes and their algorithmic
     constructions and, finally, ends up with communications ...
     Fourth, Luhmann´s focus on societal systems and his highly peculiar social adaptations of the
     “autopoiesis-vocabulary” makes it impossible to arrive at an appropriate meta-theoretical or,
     alternatively, transdisciplinary level, in which code-systems in genetics can be analyzed with
     the same basic conceptual framwork which is utilized in the case of social actions, speech
     acts, musical performances, socio-technical systems-operations and the like.

As a consequence, the following definitions and discussions, although being centered on codes, programs,
and social systems, will be undertaken in a comparatively wide cognitive distance from Luhmann´s
framework on codesL, programsL, and social systemsL, despite the striking surface identities of the basic

                             2.7.2. SOCIETAL RULES AS

                           EMBEDDED CODE-SYSTEMS

Departing, thus, radically from Luhmann´s invitation to sociological code-making, the basic units for the
new type of code-system will become, not surprisingly, encoded rules which even at the most intuitive
level share the five requirements for code-systems.

     Indifference of Code-Elements: The first condition, as will be remembered form chapter 1.2.1,
     refers to the exchangeability of specific “marks” of a basic component or “character” in a
     code-system. For social rule-systems, such indifference classes can be easily identified. Take
     two interactive games like tennis or chess as reference point, then the following indifference-
     conditions are met. In the case of tennis, high variations can occur with respect to the dresses
     of the players, the rackets, the tennis balls, the presence or absence of one or many referees,

       the access of the public, the material of the tennis court itself ... With respect to chess, the size
       of the chess-board, the shape and the colours of the chess-men, the dress of the players, the
       location ... can vary quite substantially. In both instances, the different “realizations” or
       “marks” do not distort the fulfillment of the underlying game-specific rules. Thus,
       indifference, in the case of social rule systems too, becomes a typical equivalence relation,
       being reflexive, symmetric and transitive.
       Finite Differentiation: The condition of finite differentiation demands a decidability, in
       principle, whether a given realization belongs to a specific rule-element - or not. Being strictly
       independent from the indifference condition, the requirement of finite differentiation asks for
       a code-system which, under normal conditions, possesses a very low rate of reproduction
       errors. Take, once again, tennis as a social rule system, then any sequence of returns should
       produce the same form of rule-processing, i.e. the same decisions whether a specific
       realization belongs, say, to the rule of infield returns, to the rule of outfield returns, etc.
       Combination of Code-Elements: Since the most essential requirement for code-systems, their
       differentia specifica, lies in the combinability of code-elements or characters, this condition
       does not seem to hold in the case of well defined games like chess or tennis or, more
       generally, in the case of “rule-based games”. Upon closer inspection however, rule-
       combinations occur quite naturally in social rule-systems, too.

           In a well-defined societal game like tennis players produce, seen in a rule-based
           perspective, all sorts of rule combinations. The rule of returning the tennis ball to one´s
           opponent within the confines of the tennis court, subsequently labeled as infield return
           rule, can be accomplished by returning a slice, a lob, a top-spin, a smash, a stop-ball ...76
           Two players who, due to low levels of competencies, follow the infield rule sans
           phrase, will try to return the ball somehow. Likewise, players who manage to possess a
           comprehensive repertoire of different returns will create highly differentiated return-
           sequences with a low level of predictability. Finally, players may exhibit a clear strategy
           in their returns, following a specific pattern of returns or a tennis program for short.
           Similarly, making a move in chess can be done as part of a long-term strategy in a very
           deliberate manner, choosing at random without thinking of further consequences, as a
           result of thinking about all possible combinations for ten to fifteen moves ahead, etc.
           Moreover, these rules can be applied in a sequential manner, giving rise to “inconsistent
           strategies” of very deliberate choices and random moves ... For any of these
           configurations, different rules, rule-sequences or, alternatively, “chess programs” can be
           identified giving, once again, a strong support for the combinability of rules to sets of
           rules or, alternatively, to programs.

    It should be added that the rules like “Return an infield-lob”, “Return an infield-smash” follow exactly the rule format, introduced at the
outset of the societal rule-chapter ...

     Thus, code-strings, i.e. rules, as well as programs, i.e. a set or sets of rules, can be re-arranged
     into new composite sequences through combinations and a small number of basic operations.
     Moreover, rule based strings and programs serve as an indispensable recipe or production
     function in the performance of language games, in the construction of socio-technical
     complexes, in the constitution of socio-economic systems, etc.
     Fitness functions: Not surprisingly, rule combinations become evaluable in terms of a social
     fitness function , too. Stated in a loose fashion, rules and rule combinations can be assigned
     different “degrees of fitness”, ranging from simple nominal arrangements (no fitness, small
     degree, medium degree, high degree, maximum fitness) to ordinal and, finally, to cardinal
     ensembles, defined over the interval 0,1.
     Embeddedness: Societal rule systems, by definition, must exhibit many embeddedness-
     relations which, as will be recalled, refer to the dual level character of evolutionary systems
     and, consequently, to the connections between a code-based genotype and a code-dependent
     phenotype being dynamically interwoven and entangled. In social rule-systems too, one can
     easily see a variety of dynamic embeddedness relations with a larger surrounding whose
     production and reproduction depends, however, crucially on the rule-system itself.

What will be shown in the subsequent Section III lies, inter alia, in the somewhat surprising result that
code-systems, composed of social rules, offer, in combination with action patterns and neural programs,
an interesting new way for incorporating slightly mysterious notions like “implicit knowledge” or the
seemingly self-contradictory concept of “tacit knowledge” ...

                           2.7.3. SPECIAL FEATURES OF

                           RULE-BASED CODE-SYSTEMS

In the final part on rule based code-systems a small number of characteristic features will be discussed
which exemplify not only the close affiliations between rule systems and code-ensembles, but also the
dynamic and recombinative character of societal rule-configurations.

     First, rule-systems in general, social, biological and otherwise, must meet the following three
     criteria which, due to the “multiplicity of agencies”, introduced at length in Section III, gives
     rise to a rich variety of societal rule systems within and between persons or ensembles of
     persons ...

   1. The rules must use a single system to describe all ...agents.
   2. The rule syntax must provide for all interactions among agents.
   3. There must be an acceptable procedure for adaptively modifying the rules. (HOLLAND

Second, with these three requirements as background or target area, another separation can be
brought forward for the constitution of the rule-based universes, social and otherwise.
Esapeially in the case of societal rule-systems, it seems useful, generally speaking, to
differentiate between well-defined “language games” on the one hand and weakly defined
games which do not possess a clearly defined set of exit and entrance conditions, of
evaluation measures, of termination conditions and the like. Thus, “learning to read” is a
highly complex, though weakly defined language game (GALABURDA 1989), whereas many
games in sport, theater, films, for fun or for gambling are situated within the well-defined
domain. In terms of rule-based code-systems, well-defined games are strong with respect to
three general characteristics for code systems, i.e. with respect to the indifference of code-
elements, the finite differentiation-condition as well as to the fitness function. In contrast,
weakly defined “language games” like understanding a word, a sentence, a book, a formula, a
picture, etc. do possess a restricted attribution repertoire, being focused on “strong cases” or
“paradigmatic applications” only.
Third, the embeddedness relations between rule-codes and the “corresponding” practices at
the phenotype-levels, while being discussed at length within Section III, have a special
epistemological status which can be summarized in the following quotation -

   Was nenne ich „die Regel, nach der er vorgeht‟? - Die Hypothese, die seinen Gebrauch der
   Worte, den wir beobachten, zufriedenstellend beschreibt; oder die Regel, die er beim Gebrauch
   der Zeichen nachschlägt; oder, die er uns zur Antwort gibt, wenn wir ihn nach seiner Regel
   fragen? - Wie aber, wenn die Beobachtung keine Regel klar erkennen läßt, und die Frage keine
   zu Tage fördert? - Denn er gab mir zwar auf meine Frage, was er unter „N‟ verstehe, eine
   Erklärung, war aber bereit, diese Erklärung zu widerrufen und abzuändern. - Wie soll ich also
   die Regel bestimmen, nach der er spielt? Er weiß sie selbst nicht. - Oder richtiger: Was soll der
   Ausdruck „Regel, nach welcher er vorgeht‟ hier noch besagen? (WITTGENSTEIN 1971:PU

Thus, rule-systems at the genotype level and actions and action sequences at the phenotype
level follow along a different pattern than, say, the relations between a “generative” genetic
code and the “resulting” phenotype organism. Nevertheless, the so-called “embeddedness
relations” between genotypes and phenotypes can be analyzed in the case of social rule

       systems too and they will lead to a number of unexpected surprises and counter-intuitive
       insights with respect to the status of “social genetics” ...
       Fourth, rule-systems, especially those in weakly defined domains, undergo a rapid process of
       rule generation and modification, best exemplified by the following quotation -

           Und gibt es nicht auch den Fall, wo wir spielen und „make up the rules as we go along‟? Ja
           auch den, in welchem wir sie abändern - as we go along. (WITTGENSTEIN 1971:PU 83).

       Here, in the case of making up rules “as we go along”, the area of codified rule systems has
       been left aside - and the domain of action patterns and neural programs has been entered once
       again ... Especially in the case of weakly defined “language games” one can observe a
       permanent dynamical pattern of rule-enlargements and context dependent rule adaptations -
       and a huge basis for mis-understandings and mis-representations, or, alternatively, mis-
       constructions, as well.77

To sum up, embedded code systems can be identified from the level of the genetic code up to societal rule
systems or, alternatively, “language games”. Due to the universality of the five requirements for code
systems which, apparently, turn out to be necessary for any code-system, human and otherwise, the
transdisciplinary framework on evolution and innovation has gained a widely distributed genotype basis
which, together with the phenotype levels as well as the relations between these two domains, will make
up the epigenetic “furniture of the world” (Mario Bunge) for National Innovation Systems.

                                   3. INCOMMENSURABILITIES


                                AND SCIENTIFIC PRODUCTION

Having built up a potentially rich variety of embedded code-systems with a diversity of basic components
like bases, letters, sets, words, symbols, musical notes or rules, it becomes the central focus to integrate
the code-based framework with three other concepts which, especially in the case of National Innovation
Systems, have acquired a central position, namely the concepts of “information”, “knowledge” and

   This, in turn, can be viewed as the most essential basis for the erroneous moves in probably the most important social language game
which will be discussed at length in Section III, namely the so-called “game of attribution” ...

“scientific production”. Starting with a mysterious sentence on the constitution of “post-capitalist
societies” -

     Reichtum produziert heute (fast) nur noch die Information und(!) das Wissen (DRUCKER 1993:262) -

one is led to assume that the conjunction of two entities, information and knowledge, serves as the
wealth-generating basis for today´s highly developed societies.
Upon closer inspection however, it becomes, aside from an intuitive identity relation in the form of
knowledge= information = the sum total of scientific production, extremely fuzzy what should be counted
and classified under the headings of knowledge and what under the label of information. Take the
following definitions in a recent glossary -

     Codified knowledge: Knowledge which need not be exclusively theoretical but needs to be systematic
     enough to be written down and stored. As such, it is available to anyone who knows where to look.
     Embedded knowledge: Knowledge which cannot move easily across organizational boundaries, its
     movement is constrained in a given network or set of social relations.
     Knowledge industries: Industries in which knowledge itself is the commodity traded.
     Migratory knowledge: Knowledge which is mobile and can move rapidly across organizational
     Tacit knowledge: Knowledge not available as a text and which may conveniently be regarded as
     residing in the heads of those working on a particular transformation process, or to be embodied in a
     particular organizational context.
     Technology transfer: The transmission of knowledge from universities to industries. (GIBBONS et al.
     1994:167f.) -

one is immediately confronted with questions of the following type which simply arise from the
traditional accounts of knowledge as justified true beliefs.

     Are the gross products, i.e. the set of articles, books, patents, prototypes or other output-
     genres, in the scientific manu- and neurofacturing processes to be classified as “knowledge
     carriers”, irrespective of the relevance, truth, confirmation or “transferability” of the contents
     Are the gross products, i.e. the set of articles, books, patents, prototypes or other output-
     genres, in the scientific manu- and neurofacturing processes to be classified as “knowledge
     carriers”, irrespective of the beliefs of the persons who have manu- and neurofactured these

       What are the relations between the gross products of scientific production and the contents of
       information? Is it necessarily the case that a high knowledge production, if properly defined,
       is accompanied by the proliferation of high information contents, too?
       What is the precise relation between gross outcomes of the scientific production processes and
       the net-products which enter into actual exchange processes?78
       What is the justification to extend the notion of knowledge, if properly defined, to text-free
       domains and qualify tacit knowledge, as has been shown, as “knowledge(!) not(!) available as
       Finally, how are the relations between domains of embedded or tacit knowledge - and
       information contents to be understood, in principle?

Moreover, the units of the entity which can be codified, commercialized, embedded, industrialized,
migratory, tacit or transferred remains largely in the shadow -

       Es ist zumindest bisher nicht möglich, Wissen zu quantifizieren. Sicherlich können wir abschätzen,
       wieviel es kostet, Wissen zu produzieren und zu verteilen. Wir können jedoch nicht sagen, wieviel
       Wissen konkret produziert wird oder was wir unter einem ´Ertrag aus Wissen´ überhaupt verstehen
       wollen oder sollen .. Ganz allgemein gilt: Die Quantität des Wissens, ihr rein mengenmäiger Aspekt,
       ist bei weitem nicht so entscheidend wie die Produktivität des Wissens, also ihr qualitativer Aspekt.
       (DRUCKER 1993:265f.)

Thus, the knowledge concept, especially in the scientific field devoted to knowledge production and its
economic impact, remains vague at best and highly misleading at worst, since, following once again Peter
F. Drucker, both the quantitative dimensions of knowledge and, especially important, the qualitative
aspects of knowledge diffusion and distribution are still poorly understood and analyzed domains.
The subsequent explorations will demonstrate that these three concepts will become part of a
comprehensive, well-integrated transdisciplinary framework of embedded code-systems. Even in an
intuitive understanding, a strong argument can be furnished that code systems and code-dependent
operations simply must be irreducible to notions like knowledge, information or scientific production.

       With respect to knowledge, it can be shown quite naturally that a large number of operations
       with code-systems like reading, writing a first draft, conceptual “tinkering”, etc. do not qualify
       as “knowledge” but as pre-requirements in “knowledge generation” at best.
       In relation to information, take code-systems like those described in chapter 2.7 of the present
       section, viz. rule-based systems. Here, the utilization of information concepts becomes, by
       necessity, either trivial or arbitrary since the “language of information” simply turns out to be
       insufficient to integrate highly diverse components like bodily movements or mechanical
   It belongs to the most obvious features of scientific communication and exchange processes that only small fractions of the gross-
production actually enter the exchange networks.

      instruments into its conceptual schemes. Thus, at the upper end of complexity, information
      based approaches fall short of encompassing code-systems.
      With respect to scientific production, it becomes a truism that many code-based operations
      like, once again, reading, writing notes, making preliminary calculations do not count as
      scientific output. Thus, the notion of embedded code-systems is able to account for a large
      class of scientific routines strictly independent of generating a specific scientific output.

Consequently, the irreducibility of the “code-based stance” or “code-stance” to concepts like knowledge,
information or scientific production has been successfully established. What the next four chapters are
attempting to demonstrate are the following main points.

      First, one will be led to a series of conceptual differentiations, separating three seemingly
      indistinguishable domains, namely knowledge, information and, finally, scientific production.
      Second, core-domains for auch of the three fields of investigations - for knowledge,
      information and scientific production - will be proposed.
      Third, the three groups of core-areas will be integrated under the unifying heading of
      embedded code-systems, serving as the homogeneous conceptual framework for a rich variety
      of analyses in the field of “science of science”.

With this three-fold research agenda, the next chapter will establish a clear division between scientific
production on the one hand and on ”knowledge” on the other hand. It will be shown that -

            genuine processes of scientific production are not linked to knowledge generation

and, conversely, that

              processes of knowledge production come about without scientific production.

In this manner, the necessary independence of these two basic concpets for National Innovation Systems
can be guaranteed.

                          3.1. SCIENTIFIC PRODUCTION

                                   AND KNOWLEDGE:

                                       A SEPARATION

The principal division or, more classically, the “principium divisionis” between the two domains of
scientific production and knowledge will be performed via a dynamic specification stating that the growth
processes in both domains cannot be assumed to follow a quasi-proportional direction.79 Thus, the first
clear-cut separation, expressed formally as -

                                          ( SPt/dt  0)  (Kit/dt  0)  (Kit/dt  0)

states that the growth of scientific production, as defined below, must be seen as strictly independent from
the growth of knowledge, defined in one of its possible operationalizations Ki. Hence, knowledge growth
and growth of scientific production may follow a development pattern of the type -

                                                    ( SPt/dt  0)  (Kit/dt  0)

as well as its logical counterpart -

                                                    ( SPt/dt  0)  (Kit/dt  0)

which, however, under the prevalent conditions of an ever expanding scientific system within the highly
developed OECD-regions has been hardly ever realized so far.
Before entering into a discussion on significant differences in developmental processes in both domains, it
will become a major task, first, to provide a set of operational definitions on both areas.

       Starting with scientific production from a systemic point of view, the main emphasis simply
       lies in the output of scientific institutes. Thus, articles, books, patents, prototypes,
       intermediate and final reports in contract research, transfer-activities, etc. are to be counted
       and qualified as the main empirical quantities for scientific production. Consequently, growth
       processes must be related to changes and increases in this particular output-field.
       Knowledge, on the other hand, cannot be defined with respect to the volume of scientific
       output since this output can be trivially separated into two classes, namely, on the one hand,
       innovative output and, on the other hand, replicative output. The latter consists, by and large,
       in a repetition of, loosely formulated, what has been “known” already before. Hence, it would
       be extremely problematic to associate growth in replicative research with knowledge growth.
       Thus, a different set of definitions becomes necessary which should be of relevance for
       knowledge production and, especially important, for knowledge growth
    “Quasi-proportionality” implies that a linear transformation can be identified which links a specific process x t to a second process yt in
the form of yt = a + bx.

.Summarizing the existing discussion, seven basic strategies and heuristics are open for an empirically
accessible account of knowledge processes.

Diagram 4.1: The Knowledge-Tree for the Year 2100

In the first operationalization, knowledge becomes equated with knowledge domains and
growth of knowledge, consequently, with the development patterns in these “knowledge
areas”. Take, as an interesting point of reference, the evolution of knowledge domains for the
next hundred years as proposed by Thierry Gaudin (1995:49), one is confronted with the
following potential shape of the overall knowledge tree or, alternatively, knowledge map.

   The first interesting characteristic is linked to the central position of the “cognitive
   sciences”, becoming the essential background domain for so diverse areas as medical
   research, ethology, technology or linguistics ....
   Second, the combination of “science-technology-society” (STS) on the one hand and
   “pedagogy” will serve, in conjunction with the core of the “cognitive sciences” and
   “ethology”, as the major foundational part for social, cultural and “artistic research”.
   Third, the central parts of the contemporary social sciences like economics or sociology
   will turn into relatively marginal areas, being linked as well as dependent on more
   general social science programs in the area of “General Ecology”.
   Fourth, the contemporary reductionist view in which particle physics occupies a basic
   position, will shift towards a more diversified triad in which the complex of particle
   physics and astrophysics becomes but one of three knowledge stems, the others being
   the cognitive sciences and the third consisting of the compound of “chemistry,
   materials, engineering and energy”.
   Fifth, normative sciences like mathematics, and one may add: logic and statistics, will
   retain their universal utilization contexts for any type of scientific endeavor, thus
   including and and integrating the empirical basic triad, introduced in the previous

The second path to knowledge has been followed within the philosophy of science field where
knowledge and especially knowledge growth has been associated, in its most general
Popperian form, with an increase in the degree of content -

   Scientific ... progress consisted in moving towards theories which tell us more and more -
   theories of ever greater content ... This consideration led to a theory in which scientific
   progress turned out not to consist in the accumulation of observations but in the overthrow of
   less good theories and their replacement by better ones, in particular by theories of greater
   content. Thus there was competetion between theories - a kind of Darwinian struggle for
   survival. (POPPER 1974:62p.) -

and, as a metaphysical corollary, as a slow approximation to truth in the très longue durée -

   We can explain the method of science, and much of the history of science, as the rational
   procedure for getting nearer to the truth. (POPPER 1975:58)

Thus, knowledge growth becomes linked to a single criterion like a degree of verisimilitude, a
degree of confirmation, a degree of corroboration, a degree of “problem solving-
effectiveness” (LAUDAN 1977) or other measures where a body of theories or models in a
knowledge domaini exhibits, over the very long run, an upward sloped trend. Consequently,
knowledge growth can be associated, couched in evolutionary terms, with the positive slope
of the “fitness-parameter” for the body of scientific production either in a specific domain, in
a set of different disciplines or in the scientific “production space” en bloc.
The third knowledge area defines knowledge in the traditional sense of “justified true belief”.
Taking an “ordinary language” example for an appropriate beginning -

   Wenn von Anatomie die Rede wäre, würde ich sagen:‟Ich weiß, daß vom Gehirn 12
   Nervenpaare ausgehen.‟ Ich habe diese Nerven nie gesehen, und auch ein Fachmann hat sie nur
   an wenigen Specimina beobachtet. - So wird eben hier das Wort „ich weiß‟ richtig gebraucht.
   (WITTGENSTEIN 1971, ÜG 621) -

knowledge becomes linked, on the one hand, to confirmation and justification procedures and,
on the other hand, to “agencies”, mostly, but not exclusively on the level of individuals. Here,
one of the core traditions in philosophy, centered around Kant´s question of “Was kann ich
wissen?”, has provided numerous solutions with respect to conditions of adequacy or with
respect to separating “knowledge proper” from “mere believing”.
The fourth “way of knowing” is associated with the result of learning processes and can be
summarized at best by reconfiguring the Wittgenstein quotation above by simply substituting
“Ich habe gelernt” for “Ich weiß”. Through the substitution of “having learned” for “know”, a
different emphasis of knowledge proceses can be gained -

   What we commonly understand by the word „knowledge‟ is closely related to what
   evolutionary biologists call adaptations. All adaptations are forms of knowledge. (PLOTKIN

Couched in terms of a recent theory of learning and adaptation, namely in “classifier systems”,
“genetic algorithms” and “genetic programming” (HOLLAND 1995, KOZA 1993), learning
processes consist in features like the modification of the “fitness” or the “strength paramters”
for the existing “knowledge stock”, in adding new elements to the existing stock, by
recombining existing knowledge components via crossing-over, by deleting, in a spirit of
“gracefulness” (HOLLAND 1986), specific knowledge components from the body of

knowledge, etc. The results of each of these processes, namely the new fitness parameters, or
the available knowledge components may be labeled, then, as knowledge proper, without
invoking however, as in the previous case, principles like justification or truth.
The fifth general utilization context for knowledge comes from an unusual move which will
be introduced and justified at length in Section III. With this move, knowledge will be broadly
understood as an “attribution strategy” or, alternatively, as a specific variation of an “internal
stance”. To make this point clearer, consider the following example by Ross Ashby on a
seemingly clear internal phenomenon, namely on memory -

   Suppose, for instance, that I am at a friend´s house and, as a car goes past outside, his dog runs
   to a corner of the room and cringes. To me the behavior is causeless and inexplicable. Then my
   friend says „He was run over by a car a month ago‟. The behavior is now accounted for by my
   taking account of what happened earlier. The psychologist would say I was appealing to the
   concept of „memory‟, as shown by the dog. What we can now see is that the concept of
   „memory‟ arises most naturally in the Investigator´s mind when not all of the system is
   accessible to observation, so that he must use information of what happened earlier to take the
   place of what he cannot observe now. „Memory‟, from this point of view, is not an objective
   and intrinsic property of a system, but a reflection of the Investigator´s limited powers of
   observation (ASHBY 1981:228)

Here, the emphasis shifts clearly from a focus on internal neural network-properties to an
external form of insufficient information on part of an outside observer. By analogy,
knowledge can be introduced and defined as an attribution strategy too, especially in the case
of embedded code-producing and code-retrieving systems. Thus, a small variation of the
“memory-theme” leads to a specific knowledge context, namely to an observer-dependent
“power of attribution” -

   Suppose, for instance, that I am at a friend´s house and, as a car goes past outside, the friend
   says to me, without looking out of the window, that his neighbor is heading for town. To me the
   assertion is inexplicable. Then my friend says „He always drives into town around this time.
   Besides, his engine has a combustion problem‟. The behavior is now accounted for by my
   taking account of what my friend knows about cars and neighbors. The psychologist would say
   I was appealing to the concept of „knowledge‟, as shown by my friend. What we can now see is
   that the concept of „knowledge‟ arises most naturally in the Investigator´s mind when not all of
   the system is accessible to observation, so that he must use information of what the system
   „knows‟ to take the place of what he cannot observe now. „Knowledge‟, from this point of
   view, is not an objective and intrinsic property of a system, but a reflection of the Investigator´s
   limited powers of observation.

The sixth utilization context for knowledge consists of “bodies of relevant scientific results”
as expressed in summarizing accounts of the state of the scientific art. Knowledge, thus,
becomes operationally defined as those selection procedures which have led to
“encyclopedias” or representative “textbooks” either of a particular field or in a
comprehensive set of knowledge domains. Consequently, knowledge becomes equivalent to
the stock of available encyclopedias, representative textbooks and the like. Moreover, growth
of knowledge refers to changes in the content of those highly selective summaries over time.
Finally, one of the most interesting knowledge contexts lies, seventh, in the discrepancies
between actual knowledge and yet unexplored domains. An intuitive model of the very long
run would consist in the vision of a “closing gap”, starting with huge distances at the
beginning of modern science in the renaissance, with medium discrpancies for the present
time and with a self-contained and closed body of knowledge in the distant future. An
alternative view is offered, however, by the following quotation which exhibits a highly
interesting cyclical pattern for one of the core-scientific disciplines, namely for particle
physics -

   Zu Beginn unseres Jahrhunderts glaubte man, alles könne im Sinne der Kontinuumsmechanik
   interpretiert werden. Damals meinte man, es genüge, eine gewisse Anzahl von Koeffizienten
   wie Elastizität, Viskosität, Leitfähigkeit usw. zu messen. Diese Hoffnung wurde jedoch durch
   die Entdeckung der atomaren Struktur und der Quantenmechanik erschüttert. Ende der
   zwanziger Jahre verkündete Max Born dann wieder einer Gruppe von Wissenschaftlern in
   Göttingen:‟Die Physik, so wie wir sie kennen, wird in sechs Monaten nicht mehr existieren.‟
   Dies war kurz nachdem Paul Dirac ... die Dirac-Gleichung formulierte, welche das Verhalten
   des Elektrons beschreibt ...(Ich) möchte die Möglichkeit diskutieren, daß das Ziel der
   theoretischen Physik in nicht allzu ferner Zukunft erreicht sein könnte, ungefähr zum Ende
   dieses Jahrhunderts. Damit meine ich, daß wir zu diesem Zeitpunkt eine vollständige,
   zusammenhängende vereinheitlichte Theorie der physikalischen Zusammenhänge haben
   könnten, die alle möglichen Beobachtungen beschreiben würde. (HAWKING 1991:9p.)

Following Nicholas Rescher, one can even draw a general picture of diminishing and
increasing distances which, moreover, is not confined to particle physics alone. Here, the first
peak period for the body of scientific knowledge, coming close to an age of “objective” or
“finalized” knowledge, occured around the time of Kant and Diderot and the second peak-
phase towards the fin de siècle. On the downward side of the cycle, the lower turning points
lie, according to Rescher, in the last quarter of the seventeenth century, associated with the
emergence of the Newtonian program, around the middle of the 19th century and, finally, in
the 1950´s. (RESCHER 1982:31) Thus, knowledge spaces can be characterized, on the one

      hand, by periods of wide-spread proportions of “terrae incognitae” like in the case of the
      infant stages of the Newtonian program and, on the other hand, by phases of common “terrae
      cognitae” where, like in the case of the Dirac equations the only main problem to be solved
      was the specification for a similar equation for the proton, the second elementary particle
      known in the twenties. (HAWKING 1991:10) In the long run, the structure of scientific
      revolutions leads, by necessity, to a cyclical pattern of close and wide distances with respect to
      a “finalized” body of knowledge ...

From the sevenfold way to knowledge and knowledge growth, it becomes relatively easy to see that the
dynamic independency-relations between knowledge and scientific production can and must be upheld in
various forms of interpretation and contexts. Scientific production, defined as the accumulated scientific
program-output over time, can follow different trajectories than knowledge in at least one of the following
five utilization contexts, namely with respect to -

      the evolution of knowledge domains where a radical upward trend in the proliferation of
      scientific output may be accompanied with a marginalization and a peripheralization of the
      number of specific knowledge domains ...
      the fitness of the respective output, evaluated in one of the many measures suggested within
      the philosophy of science debates over the last decades since a decrease in overall or special
      fitnesses may correspond with a highly active and growing output volume ...
      the discrepancy between a fast growing scientific output and a discontinuous drop in the
      “belief systems” of scientists leading to postmodernist attitudes of “anything can be
      recombined” and “anything is accepted” ...
      the rapid decrease in the utilization of a “knowledge stance” for the description of the
      behavior and “attitudes” of “minds and machines” whereas the scientific production grows at
      a rapid annual pace ...
      the construction and the production of encyclopedias or representative textbooks where a
      radical substitution process over time may lead to a constant or even shrinking volume of the
      encyclopedic content despite the fact that scientific production remains still in the lower part
      of a logistic curve ...
      the widening of distances between actual and potential knowledge despite a rapid increase in
      scientific production ...

Thus, the first divergence, the one between scientific production and knowledge, has been successfully
established. Moreover, it is hoped that the preceding set of divergent growth patterns of knowledge and
scientific production has demonstrated, once again, the strict independence of the development processes
in both areas.

                            3.2. SCIENTIFIC PRODUCTION

                                       AND INFORMATION:

                            A SECOND DIFFERENTIATION

Again, a dynamic divorce between information and scientific production stands at the beginning of the
present chapter. Consequently, the second separation, expressed formally once again as -

                                   ( SPt/dt  0)  (Iit/dt  0)  (Iit/dt  0)

demands that the growth of scientific production be strictly independent from the growth of information,
as defined below in a variety of different semantic fields. Thus, information growth and growth of
scientific production may follow a contradictory development pattern of the type -

                                          ( SPt/dt  0)  (Iit/dt  0)

as well as its logical counterpart -

                                          ( SPt/dt  0)  (Iit/dt  0)

Thus, the main objective at the present stage, after having introduced a definitional account of scientific
production in the previous chapter, will be to find a suitable account for the information concept. Looking
at some conventional definitions of information as summarized in the following list -

      Information kann in vieler Hinsicht als Ergebnis der Interpretation einer Nachricht aufgefaßt werden.
      (BAUER/GOOS 1973:2)
      Information ... bedeutet in der Alltagssprache Wissensgewinn. Die Informationstheorie versteht unter
      Information ... ein rein technisches Maß, das der Zeichenfolge einer Nachricht zugeordnet werden
      kann. (BREUER 1995:33)
      Eine gewisse Sequenz aus Symbolen trägt Information. In der Informationstheorie versuchen wir, ein
      Maß für die Größe der Information aufzufinden (HAKEN 1982:47)
      Information (Informationsgehalt): Maßgröße für die Ungewißheit des Eintretens von Ereignissen im
      Sinne der Wahrscheinlichkeitsrechnung (KLAUS/LIEBSCHER 1979:278)
      Information ist alles, was nicht Stoff oder Energie ist. (VÖLZ 1994:4) -

one may conceptualize information on a very general level as a measure which is based on a well-defined
system, coded or otherwise, and which is focused on one of three different systemic areas, namely, in
alphabetical order, -

      distribution of information (information1), a measure on the probability for a specific
      configuration of a simple or complex system, being composed of a large number of sub-
      components (entropy, order, etc.)
      information content (information2), the number of decision steps for finding and identifying a
      special systemic component in a sequence of components.
      transmission of information (information3), the magnitude of the flow of characters of a code-
      system between at least two embedded contexts (exchange of a code between machines
      (transmitter and receiver), of words and sentences between two persons, of symbols between
      man and machine, etc.)

The three different fields within the seemingly unitary field of information can be supported further with
the help of paradigmatic examples.

      For the information content, the main formula is the well known -

                                           Ic = log2 (1/p) = - log2 p

      where Ic is the information content of a code-element c, where log2 is the logarithm with basis
      2 and where p denotes the probability of the occurrence of c. Thus, for a code with eight
      elements, the information content for a specific element is 3 bits where bits are the units for
      measuring the information content. For the ten numbers of the decimal system, the
      information content becomes for example 3,322 ...
      Second, the mean value of the information content can be built up in the following way,
      starting from the initial condition of -

                                       H = p1I1 + p2I2 + p3I3 + ... + pnIn

      and summing over all i´s (i = 1,2,...,n) -

                                               H = -  pi log2 pi

      H is directly related to the entropy-concept in physics and is called entropy or negentropy
      since an inverse ordering of extremal points has taken place. The maximal degree of order,
      certainty, has the value 1, whereas absolute ignorance with respect to the “state of the world”

       is assigned the value zero. In contrast, maximum entropy in thermodynamics is associated
       with disorder and zero-entropy with perfect order.
       Finally, one can develop very easily a measure c for the transmission of information, namely -

                                                                   c = I/t

       c is measured in bits/second. For a b/w television set, consisting of 625 x 800 points, 15
       different shades of gray and, finally 25 pictures per second, on arrives at a c-value of 5 x 107
       bits/second which is high compared to the 50 bits in reading the present text ...

Through this definitional exercise, it becomes quite obvious that growth processes in the fields of
information are not to be conceived as quasi-proportional to development patterns in the area of scientific
production. Three examples will be sufficient to demonstrate the basic dissimilarities.80

       First, a radical drop in negentropy may go hand in hand with a steep increase in scientific
       production since the growth of scientific production, far from being concentrated in special
       domains of “hot topics”, is evenly distributed over the whole domains of scientific discourse.
       Second, the information content may be, albeit in a highly trivial manner, strictly independent
       from any change in scientific productions, be they increasing or decreasing, since the code-
       repertoire for scientific outputs has been set constant over time.
       Third, growth of scientific production can be accompanied by severe losses in written
       information-transmission. It might very well be the case that an ever-increasing scientific
       output leads to a clear-cut degradation in and to processes of a “second orality”, i.e. on an
       increasing reliance on personal exchanges.

These three examples must be sufficient for showing the possibly divergent dynamics in the field of
scientific production and in the information area. Due to the unclear status of the transitivity of the
divergence relation - “If A is different from B and B is different from C, then A is different form C” does
not hold within the practices and routines of human reasoning (KAHNEMANN, TVERSKY 1973,
KAHNEMANN/SLOVIC/TVERSKY 1982) - it seems safer to establish, on independent grounds, the
separability between knowledge and information, too.

    It should be remarked, at least in a footnote, that information and scientific production can be related in a trivial manner by “measuring”
the quantity of scientific output in pages, sentences, words or even letters and by classifying the resulting numbers as “amount of
information”. In this manner, the genuine perspective relevant to the information concept gets entirely lost. Moreover, no cognitive gain has
been achieved since scientific production has been provided with just another, highly un-informative indicator.

                                 3.3. INFORMATION AND


                                      A THIRD DIVISION

The third important differentiation, once again defined in dynamic terms as -

                                 ( I jt/dt  0)  (Kit/dt  0)  (Kit/dt  0)

sees the growth of knowledge in all its seven basic instantiations as strictly independent from the growth
of information in its three different instantiations. Moreover, knowledge growth and growth of
information may exhibit all possible dynamic development patterns, especially the two diverging cases -

                                         ( I jt/dt  0)  (Kit/dt  0)

as well as its historically more frequent counterpart -

                                         ( I jt/dt  0)  (Kit/dt  0)

What must be shown therefore, lies in the construction of a list of significant examples which exhibit
clearly diverging development patterns. Surprisingly, a large number of historical as well as contemporary
configurations come to mind.

      First, a knowledge1 increase with respect to scientific disciplines can be accompanied by a
      decrease in information, defined as neg-entropy.
      Second, an increase in the distances between actual and potential knowledge spaces and thus,
      a drop in knowledge7 can be brought about by an increase in information, defined once again
      as neg-entropy.
      Third, the transmission rates within the scientific system may increase, whereas the
      knowledge3, expressed in “justified true beliefs” is shrinking simply because the transmission
      increase has created, inter alia, a significant lowering in the credibility of the available
      knowledge alternatives.
      Fourth, the transmission and thus the rate of information may significantly decrease, but the
      knowledge domains may be spreading in a heterogeneous manner related to the fact that
      communication and exchange processes occur at a much lower rate.

Thus, the third independence between three seemingly similar domains brings an essential clarification,
central for the conceptualizations of National Innovation Systems, to an end. Concluding with a quotation,
already introduced at the beginning of the second chapter -

     Reichtum produziert heute (fast) nur noch die Information und(!) das Wissen (DRUCKER 1993:262)

it should become obvious by now that, due to the overwhelming importance of concepts like knowledge,
information and scientific production for a “knowledge based-economy”, a substantial amount of re-
designing is required. The most important step in this re-design process is accomplished within the next
chapter which will offer, finally, a comprehensive and well-integrated framework for the transformation
of Drucker´s original quotation into the following sentence -

     Wealth is produced today almost exclusively by scientific and technological programs, i.e. by
     economically relevant scientific, technological or, finally, social recipes for production and

                                    3.4. INFORMATION,

                              SCIENTIFIC PRODUCTION

                                    AND KNOWLEDGE:

                                A TRANSDISCIPLINARY

                            CODE-BASED PERSPECTIVE

Having followed, so far, a path of differentiation between scientific production, information and
knowledge, the final chapter will be devoted, in a spirit of establishing a unity among a newly established
diversity, to integrate the three main research areas under the single heading of embedded code-systems
(ECS). For this target, the heterogeneous sets of knowledge domains, of information fields and of
scientific production areas have to become distinctive elements within a single, homogenous ECS-class.
Three initial examples, one from each of the three domains of knowledge, information and scientific

production, must be sufficient for the feasibility of the desired integrated conceptual code-based
Table 3.1:    A Unified Code-Based Perspective for a Heterogeneous Class
              of Cognitive Growth Processes in Science

SPECIFIC           GROWTH &                                  ESC-
DOMAIN             DEVELOPMENT                               ANALOGY

                   Growth in                                 Growth in Scientific
                   Scientific Output                         Programs
SCIENTIFIC         Growth in                                 Growth in the Co-Author-
PRODUCTION         Co-Publications                              ship of Programs
                   Growth in National or                     Growth in National or
                   International Linkages                       International Linkages
                   with Respect to Scientific                   in Science Code-Based
                   Exchanges                                 Exchanges

                   Growth in “Justified True                 Growth in “Justified True
                   Belief”                                   Meme-Systems”
                   Development of Knowledge                  Development of Scientific
                   Domains                                   Language Games
KNOWLEDGE          Increase in Encyclopedic                  Increase in Programs within the
                   Traditions                                Encyclopedic Genre
                   Growth in the Distances between           Growth in the Distances between
                   Actual and Potential Knowledge               Actual and Potential Intellectual Code-
                   Spaces in Scientific Disciplines          Spaces within a Specific Scientific Domain

                   Growth in the Neg-entropy of              Growth in the Overall Fitness (“Degree of
                   Scientific Production in a Specific       Order”) in a Fitness Landscape of a Particular
                   Domain                                    Scientific Area
                   Upward Development in the                 Increase in the Speed of Transmission for
INFORMATION        Transmission of Scientific Results        Scientific   Programs    (or   for   Condensed
                   or of Scientific Information in General      of Scientific Programs)
                   An Increase in the Necessary Steps        An Increae in the Necessary Steps
                   of Identifying New Scientific Results     of Identifying New Scientifc Programs (or of

                    or of Scientific Information in General       Condensed     Programs     of    Scientific

     Take the case of scientific production as scientific output per unit of time, then a code-based
     transformation is accomplished by defining the scientific output - articles, books, patents,
     prototypes, tests-reports, research paper, etc. - as a heterogeneous set of “programs”, written in
     a specific code-system, namely the scientific code for short.
     Using knowledge under the heading of knowledge domains, the varieties of scientific
     disciplines as well as their respective programs may be used for the identification of
     disciplinary program areas.
     Finally, information as an order-concept of a given system, can be easily transformed into a
     corresponding code-concept, referring to the degree of order in a system by arranging the
     systemic components as code-elements and by applying the well-defined negentropy function.

To sum up, the three different classes of scientific growth and development processes, based either on
scientific production, on knowledge and, finally, on information, can be conceptualized in a code-based
perspective in a unified code-based way, exemplified by Table 3.1 (previous page)
The short list of elementary growth and development processes within science can be extended, by
inductive reasoning, to the general case of conceptualizing, under the heading of embedded code-systems,
all knowledge-, information- and scientific production-relevant domains of discourse into their code-
based analogues. The main rationale for the inductive generalization comes from two results:

     First, evolutionary systems, as will be recalled from Section I, should be considered as dual-
     level ensembles, with the geoetype domains and the genetic code spaces as one of its two foci.
     Since the quintessential substrate for most of the growth processes in knowledge and
     information-areas is language based, it must be possible to re-conceptualize such issues almost
     naturally as a code-system, too.
     Second, the transdisciplinary framework for programs, scientific, technological, social,
     ecological, biological and otherwise, makes it (almost) tautological to replace any scientific
     output problem into an isomorphic ECS-issue.

Consequently, the essential requirement for the applicabilities of “complex embedded systems” (CES) lies
in the important genotype fact that “languages”, or more generally, “code systems matter”. And it can be
safely argued that the ECS-frameworks are, almost by necessity, applicable to any of the problem areas of
information, knowledge or scientific production, past and present ...

                          4. HELICES AS “KNOWLEDGE BASES”

                         OF NATIONAL INNOVATION SYSTEMS

Following the integrative capacities of the new ECS framework, it becomes relatively easy to identfy,
then, a central genotype linked theoretical concept for the constitution of National Innovation Systems.
Moreover, this new concept has come about as a recombinative operation of merging the well defined
genetic concept of “double helix” with the metaphor of a “triple helix” (ETZKOWITZ/LEYDESDORFF
1995) which has been introduced to describe the relationships between science, economy and the state
sector. The result of this merging procedure leads, then, to a generalized and transdisciplinary notion of
“helices” which, at the meta-theoretical level, comprise the sum total of (re)productive and non-
reproductive programs for the production and the maintenance of the embedded ensembles of code-
systems. Thus, the double helix at the genetic level and, as will be shown in the subsequent two chapters,
different types of helices at the societal level can be considered as the “instruction, recipe and operation
basis” for the construction and the maintenance of phenotype systems - organisms and species, products,
socio-technical systems, organizations, households and the like ...

                                              4.1. SINGLE HELICES


From the transdisciplinary helix-specification, societal “helices” can be introduced for the NIS-context as
a set, consisting on the one hand, of all programs which play a non-trivial role in the (re)production of
phenotype-components of a specific societal system and, on the other hand, of those programs which have
not entered into meaningful (re)productive relations so far. 81Moreover, societal systems are introduced
here as a set of large-scale systems, comprising essential societal domains like the economic realm, the
scientific area, the political-administrative field, artistic and cultural aspects, sports and recreation
territories, household activities, media and the like.82 Contrary to a classical tradition, focusing on a
specific number of societal systems in an a priori manner or on specific well defined societal “functions”
and unique goals, associated with each of these ensembles, the number of societal systems is neither

   In Section III, it will become clear however, that programs are not the only elements of helices since at least two additional components
must be mentioned, namely, first, special program “chunks” called “memes” and, second, “neural groups”.
   For more details, see the chapter on “basic defifinitions” (4.1) in Section IV.

bound to a fixed number or to an upper limit, nor is the functional perspective upheld. Rather, an
irreducible multiplicity of different systemic constructions for, say, the economic system, with some well-
established paradigmatic or core specifications for markets, supply and demand, prices, micro-economic
units like firms and households or rational decision procedures, will be considered as the only a priori
requirements for the analysis of a large-scale societal ensemble. Consequently, a singular helix like the
“science helix” can be described in its basic structures and configurations in the following manner:

       First, an important Caveat must be stated right at the beginning. The science helix should not
       be equated with a popular notion in the sphere of scientific production, namely with Popper´s
       “World 3”. Here, a highly problematic differentiation yields, following Mario Bunge, an
       illusive separation between well-defined areas, namely World (1,2)83, and a separated World

           When someone finishes a paper or a drawing, these pieces of matter can be detached and seen
           by somebody else: even their creator can stand back and contemplate them as if they were self-
           existing, while in fact their „content‟ depends on their being perceived and understood by some
           brain. This creates the illusion that we are in the presence of three separate items: the neural
           (and motor) process resulting in the writing or drawing, the cultural artifact, and the knowledge
           or feeling encoded in the latter. The next step is to collect all such bits of knowledge detached
           from brains - i.e. all the problems and data, theories and plans „in themselves‟ - and endow such
           a collection with a life of its own. The final step is to give a name to such a collection of items
           allegedly hovering above brains and society - e.g. the „realm of ideas‟, the „objective spirit‟ or
           „world 3‟. (BUNGE 1983: 65)

       To make the dimensions of Popper´s “World 3” more visible, a direct comparison should be
       helpful: “World 3” includes even Borges´ already infinite “Library of Babel” -

           In dieser ungeheuer weiträumigen Bibliothek gibt es nicht zwei identische Bücher. Aus dieser
           unwiderleglichen Prämisse folgerte er, daß die Bibliothek total ist und daß ihre Regale alle
           irgend möglichen Kombinationen der zwanzig und soviel orthografischen Zeichen (deren Zahl,
           wenn auch außerordentlich groß, nicht unendlich ist) verzeichnen, mithin alles, was sich irgend
           ausdrücken läßt: in sämtlichen Sprachen (BORGES 1974: 51) -

   The notion of “World (1,2)” should be seen as a reminder that, following Section III, no principal divisions should be undertaken
between the phyical universe and its psychological domains ...

       as a subset proper, since Borge´s “Babel-Library” would fill but one department in Popper´s
       “World-3 Labyrinth”, the others consisting, inter alia, of possible paintings, possible
       partitures, possible sculptures, possible designs and the like ...84
       Second, the internal structures and sequences of helices for a specific societal system do not
       become, unlike their biological counterparts, a particularly important issue. Under no
       circumstances, specific complex arrangements like the ones identified in the previous sections
       on the genetic code play a decisive role in the case of societal helices. On the contrary, the
       notion of “programs for free” points to the vaguely structured and un-differentiated character
       of (re)productive programs. Thus, the helix for a specific societal system, far from being a
       homogenously interlinked ensemble, should be considered as a widely distributed and trivially
       interlinked arrangement which, however, plays an integral part in the (re)production of
       societal systems ...
       Third, the spatio-temporal domains for societal helices is not, once again totally unlike the
       biological case, restricted to very specific areas and highly standardized embeddedness
       relations but is widely distributed throughout the socio-economic system under consideration,
       exhibiting, as will become clear in Section III, a large number of heterogeneous
       embeddedness relations. Take, as a concrete example, the present report which will be
       distributed over some national and international places, then one can identify the following
       components which could become relevant in the (re)production and in the adaptation of the
       Austrian science system. For instance, the programs for science reforms (Volume VI) may
       become part of national or international science policy institutions, some catch phrases like
       the “genotype of National Innovation Systems” or “embeddedness relations” in Volume I may
       eventually become part of the “meme-repertoire” of NIS-researchers, the empirical results of
       the Austrian Survey of Innovation and Transfer in Volume IV may induce some research
       institutes to increase their transfer outputs or their research orientation toward a mixture of
       applied and basic research and the like ...

   It can be shown very easily that any interpretation of intellectual spaces, fitness landscapes as causally relevant domains sui generis is
bound to fall under the verdict of Occam´s razor. The main reason has been provided by Karl Popper himself, since he insisted, over and
over, on the inexplainable character of future knowledge -

       Wir können mit rational-wissenschaftlichen Methoden das zukünftige Anwachsen unserer wissenschaftlichen Erkenntnisse
       nicht vorhersagen (POPPER 1971:XI).

Since the results of future explorations in knowledge spaces cannot be anticipated at the present time and since, at present, only the space of
possible configurations can be defined, it becomes impossible to place Popper´s “World 3” and similar ensembles into any meaningful
context of explanation or prediction. A space of all possible configurations cannot exert a “causal influence” on specific outcomes. This
impossibility, however, runs counter to Popper´s own vision of causal transport from “World 3” to “World 2” -

       There is no doubt in my mind that Word 2 and 3 do interact If we try to grasp or understand a theory, or to remember a
       symphony, then our minds are cuasally influenced; not merely by the memory of noises stored in our brains, but at least in
       part by the work of the composer, by the autonomous inner structure of the World 3 object which we try to grasp. All this
       means that World 3 can act upon the World 2 of our minds. (POPPER 1982:128p.)

In this sense, Popper´s “World 3” as well as other “objectified assemblies” have to be abandoned via Occam´s razor, since a domain devoid
of explanatory and predictive contexts becomes, almost by necessity, empty and blind...

       Fourth, the helix composition of the scientific system is accomplished through the integration
       of a variety of components which, originally, have been developed both in their genotype and
       phenotype-forms in other societal segments. Thus, Steven Shaffer´s reminiscence on the
       societal constitution of science85 can be given a straightforward additional support, since even
       at the genotype level one finds, aside from the scientifically produced programs, rules for
       bargaining and coalition building which have come into existence long prior to the
       “differentiation” of a science system or a rule-system of “gossiping” which has been labeled
       rightly as the oldest societal mass medium (KAPFERER 1996). In addition, one may add
       rules for special communication processes like “greeting” and, of course, the whole rule-
       repertoire of “everyday tasks” or “lifeworld-drafts” like walking, opening doors, sitting,
       making telephone calls, buying electronic equipment, driving, etc. which are performed within
       the spatio-temporal confines of a large-scale science system ...
       Fifth, a highly interesting segment of the science helix, undergoing rapid changes from a
       cognitive point of view, comes from the programs on the science system itself, its
       organizational deep-structures, its cognitive dynamics and the like. In particular, programs on
       National Innovation Systems, especially volume VI of the present report, offer, at the
       genotype level, an interesting array of “recipes” or “instructions” for reforms, for increasing
       the network densities, for speeding up exchange processes and the like. Without exaggeration,
       this specific helix segment can be described as the differentia specifica, separating the science
       helix from other societal helices ...
       Sixth, the embedded code-systems and their respective programs, even in the case of the single
       science helix, are not restricted to, say, natural languages or “scientific languages” like
       mathematics alone, but comprise other code-systems as well, including symbolic codes,
       musical codes and the like ...
       Seventh, an essential element for the (re)production of the science system itself, comes, quite
       naturally, from the genetic codes for organisms, especially from the genetic code for actors
       within the science system itself. Since a sudden dramatic change in the genetic outfit of
       humans could even bring the program-proliferation within science to a halt, the genetic code
       relevant for “science actors” will be included into the ensemble, called “science helix”.

   Steven Shaffer´s “Es gibt ebensoviel Gesellschaft in der Wissenschaft wie außerhalb” (quoted after NOWOTNY 1996) may serve as the
“dominant metaphor” for an intergrated perspective. In a similar fashion, Helga Nowothny has come to the conclusion that in contemporary
anaylyses of science a remarkable as well as radical shift has been accomplished which, inter alia, runs through the present theoretical
foundation as well -

       Die Trennung in wissenschaftsinterne oder kognitive und wissenschaftsexterne, soziale Faktoren, die lange Zeit auch die
       Reflexion über Wissenschaft beherrschte, hat ihre Gültigkeit verloren. In der Wissenschaftsforschung und in der neueren
       Wissenschaftsgeschichte wurde sie wietgehend aufgegeben. (NOWOTNY 1996:39)

In this manner, some striking similarities and, more importantly, dissimilarities between biological and
societal helices have been identified. What follows next, lies in the straightforward extension of single
societal helices to more general cases.

                    4.2. LINKAGES BETWEEN NIS-HELICES

Societal helices can be identified for any type of large-scale systemic configurations. Thus, for two
societal systems, one can define a “double helix”, for three societal systems a “triple helix” and, more
generally, for n societal systems an “n-tupel” helix. What remains to be shown at this point, is the specific
linkage structure between societal helices. In sum, three probably unexpected and surprising
characteristics can be identified:

      The first counter-intuitive feature with respect to helix linkages lies in the fact that an increase
      in the number of societal systems is accompanied by a diminishing number of additional helix
      components. The main reason for the “law of decreasing helix components” has to do with the
      embeddedness-relations discussed at length in Section III. In short, a single helix like the
      scientific one will and must comprise (re)productive programs not only from the scientific
      domain but from the other societal domains as well. Clearly, this collection will not consist of
      all of the (re)productive programs outside the specific systems domain. Consequently, a
      second societal system will add only a fraction of the (re)productive programs which become
      necessary for the (re)production of the large-scale system in question itself.
      The second character of “double helices” refers to their specific shapes. Instead of invoking
      the metaphor of a well-organized “double-helix”, it seems far more appropriate to utilize the
      analogy of a “helix pool” where no separate compartmentalizations should be built up,
      differentiating between program strands of a “science helix” and the program strands of an
      “economy helix”. One may separate the program sets into three disjunctive classes, namely
      into programs used exclusively in the reproduction of the scientific system, programs reserved
      for economic reproduction alone and mixed programs, relevant both for the economic and the
      scientific system. But such a division turns out to be entirely context dependent, being directly
      related to the spatio-temporal boundaries drawn for the NIS-configuration under investigation.
      Third, another highly interesting feature with respect to the linkage structure of multiple
      helices lies in the fact that the very act of adding a new helix to a specified helix pool will
      alter, in all probability, the existing linkages within the original ensemble. Thus, a triple helix
      for the societal systems of economy, science and the state will and must change the program

        separations at the level of dual helices since parts of the programs reserved for the
        reproduction of the scientific system alone will turn into mixed programs, decreasing, thus,
        the number of helix-components bound entirely to the scientific system. Consequently, the
        form of the helix pool and its partitionings remain sensitive to the context-specifications like
        the number of societal systems under consideration, the spatio-temporal demarcations and the

In this manner, the introductory remarks on the genotype-constitution of National Innovation Systems can
be finished. The two important points for the subsequent construction of NIS-families undertaken in
Section IV, namely the focus of helices on the genotype-level and the combinability of societal helices
into more comprehensive “pools”, have been put forward in a condensed and, hopefully, plausible

                                     5. SCIENTIFIC CODES AND

                              RECOMBINATIONS: THE RIDDLES


Having arrived at a clear separation between patterns of scientific production and reception, information
and, finally, knowledge, and, moreover, at a satisficing characterization of the genotype bases of National
Innovation Systems the final chapter will bedevoted to a “second order-operation”, namely to a
recombination of recombinations. One of the most interesting variations on the theme of recombinations
in embedded code-systems lies in the area of scientific changes, culminating in the present analyses and
discussions of scientific innovations or, alternatively, of scientific “creativity”.86 Thus, before entering
into an appropriate recombination of recombinations for the scientific fields, a brief summary of the
existing stock of perspectives on the elusive phenomenon of scientific creativity must be presented.

    While the concept of scientific innovation must be considered as the more general one, the notion of “scientific creativity” captures an
important aspect in the overall innovation process, namely the operations as well as characteristic dispositions on the side of scientists. In the
language of embedded code-systems, creativity studies are concentrated on genotype and phenotype changes within the scientific system,
whereas innovation, as has been pointed out in Section I, covers the whole range of change processes from biology to machines and socio-
technical ensembles, organizations ...

                                   5.1. SCIENTIFIC CREATIVITY -

                                     A THEORETICAL OVERVIEW

The available literature on scientific creativity presents an astonishing diversity of generic traits and
indispensable requirements which, though mutually inconsistent, must be present in the in the
emergence1,2,3 of new genotype programs.87 As an interesting stage-setting account of scientific creativity,
Robert Sternbergs “The Nature of Creativity“ will be used, to present six basic ingredients and
characteristic features of creative processes in science.

       First, successful creative processes take a relatively large amount of time and are not to be confused
       with spontaneous “Aha-experiences” - „the very nature of creativity depends on the time constraints
       involved      and     the    opportunity       to    revise,     or   nurture,     the    outcomes       once     produced.“
       (TARDIF/STERNBERG 1988:430)
       Second, creative processes may be due to a variety of reasons, starting as an active exploration for
       problem solutions, as random variations, as a reaction of persistent failures, deficiencies and
       shortcomings or as „a general drive toward self-organization through the reduction of chaos.“(IBID.)
       Third, creativity, due to a variety of reasons, finds itself in a close family resemblance with notions
       like cognitive dissonances, intellectual conflict and the like. „First, one may be faced with conflict
       between staying with tradition and breaking new ground ... Second, tension may lie in the ideas
       themselves ... Finally, it may exist in the constant battle between unorganized chaos and the drive to
       higher levels of organization and efficiency within the individual, or the society at large.“(IBID:431)
       Fourth, the major attributes relevant for creativity processes can be associated with processes like the
       “formation of analogies”, “redefinition of problems”, an application of established patterns and search
       procedures for new domains „to make the new familiar and the old new.“(IBID.)
       Fifth, creativity in persons should not be considered as an encompassing capacity in a variety of
       essential domains, but a s a capacity which operates within a specific field and not in other domains as
       well. Creative persons in a specific area turn out to be, with a high degree of probability, as non-
       creative in other domains.
       Finally, creative persons are regarded, sixth, as exhibiting a high concentration of the following list of
       attributes - „originality, articulate and verbally fluent, thinks metaphorically, uses wide categories and
       images, flexible and skilled decision maker, makes independent judgments, builds new structures,
       finds order in chaos,, questions norms and assumptions, alert to novelty and gaps in knowledge, uses
       existing knowledge as base for new ideas.“(IBID:434)

    Even in 1990, a summary of a collection of articles on scientific creativity produces, in the end, just truisms of an indisputable, though
tautological variety like the following - „the slow organic fusion of talents into tradition and genre is a historical-cultural phenomenon that
takes place over longer-than-life slices of time and space.“ (ALBERT/RUNCO 1990:258).

Creative processes in science require, in a nutshell, on the micro- as well as on the macro-levels, first, a
high degree of competence or, in more general terms, a sufficiently rich and up to-date repertoire, second,
a high ability and willingness for change, as expressed in a frequent application of non-standard
approaches and, finally, third a high competence in the analogy-formation, expressed in a permanent
activation of inference- and inductive operations..88
Having established a common plattform for creative processes in the scientific domain, Table 5.1 presents
a morphological summary of current model- and theory families which occupy the main cognitive space
on this specific topic:

Table 5.1: A Morphological Frame for Approaches to Scientific Creativity

                           MICRO-LEVEL                                      MACRO-LEVEL
INDEPENDENT                Theory-Group I                                   Theory-Group II
DEPENDENT                  Theory-Group III                             Theory-Group IV

The four sets of approaches summarized in Table 5.1 can be described in greater detail by referring to the
following list of characteristic attributes.

      The first group which incorporates a large number of different approaches, ranging from the
      historiography of science to “New Age” offsprings, stresses the unsystematic, inexplicable or,
      depending on the theological roots, the divine character of creative production. Generally
      speaking, Family I sees scientific creativity as a phenomenon with few or even no possible
      ways for explanations, let alone “mechanisms”.89
      Within the third theory-nets, one finds predominantly approaches from the cognitive science-
      field or, more generally, from the “neuro-sciences” (GAZZANIGA 1995) where a
      comparatively large number of models has been assembled over the last decade to account for
      the, horribile dictu, micro-mechanisms of scientific creativity. 90
      Theory-group II is situated mainly on the macro-level and stresses, above all, the role and
      function of scientific communities or of, to use Ludwig Fleck´s term, “Denkkollektive” whose

    On this point, see especially HOLLAND et al. (1989), HOLYOAK/THAGARD (1989), ANDERSON/THOMPSON (1989) or
    On this tradition, see, aside from RUNCO/ALBERT (1990), also ARIETI (1976), KOHN 1989 or PACEY (1992). For an appropriate
antidote, see esp. WEISBERG 1989.
    Aside from STERNBERG (1988), see also BODEN (1990), FINKE et al. (1992), GARDNER (1993), HOFSTADTER (1985, 1995),
LANGLEY et al. (1987), SIMON (1977), STERNBERG/FRENSCH (1991), STERNBERG/WAGNER (1994) or WEST 1991.

        internal linkages and network-relations determine, to a large extent, the fall and the rise of
        macro-scientific creativity.91
        Finally, a fourth model-class, stressing the context dependent character of creativity on macro-
        levels too, emphasizes a variety of different external factors and ensembles as pre-conditions
        or as necessary bases of scientific creativity. This external factor set ranges from traditional
        class-positions to more recent developments which stress the determining role of gender and
        gender-relations for processes of scientific creativity.92

Thus, via Table 5.1 a large group of recent approaches on scientific creativity has been captured which
serves as the theoretical background for the subsequent discussions. At the end, a new group of theories,
linked to the groups III and IV of Table 1, will emerge which, moreover, has the distinctive advantage of
being a special case of the more general phenomenon of recombinations in embedded code-based

                                5.2. SCIENTIFIC CREATIVITY -


Starting, once again, with the general definition for a code-based innovation potential of an ECS -

        Full-scale innovation potential for an embedded code-system consists in having a rich repertoire for
        recombinations with comparative advantages, following them recursively, applying them at the meta-
        level, and modifying them accordingly.

one is led to the following small modifications which become necessary within the scientific field.

        The first requirement - „availability of a rich repertoire“ - may be easily adapted to the
        scientific arena by requiring a sufficiently competent integration of the current state of the art
        in a given scientific field. Thus, the ability to manipulate the latest available scientific
        programs in a specific domain must be formulated as the first essential requirement for
        achieving a creative scientific result or product.

     See on this point esp. ROOT BERNSTEIN (1989), THAGARD (1992) or ZIMAN (1995).
     On this tradition, see, inter alia, SCHIEBINGER (1993, 1995) or YOUING-BRUEHL (1991:215pp.)

       The notion of a rich repertoire can be extended, second, to the availability of a meta-level
       control element, i.e. to the existence of cognitive maps on “intellectual spaces”93, which are
       characterized by three basic properties: first, by a comparatively large number of unexplored
       territories so that scientific recombinations have a sufficiently broad domain for potential
       applications („explorations in cognitive space“), second, a fully contradictory configuration
       which require solutions or clarifications („dissonances in cognitive space“) and, finally, a
       highly entropic overall configuration or, to use a famous concept by Jürgen Habermas, “neue
       Unübersichtlichkeiten” which are in urgent need of new categorizations and separations of a
       distinctively lower-entropic type („orderings of cognitive space“).
       The central area for scientific creativity consists, like in the general case, in the application of
       a set of distinctive “creativity-” or “innovation-operators”, which alone or in combination
       generate, in a recursive manner, the sought for creative recombinations. Using, once again, an
       enlarged list from Douglas R. Hofstadter (HOFSTADTER 1995:77), one can identify at least
       ten recursive creativity operators within the scientific fields -

           Adding, the integration of new building blocks into an existing scheme ...
           Breaking, the differentiation of at least one scheme into two disjunctive building blocks ...
           Crossing-over, the breaking of at least two schemes and their merging into a new ensemble ...
           Deletion, the destruction of a specific building block from a set of schemes ...
           Duplication, the repeated insertion of at least one identical scheme ...
           Inverting, the making of copies with an opposite scheme ...
           Merging, the integration of at least two existing schemes into a new one ...
           Moving, the shifting of schemes out of its established boundaries ...
           Replacing, the substitution of a particular sub-scheme by another one ...
           Swapping, the vertical movement of a particular scheme from a level Li to a different level Lj ...

       The requirements four and five demand, also in the case of scientific creativity, a sufficient
       degree of flexibility and a cognitive capacity to salient adaptations (requirement four) as well
       as the approaching of target domains within a relatively small amount of time (requirement
       five). Finally, a sufficient control-capacity must be able to map the adaptations reached so far
       in a cognitive space of possible problem solutions - „applying it at the meta-level“
       (requirement six) - ...94

With this small sample of essential requirements, the basic ingredients for scientific recombinations can
be set in motion ...

   For a history of science-conceptualization on the notion of intellectual spaces, see esp. SHAPIN/SCHAFFER (1985).
   It should be added, that this type of creativity analysis allows for an easy utilization of complex network models where the creativity
operators are building the network nodes and where the existing schemes undergo a set of recursive transformation processes.

                          5.3. SCIENTIFIC CREATIVITY -

                         A SMALL SET OF CASE-STUDIES

For an exemplar-based presentation of the six creativity-requirements, four internationally well-known
cognitive achievements from the science system in Austria in the First Republic (1918 - 1933/38), as well
as a local recombination effort during the First Republic and after have been selected. For the first set,
four renowned examples have been chosen, namely -

     Kurt Gödel : Incompleteness-Theorems
     Rudolf Carnap, Otto Neurath et al.: Criterion of Verification
     Karl R. Popper : Criterion of Falsification
     Rudolf Brunngraber, Otto Neurath : Statistical Novel

In these four instances, the conceptual apparatus on recombinations, which, inter many alia (Clifford
Geertz), can be understood as an approach for the study of innovations and scientific creativity too, yields
a series of partially unexpected and counter-intuitive results.
For Gödel´s “Incompleteness Theorems”, analyzed according to the list of creativity operators, the
following transformation steps have become essential:

     Dissonance in cognitive space, the foundational debate in mathematics (logicism, intuitivism),
     the challenges posed by the Hilbert-program ...
     Replacing, the substitution of an established mode of numerical coding through a new coding
     procedure, the Gödel-numbering ....
     Swapping, the utilization of the Gödel-code for numbers for axioms and theorems of number-
     theory itself ...
     Meta-Level-control, the selection of a research-heuristics which, within the history of modern
     mathematics, has proven to be very successful in the generation of paradoxes, namely the
     route of self-referential constructions ...
     Merging, the integration of multiple levels and a homogeneous coding-scheme with self-
     referentiality ...

It should be added that the recombination or creativity operations described above capture apparently the
essence of Gödel´s proofs.

     The first key-idea is the deep discovery that there are strings of TNT (Typographical Number Theory,
     a notation scheme developed by D.R. Hofstadter, K.H.M.) which can be interpreted as speaking about
     other strings of TNT; in short, that TNT, as a language, is capable of „introspection“ or self-scrutiny.
     This is what comes from Gödel-numbering. The second key-idea is that the property of self-scrutiny
     can be entirely concentrated into a single string; thus that string´s sole focus of attention is itself ... In
     my opinion, if one is interested in understanding Gödel´s proof in a deep way, then one must
     recognize that the proof, in essence, consists of a fusion of these two main ideas. Each of them alone
     is a master stroke; to put them together took an act of genius. If I were to choose, however, which of
     the two key ideas is deeper, I would unhesitatingly pick the first one, the idea of Gödel-
     numbering.(HOFSTADTER 1982:438)

The second creative achievement from Austria´s First Republic is given by the “unified science”program
of the Vienna Circle, as it has been developed from the late twenties onward and as it has been
institutionalized in six large conferences on the “unity of science” from 1935 in Paris to 1940 in
Cambridge, Mass. Here, verificationist or testability linkages have been demanded between a basic or
protocol-language on the one hand and higher theoretical terms, expressions, relations or functions on the
other hand. Moreover, this so-called physicalist program, named, inter alia, to honor the ongoing
revolutions in physics, has been forcefully applied in the areas of cultural sciences and psychology for
which a time-space-based protocol-language has been, especially within the intellectual backgrounds of
the German speaking world, a typical non-entity. Relying on the creativity-operators specified above, the
following transformation steps can be distinguished.

     Exploration in cognitive space, the extension of a new tool, i.e. the “new logic” of Frege,
     Rusell and Whitehead, for the analysis of the languages of science ...
     Ordering of cognitive space, the search for a set of universally applicable principles, criteria
     and standards which, despite the splitting of cognitive spaces into “idiographic domains” and
     “nomothetic areas”, can be utilized within the natural sciences and within the social sciences.
     Merging, the integration of natural sciences and social sciences under the heading of a
     universal scientific language ...
     Adding, the specification of a special class of empirical sentences, the so-called protocol-
     sentences which should serve as the basic language both for the natural as well as the social
     sciences ...
     Moving, the shifting of a space-time language into the domains of the cultural sciences and
     psychology ...
     Breaking, the partitioning of statements into empirically meaningful ones and their
     syntactically or semantically ill-defined counterparts. As a special point of controversy, large
     segments of contemporary metaphysics, especially of proponents of right wing-Hegelianism
     or Heidegger´s Sein und Zeit, have been included in the second class. Moreover, wide

      segments of the psychology of the time has been deemed in need of reformulation and
      reconfiguration according to the physicalist approach ...

Given the Vienna Circle-program, the transition to the third major intellectual achievement is easy to
undertake, since a single property in cognitive space and only two creativity operators are needed to
account for it:

      Dissonance in cognitive space, the irrelevance, at least in Popper´s view, of the criterion of
      verification for problems of theory selection or scientific growth ...
      Inverting, the shift from verification to falsification ...
      Breaking the separation of the domain of “meaningful statements” into two disjunctive
      domains, namely into falsifiable statements like the “General Theory of Relativity” on the one
      hand, and into unfalsifiable statements like those to be found in Marxism, in psycho-analysis
      of the Freud-variety, Adler-offspring, Jung-adherence ...

At the same time, the case of Karl R. Popper offers an interesting illustration that scientific creativity is
under no circumstances to be equated with sudden mystical experiences of “enlightenment” or
spontaneous “Aha-experiences”. Consider, first, the following autobiographical remarks by Karl R.

      It happened shortly before my seventeenth birthday. In Vienna, shooting broke out during a
      demonstration by unarmed young socialists who, instigated by the communists, tried to help some
      communists to escape who were under arrest in the central police station in Vienna. Several young
      socialist and communist workers were killed. I was horrified and shocked at the police, but also at
      myself. For I felt that as a Marxist I bore part of the responsibility for the tragedy - at least in
      principle. Marxist theory demands that the class struggle be intensified, in order to speed up the
      coming of socialism ... I now asked myself whether such a calculation could ever be supported by
      „science“. The whole experience, and especially this question, led me to a life-long revulsion of
      feeling.(POPPER 1974:25)

It should be seen as highly remarkable and astonishing that roughly fifteen long years have been needed
for tangible linkages between a „life long revulsion“ and a corresponding manifestation, namely the
publication of Poppers “Logik der Forschung” in the fall of 1934. Likewise, it is very strange to note that
Popper´s solutions to epistemological problems follow very closely, albeit in an inverse fashion, the on-
going discussions of the Vienna Circle on the criterion of verification or on the debate on protocol-
sentences. Generally speaking, one must be extremely careful to take ex-post accounts of life-long
development patterns, revulsions or cycles at face-value. One is even tempted to ask whether these

“stylized configurations” should be taken seriously at all since they have, almost by necessity, a very small
capacity to integrate or aggregate the full sequence of biographical events.95
The fourth case of a creative solution leads away from the narrow confines of the scientific system and
opens up the domain of literature and novels where around Christmas-time of 1932 a book with the
slightly mysterious title of “Karl und das 20. Jahrhundert” has been published. Its author, Rudolf
Brunngraber, presents the reader with the biography of a Karl Lakner who, from a typical working class
background, manages to become a teacher and who, in the end, suffers from the consequences of the
Great Depression of 1929ff. and commits suicide. The interesting point in this not particularly exciting
story lies in the integration of literary elements with elements form the social science program of the
Vienna Circle, as it has been formulated and developed by Otto Neurath. Thus, elements of a well known
literary genre, the biographical novel, become intermixed with components of Neurath´s system of social
indicators on “Lebenslagen”, on economic indicators and, more generally, with Neurath´s idea of “cosmic
aggregations”. Thus, Brunngraber´s novel offers a clear-cut projection of intra-scientific components into
the area of literature. Looking at this achievement in terms of creativity operators, one can identify the
following transformation steps:

        Exploration in cognitive space, the utilization of other cognitive components of the Vienna
        Circle program for the production of a literary opus ...
        Merging, the mixing of a traditional literary scheme with the scientific scheme of social
        reporting ...
        Adding1, the integration of a large body of socio-economic data into the field of literature ...
        Adding2, the usage of a space-time language, closely related to the protocol-language of the
        Vienna Circle-program ...

Thus, the first set of examples should establish a sufficient understanding that the basic requirements for
recombinations in general and the specifications for scientific recombinations in particular offer an
interesting way of analyzing creative breakthroughs by reproducing the essential creativity operators and
the recursive manner of transformations leading from an initial configuration to the final innovative
The second set of examples is aimed to demonstrate that the missing of one of the essential requirements,
especially the “rich repertoire-condition”, can be considered as sufficient to explain the highly
recombinative, but denovative character of a research program in the social sciences which has become of
utmost importance for Austria in the years following 1945. More concretely, the “universalist program” of
Othmar Spann and his followers who dominated several Austrian universities during the 1950´s, has
produced, especially through its mentor Othmar Spann, a large number of cognitive recombinations,

     One of the easiest ways to demonstrate the inadequcies of ex-post syntheses, comes from a thought experiment which can be performed
via self-inspection: Think of yourself at the age of twenty and define core areas for achivements, goals, ambitions, life-contexts, locations
which could be termed, then, as essential, desirable, etc. to be reached within the next ten years. Then repeat this procedure at the age of
thirty, forty, etc. Normally, one will find via simple introspections that expectations for the future and syntheses from the past are very fragile
in character and exhibit all kinds of breaks, discontinuities or family-unresemblances ....

integrations and, thus, achievements, although the first creativity requirement, namely the demand for a
rich recombination repertoire, was largely absent. The following quotation will make this deficiency
abundantly clear:

       Auf dem Gebiet der Logik und Mathematik fühlte sich Spann besonders Aristoteles und der
       Schlußfigur des modus Barbara verpflichtet; hingegen wurde die „neue Logik“ durch Frege, Russell
       oder Carnap nicht zur Kenntnis genommen.
       Im Philosophischen führte eine Ahnengalerie von Platon via Meister Eckhart bis hin zum deutschen
       Idealismus; aber für die Anfänge der analytischen Wissenschaftsphilosophie wie auch für die
       österreichische sprachphilosophische Tradition war im großen und ganzen kein Platz vorgesehen.
       Auf naturwissenschaftlich-physikalischem Terrain war es das Newtonsche System, dem Spanns
       universelle Aufmerksamkeit galt; im Gegensatz dazu firmierten Maxwell, Mach, Boltzmann oder
       Einstein für Spann weitestgehend als Nichtentitäten.
       Ambivalent erwies sich die Haltung Spanns zur neueren Biologie, und hier vor allem zu Darwin, der
       nur zu geringen Teilen akzeptiert wurde.
       Und auf den sozialwissenschaftlichen Feldern werden Adam Müller, Lorenz Stein oder sein Tübinger
       Lehrer Albert Schäffle hochgehalten, wogegen die Nationalökonomie mathematisch-formalerer
       Observanz, die frühe empirisch-kritische Sozialforschung jener Jahre, die Psychoanalyse und so vieles
       andere keinerlei Beachtung erfuhren.

Through this enumeration of relevant intellectual fields of reception and intellectual spaces generated via
a reliance on “modus Barbara”, Platonic epistemology, Meister Eckhart, Isaac Newton or dialectical
movements in Hegelian spirits, it becomes comparatively easy to see that recombinations within such an
intellectual space, without the availability of contemporary schemes in the natural sciences, in the
philosophy of science or in the social sciences, must lead, by necessity, to new schemes of the denovation
type since they must be characterized, in all probability, by clearly recognizable comparative
disadvantages.97 From the viewpoint of the recombination effort or from the amount of reorganizations
and adaptations, it might turn out to be extremely difficult, to transform a framework consisting of
„Ausgliederungsordnungen“, of „Stufenordnungen“ or of „teilinhaltlichen Begriffe“ so that it fits to the
demand for a virile, strong state, to the abandonment of democratic forms of legitimation and to a
hierarchical, unbalanced configuration of controls, reserved mainly for the lower regions of the societal
pyramid. Nevertheless, no reconfiguration of the universalist intellectual space will produce, even in an
perennial performance of recombinations and permutations, something scientifically new, exciting, and
innovative since the requirement of a sufficiently rich repertoire has remained unfulfilled in the case of the
universalist research program ...

    For more details, see MÜLLER (1987:243pp.)
    The only intellectually interesting results which can be achieved via recombinations in old cognitive spaces may consist in “great
visions” of development, change or stability, in newly produced paradoxes and the like ...

                                     6. AT THE END -

                                   A NEW BEGINNING

Section II has been concentrated on a central domain for National Innovation Systems, namely, in a
formal mode of description, on the genotype-levels of National Innovation Systems or, in a material
mode, on areas like knowledge, knowledge production, knowledge transfers, information gathering,
scientific production and the like. It has become the main task of Section II to provide a slightly
unconventional framework in which notions of “code-systems” and “embedded code-systems” (ECS) at
various levels have occupied the center stage and in which “knowledge”, aside from “information” and
“scientific production,” has turnd out to be but one of at least three different genotype-level domains for
growth and development processes within the scientific system. Thus, the three key notions of
“knowledge”, “information” and “scientific production” have been unified under the new heading of
“embedded code-systems”. Moreover, the new evolutionary or, more concretely, epigenetic framework
has allowed to focus more clearly on those areas in “knowledge production”, in “information processing”
or in “scientific output” which like -

      other products of the economic system, can be manufactured, stored, sold or exchanged ...
      other services of the economic system, cannot be be manufactured, stored or sold ...

Similarly, the production and manufacturing processes within scientific research laboratories, within
universities, within firm-specific R&D departments or within the confines of small rooms at a research
institute have ben summarized under the heading of strings, grammars, programs and recombinations
which at least in the traditional understanding can be classified, to modify a phrase from Stuart Kauffman,
as “programs for free”. (KAUFFMAN 1995) In turn, the comparatively wide distributions of “programs
for free” facilitate in a mode of self-organization the emergence1,2,3 of highly ordered secondary, tertiary,
etc. structures. Likewise, the notion of “(re)productive” as well as of “non (re)productive programs” have
served as an additional evolutionary fitting entity to account for the traditional domains of knowledge-,
information- or scientific output-transfers ...
The final chapter, so far, has used a lage number of meanwhile well-known concepts which are situated,
by now, far from the point of intersubjective non-accessibility and non-understanding. The introduction to
Section II, which has been repeated here at the end in a slightly altered version, has moved, hopefully, to
the core-domains of intersubjective accessibilities and understandings ....

          SECTION III:





Section III will bring another essential “building block” for National Innovation Systems in proper
perspective, namely the so-called “embeddedness relations” which occupy a central role for the
production, the reproduction and, especially important, for dual level-combinations and recombinations in
the case of National Innovation Systems. “Embeddedness”, as will be recalled from Section II, signifies
the relations and development patterns between a code-system and its wider environment or, alternatively,
its utilization contexts.
In the case of National Innovation Systems, “embeddedness-relations” can be identified in numerous
ways. Below, seven paradigmatic examples are listed where a significant interplay between genotype
level components and phenotype level elements takes place:

       First, at the level of socio-technical systems, machines and man-machine interactions, under
       the heavy assistance of socio-technical programs, are constantly adapted, redesigned,
       recombined into novel forms ....
       Second, again at the level of socio-technical systems, organization-schemes, utilizing social
       rule-systems, are in a continuous process of re-shuffling and adaptation ...
       Third, in the scientific arena, the “structures of scientific revolutions” generate, following
       Thomas S. Kuhn, a cyclical pattern of new scientific programs, of diffusion processes within
       the scientific communities, of a growing number of anomalies at the genotype-level of
       programs as well as of the emergence of new scientific groups at the phenotype level with still
       newer programs, free at least of some of the accumulated anomalies ...98
       Fourth, in the infant science and technology segment of “applied molecular evolution”
       (KAUFFMAN 1995:132ff.) a genetic code  language code-transformation establishes the
       basis for a new industry on useful molecules for medical purposes.

            Applied molecular evolution is based on the fundamental recognition that the near and outer
            reaches of protein space, DNA space, or RNA space are fully likely to harbor novel molecules
            of enormous practical benefit. The central idea is simple: generate billions, trillions, billions of
            trillions of random DNA or RNA or protein sequences, and learn to search this enormous,
            staggering diversity of moldecular forms for the useful molecules we need. For better or worse,
            we will find fabulous new medicines. For better or worse, we will find frightful new toxins. For
            better or worse - so it has always been. (KAUFFMAN 1995:134)

       Fifth, within the “human actants” in National Innovation Systems, a permanent internal re-
       shaping of their cognitive, code-based organization takes place which, due to its multiplicity
       of different code-systems, must be qualified as hyper-complex ...

    Alternatively, one could describe the Kuhnian vision of scientific development as an evolutionary pattern, starting with a program in a
particular domain of the intellectual space which, after a prolonged period of exploration and decreasing returns in cognitive fitness, leads to
the proliferation of a new program, sometimes situated in a radically different part of the intellectual space - and so on ...

        Sixth, the reproduction of human organisms undergoes, seen in the very long run of centuries,
        a genetic code-based drift towards greater height, greater intelligence ...99
        Seventh, in the wider environments of National Innovation Systems, in their ecological and
        biospherical     surroundings,     the     long-standing   code-based       “Games      of
        Life”(ATTENBOROUGH 1995, SIGMUND 1995, WEIZSÄCKER 1995) operate in a
        perennial rhythm of a dual level “Werde und vergehe”...

Thus, the present section will be devoted to identify, sticking to the overall meta-theoretical framework,
the appropriate embeddedness-domains and to specify different and, one must immediately add,
unexpected types of “embeddedness-relations” in biological, ecological, social, scientific or technical
systems. The main thrust of the present section though lies in alternative foundations for a seemingly
simple and innocent question, namely on the actors and the agencies in evolutionary, code-based systems.
It will become one of the central tasks, inter alia, to depart from apparently trivial and self-evident
characterizations of “agency” within the contemporary social sciences like the following one -

        We have traced the wanderings of the idea of agency through the labyrinth of social and sociological
        thought. At the entrance it was entirely superhuman and extrasocial. At the exit it appears as fully
        human and fully social, in the two forms of individual actors and collective agents. (SZTOMPKA

Thus, the traditional accounts of agency is bound to two domains, namely to the level of “individual
actors” and of “collective agents”. Throughout Section III it will be seriously questioned whether this type
of focus for actions is to be qualified a heuristically fruitful and rewarding basis for studying processes of
innovation and diffusion or, more generally, of societal evolution. As a consequence, a counter-intuitive
framework will be established which will come to a strikingly different set of agencies as well as of
Finally, at the end of Section III, some additional remarks and comments will be devoted to evolutionary
systems with a multiplicity of code-systems and embeddedness-relations. This, in turn, will lead to the
notion of hyper-complex systems which will become of central importance for the constitution of
National Innovation System, developed in Section IV...

     On this point, see e.g. GRIBBIN 1993:217pp.

                               1. BARRIERS AND LIMITATIONS

                               IN THE ANALYSIS OF AGENCIES

Before entering into a detailed discussion of the “embeddedness-relations” for the rich variety of code-
systems introduced in Section II, a disgression will be made in which the conventional obstacles for a
unified, code-based theory of the embedded trias of knowledge-information-scientific production will be
removed. From the current philosophy of the social science-literature, especially in the German speaking
countries, a “horror machinarum” as well as a “horror bestiarum” have prevented, so far, an integrative
perspective which would guarantee a homogeneous framework for evolutionary, code-based processes in
humans, machines and animals. In order to achieve this highly ambitious integration-goal, a detailed
discussion on the removability of these barriers as well as on the substitutability of the conventional
accounts with more general alternatives becomes strictly unavoidable. More concretely, the main foci of
the subsequent analysis will lie in two fields.

       First, under the headings of “What intentions (mental representations, conscious experiences,
       habitus, values ...) are not” an inversion of the status of the “internal vocabulary”, be it
       intentionalist, value-based or, following Pierre Bourdieu (1982/85/87/91), habitus-
       generative100 in nature, will be undertaken.
       Furthermore, a second domain of discussion will be directed to the detection and
       identification of trivial explanation schemes with respect to actions and behavior, human and
       otherwise. Thus, in the course of the present chapter, a phase-transition from trivial schemes
       with respect to explanatory frameworks to their non-trivial-counterparts will be proposed.

In a metaphorical manner, the subsequent chapters serve as preparation for a journey far away from the
playground of the centralized “Cartesian Theatre” (DENNETT 1991, DAMASIO 1994) -

       The Cartesian Theater is a metaphorical picture of how conscious experience must sit in the brain
       ...As we shall see, it keeps reasserting itself, in new guises, and for a variety of ostensibly compelling
       reasons. To begin with, there is our personal, introspective appreciation of the „unity of
       consciousness‟, which impresses on us the distinction between „in here‟ and „out there‟ ... Doesn´t it
       follow as a matter of geometric necessity that our conscious minds are located in the termination of
       all the inbound processes, just before the initiation of all the outbound processes that implement our
       actions? Advancing from one periphery along the input channels from the eye, for instance, we ascend
       the optic nerve, and up through various areas of a visual cortex, and then ...? Advancing from the
    The strange notion of “habitus-generative” refers to a very special critique of Bourdieu´s approach, since it leaves the conceptual
framework unaffected and intact and is simply directed toward a far too strong “generative” interpretation of the habitus concept.

      other periphery by swimming upstream from the muscles and the motor neurons that control them, we
      arrive at the supplementary motor area in the cortex and then ...? These two journeys advance toward
      each other up two slopes, the afferent (input) and the efferent (output). However difficult it might be
      to determine in practice the precise location of the Continental Divide in the brain, must there not be,
      by sheer geometric extrapolation, a highest point, a turning point, a point such that all tamperings on
      one side of it are pre-experiential, and all tamperings on the other are post-experiential? (DENNETT

Consequently, an appropriate continuation of the chapter lies in the shift to the first domain of discourse,
namely to the problem of what alternative vocabularies outside the intuitively plausible settings of the
Cartesian Theatre might look like. As a “Caveat lector”, one should add that the next thirty pages are of
immediate relevance to current discussions in the philosophy of the social sciences only. Thus, for readers
mainly interested in the constitution of National Innovation Systems, an immediate move to the
beginnings of the next main chapter on agencies (chapter 2) is strongly suggested ...

                         1.1. TRADITIONAL FRAMEWORKS

                                            FOR AGENCY:

                                   FOLK-SOCIAL SCIENCE

The traditional terrain to be abandoned within the discussion of the first chapter will carry the label of
“folk-social science” with its departments like “folk-sociology”, “folk-psychology”, “folk-economics”.
“Folk social science”, this includes, inter a lot of alia, the reliance on an “internal vocabulary”, using a -

      variety of commonsense psychological terms including 'believe', 'remember', 'feel', 'think', 'desire',
      'prefer' (including „decide‟, K.H.M.), 'imagine', 'fear' and many others. (STICH 1983:1)

Moreover, “folk social science” is a particular view on the dual constitution of the social universes,
separating, on the one hand, between a reclusive area of privacy, accessible only to an epistemologically
privileged subject and, on the other hand, a public domain of mutually accessibilities, but of secondary
relevance only. Thus, the traditional domain of folk social science has become one in which,

      dann ich der Mittelpunkt des Universums (bin), meine Wirklichkeit meine Träume und meine
      Alpträume (sind), meine Sprache ein Monolog (ist), meine Logik eine Monologik. (FOERSTER

With the help of this basic separation, a sophisticated methodological as well as theoretical intellectual
space has been erected which has become the conventional wisdom especially with respect to the study
and explainability of human actions or, alternatively of the mundus socialis. For an early manifestation of
this profound division, one may quote Wilhelm Dilthey, one of the leading proponents of the modern
“bifurcation” between the social sciences and their natural counter-parts.

      Das eigentliche Reich der Geschichte ist zwar auch ein äußeres; doch die Töne, welche das
      Musikstück bilden, die Leinwand, auf der gemalt ist, der Gerichtssaal, in dem Recht gesprochen wird,
      das Gefängnis, in dem Strafe abgesessen wird, haben nur ihr Material an der Natur; jede
      geisteswissenschaftliche Operation dagegen, die mit solchen äußeren Tatbeständen vorgenommen
      wird, hat es allein mit dem Sinne und der Bedeutung zu tun, die sie durch das Wirken des Geistes
      erhalten haben; sie dient dem Verstehen, das diese Bedeutung, diesen Sinn in ihnen erfaßt ... Dies
      Verstehen bezeichnet nicht nur ein eigentümliches methodisches Verhalten, das wir solchen
      Gegenständen gegenüber einnehmen; es handelt sich nicht nur zwischen Geistes- und
      Naturwissenschaften um einen Unterschied in der Stellung des Subjekts zum Objekt, um eine
      Verhaltungsweise, eine Methode, sondern das Verfahren des Verstehens ist sachlich darin begründet,
      daß das Äußere, das ihren Gegenstand ausmacht, sich von dem Gegenstand der Naturwissenschaften
      durchaus unterscheidet. Der Geist hat sich in ihnen objektiviert, Zwecke haben sich in ihnen gebildet,
      Werte sich in ihnen verwirklicht, und eben dies Geistige, das in sie hinein gebildet ist, erfaßt das
      Verstehen. Ein Lebensverhältnis besteht zwischen mir und ihnen. Ihre Zweckmäßigkeit ist in meiner
      Zwecksetzung gegründet, ihre Schönheit und Güte in meiner Wertgebung, ihre Verstandesmäßigkeit
      in meinem Intellekt. Realitäten gehen ferner nicht nur in meinem Erleben und Verstehen auf: sie
      bilden den Zusammenhang der Vorstellungswelt, in dem das Außengegebene mit meinem
      Lebensverlauf verknüpft ist: in dieser Vorstellungswelt lebe ich, und ihre objektive Geltung ist mir
      durch den beständigen Austausch mit dem Erleben und dem Verstehen anderer selbst garantiert;
      endlich die Begriffe, die allgemeinen Urteile, die generellen Theorien sind nicht Hypothesen über
      etwas, auf das wir äußere Eindrücke beziehen, sondern Abkömmlinge von Erleben und Verstehen.
      Und wie in diesem die Totalität unseres Lebens immer gegenwärtig ist, so klingt die Fülle des Lebens
      auch in den abstraktesten Sätzen dieser Wissenschaft nach (DILTHEY 1981:140f.) -

Here, one can clearly distinguish an internal as well as an external domain, the former belonging to the
social or cultural sciences, the latter to the natural science-world. Moreover, on the basis of this division,
an extensive repertoire of methodological as well as theoretical devices has been identified, operations
like “verstehen”, “values”, “intentions” which, as their “differentia specifica” belong to the inner world of

generating and producing human actions and interactions. One of the early culmination points along this
dual trajectory in the disciplinary space has become, then, Max Weber who developed a large number of
heuristics particularly suited for the social domains. Four of these requirements for social research designs
have become highly successful obstacles to a more comprehensive view of the embeddedness relations.

      The first directive consists in the demarcation of the sociological domain proper, namely the
      limitation to actions and social actions.

         Soziologie ... soll heißen: eine Wissenschaft, welche soziales Handeln deutend verstehen und
         dadurch in seinem Ablauf und Wirkungen ursächlich erklären will. 'Handeln' soll dabei ein
         menschliches Verhalten ... heißen, wenn und insofern als der oder die Handelnden mit ihm
         einen subjektiven Sinn verbinden. 'Soziales' Handeln aber soll ein solches Handeln heißen,
         welches seinem von dem oder den Handelnden gemeinten Sinn nach auf das Verhalten anderer
         bezogen wird und daran in seinem Ablauf orientiert ist. (WEBER 1982:542)

      Second, the realm of actions exhibits differing degrees of rationality, from combinations of
      low rationality and high routine to the opposite version of high rationality and low

         Wie jedes Handeln kann auch das soziale Handeln bestimmt sein 1. zweckrational: durch
         Erwartungen des Verhaltens von Gegenständen der Außenwelt und von anderen Menschen und
         unter Benutzung dieser Erwartungen als Bedingungen oder als Mittel für rational, als Erfolg
         erstrebte und abgewogene eigene Zwecke, - 2. wertrational: durch bewußten Glauben an den -
         ethischen, ästhetischen, religiösen oder wie immer sonst zu deutenden - unbedingten Eigenwert
         eines bestimmten Sichverhaltens rein als solchen und unabhängig vom Erfolg, - 3. affektuell,
         insbesondere emotional: durch aktuelle Affekte und Gefühlslagen, - 4. traditional: durch
         eingelebte Gewohnheit. (IBID:565)

      Third, the rationality pyramid has, according to Weber, definite boundaries both in the upper
      and in the lower dimensions. For the upper limits, Weber points to the following reduction-

         Begriffe wie 'Staat', 'Genossenschaft', 'Feudalismus' und ähnliche bezeichnen für die Soziologie,
         allgemein gesagt, Kategorien für bestimmte Arten, menschlichen Zusammenhandelns, und es
         ist also ihre Aufgabe, sie auf 'verständliches' Handeln, und das heißt ausnahmslos: auf Handeln
         der beteiligten Einzelmenschen, zu reduzieren. (IBID:439)

     Fourth, the lower-bound limit is marked, following Weber, by a non-integration of the
     cognitive and neural science-domains.

         Die Aufgabe anderer Betrachtungsweisen kann es sehr wohl mit sich bringen, das Einzelindivi-
         duum vielleicht als einen Komplex psychsicher, chemischer oder anderer 'Prozesse' irgendwel-
         cher Art zu behandeln. Für die Soziologie aber kommt alles die Schwelle eines sinnhaft deut-
         baren Sichverhaltens zu 'Objekten' (inneren oder äußeren) Unterschreitende nur ebenso in
         Betracht, wie die Vorgänge der 'sinnfremden' Natur: als Bedingung oder subjektiver Bezogen-
         heitsgegenstand des ersteren. (IBID)

For a more comprehensive perspective on embeddedness-relations, historical limitations, drawn a century
ago, should have been superseded and canceled by contemporary lines of reasoning. Surprisingly
however, also the present demarcations in the methodology and philosophy of the social sciences, though
in a profoundly changed and altered terminology, run along the same Weberian or Parsonian lines
(PARSONS 1961/1964/1994). Take, for example Niklas Luhmann and his conceptual apparatus of codes,
communication, autopoietic systems and the like, then one finds the following assertions.

     Angesichts ihrer Umweltlage kann kein(!) Zweifel bestehen, daß psychische Systeme autopoietische
     Systeme sind - und zwar nicht auf der Basis von Leben(!), sondern auf der Basis von Bewußtsein(!).
     Sie verwenden(!) Bewußtsein(!) nur im Kontext ihrer eigenen Operationen, während alle
     Umweltkontakte (einschließlich der Kontakte mit dem eigenen Körper(!!!)) durch das Nervensystem
     vermittelt(!!) werden, also andere(!) Realitätsebenen(!!) benutzen müssen(!). Schon das Nervensystem
     ist ein geschlossenes System, und schon deshalb(!!) muß auch das mit Bewußtsein operierende
     psychische System ausschließlich auf selbstkonstituierten Elementen aufbauen. Wie immer man die
     Elementareinheiten des Bewußtseins bezeichnen will ..., kann nur(!) das Arrangement dieser
     Elemente neue Elemente produzieren. Vorstellungen(!!) sind nötig(!), um zu Vorstellungen(!!) zu
     kommen (LUHMANN 1983:355p.)

Representations, although largely undefined, are, by a sort of Luhmannian necessity, “necessary” to
generate other(!) representations, a highly ambiguous term like “consciousness” remains both undefined
and, like in the good old Weberian days, separated from “life” in general and biology in particular. On the
other hand, a well-known opponent of Luhmann, Jürgen Habermas, operates with concepts from a very
traditional philosophical tradition since terms like “will”, “autonomous will”, “the self”, “identity” are
combined and reconfigured in a manner as if the cognitive science-domains would still reside in their
knowledge stocks of the 17th or 18th century.

     Zu den allgemeinen und unvermeidlichen Präsuppositionen verständigungsorientierten Handelns
     gehört es, daß der Sprecher als Aktor beansprucht, zugleich(!) als autonomer Wille(!!) und

     individuelles Wesen anerkannt zu werden. Und zwar kommt das Selbst, das sich seiner im
     Anerkanntwerden(!) dieser Identität durch andere vergewissern kann, in der Bedeutung des
     performativ gebrauchten Personalpronomens der ersten Person zur Sprache. Wie weit diese
     Bedeutung unter den beiden Aspekten der Selbstbestimmung und Selbstverwirklichung im konkreten
     Fall jeweils artikuliert zum Vorschein kommt oder implizit bleibt, ja neutralisiert wird, hängt freilich
     von der Handlungssituation und dem weiteren Kontext ab. Die allgemeinen pragmatischen
     Voraussetzungen kommunikativen Handelns bilden semantische Ressourcen, aus denen historische
     Gesellschaften    je    auf   ihre     Weise   Geist-    und    Seelenvorstellungen,     Personenkonzepte,
     Handlungsbegriffe, moralisches Bewußtsein usw. schöpfen und artikulieren (HABERMAS 1988:232)

To conclude this short presentation of a contemporary cognitivist local minimum within the international
social sciences, one may add one of the leading exponents of social science-methodology in the English
speaking areas, namely Anthony Giddens, where one can identify, despite the coming of post-modernity,
the traditional Weberian boundaries, although in genuinely post-modernist terminological cloths.

     In the post-traditional order of modernity, and against the backdrop of new forms of mediated
     experience, self-identity(!) becomes a reflexively organised endeavor. The reflexive project of the
     self(!), which consists in the sustaining of coherent, yet continuously revised, biographical
     narratives(!), takes place in the context of multiple choice as filtered(!) through abstract systems(!!!)
     ... Because of the 'openness' of social life today, the pluralisation of contexts of action and the
     diversity of 'authorities', lifestyle choice is increasingly important in the constitution of self-identity(!)
     and daily activity. Reflexively organised life-planning, which normally(!!!) presumes consideration of
     risks(!) as filtered(!) through contact(!) with expert knowledge(!), becomes a central feature of the
     structuring(!) of self-identity(!) (GIDDENS 1991:5)

With an apparently world-wide maladie sociale, the next two steps consist in two radical departures, the
first one from the conventional understanding of internal and external dimensions or mechanisms, the
second one from the triviality of conceptual schemes, utilized for the study of actions, human and
otherwise. At the end, a highly mysterious sentence like the following -

     Die Sätze der Hirnfunktionen - oder noch allgemeiner - die Sätze der Biologie müssen so geschrieben sein,
     daß das Schreiben dieser Sätze von ihnen abgeleitet werden kann, das heißt: Sie müssen sich selber
     schreiben. (FOERSTER 1985:67) -

should become entirely intelligible. And how this can be done, will be shown within the next chapter,
where, by the recombinative operation of inversion, -

                                          Und was drinnen ist, ist draußen -

the conventional relations between -

          {Subject  Meaning  Internal}:{Observable Actions  Understanding  External}

will be transformed into an unusual, though evolutionary highly fruitful configuration of -

  {Human Systems  Recursive Operations  Evolution}:{Observer  Attribution  Form of Life}

with a universal mode of operation, namely -

                                    Operate recursively 
                                                               

where -

     the ... theory, in itself, is an Eigen-Function and is to be understood through itself. (FOERSTER

                           1.2. THE EXTERNAL ASPECTS

                                       OF INTERNALITY

Thus, the most prominent obstacle for an integrated perspective on the embeddedness-relations of code-
systems has been erected with respect to the accessibility of human code-operations especially in the area
of natural languages as well as with respect to the relations between observable performances and the
internal ways of code-generation. Here, a rich protective belt has been developed which has acquired the
status of a persona obscura where the important and essential code-processing activities, the primary
components like decisions, desires, value-adherences, habitus-distinctions ... are located within persons
and where the secondary elements like actions, the observable practices and routines, have a derivative,
second-best status only. The important moves are within a sphere of private code-operations which
manifest themselves only in small traces and remnants - like “das Flügelrauschen der Gottheit”
(NEURATH 1981:359).

The alternative which will be proposed during the first radical departure in chapter 1.2.101 can best be
summarized with the help of a guiding metaphor, first proposed by Ludwig Wittgenstein, the tale of
beagles, boxes and the conditions of the possibility of communication processes:

          Angenommen, es hätte Jeder eine Schachtel, darin wäre etwas, was wir 'Käfer' nennen. Niemand kann
          je in die Schachtel des Andern schaun; und Jeder sagt, er wisse nur vom Anblick seines Käfers, was
          ein Käfer ist, - Da könnte es ja sein, daß Jeder ein anderes Ding in seiner Schachtel hätte. Ja, man
          könnte sich vorstellen, daß sich ein solches Ding fortwährend veränderte. - Aber wenn nun das Wort
          'Käfer' dieser Leute doch einen Gebrauch hätte? - So wäre er nicht der der Bezeichnung eines Dings.
          Das Ding in der Schachtel gehört überhaupt nicht zum Sprachspiel; auch nicht einmal als ein Etwas:
          denn die Schachtel könnte auch leer sein. - Nein, durch dieses Ding in der Schachtel kann 'gekürzt'
          werden; es hebt sich weg, was immer es ist.(WITTGENSTEIN 1971:PU 293)

With the help of the Wittgensteinian metaphor, an alternative foundation of embedded code-systems
comes into view which assumes an inverse relation on “meaning”, “accessibility”, “intentional stance”,
“values”, and the like. Here, an inter-subjectively distributed “consensus” as well as highly specific
clusters of action-sequences or, in Wittgenstein´s words, a “form of life” become the important
“stabilization” component which guarantees the success of exchange operations, communicative and
otherwise. The conventional construction, relying on the private view on private objects or sensations,
will be confronted, upon closer inspection, with the strictly un-solvable problem of finding proper
semantic identifications and equivalencies in a sufficiently small amount of time ...102
Thus, what will be presented next, lies in the detection of a subtle, but widely made “categorial mistake”
which, in the end, will lead to a complete externalization of the allegedly “internal vocabularies”.
First step: The starting point seems totally unrelated to the problem at hand, for the exploration begins
with a special problem of dispositional terms, namely with the question of the relations between
dispositions and their constitutive actions. Over the last decades, almost an unanimous judgment has been
passed that the “disposition-relation” should be considered as “quasi-analytical”. Intelligence, for
example, is not to be viewed as an internal force for generating intelligent behavior, but an external
attribution with the following characteristics. If intelligence is attributed to a person, this person should

    As a general orientation, the planned departure may be seen, , as a typical Austrian expedition which takes three typical strong statements
of the “linguistic turn” very seriously, namely Otto Neurath´s vision -

                                         In der Wissenschaft gibt es keine 'Tiefen'; überall ist Oberfläche -

Ludwig Wittgenstein´s specific “fact finding mission” -

                                           Die Welt ist die Gesamtheit der Tatsachen, nicht der Dinge -

and, finally, Heinz von Foerster´s descriptive directive -

                                           Die Logik der Welt ist die Logik der Beschreibung der Welt -

      For more details especially on this point, see the last chapter in Section IV.

produce comparatively quick and successful problem-solutions in the area of standardized and non-
standardized cognitive problems. Thus, the attribution of intelligence has to be linked to a complex web
of behavior patterns and not to an internal mechanism called “intelligence”. Only a typical scholastic
response, like in Moliere´s quia est in eo103, would establish a causal efficacy for the “disposition

          Unser Fehler ist, dort nach einer Erklärung zu suchen, wo wir die Tatsachen als 'Urphänomene' sehen
          sollten. D.h., wo wir sagen sollten: dieses Sprachspiel wird gespielt. (WITTGENSTEIN 1971:PU

Thus, the first step, stressing the “quasi-analytical” nature and the causal inefficacy of the “disposition
relation” should pose no serious problems.
Second step: The main reason for the “internally closed nature” of internal descriptions comes, aside from
a philosophical tradition of the longue durée, from a perfectly legitimate distinction, introduced in its
weaker version by Hilary Putnam. Here, a separation is undertaken with respect to the material, physical
or observable inner states of an actor, human or otherwise, and the logical states.

          The proposition
          (1) I am in state A if, and only if, flip-flop 36 is on
          is clearly a 'synthetic' proposition for the machine ... And just as philosophers have argued from the
          synthetic nature of the proposition
          (2) I am in pain if, and only if, my C-fibers are stimulated
          to the conclusion that the properties (or 'states` or 'events') being in pain, and having C-fibers
          stimulated, cannot possibly be the same ...; so one should be able to conclude from the fact that (1) is
          synthetic that the two properties (or 'states' or 'events') - being in state A and having flip-flop 36 on -
          cannot possibly be the same ... The logical aspects of the Mind-Body Problem are aspects of a
          problem that must arise for any computing system satisfying the conditions that (1) it uses language
          and constructs theories; (2) it does not initially 'know' its own physical make-up, except superficially;
          (3) it is equipped with sense organs, and able to perform experiments; (4) it comes to know its own
          make-up through empirical investigation and theory construction. (PUTNAM 1984:363p.,388)

To achieve a full-fledged separation, the following two propositions will be introduced, the first one on
the non-identity relation between physical and logical states -
      In Moliere, the scholastic response and the causal efficacy of the “disposition relation” takes the following form:

                  Mihi a docto doctore
                  Fragatur causam et rationem quare
                  Opium facit dormire?
                  Quia est in eo
                  Virtus dormitiva,
                  Cujus est natura
                  Sensus betäubare ( MOLIERE (no year:1024)

         If two systems of different physical set-up and constitution exhibit an isomorphic relation at the level
         of machine behavior, then the logical states cannot be identical with the physical states in which they
         are incorporated -,

the second one, under the isomorphism-assumption for physical systems, on the general independence
between the two levels:

         Logical states are not identical with the physical states in which the logical states are incorporated.

Consequently, the mental corollary must be read as follows:

         Mental states are not identical with physical states in which they are incorporated. 104

What remains to be done, is to find a satisficing solution for the interpretation of the dual nature of
logical and mental states. This problem leads, in turn, to the third step.
Third step: Having established the independence of the mental setting aside from its neural substrate, the
next operation lies in the proliferation of conceptual schemes, explaining mental states. Thus, concepts
like “values” and “value change”, “habitus”, “instincts” and the like have been proposed in the social
science literature to act as explanatory schemes sui generis. Couched in terms of an internal/external
distinction, a rich internal “vocabulary” as well as an internal explanatory repertoire has been developed
to account for internal revulsions and, ultimately, for changes in behavior.
What the next steps, from step four to step seven, will demonstrate lies on two different levels:

         For the descriptive internal vocabularies, it will be shown that it is far more fruitful and
         problem-deflating to regard them as external attribution strategies.
         For the explanatory schemes available for “mental states”, the discussion will be aimed at the
         demonstration of the causal inefficacy and, thus, of the explanatory irrelevance of the
         proposed frameworks.

Quod erit demonstrandum.
Fourth step: To start the two-fold inversion, a short reminder on an intensive debate in the philosophy of
science field in the 1960ies will be helpful. In the controversies with respect to the appropriate
explanation schemes in the social sciences, particularly in historiography, a large number of frameworks
has been proposed to the standard deductive-nomological scheme of the Hempel-Oppenheim type.
“Teleological explanations”, “rational explanations” and the like have been suggested as genuine
alternatives to the four explanation requirements by Hempel and Oppenheim. Upon closer inspection

      On this point, see especially NELSON (1982).

however, these alternatives turned out to be sufficiently close variants of the original covering law-
configuration -

      All scientific explanation involves, explicitly or by implication, a subsumption of its subject matter
      under general regularities; ... it seeks to provide a systematic understanding of empirical phenomena
      by showing that they fit into a nomic nexus. (HEMPEL 1965:488).

Transformed into the realm of internal explanatory schemes, the contribution of these schemes for the
question Why did agent A do X must stay at a global minimum, simply because these schemes like “value
changes”, “intentions”, etc. cannot possibly explain why A did X. Take, once again, Hempel´s verdict on
the insufficient nature of many rational explanations -

      Any adequate answer to the question why a given event occurred will have to provide information
      which, if accepted as true, would afford good grounds for believing that the event did occur. Now, the
      information that agent A was in a situation of kind C and that in such a situation the rational thing to
      do is x, affords grounds for believing that it would have been rational for A to do x, but no grounds
      for believing that A did in fact x. To justify this latter belief, we clearly need a further explanatory
      assumption, namely that - at least at the time in question, A was a rational agent and thus was
      disposed to do whatever was rational under the circumstances. (IBID:470p.)

But, so the spontaneous counter-argument, the additional explanatory assumption can be identified,
especially with respect to schemes like “value-changes” very easily since these concepts are very directly
linked to empirical social research, to item-batteries and scales of “materialistic” and “postmaterialistic”
value-orientations and the like which form an important basis for the problem Why did agent A do X ...
Thus, the following steps will have to show that despite an apparently firm grounding in empirical
research, the essential additional explanatory assumptions -

      At least at the time in question, A was a rational agent and thus was disposed to do whatever was
      rational under the circumstances ...
      At least at the time in question, A was in a specific intentional configuration and thus was disposed to
      do whatever was intentionally appropriate under the circumstances ...
      At least at the time in question, A was adhering to a specific value-system and thus was disposed to
      do whatever was required by this particular system of values under the circumstances ... -

run into the serious problem of being (almost) totally circular ...
Fifth step: For the subsequent demonstration of an inherent circularity, three examples from the “internal
explanatory vocabulary”, one couched in terms of a misguided interpretation of habitus-operations, one

in terms of “value-orientations”, one in the “intentional stance”, will be used. Thus, the starting point lies
in quotations like the following ones -

      Der Aufbau der verschiedenen Präferenz-Räume (folgt) in bezug auf Nahrung, Kleidung und Kosmetik
      derselben(!) Grundstruktur - der des von Umfang und Struktur des Kapitals determinierten(!) Sozialraums.
      Zur umfassenden Konstruktion des Raumes der Lebensstile, innerhalb derer sich kultureller Konsum
      definiert, wäre für jede Klasse und Klassenfraktion, d.h. für jede Kapitalkonfiguration, die generative
      Formel(!!!) des Habitus(!) zu ermitteln, die die für eine jeweilige Klasse (relativ homogener!))
      Lebensbedingungen charakteristischen(!) Zwänge und Freiheitsräume(!) in einen spezifischen Lebensstil(!)
      umsetzt(!). Es wäre daran anschließend im einzelnen auszumachen, auf welche Weise sich die
      Dispositionen(!) des Habitus(!) im Rahmen jedes(!!!) größeren Bereichs der Praxis derart(!) spezifizieren,
      daß sie bestimmte(!), von jedem(!!!!!) einzelnen Feld - Sport wie Musik, Nahrung wie Inneneinrichtung,
      Politik wie Sprache, etc.(!!!!!) - angebotene stilistische Möglichkeiten verwirklichen(!). (BOURDIEU 1982:
      Ein Prozeß des intergenerationellen Wertwandels verändert(!!!) langsam Politik und Normen(!) in den
      entwickelten    Industriegesellschaften.   Eine     Gewichtsverlagerung     von    materialistischen   zu
      postmaterialistischen Werten hat neue politische Anliegen ins Zentrum gerückt(!!!) und oftmals neue
      politische Bewegungen in Gang gesetzt(!!!) ... (INGLEHART 1989:90)
      (Dem) Verstehen kommt auch Erklärungswert(!) hinsichtlich der Elemente und Phasen der Handlung
      zu. Die Intention 'Geld von der Bank abheben' erklärt(!) die dazu erforderlichen Einzelschritte und
      Körperbewegungen. (BALOG 1988:65)

Here, one is confronted with an extremely interesting, a very strong and, given the three examples, a
homogeneous assumption which can be summarized in the following manner.

      For explanatory schemes applicable for the “action potential” of mental states, a
      comparatively rich repertoire is currently available which is located, inter alia, in a
      misinterpreted habitus-domain, in the value-fields or, finally, in the intentional area and which
      is characterized, above all, by a causal efficacy and, thus, by an explanatory relevance of
      “internal mechanisms” in the production of social actions, practices and routines,.

As will be remembered, this conclusion runs counter to the proposed result of the causal inefficacy and,
additionally, of the explanatory irrelevance of traditional “internal” frameworks -

      For the explanatory schemes available for the “action potential” of mental states, the
      discussion will be aimed at the demonstration of the causal inefficacy and, thus, of the
      explanatory irrelevance of the proposed frameworks.

Thus, a brief discussion of the three “internal” explanatory schemes will and must show their special
causal role.

  Starting with Pierre Bourdieu, one is confronted with a highly interesting general scheme of the
  format -

                   {[(Habitus)(Capital)] + Field = Praxis} (BOURDIEU 1982:175)

  Upon closer inspection however, circularity enters Bourdieu´s framework when the generative
  operations of the habitus are specified in closer detail. Take the following characteristic summary
  from Bourdieu:

        Der Habitus ist Erzeugungsprinzip(!) objektiv klassifizierbarer Formen von Praxis und(!)
        Klassifikationssystem (principium divisionis) dieser Formen. In der Beziehung dieser beiden
        den Habitus definierenden Leistungen: der Hervorbringung(!) klassifizierbarer Praxisformen
        und Werke zum einen, der Unterscheidung und Bewertung der Formen und Produkte
        (Geschmack) zum anderen konstituiert sich die repräsentierte soziale Welt, mit anderen Worten
        der Raum der Lebensstile ... Der Habitus bewirkt(!), daß die Gesamtheit der Praxisformen eines
        Akteurs (oder einer Gruppe von aus ähnlichen Soziallagen hervorgegangenen Akteuren) als
        Produkt(!) der Anwendung identischer (oder wechselseitig austauschbarer) Schemata zugleich
        systematischen Charakter tragen und systematisch unterschieden sind von den konstitutiven
        Praxisformen eines anderen Lebensstils ... Der Habitus ist nicht nur strukturierende(!), die
        Praxis wie deren Wahrnehmung organisierende Struktur(!), sondern auch strukturierte(!)
        Struktur(!) (BOURDIEU 1982: 277ff.)

     To make the generative over-interpretation in Bourdieu´s scheme of -

                                  {[(Habitus)(Capital)] + Field = Praxis}

     more transparent, a methodological analysis on typical normal science approaches might be
     helpful. Here, three different research-steps can be distinguished. First, different types of
     capital (economic, social, cultural, ...) may become the main components in a “capital space”
     which, in turn, can be partititioned into different clusters of life-styles. Using the symbol  for
     a separation into several domains of the capital space, one can write -

                              Space of Capital Types  Clusters of life-styles

Second, different types or clusters of life-styles can be used as “explanandum” and a small set
of capital types as “explanans” which produces the following explanatory scheme -

                    Combination of Capital Types  Clusters of life-styles

Third, the various dimensions of life-styles can be aggregated via cluster-, correspondence or
factor-analysis, into a small number around the limit-number nine of “characteristic life-
styles”(SPELLERBERG 1994/1995) -

                                     Clusters of life-styles
                               Routines in Life style dimensions

An unusual and totally unwarranted move, however, would be a combination of these three
research designs, albeit one with an inversion of the above scheme, yielding a “generative”
pattern of -

  Combinations of Capital Types  (Clusters of life-stylesHabitus)  Routines + Fields

Thus, in Bourdieu´s case, an extremely interesting, but non-generative explanatory framework
on the relations between -

            {Combinations of Different Types of Capital, Habitus, Fields, Praxis}

remains intact and, moreover, most of Bourdieu´s empirical results are not affected at all by
the present critique of his specific perspective of an overall “generative mechanisms”.
What has to be changed drastically though, is the interpretation of the concept of habitus and
the nature of Bourdieu´s internal demon for selections and distinctions. Following the present
arguments, the “habitus” like “intelligence” or “aggression”, is a disposition term, subject to
the common logic of dispositions. In this sense, the relation between habitus and its
“constitutive” or “generative” actions has to be transformed into the trivial “disposition
relation”, as it has been defined during the first step. Thus, a special habitus does not “effect”
or “produce” specific practices or routines, but, conversely, special routines and actions
become necessary conditions for the attribution of a special type of habitus. Moreover, the
external attribution of a specific habitus does not imply that a specific mechanisms acts and
operates as principium divisionis within. (On this point, see especially ASHBY 1981,
FOERSTER 1985) Moreover, the relationship between a space of capital types and

corresponding habitus families, though forming under appropriate specifications a causal link,
does not render a generative chain, leading from capital types via internal intermediary steps
to specific routines and distinctions in everyday life. Instead, one is confronted with an
additional explanatory level for a special disposition term. Take an analogy from science
fiction, namely a causal linkage from the space of genes to different levels of human
intelligence, then a causal sequence of the format -

   Space of Genes  (Clusters of Problem SolutionsIntelligence)  Routines + Fields

would have to be replaced by a drastically less “generative” arrangement -

            Space of Genes  (Capacity Levels of Problem SolutionsIntelligence)
                                                             Routines + Fields

Here, genes do not “generate” or “produce” problem solutions in everyday life but become an
essential element in explaining different capacity levels for problem solutions, identified in
everyday life or in specific “intelligence tests”.
A drastic parallel may be helpful to strengthen this result of a trivial, i.e. “quasi-analytical”
relation between “habitus” and “habitus selections” on the one hand and forms of “praxis”. on
the other hand. Take a concept like “Scandinavia” which is defined by simple enumerations,
namely as the set of {Denmark, Finland, Norway, Sweden}. An evident categorial mistake
would consist in using a simple enumeration as an explanatory and, thus, causally relevant
scheme of the generative form which operates as a “selection demon” deep within.

   Scandinavia is within the countries of Northern Europe a generative principle(!) of objectively
   classifiable forms of praxis and(!) a system of classifications (principium divisionis) of
   practices. With these two achievements: the production(!) of classifiable forms of practices on
   the one hand and the distinction and evaluation of these forms and products on the other hand
   which, in combination, define Scandinavia, the constitution of the social world in Northern
   Europe, in other words: the space of life-styles, is brought about ... Scandinavia, in effect, is
   causally responsible(!) that the totality of a North European country, conceptualized as the
   product of the application of identical (or mutually exchangeable) schemes, exhibits at the same
   time a systematic character(!) and is systematically differentiated from the constitutive(!) forms
   of praxis in another North-European country ...

Thus, “Bourdieu´s demon” of selecting and producing actions should be abandoned, leaving,
instead, a highly fruitful and interlinked research program for the social universes next door
and beyond in the format of -

           Combination of Capital Types  (Clusters of Life StylesHabitus)
                                                     Routines + Fields

With respect to value-changes, a much more radical demonstration on the external nature of
the explanation-sketch offered by Ronald Inglehart and, moreover, on the circular nature of
the “value-relations” can be made. Due to restrictions in space, only the three main-moves
will be sketched very briefly.

  First, an explanatory scheme, having values as its “explanandum” and a specific socio-
  configuration after 1945 within the developed world in general as its “explanans” must
  be considered, quite naturally, as a highly legitimate scientific endeavor.
  Second, the specific practices in a variety of everyday affairs can be clustered and
  combined into generalized value-orientations. Quite logically, a materialistic and post-
  materialistic “cluster” can be identified via standard statistical procedures and,
  moreover, specific scales for the identification of postmaterialistic or materialistic
  values can be constructed.
  Third, the categorial mistake enters into the value-game, as soon as the following
  hierarchy, especially its second part of values  actions, is assumed -

     Socio- economic configurations  Value preferences  Routines in everyday-life

Here, the verdict of the first step, i.e. the quasi-analytical character of the “disposition-
relation” (DPR) must be recalled and, thus, the explanatory relevance as well as the causal
efficacy of value-schemes be removed via the non-generative arrangement of -

                   Socio- economic configurations  Value preferences
                                                               (DPR)
                                                  Routines in everyday-life

For the “intentional” example Nr. 3, a preliminary question will turn out to be helpful, namely
the question, under which peculiar circumstances an explanatory scheme can be established,
in principle, between the “intention” of withdrawing money from a bank and its appropriate
activities. One of the very rare examples comes, at least in nuce, from learning and training
processes where a new practice, a novel habit or the rules for a new language game are at
stake. In the process of adapting one´s behavior to a new set of rules, one could be tempted to
construct an explanatory framework of the following type -

   A has learned the language game „to withdraw money from a bank‟ right now (Initial condition)

as well as the nomic proposition -

   Persons who have been trained in new language games have a strong tendency to execute the
   moves in the newly trained language games by themselves.

This lawlike statement yields, almost as a logical consequence, the following outcome -

   Due to A´s “intention training”, A has developed a strong tendency to execute the moves in the
   newly trained language game by herself.

This example however, is clearly a far cry from the desired claim of causal efficacy with
respect to the “intentional stance” -

   Die Intention 'Geld von der Bank abheben' erklärt(!!!) die dazu erforderlichen Einzelschritte
   und Körperbewegungen.

The main reason for the causal inefficacy can be found by comparing the above claim with an
adequate explication of the basic requirements for “intentional explanations”.

   Unter Zugrundelegung der Annahme, daß das Ereignis B eintritt (oder wahrscheinlich eintritt),
   wenn das Ereignis A geschieht, macht jemand A, um B zu erreichen; das heißt: B wird von ihm
   intendiert. Natürlich reicht der Hinweis darauf, daß A ausgeführt wurde, um B zu erreichen,
   noch nicht hin, um einsichtig zu machen, daß und warum diese Erklärung zutreffend ist. Nicht
   immer sind allerdings Erklärungen intentionaler Art so zu 'vervollständigen', daß sie diese
   zusätzliche Einsicht vermitteln. Die Struktur einer derartigen 'vollständigen' intentionalen
   Erklärung wäre: 'Die Person X tat A, weil sie B erreichen wollte', wobei diese Erklärung auf der
   Verallgemeinerung beruht: 'Leute, welche B erreichen wollen, tendieren dazu, A unter
   bestimmten Bedingungen U zu tun.' So sind auch intentionale Erklärungen in ihrer logischen
   Struktur nomologischer Art; sie unterscheiden sich von anderen nur hinsichtlich der Art der in
   ihnen verwendeten Begriffe und hinsichtlich der Art, wie die in ihnen figurierenden
   Verallgemeinerungen gebildet werden (ACHAM 1974:196f.).-

From the above quotation it becomes almost “self-evident” that the general law (L) in
question for a generalized “intentional explanation” must have a format like -

   (L): People who want to achieve B, operate, under normal circumstances U, via an
   action pattern A.

L, however, should be considered not more than a “quasi-analytical” generalization for the
combination of intention B and its “manifestations” A. Why? Simply because the events or
processes in the action pattern A are, at the same time, necessary conditions for the attribution
of the intention B. It becomes, at least partially, a matter of taste whether a listing of necessary
attribution conditions within a given socio-economic ensemble should be qualified as an
empirical law. If, however, a statement of the form “People who want to achieve B tend under
normal circumstances U to “execute” an action pattern A” is qualified as lawlike, then the
normal interactions between causes and effects cannot be applied for this specific lawlike
More generally, what has to be removed in “intentional stances” is, once again, the traditional
philosophical or psychological accounts of the intentional efficacy. Take, finally, the
following external description -

   a child discovers, as so many do, that by hitting a vase with a stone he can smash it (SEARLE

then the subsequent intentional transformation steps, undertaken, inter alia, by John Searle,
should be considered as ill-founded and un-warranted.

   The child has discovered that this intention in action(!) results in this movement of the hand
   and the arm, which results(!) in this movement of the stone which results(!) in this smashing of
   the vase ... The causality of the intention in action can carry through to the final step, the
   smashing of the vase, because it goes through each of the intervening steps of the by-means-of-
   relation. Each step is a causal step, and the transitivity of the by-means-of-relation enables the
   intention in action to encompass all of them. It is part of the content(!) of the child's intention in
   action that this intention cause(!) this movement of the arm, but also that this movement of the
   stone cause the smashing of the vase, because that is what the child is trying to do: cause the
   smashing of the vase by hitting it with the stone. (IBID.)

“... that this intention cause this movement of the arm ...” Can this be? The answer is clearly -
a trivial yes and a non-trivial no. Yes, since external attribution strategies and sequences of
external ascriptions can be linked to external causal chains like in the following “emotional
stance”-version -

   Person A did not go to the conference because A was in an entirely unhappy and miserable mood -

      or in the subsequent “intentional variation -

         Person A did not go to the conference in X because A wanted to meet B in Y.

      The far more important part of the answer, however, stresses the point that external ascription
      sequences should not be interpreted as “intentional demons”, generating and producing action

What has happened in all three examples of allegedly “generative internal mechanisms”, is a result, by
and large, of Beulen, die sich der Verstand beim Anrennen an die Grenze der Sprache geholt hat
(WITTGENSTEIN 1971:PU 119), since a quasi-analytical relation between dispositions and events has
been mis-interpreted and mis-used as an explanatory scheme of “internal” causes (dispositional terms like
habitus, values, intentions ...) and “external” effects (actions, practices, routines, evaluations ...).
More systematically, a causal relation, as Gilbert Ryle (1969) and in variations Eike von Savigny (1970)
or Karl H. Müller (1989) have noted, fulfills at least one central condition, namely the independence of

      So wie Ursache und Wirkung verhalten sich die Ereignisse in der folgenden Kette: Weil es regnete,
      wurde die Straße naß. Weil die Straße naß wurde, geriet der Wagen ins Schleudern ... Charakteristisch
      für diese Kette ist zweierlei: Erstens kommen die Sachverhalte nacheinander; und zweitens (das ist
      wichtiger) kann man jeweils die Ursache oder die Wirkung beschreiben, ohne die zugehörige
      Wirkung beziehungsweise Ursache zu beschreiben. (SAVIGNY 1970:86)

With seemingly “causal” sequences in internal notation, be they “intentional” or “value-based” and the
like, the independence of description-postulate does not hold. In a slight variation to the above quotation -

      kommen (hier) ... nicht verschiedene intentionale Akte nacheinander, und zweitens kann man die eine
      Intention nicht beschreiben, ohne die andere zu beschreiben ... Es handelt sich jeweils um die
      Einordnung einer speziellen Disposition in eine allgemeinere. Was bei der Erklärung der Handlung
      durch Intentionen vorliegt, ist eine dispositionale Erklärung; und dispositionale Erklärungen geben
      keine Ursachen an.

A similar point has been stressed, surprisingly enough, by Robert Cummins, too -

      The attributions of analyzing properties should be justifiable independently of the analysis that
      features them. If, for example, we analyze the capacity of a child to solve division problems into the
      capacity to copy numerals, to multiply, and to subtract, we must know, or be able to find out, that the

         child can copy numerals, multiply and subtract without simply inferring this from the capacity to
         divide, and we must know, or be able to find out, that these capacities are in fact organized as the
         analysis specifies. (CUMMINS 1985:26)

Sixth step: It would lead to far away from the current discussion of embeddedness-relations to sketch the
consequences of an external perspective for the seemingly reclusive “internal vocabulary” which, so far,
has dominated the traditional domains of empirical social research, too. Suffice it to say that a large class
of pseudo-problems on the consistency between value-orientations and everyday activities could be
eliminated since this type of problems can be subsumed under the heading of insufficient measurements.
An analogy from a more obvious domain might be helpful, namely “intelligence” and “intelligence
testing” where a clear separation has been undertaken between two forms of inquiry:

         At the level of self-evaluations, very few questions and item batteries like -“All in all, how
         would you grade your IQ?” - have been constructed so that the area of self-estimations has not
         acquired a prominent role in determining degrees of personal “intelligence” ...
         On the other hand, a large class of different test-domains for intelligent behavior has been
         built up which, in combination, yield a specific value, namely the intelligence quotient ...105

Thus, in the area of intelligence, consistency problems of the value-type like “Why do persons with a high
self-evaluation of their intelligence act in a non-intelligent manner?” do not arise at all. And this analogy
should serve as an illustration that in explanation schemes especially of the value-variety the required
non-circular additional assumptions have not come into existence yet and, due to the logic of
dispositional terms, will have tremendous difficulties of entering into an “existence-status” in the distant
future ....
At this point, the elliptic and fragmented character of conventional research designs in the field of value-,
emotion- or intention based research for that matter should have become more transparent. Put in a
nutshell, these types of research strategies need at least two significant extensions.

         On the one hand, “values” or “intentions” need - like in the case of “intelligence” - a rich
         variety of questions and items at the level of everyday routines, of life-style practices, of
         preferences in everyday contexts and the like independent from the self-evaluations of
         respondents. Even then, a puzzling problem will still persist since, again like in the case of
         intelligence, the actual “life world”-practices might differ significantly from the question and
         answer-context of empirical social research ...
         On the other hand, value-systems or intentional schemes need a historical record since a
         single point of observation is, in all probability, an insufficient basis for the attribution of

      On the conceptual problems and difficulties in this area, see, e.g. GOLEMAN 1995.

         value- or habitus styles. Again, like in the case of intelligence, interesting intra- as well as
         interpersonal differences and patterns are, as yet, totally unresolved.

In a beautiful quotation, the argument for a necessary perspective of the long run in the attribution of
features, processes or functions can be generalized and can be put, moreover, in an impressive
cosmogonic context -

         Suppose that by some cosmic accident a collection of molecules formerly in random motion were to
         coalesce to form your exact physical double. Though possibly that being would be and even would
         have to be in a state of consciousness exactly like yours, that being would have no ideas, no beliefs,
         no intentions, no aspirations, no fears, and no hopes ... This because the evolutionary history of the
         being would be wrong. For only in virtue of one's evolutionary history do one's intentional mental
         states have proper functions, hence does one mean or intend at all, let alone mean anything
         determinate ... That being would also have no liver, no heart, no eyes, no brain, etc. This again,
         because the history of the being would be wrong. For the categories 'heart', 'liver', 'eye', 'brain' and also
         'idea', 'belief', and 'intention' are proper function categories, defined in the end by reference to long-
         term and short-term evolutionary history, not present constitution or disposition. Were this not so,
         there could be no malformed hearts or nonfunctioning hearts nor could there be confused ideas or
         empty ideas or false beliefs, etc. Ideas, beliefs, and intentions are not such because of what they do or
         could do. They are such because of what they are, given the context of their history, supposed to do
         and of how they are supposed to do it. (MILIKAN 1984:93)

In this manner, a problem domain comes slowly into sight which stood at the beginning of the current
exploration into the external aspects of “internal descriptions”, namely the ...
Seventh = first step, where the precise relation between dispositional concepts and their constitutive
events have been put forward in the conventional manner, stressing the “quasi-analytical” and causally
ineffective character of the “disposition-relation” ...
As an important consequence of the radical departure on inaccessible language privacies and subjective
meanings as embeddedness domains, one is led to another famous quote by Ludwig Wittgenstein,
stressing the gradual differences with respect to one´s internal mental states and “other minds”:106

         Daß, was ein Anderer innerlich redet, mir verborgen ist, liegt im Begriff 'innerlich reden'. Nur ist
         'verborgen' hier das falsche Wort; denn ist es mir verborgen, so sollte es ihm selbst offenbar sein, er
         müßte es wissen. Aber er 'weiß' es nicht, nur den Zweifel, den es für mich gibt, gibt es für ihn nicht.

      On this point, see especially D.O. Hebb´s warning sign on the allegeded immediacy of “self-knowledge”. One should not presuppose -

         that there is immediate knowledge of one's own mental processes ... We do of course know something of what goes on in our
         minds but ... the knowledge depends on inference. I look out of the window and perceive a tree, but I do not perceive the
         perception, to know how perception occurs; I imagine a monster and know what the imagined monster would look like, but
         not how I imagine it ... That kind of self-knowledge almost certainly does not exist. (HEBB 1980:4p)

       'Was Einer zu sich selbst im Innern spricht, ist mir verborgen' könnte freilich auch heißen, ich könnte
       es zumeist nicht erraten, noch auch (wie es ja möglich wäre) aus den Bewegungen seines Kehlkopfs
       z.B. ablesen.
       'Ich weiß, was ich will, wünsche, glaube, fühle, ... ' (usf. durch alle psychologischen Verben) ist
       entweder Philosophen-Unsinn, oder aber nicht ein Urteil a priori. (WITTGENSTEIN 1971:p. 257)

Thus, the differences between neural descriptions, behavior descriptions and the variety of “mental
ascriptions” do not lend themselves to causal divisions and to “causality arrows” between, say, the
“intentional vocabulary” and behavior descriptions. This “causal fallacy” becomes more transparent if it is
used in the case of machines like computers which can be described in high chunk form (Douglas R.
Hofstadter) with respect to program performances, which can be characterized at the meso-level of
assembler-languages and which, finally, can be analyzed on the lowest level in terms of electronic
circuits. It should be added that the computer-analogy is just a contemporary way of making the main
point of different levels of descriptions as well as ascriptions more intelligible, since exactly this point has
been raised and established already a century ago.

       Ein Organismus ist ein System von Molekülen. Elektrische Ströme laufen hinein von außen angeregt
       und kommen wieder zurück in die Muskel. Alles läßt sich physikalisch erklären. Nur daß dieser
       Mensch empfindet, nicht .... Wir haben also die Schwierigkeit, etwas fundamental Neues in einer
       Gruppierung zu haben, was in den Teilen nicht ist. Dieser Schwierigkeit entgehen wir, wenn wir die
       Empfindung als eine allgemeine Eigenschaft der Materie betrachten ... Welchen wissenschaftlichen
       Wert diese Annahme einer allgemeinen Empfindung der Materie hat, das kann nur dadurch
       entschieden werden, ob wir physikalische Erscheinung durch dieselbe besser ableiten und verstehen
       können. Ableitungsregeln für unsere Empfindungen mit Hilfe anderer hinzugedachter und im
       Kausalzusammenhang gedachter Empfindungen. (MACH 1988:173f.)

Consequently, “mental states”, following Ernst Mach, should be treated as a typical problem of
emergence1,2,3 in evolutionary systems. 107
Quod erat demonstrandum.

   In this sense however, the problem domains for the social sciences become far more complex than the following perspective of an
“ontological U-turn” of the Parmenides-variety suggests:

       There is an external universe 'out there' which exists independently of our conceptualizations of it; this universe reveals
       certain timeless, universal and invariant properties; the goal of sociological theory is to isolate these generic properties and
       understand their operation. (TURNER 1987:156)

                         1.3. NON-TRIVIAL FRAMEWORKS

                        FOR THE ANALYSIS OF AGENCIES

For the necessary departure of the second type, namely a disgression from trivial schemes of explanation
schemes for actions, human and otherwise, the starting point is marked by a short kaleidoscope of present-
day explanation frameworks which aside from a heavy emphasis on concepts like values, value-systems,
habitus, life-style and the like exhibit a logical structure which will be identified, subsequently, as trivial.
Take the following general explanatory relation for “values” -

      Man kann, ganz grob gesagt, davon sprechen, daß Werte - als Führungsgrößen(!) - das menschliche
      Verhalten 'steuern'(!!!) ... Nichtsdestoweniger lassen die Werte(!) dem Menschen(!) - ungeachtet der
      Stärke, die sie 'eigentlich' haben(!!!) - in großen Teilen des 'normalen' Tagesablaufs einen mehr oder
      minder breiten Spielraum(!) für ein Verhalten ... (KLAGES 1984:10f.) -

The main point of criticism for explanation types like the one quoted above lies in their basic input-output
character and in their reliance on the following generalized framework -

        y (observable behavior) =  (input variables: “values”, “intentions”, “character”, “will” ...)

which will be qualified, following Heinz von Foerster, as a trivial, though highly successful scheme.

      Der überzeugende Erfolg dieses Erklärungsschemas in Physik und Astronomie hat Denker in anderen
      Disziplinen überredet, diesen Modus zur Basis ihrer Argumentation zu nehmen: Kausalität wurde zum
      Fundmanet westlichen Denkens. So wie man in der Physik die Triade Ursache/Naturgesetz/Wirkung
      hat, so hat man in der Psychologie Reiz/Organismus/Reaktion, in Zweigen der Psychologie
      Motiv/Charakter/Verhalten, usw. usw. Deutlicher wird dieses Schema in der Algenra, wo die
      unabhängige Variable x über die Funktion  die anhängige Variable y bestimmt: y =  (x); aber am
      deutlichsten sieht man wohl die Fundamentalstruktur der Kausalität am Beispiel der modernen
      Rechner mit der Triade Input/Operation/Output. (FOERSTER 1990:80)

What is needed however, would be an alternative explanatory framework which, again sticking to the von
Foerster terminology, may be classified as non-trivial schemes or alternatively, a state-dependent or state-
determined approach. In state-determined schemes the input/operation/output relation acquires the
following format -

                 input  state-determined operations  state-determined operations  output

The consequences of non-trivial machines or schemes are striking in at least two respects. On the one
hand, the number of possible configurations increases exponentially, depending on the number of internal
states.108 On the other hand, the problem of identifying ensembles with an unknown number of internal
states belongs to the class of non-identifiable and un-solvable problems.

       Die Situation ist aber völlig anders, wenn weder wir noch unser Forscher die Transformationsregeln
       und die Anzahl der inneren Zustände wissen. Das Identifikationsproblem für nicht-triviale Maschinen
       ist dann ebenso nicht-trivial. Beschränken wir uns wieder auf nur vier Inputs und Outputs, dann wird
       N, die Zahl der möglichen Maschinen:

                                                                 44      4
                                                          N=2         . 24 = 2512

       ... Das Problem ist jenseits der Errechenbarkeit , es ist „transkomputabel‟. Und was schon vorhin über
       das rasche Anwachsen möglicher Maschinen bei Vermehrung innerer Zustände angedeutet war, gilt a
       fortiori für die Transkomputabilität. (FOERSTER 1990:87)

Focused on socio-economic domains, the required non-trivial frameworks (NTF) have, by necessity or by
chance, a peculiar feature since they exhibit the following relation -

                                                              CM  NTF

i.e., the class of complex or transdisciplinary models (CM) which will be introduced at greater length in
Part II of Volume II is a subset of the set of non-trivial frameworks (NTF). Table 1.1 provides an overall
survey of the non-trivial elements within the complex model-families. Three additional comments seem

       First, non-triviality can be achieved in a variety of ways, ranging from state-determined input-
       output configurations like in the case of “Classifier systems” and “Genetic Algorithms” or
       “Neural Networks” to the case of non-trivial aggregation procedures, as exhibited via
       applications with “Cellular Automata”, “Artificial Life” or complex “Rational Choice”-
       Second, almost all of the CM-classes allow for the possibility of self-referentiality by
       analyzing processes in which the scientific observer becomes an integral and indispensable
       part. Take the case of master-equations which will turn out to be of considerable importance
       in Volume V, then the scientific actor networks which can be constructed in this framework
    For four inputs, four outputs and two internal states, the number of possible ensembles becomes 44.44 = 2562 = 65536, for three internal
states the number increases to 44.44.44 = 16 777 216, etc.

       comprises, inter alia, a network, consisting of scientific groups and different types of reserach
       programs, including master equations. Thus, the diffusion potential for master-equations can
       be studied within a master equation design. In this specification (for more details see
       MÜLLER 1994), no outright and direct paradoxes are generated although upon closer
       inspection some interesting cases may result out of a self-referential application of a complex
       model onto itself.
       Third, scientific models, no other than words or sentences, depend on specific contexts. Take
       as reference case two master equation models on employment and education
       (HAAG/MÜLLER 1992, MÜLLER/HAAG 1994), then it becomes quite obvious that
       mobility processes between network components are far more intense in the case of the
       employment sphere than in the education domain. Why? Within the master equation context,
       mobility, i.e. a transfer of people between employment segments like agriculture, industry,
       firm related services, etc., occurs, at least in principle, between all components in the
       employment model whereas severe restrictions have to be imposed in the case of mobilities
       between pupils across various school-types. Furthermore, only specific segments of a school-
       type are legally allowed to change into a hierarchically higher type, whereas, in principle, the
       whole population may, from a legal point of view, move from one sector, say agriculture, to
       industry. Thus, one would be led to the conclusion that mobility is an essential feature for
       employment systems and a marginal phenomenon for school ensembles. However, an
       alternative way of “Rational Choice”-modeling, taking the relations between entrances and
       exits as the main criterion of mobility, leads to opposite conclusions since a large proportion
       of unskilled and semi-skilled persons leaves the employment sphere, after thirty or fourty
       years, still in the status of unskilled or semi-skilled groups. On the contrary, the school system
       must be judged as a far more mobile arrangements, since (almost) everyone leaves the school
       system in a different type than the entrance form of the Elementary School. Thus, different
       types of models generate a model-context-dependent distribution of features, leading to the
       non-contradictory conclusion that mobility, in the conetxt of master-equations, is strong in the
       case of employment and comparatively weak in the education realm, whereas mobility, in the
       case of Rational Choice-modeling, plays a strong role in the school system and a relatively
       weaker one in employment areas.

To sum up, the non-triviality requirement, which should have become a core-condition for modeling and
explanations especially in the social sciences a long time ago already109, can be fulfilled in an almost
trivial manner by the utilization of the stock of complex models outlined in Table 1.1.

    It would be an extremely fruitful reserach task to classify the conventional arguments against a naturalized version of the social sciences
along a trivial - non-trivial dimension. In all probability, a large number of arguments, especially those on the internal nature of the social
universes, could be grouped in the non-trivial domain.

Table 1.1: Set of Complex or Transdisciplinary Models
           with Non-Trivial Modeling Schemes

                             SUCCESSFUL APPLICATIONS            COMPONENTS
Adaptive Systems             Development Patterns               Non-trivial, i.e. circular re-
and Control                  with Emphasis on Steering          lations between controling systems
Theory                       Units or Internal Re-              and the controlled ones ...
                             presentations et al.

Artificial Life              Development Patterns               Non-trivial aggregation pro-,
                             of Living Units with an            cedures, yielding un-expected and
                             Emphasis on Cognition- and         un-forseeable (emergent1)
                             Adaption Processes et al.          results ...

Autopoiesis                  Development Patterns with          Non-trivial, state-determined
                             in a Direct Connection of          perspectives on biological pro-
                             Interaction and Reproduction       cesses;      special      non-trivial
                             as well as with an Observer-       on thje role of the “observer” in
                             Dependent Reference Frame et al.   descriptions and models ...

Cellular                     Development Patterns, Resul-       High Potential for non-trivial
Automata                     ting from the Interaction of       aggregation schemes via non-linear
                             Spatially Ordered NIS-             interaction patterns ...
                             Components (Scientists, Insti-
                             tutes) et al.

Chaos-theory                 NIS-Processes with a Com-          Non-trivial results through
                             plex Dynamical Pattern             the specification of high-dimen-
                                                                sional,     non-linear      dynamic
                                                                leading to “strange attractors” ,,,

Classifier Systems           Development of Learning            Non-trivial through a “nice”
and Evolutionary             Patterns via an Evolutionary       bidding process in which a variety
Programming                  (Crossover) Process of Variation   of rule-components are involved;
                             and Selection et al.               random processes of recombina-
                                                                tions (“cross-over”) ...

Complexity                   Identification of Types and        Non-trivial through the unresolved
Theory                       Degrees of Possibilities of          status of decidable and non-.

Problem Solutions in a NIS-   decidable problems in polynomial
Context et al.                time (NP-problems) ...

Table 1.1: Set of Complex or Transdisciplinary Models

              with Non-Trivial Modeling Schemes (continued)

                                SUCCESSFUL APPLICATIONS              COMPONENTS

Dissipative Struc-              Development Processes                Non-linear                   modeling
tures                           Far from the Thermodynamic              giving rise to non-trivial, i.e.
                                Equilibrium, et al.                  “chaotic” behaviour ...

Evolutionary                    NIS Interaction Patterns             Non-trivial aggregation pro-
Gametheory                      with a Well-Defined Decision            cedures yielding differentiations
and Rational Choice             and Preference Set, et al.           between evolutionary stable and
                                                                     evolutionary unstable dynamics ...

Group Theory                    Changes, Resulting from              Non-triviality through the speci-
                                a Well-Defined Exchange              fication of groups and group opera-
                                in a NIS Ensemble,                   tions, generating complex patterns
                                et al.                               of symmetry and symmetry
                                                                     breaking ...

Hypercycles                     Developmental Dynamics,                 Non-triviality through the use of
                                Resulting in the Autocatalysis       non-linear equation systems
                                of Relatively Complex NIS-           with a rich variety of dynamic
                                Ensembles from Simpler               behaviors ...
                                Ones, et al.

Neural                          Indirect and Highly Connected        Very dense and non-linear inter-
Networks                        NIS-Networks with Clear-          linkages between input-hidden
                                Differentiations between Input-      layer(s) and output, giving rise to
                                and Output-Domains, et al.           a non-trivial, sequential learning
                                                                     process ...

Population                   Interaction- and Development         Non-linear systems of equations,
Dynamics                        Patterns between Several            producing, as a non-trivial out-
                                NIS Groups et al.                   come, chaotic behavior ...

Synergetics                     Dynamics of a NIS-Ensemble              Non-trivial             aggregation
                                Resulting from the Interactions      (“enslaving      principles”),    giving
                                of Comparatively Smaller            to few “order parameters”;
                                NIS-Groups et al.                 dynamical, at times “chaotic” be-
                                                                    havior of systems ...

Theory of                       Abrupt, Discontinuous                Non    trivial    dynamics       due   to

Catastrophes                     Changes, Resulting from              tinuous changes and sudden “phase
                                 a Highly Specific Functional         transitions” ...
                                 Relationship, et al.
Moreover, it could be shown that “the eye of the observer” is ubiquitously present in the area of complex
modeling, too. This, in turn, leads to a variation of a well-known poem110 on the “tangled networks” of
“observing systems” -

                                                     Big systems have observing ones,
                                                        In a network that binds „em.
                                                     The little ones have observers, too,
                                                             And so ad infinitum.

Finally, the implications of this peculiar poem will become the central theme for the next chapter in which
the observer-dependent approaches to agency will be outlined in greater detail.

                                           1.4. “INTERNAL STANCES”


                                          ATTRIBUTION STRATEGIES

One of the most important additional moves which is not only directly linked to the above discussion, but
which also guarantees the successful continuation of the transdisciplinary approach on codes and
evolution must be undertaken right at the beginning. It consists in a “radically constructivist turn” by
pointing to two requirements, namely, first, to the unavoidable and necessary role of the observer in any
type of scientific output and, second, to the attribution character of the so-called internal vocabulary,
intentional and otherwise.

          First, following recent advances initiated by Heinz von Foerster (1982, 1985, 1993), Ernst
          von Glasersfeld (1986), Humberto R. Maturana (1985), Jean Piaget (1973, 1983, 1985, 1992),
          S.J. Schmidt (1987) or Francisco J. Varela (1989), the role of the observer will be transformed

      In its original version, the poem reads as -
                                                         Big operations have little ones
                                                          In a network that binds „em.
                                                         The little ones have lesser ones
                                                              And so ad infinitum.

For a bibliographical identification, one may find this well-known doggerel in SINGH (1972:87).

          from an unavoidable nuisance of the “white noise” family backstage to that of a central main
          Diagram 2.1: Observer-Dependencies and the Modeling Relations111

                                      PRAGMATICS                   SEMANTICS                   SYNTAX

                                                           decoding                decoding          

                                                 O                         N                         F

                                                           encodings                 encodings 

          According to this turn of Bringing the observer back in (WATZLAWICK/KRIEG 1991), any
          account of the socio-economic worlds, by necessity, is bound to be observer-dependent. A
          schematic representation of observer-dependencies can be given, following (CASTI 1994) and
          (ROSEN 1991), in the small diagram above. Here, the relations between a “concrete system”
           in the domain N and a formalized model MOD in area F can be summarized by the
          following relations. In Diagram 2.1 -

              we see a natural (read:real world112) system N characterized by observations and relations
              stated in everyday language. The formalization process then involves encoding these
              characterizations of N into the symbols and strings of a formal logic (read:mathematical)
              system F. The key to understand this process of formalization is to recognize that all notions of

      It should be added that the plural of “modeling relations” refers, inter alia, to the possibility for “second order” ensembles of the format -

                                      PRAGMATICS                SEMANTICS                  SYNTAX

                                                         decoding               decoding

                                   O(O ...(O))                       N                       F(F ...(F))

                                                     encodings                 encodings 

and to the following variation -

              we see an observing (read:real world) system O by using our encoding capacities, especially the everyday language
              code and its contexts. The observing process then involves our encoding of the encoding characterizations of O. The
              key to understand this process of encoding of encodings is to recognize that all notions of contexts (i.e., pragmatics)
              reside on the left hand side of the diagram ... N consists of mere semantic symbols, together with rules (a grammar)
              for how strings of these symbols can be combined to form new strings. Whatever utilization might inhere in these
              strings is then brought out by the decodings of the decodings of the strings back into O(O).
   For a precise notion of “real world systems” and the like, see Part II of Volume II where several differentiations along the modal
dimension of “actual worlds” and “possible worlds” are introduced.

   meaning (i.e., semantics) reside (in the middle) of the diagram ... F consists of mere abstract
   symbols, together with rules (a grammar) for how strings of these symbols can be combined to
   form new strings. Whatever meaning might inhere in these strings is then brought out by the
   decoding of the strings back into N. (CASTI 1994:274)

Likewise, the relations between O and N can be summarized by employing just a few
recombinative operations, since in Diagram 2.1 -

   we see (O) a natural (read:real world) system N by using our encoding capacities, especially the
   everyday language code and its contexts. The observing process then involves encoding
   characterizations of N. The key to understand this process of encoding is to recognize that all
   notions of contexts (i.e., pragmatics) reside on the left hand side of the diagram ... N consists of
   mere semantic symbols, together with rules (a grammar) for how strings of these symbols can
   be combined to form new strings. Whatever utilization might inhere in these strings is then
   brought out by the decoding of the strings back into O.

In this manner, observer dependencies become a necessary feature of any scientific output,
even and especially in the case of descriptions or models of hyper-complex configurations like
a National Innovation System. The fourth section of the present volume as well as the first
part in Volume V will demonstrate from a methodological point of view that, by taking
“observer dependencies” seriously, one arrives at a large number of non-standard features on
the constitution of socio-economic ensembles.
Second, the traditional internal vocabulary as well as seemingly internal models like rational
decision theory (see, e.g. BACHARACH/HURLEY 1994) or similar offsprings will be
perceived as an external mode of attribution or, following Daniel C. Dennett
(1987/1991/1995) as an “internal stance” which can be used as an attribution strategy for the
whole animate and, at least partly, for the inanimate world, too. Take Dennett´s
characterization of one of the typical internal stances, namely the intentional one, then one is
confronted with a stunning variety of applications.

   Do people actually use this strategy (i.e., the intentional stance, K.H.M.)? Yes, all the time.
   There may someday be other strategies for attributing belief and desire and for predicting
   behavior, but this is the only one we all know now. And when does it work? It works with
   people almost all the time ... The strategy also works on most other mammals most of the time.
   For instance, you can use it to design better traps to catch those mammals, by reasoning about
   what the creature knows or believes about various things, what it prefers, what it wants to
   avoid. The strategy works on birds, and on fish, and on reptiles, and on insects and spiders, and
   even on such lowly and unenterprising creatures as clams ... It also works on some artifacts ...

           The strategy even works for plants ... It even works for such inanimate and apparently
           undesigned phenomena as lightning. An electrician once explained to me how he worked out
           how to protect my underground water pump from lightning damage: lightning, he said, always
           wants to find the best way to ground, but sometimes it gets tricked into taking second-best
           paths. (DENNETT 1987:21f.)

       In this manner, a rich variety of other “internal” attribution modes can be proposed. Make
       small variations in Dennett´s intentional stance-quotation and one arrives, in congruence with
       Jon Elster (1989), at an emotional stance, empowered with all the virtues and insignia of the
       intentional stance.

           Do people actually use this strategy (i.e. the “emotional stance”, K.H.M.)? Yes, all the time.
           There may someday be other strategies for attributing emotions and feelings and for predicting
           behavior, but this is the only one we all know now. And when does it work? It works with
           people almost all the time ... The strategy also works on most other mammals most of the time.
           For instance, you can use it to design better traps to catch those mammals, by reasoning about
           how the creature feels about various things, what it fears, what it wants to avoid. The strategy
           works on birds, and on fish, and on reptiles, and on insects and spiders, and even on such lowly
           and unenterprising creatures as clams ... It also works on some artifacts ... The strategy even
           works for plants ... It even works for such inanimate and apparently undesigned phenomena as
           lightning. An electrician once explained to me how he worked out how to protect my
           underground water pump from lightning damage: lightning, he said, always fears to loose
           momentum, but sometimes it gets tricked into taking second-best paths.

       In similar moves, other modes of attribution like a normative stance - the behavior attribution
       in terms of social rules and obligations - can be built up which share, by and large, the
       comparative advantages of Dennett´s allegedly single “intentional stance”.113
     In a mode of analogy-building, an uncommon variety of additional modes of attribution can be constructed. Following Ross W. Ashby´s
criterion for “memory” as an observer dependent “explanatory principle” (ASHBY 1981), one is led to the possibility of a “memory stance” -
and to the following variation to Dennett´s “intentional stance”.

       Do people actually use this strategy? Yes, all the time. There may someday be other strategies for attributing history and
       memory and for predicting behavior, but this is the only one we all know now. And when does it work? It works with people
       almost all the time ... The strategy also works on most other mammals most of the time. For instance, you can use it to design
       better traps to catch those mammals, by reasoning about how the creature has acted in the past, what it has done repeatedly,
       what it has not done at all in the past. The strategy works on birds, and on fish, and on reptiles, and on insects and spiders,
       and even on such lowly and unenterprising creatures as clams ... It also works on some artifacts ... The strategy even works
       for plants ... It even works for such inanimate and apparently undesigned phenomena as lightning. An electrician once
       explained to me how he worked out how to protect my underground water pump from lightning damage: lightning, he said,
       has a strong historical record to find the best way to ground, but sometimes it gets tricked into taking second-best paths.

Following the current literature on National Innovation Systems, one is tempted to introduce an “adaptation and learning stance” as an
appropriate universal mode of ascriptions which, then, results in the following recombinations of Dennett´s “intentional stance” ...

       Do people actually use this strategy? Yes, all the time. There may someday be other strategies for attributing learning and
       adaptation and for predicting behavior, but this is the only one we all know now. And when does it work? It works with

       Finally, the behavior of systemic ensembles can be couched in an “agency stance”, too. Here,
       however, it has always been clear that this type of attribution is clearly external in nature,
       referring to observable and measurable properties only. For the last time, a variation on
       Dennett´s “intentional stance” will be performed:

           Do people actually use this strategy? Yes, all the time. There may someday be other strategies
           for attributing routines and practices and for predicting behavior, but this is the only one we all
           know now. And when does it work? It works with people almost all the time ... The strategy
           also works on most other mammals most of the time. For instance, you can use it to design
           better traps to catch those mammals, by reasoning about how the creature behaves and acts,
           what it does with high frequency or what it rarely does. The strategy works on birds, and on
           fish, and on reptiles, and on insects and spiders, and even on such lowly and unenterprising
           creatures as clams ... It also works on some artifacts ... The strategy even works for plants ... It
           even works for such inanimate and apparently undesigned phenomena as lightning. An
           electrician once explained to me how he worked out how to protect my underground water
           pump from lightning damage: lightning, he said, always finds the best way to ground, but
           sometimes an arrangement can be constructed so that it takes second-best paths.

Due to these two unconventional moves, namely to universal observer-dependencies as well as to the
construction of intentional, emotional, normative “stances” or, alternatively, of external “modes of
attribution”, it will become possible to arrive at a new, multi-level, multi-domain foundation for agencies
which, together with the multiplicity of code-systems, allows completely new insights into the making and
constitution of complex and hyper-complex configurations like National Innovation Systems.

       people almost all the time ... The strategy also works on most other mammals most of the time. For instance, you can use it
       to design better traps to catch those mammals, by reasoning about how the creature learns and adapts itself to its
       environment, what it learns quickly, what it does not learn at all. The strategy works on birds, and on fish, and on reptiles,
       and on insects and spiders, and even on such lowly and unenterprising creatures as clams ... It also works on some artifacts
       ... The strategy even works for plants ... It even works for such inanimate and apparently undesigned phenomena as
       lightning. An electrician once explained to me how he worked out how to protect my underground water pump from
       lightning damage: lightning, he said, has learned to find the best way to ground, but sometimes it gets tricked into taking
       second-best paths.

Hopefully, the preceding examples have made it clear that the everyday vocabulary is full of different modes of attributions which, moreover,
can be combined into comparatively more complex modes like the following recombination from Dennett´s original “intentional stance”:

       Do people actually use this strategy ? Yes, all the time. There may someday be other strategies for attributing belief, memory
       and emotion and for predicting behavior, but this is the only one we all know now. And when does it work? It works with
       people almost all the time ... The strategy also works on most other mammals most of the time. For instance, you can use it
       to design better traps to catch those mammals, by reasoning about what the creature knows or believes about various things,
       what it has done in the past, what it wants to avoid. The strategy works on birds, and on fish, and on reptiles, and on insects
       and spiders, and even on such lowly and unenterprising creatures as clams ... It also works on some artifacts ... The strategy
       even works for plants ... It even works for such inanimate and apparently undesigned phenomena as lightning. An electrician
       once explained to me how he worked out how to protect my underground water pump from lightning damage: lightning, he
       said, fears to loose momentum and wants to find the best way to ground, but sometimes it gets tricked into taking second-
       best paths.

                                             2. “AGENCIES”

                                       AT THE GENOTYPE-

                         AND AT THE PHENOTYPE-LEVELS

In order to stick to the format of a small theoretical volume and not to the common reference frame of a
large and voluminous book on social actions and agencies114, only a bare sketch can be presented to arrive
at satisficing answers for the many facets of agencies in complex socio-economic ensembles like a
National Innovation System. Thus, the second chapter will introduce in a very brief manner a general
definition of “agency”, by relying on the evolutionary framework developed so far. Accordingly, the
concept of “agency” can be used, loosely formulated, for the behavior of any evolutionary system at its
phenotype-level - and at its genotype-level, too. Thus, the proposed agency-category is characterized by a
series of surprising features:

      First, “agency” can be applied to a comparatively large domain, from the genetic code to
      languages up to institutions or global actors. As long as the dual level-separation and the code-
      based (re)production requirement is fulfilled, the “agency-stance” can be attributed to any type
      of system along the evolutionary ladder. To make this wide applicability more transparent, the
      following quotation will serve as a useful reference point -

         Man könnte auch von einer Tätigkeit der Butter reden, wenn sie im Preis steigt; und wenn
         dadurch keine Probleme erzeugt werden, so ist es harmlos. (WITTGENSTEIN 1971:PU 693)

      The important restriction on a problem-deflationary account of “agency” can be interpreted in
      such a way that the conceptualization of agencies should be performed far away from
      “animistic” connotations. Thus, the universal mode of agency attributions should not be
      confused with the totally unwarranted assertion of a close family resemblance in different
      agency environments.
      Second, innovation and diffusion processes within National Innovation Systems must be
      considered as the result of “multiple agencies”, producing and generating the innovation and
      diffusion processes in question. Thus, machines, the genetic code, technicians, plants, the firm
      management, the emergence of a new research program, the (re)production of species and the
      like become entangled in the dual chains of acting, becoming and evolving. In this sense, a

  For the apparently common format of foundations above 700 pages   in the field of social actions, see e.g. PARSONS (1937),
LUHMANN (1984), MÜNCH (1988) or HABERMAS (1981).

       National Innovation System is to be understood in a radically more complex manner than the
       conventional accounts of linkages between firms and institutes suggests ...
       Third, even human agency turns out to be a highly non-trivial concept, since under the present
       account human agency can be identified at the P-level of individuals, at the P-level of
       collective actors, economic sectors, global institutions - and at the G-level of cells, neural
       groups, groups of neural groups and the like. Consequently, National Innovation Systems,
       even if they could be centered on interaction patterns between groups of scientists and firm
       employees alone, would have a far more complicated agency structure than a conventional
       micro-account at the P-level alone.

Thus, for a construction of agency domains, human and otherwise, a morphological space with at least
eight main areas can be built up. Moreover, this contemporary partitioning will proceed, concentrating on
the area of human agency alone115, in the following manner:

       Along the first dimension, descriptions, a classical distinction, dating back to David Hume,
       between the empirical and the normative realms (HUME 1989, STREMINGER 1994), can be
       introduced which leads, as a special instance, to the subsequent differentiation between two
       areas of decision theory.

           Normative decision theory ... is deductive. It postulates certain criteria of optimality or
           rationality or equity and derives strategies or methods of allocation or methods of aggregating
           preferences that are supposed to satisfy these criteria. A descriptive theory starts with
           observations of how actors choose in given classes of decision situations and attempts to
           describe their behaviour as systematically as possible. (RAPOPORT 1989:5f.) -

       Along the second dimension, relations between an observer and her or his P-actions can be
       qualified as self-referential whereas relations between a scientific observer and her or his
       fields outside one´s own environment is to be categorized as referential.
       Finally, the third dimension will be devoted to the central distinction throughout the present
       volume, namely to the differentiation between genotype and phenotype levels. By this move, a
       multiplicity of domains for human agency can be opened up which, contrary to the classical
       demarcations built up by Max Weber, distribute human agency both at the genotype and
       phenotype levels.

In this manner, a 23 matrix can be identified which exhibits eight basic areas for the distributed concept of
agency, human and otherwise, where the genotype level starts from single cells or even from smaller sub-

   The main reason for a focus on human agency lies simply in the restriction that a discussion of several applications in both the “animate”
and “inanimate” universes would be a too long disgression before the central chapter on the “embeddedness relations” ...

cellular units - bases, enzymes, proteins, etc. - and proceeds to cell-assemblies, neural groups, the central
nervous system - and their corresponding phenotypes agency patterns ...

Table 2.1a: A contemporary morphological space for the fields of human micro-agency116

                                                             OBSERVER/                                   OBSERVER/
                                                             OBSERVER                                    OBSERVED
                                                           (self-referential)                          (referential)
                                                           DIMENSION3                                  DIMENSION3
                                                         G-LEVEL        P-LEVEL                      G-LEVEL        P-LEVEL

                     EMPIRICAL                           Area I              Area II                 Area III               Area IV


                     NORMATIVE                           Area V              Area VI                 Area VII          Area VIII

Not surprisingly, a wide array of research topics and of different modeling approaches can be used for
these eight domains of human micro agency. Starting with the traditional problem classes, the following
specific additions and qualifications become necessary.

         AREA IV: Classically, Area IV belongs to the core-domain of human agency, i.e. of micro-
         sociology, micro-economics, micro-political science and the like. With respect to explanatory

      It should be added that a similar table can be constructed for the fields of human macro-agencies, too.

Table 2.1b:          A contemporary morphological space for the fields of macro-agency

                                                             OBSERVER/                           OBSERVER/
                                                             OBSERVER                            OBSERVED
                                                            (self-referential)                  (referential)
                                                            Dimension3                          Dimension3
                                                         G-LEVEL         P-LEVEL             G-LEVEL         P-LEVEL

                     EMPIRICAL                           Area I           Area II            Area III              Area IV


                     NORMATIVE                           Area V           Area VI            Area VII           Area VIII

One should add that the macro-relation between observer/observer simply refers to any type of scientific investigation into the macro-
structures and macro-processes of scientific systems - like the current explorations into the wide world of National Innovation Systems ...

       contexts, complex models of the homo oeconomicus variety have been most efficiently
       utilized within this external area where they serve as useful “paradox detectors” on the limits
       of rationality or on the un-intended consequences of rational decisions.117 Transdisciplinary
       models, focusing on rational decision procedures or on frameworks of evolutionary game
       theory, should and can be employed only very seldom, however, for the explanation of
       observable patterns in communication processes, in interactions, etc.
       AREA III: For these fields, a cognitive revolution on cognitivism or, to use a book-title by
       Michael Gazzaniga, on the cognitive neuro-sciences has occurred (GAZZANIGA 1995)
       whereby the internal neuro-states of individuals become subject to a rapidly increasing variety
       of explorations, based on the cognitive apparatus within persons. Taking a separation from the
       domain of Artificial Life (LANGTON 1989, LANGTON et al. 1992, LANGTON 1994,
       VARELA/BOURGINE 1992), it becomes useful to separate the research-areas, aside from the
       traditional micro-level of human actions and interactions, into two additional domains. On
       the one hand, a basic or fento-area can be distinguished, where -

           tasks like ... wandering, avoiding obstacles, wall following, looking for a certain object,
           delivering some object, cleaning the floor, following someone, etc. (BROOKS 1992:436)                            -

       become the central focus of genotype based investigations. On the other hand, a meso or pico-
       domain (see, e.g. AINSLIE 1992) can be identified whose main research interest lies in the
       problem of task-integration at the neural level. Thus, the following three levels of G-
       investigations can be put forward for this relatively recent area of neural-based social

           First, tasks (Rodney A. Brooks), drafts (Daniel C. Dennett) or agents (Marvin Minsky),
           at the fento-level, task-, draft- or agent-integration at the pico-level and sequences of
           integrated tasks, drafts or agents at the micro-level, especially for senso-cognitive-
           motoric processes like walking, drawing, seeing, speaking, grasping, hearing, reading
           and the like ...
           Second, recursive couplings, especially, but, pace Luhmann (1984, 1990), not
           exclusively communications, at the fento-, pico- and micro-level ...

     Thus, the verve of contemporary criticism against Rational Choice-modeling, like the one from ETZIONI (1994) or, to a lesser extent,
from FRIEDBERG (1995), must be seen as valid with respect to its comparative disadvantages in “actual world”-applications and as highly
invalid with respect to the successes within “possible worlds”-investigations. (On the precise pragmatic notions of the terms “actual
world/possible world utilizations, see Part II of Volume II)
     It must be stressed, however, that in the articles by Rodney Brooks one finds a separation into micro-domains, macro-areas and the
ecological level which, following the terminology introduced here, corresponds to the fento- pico- and micro-distinctions. The new
terminology has been chosen for two reasons. First, Ainslie´s book on pico-economics (1992) has become a well-known social science
standard for problems of conflicts within persons. Second, the micro- macro-dualism is very much entrenched in the current social science
literature and has acquired, by now, relatively clear boundaries.

        Third, disturbances, again on all three levels of investigation, in the implications for the
        neural settings. (For more details, see MÜLLER 1991)

     Thus, the internal linkages and the “fine distinctions” between the “sensorium” and the
     “motorium” play within Area III the most decisive role and stand at the center stage for the
     explanatory frameworks for the interaction patterns of neural groups. (See esp. EDELMAN
     1989, 1992, 1993)

The remaining two empirical domains belong to the discipline of science of science since they focus on
the actions and practices of scientists (Area II) or on the genotype domains of science, namely on
scientific programs and the neuro-settings of scientists (Area I). More precisely, the following research
topics can be identified.

     AREA II: Once again, the same model-types which are at the disposal for Area IV, can be
     applied to the scientific realm, too. Moreover, extremely interesting moves toward self-
     referentiality could be accomplished since modeling frameworks can be used for purposes of
     self-explanations as well. Take, for example, a model which is couched in a master-equation
     scheme and which is specified to capture the cognitive dynamics within a scientific domain or
     discipline (MÜLLER 1994b), then, via a consistent process of self-specification and data-
     collection, an explanatory scheme for the most likely diffusion trajectories of this type of
     model can be built up. Like in the case of Area IVI, homo oeconomicus-variations will play an
     essential role in explorations on the economics of research, on the detection of possible
     critical limits in diffusion processes, on paradoxical results with respect to innovation
     patterns and the like ....
     AREA I: Finally, establishing links between advances in the cognitive sciences (see, e.g.
     HANSON/OLSON 1990, KOCH/SEGEV 1991 or WISE 1987) with an in-depth analysis of
     the neural ensembles of scientists might turn out, in the future, as an extremely valuable
     research road. Again, typical internal approaches like the utilization of genetic algorithms
     (HOLLAND et al. 1986, HOLLAND 1992, KOUZA 1993) or, alternatively, PET-tomography
     and similar experimental routes should become a frequently used research tool for neuro-
     based investigations of science in action (Bruno Latour). Aside from the neural bases, a large
     number of additional genotype approaches fall under Area I, too. For example, the written
     exchanges or controversies between two scientists on a single issue can be studied as an Area
     I problem where the focus of analysis could lie in the identification of the “hidden order” in
     this flow of exchanges by studying, above all, their fine grained surface features. (For an
     interesting set of “surface characteristics”, see especially the next chapter on “embeddedness
     relations” and, above all, HOFSTADTER 1995:489p.).

In a similar manner, the remaining four agency areas at the normative level can be introduced and defined.
At this point, only some exemples will be provided which should be able to show both the applicability
and the usefulenesss of the area-separations undertaken in Table 2.1a.

     The meanwhile widely distributed field of rational decision theory belongs to the core domain
     of Area VIII where questions of optimal strategies, optimization methods, evolutionary stable
     strategies and the like dominate the cognitive “scenes and scripts”. (See e.g.,
     A rationality analysis of “observing scientists” during scientific observations might be
     qualified as an Area VI topic where the observed practices of scientists are brought into close
     contact with some well-defined optimality criteria, with problems of economic or design
     efficiency and the like.
     The neural analysis of rational decision procedures in everyday contexts, if couched in a
     genotype framework on the relations between the development of neural capacities and the
     required changes in the normative action patterns, fall typically into the future Area VII since,
     at the present time, (almost) no research designs can be identified which would be directly
     applicable to this problem..
     Likewise, the problem of design optimality or “scientific rationality” in the course of the
     codified correspondence between two scientists over a specific research problem must be
     qualified as an Area V issue. Moreover, problems of optimal neural group arrangements for
     processes of scientific creativity may become examples of a future Area V study, which, for
     the time being however, is mostly located in science-fiction genres only.

Thus, the concept of agency both of the human and on the non-human variety, can be defined and
investigated both on a multiplicity of phenotype or genotype levels and in a multitude of perspectives
which, not surprisingly, turn out to be largely irreducible.

                               3. THE EMBEDDEDNESS -


Having removed, on the one hand, the main barriers for a comprehensive and transdisciplinary view of
embeddedness relations and having added, on the other hand, a rich variety of agency-contexts for
multiple descriptions of these embeddedness relations, it will become the central target in one of the

theoretical core-chapters of Section III to summarize the main relations between phenotype-level
processes and their genotype-level counterparts. Before entering into the discussion of embeddedness
relations however, another problem area has to be dealt with first, namely the problem of “embeddedness
domains”. Embeddedness domains refer, very generally and largely metaphorical, to “fixed points”,
restrictions, limitations or “constants” in the utilization contexts of embedded code-systems.

     In biological domains, for example, “embeddedness domains” can be safely associated with
     the realm of phenotype selection which restricts and limits the future range of genotype
     variation and recombination. Likewise, the interaction patterns and, more generally, the
     phenotype-phenotype fabric offers an essential “embeddedness domain”, restricting or
     enhancing the future potential for genotype variations.
     With respect to human organisms, the quest for “embeddedness domains” becomes
     complicated for two different reasons.

        On the one hand, one is confronted with a multiplicity of code-systems, a multiplicity of
        integrated code-systems and, thus, with hyper-complex ensembles which clearly
        transcend the established biological embeddeness domains.
        On the other hand, the “embeddedness domains” for code-utilization seem to lie, above
        all, within distributed control and steering units, namely the individual I´s (see also
        HUMPHREY 1995) Thus, “embeddedness domains” between human code-systems are,
        contrary to the biological counterparts, internal in nature ...

Following the “external aspects of internality” within the previous chapters, an external and, thus, a
homogeneous transdisciplinary perspective on embeddedness relations is well within the cognitive
reachability. The subsequent chapter will demonstrate then in a concrete fashion, how a unified view on
both embeddedness domains as well as on embeddedness relations can be upheld.

                         3.1. EMBEDDEDNESS-DOMAINS

One of the most interesting starting-points for the subsequent discussion can be identified by looking at
the intricate ensembles of everyday life-events and instances, associated with the complex phenomenon of
“agency and code-following”, including the mysterious process of codified rule following. As so often in
Volume I, the initial point of departure comes from a quotation from Ludwig Wittgenstein who, also in
the case of rule-following, provides powerful arguments against a causal “internal-external relation”

whereby internal “intentions”, “decisions” and the like for rule following cause a specific outside
behavior, namely the rule-following.

       Denken wir an das Erlebnis des Geführtwerdens! Fragen wir uns: Worin besteht dieses Erlebnis,
       wenn wir z.B. einen Weg geführt werden? - Stelle dir diese Fälle vor:
       Du bist auf einem Spielplatz, etwa mit verbundenen Augen, und wirst von jemand an der Hand
       geleitet, bald links, bald rechts; du mußt immer des Zugs seiner Hand gewärtig sein, auch Acht geben,
       daß du bei einem unerwarteten Zug nicht stolperst.
       Oder aber: du wirst von jemandem an der Hand mit Gewalt geführt, wohin du nicht willst.
       Oder: du wirst im Tanz von einem Partner geführt; du machst dich so rezeptiv wie möglich, um seine
       Absicht zu erraten und dem leisesten Drucke zu folgen.
       Oder jemand führt dich einen Spazierweg; ihr geht im Gespräch; wo immer er geht, gehst du auch.
       Oder: du gehst einen Feldweg entlang, läßt dich von ihm führen ...
       Aber nun merke dies: Während ich mich führen lasse, ist alles ganz einfach, ich merke nichts
       besonderes; aber danach, wenn ich mich frage, was damals geschehen ist, so scheint es etwas
       Unbeschreibbares gewesen zu sein. Danach genügt dir keine Beschreibung ... Aber erinnere ich mich
       denn an etwas anderes? Nein; und doch kommt mir vor, als müsse etwas anderes gewesen sein; und
       zwar dann, wenn ich mir das Wort 'führen', 'Einfluß' und derlei vorsage. 'Denn ich bin doch geführt
       worden', sage ich mir. - Dann erst tritt die Idee jenes ätherischen, ungreifbaren Einflusses auf.
       (WITTGENSTEIN 1971:PI 172 - 175, passim)

Abandoning the causal efficacy of the “internal vocabulary” however, leads to a well-known conundrum,
philosophical and otherwise, with respect to the necessary attribution conditions for rule-follwowing -

       Wie kann mich eine Regel lehren, was ich an dieser Stelle zu tun habe? Was immer ich tue, ist doch
       durch irgend eine Deutung mit der Regel zu vereinbaren. (WITTGENSTEIN 1971:PU 198)

Thus, the main parts in the present chapter will be entirely devoted to the problem of “embeddedness
domains” or, more generally, to the “second-order”-problem of the “embeddedness” of “embeddedness-
relations” ... 119
The first point for a successful answer comes from the neuro-sciences where one has furnished an
astonishing and highly counter-intuitive experimental result, namely the interpretative role of the
Cartesian demon alias “interpreter” (Michael S. Gazzaniga) even in the total absence of any support from

   The corresponding scientific exploration into embeddedness domains can be undertaken for a variety of domains, human and otherwise.
Moreover, due to the multiple agency-principle, even non-trivial machines, following John L. Pollock -

       If the concept of a person is to be such that it can be in principle include creatures with different physiologies, then the
       concept of a person must simply be the concept of a thing having states that can be mapped onto our own in such a way that
       if we suppose the corresponding states to be the same, then the thing is for the most part rational. In other words, something
       is a person iff it has states whose interactions appropriately mimic our rational architecture (POLLOCK 1989:111) -

may and eventually must exhibit embeddedness domains similar to the ones which will be introduced in the course of chapter 3.1.

perceptual data ... To make the last sentence, especially the last passage, more legible, a reference will be
made to the so-called “split brain-research” where patients, suffering from epilepsy, have been undergone
a complete removal of the corpus callosum, i.e. of the nerve fibers connecting the two brain hemispheres.
Experimental research with split brain patients or split brain animals has produced a large number of
astonishing and highly counter-intuitive results. The experimental setting in the specific example to be
discussed assumed the lateralization of pictures to one hemisphere only. More concretely, the left
hemisphere, associated with the right eye, “saw” the picture of a claw whereas the right hemisphere
“recognized” the picture of a snow scene. In turn, the probands were asked to solve two simple selection
problems. They were confronted with a set of additional pictures from which they had to choose the two
ones that fits best to the context they had seen previously. The most appropriate answer for the left brain
scene was a chicken, the most similar response to the right brain-context had been a shovel.

      After the two pictures are flashed to each half-brain, the subjects are required to point to the answers.
      (GAZZANIGA 1985:72)

And the results?

      A typical response is that of P.S., who pointed to the chicken with his right hand and the shovel with
      the left. After his response I asked him, 'Paul, why did you do that?' Paul looked up and without a
      moment's hesitation said from his left hemisphere, 'Oh, that's easy. The chicken claw goes with the
      chicken and you need a shovel to clean out the chicken shed.'
      Here was the left half-brain having to explain why the left hand was pointing to a shovel when the
      only picture it saw was a claw. The left brain is not privy to what the right brain saw because of the
      brain disconnection. Yet the patient's very own body was doing something. Why was it doing that?
      Why was the left hand pointing to the shovel? The left brain's cognitive system needed a theory and
      instantly supplied one that made sense given the information it had on this particular task. It is hard to
      describe the spell-binding power of such things. Manipulating mind variables is awesome. (IBID.)

The special point of this experiment lies in its peculiar variation of an “other minds” problem within one
and the same person. Visual inputs which are processed within an isolated half-brain without language
competencies and a corresponding selection-behavior lead, quite naturally, to an ad-hoc interpretation of
this action on part of the other half-brain with language competencies ... It can be asked whether this
special configuration of an isolated “bicameral mind” (Julian Jaynes) can be generalized into a basic
vision of a modular or a multi-cameral mind in which language operations become entangled in a far
more “superfluous” and “interpretative” manner.

      We're always making up stories ... and relating only the best candidate - which is, if things are going
      well, the true story. The patients who confabulate aren't lying, in the usual sense of the word. They

       probably relate the best story that they were able to construct from the data available to them and
       think it true ... It shows you that our memories are not like tape recordings which keep the events in an
       immutable sequence; we remember the elements by recognition, and recall the sequence by ties
       between these elements. But each recognized element has usually been seen before, and combinations
       have likely occurred in various orders ... (CALVIN 1990:53p.)

The above quotation should not seen as ammunition for the “unreliability” of memories, human or
otherwise. On the contrary, the whole argument by Calvin can be interpreted as support for far-reaching
revisions and revulsions in the overall perspective for phenomena like memory, meaning, understanding
or, in general, embeddedness domains. A similar point has been raised a long time ago by Ross W. Ashby
or Heinz von Foerster already, since, in their view, memory should be interpreted as -

       the irreducible uncertainty of an observer with incomplete knowledge of the present internal state of a
       non-trivial machine (say, a living organism), which the observer interprets as a property of the
       machine. (FOERSTER 1995:312)120

The upshot of the Foerster-quotation lies, similar to Calvin´s emphasis on “story telling” in the demand
for a radical revision with respect to the status of internal performances and internal mechanisms like
“memory”, “decision making” and the like ...
Second, the “intentional stance” must be considered, following once more Daniel C. Dennett, as a highly
successful way of external ascriptions when evaluated in terms of its predictive or retrodictive surplus
value. Thus, the “intentional vocabulary” will in all probability retain its everyday status in a strikingly
similar manner as, say, the Ptolemean world-view which despite being overcome and replaced within
astronomy, is still widely used in everyday contexts - metaphors of the sun going down or the bad moon
rising are still strongly entangled in everyday descriptions. In a similar manner, other “internal stances”
like the emotional, normative, learning or decision making ones, will retain a prominent role in everyday
communications, including the scientific ones.
Third, what has to be searched for is, then, the legitimate successor for the lost internal “embeddedness
domains”. If the internal processes within subjects and their “maufacturing of meaning” cannot be
considered as a sufficiently reliable basis for embeddedness domains, then where is one to find the
necessary substitutes and alternatives? The answer to this question leads from the worlds “within” to the
worlds “out there”; to well-structured and organized socio-economic ensembles - to specific routines in
small social groups, to “language games” in institutions, to behavior patterns in large societies, in sum, to
different “forms of life” (Ludwig Wittgenstein).121

    For a similar point of view, see also WATZLAWICK (1992:17).
    “Forms of life” can be conceptualized at different P-levels within a specific species (or in a more subtle understanding at different G-
levels, too). Thus, “form of life” must not be equated with just the sum total of a gemeinsame(n) menschliche(n) Handlungsweise.
(WITTGENSTEIN 1971:PU 206)..

Following the Wittgenstein-Kripke solution to the problem of rule-attribution, one finds three basic
“embeddedness domains” as will become clear from the subsequent quotation.122

       First, the entire 'game' we have described - that the community attributes a concept to an individual so
       long as he exhibits sufficient conformity, under test circumstances, to the behavior of the community -
       would lose its point outside a community that agrees in its practices ...
       The set of responses in which we agree, and the way they interweave with our activities, is our form
       of life ...
       Finally, criteria ... Wittgenstein's skeptical solution to his problem depends on agreement, and
       checkability - on one person's ability to test whether another uses a term as he does. In our own form
       of life, how does this agreement come about? In the case of a term like 'table', the situation, at least in
       elementary cases, is simple. A child who says 'table' or 'That's a table' when adults see a table in the
       area (and does not do so otherwise) is said to have mastered the term 'table': he says 'That's a table',
       based on his observation, in agreement with the usage of adults, based on their observation. That is,
       they say, 'That's a table' under like circumstances, and confirm the correctness of the child's
       utterances. (KRIPKE 1985:96pp.)

Aside from these three areas - {community specific consensus, forms of life, criteria} - one can
immediately find another set of embeddedness domains being linked to the formal as well as the informal
infrastructure of socio-economic ensembles:

       First, one has to add standardized forms of attributions and exchanges, including, for
       example, money, measurements, social rules, norms or, highly interesting, constants
       especially in the natural sciences which have found their way into constitutions, dictionaries,
       laws, encyclopedias, thesauri, weight and measurement-tables, style manuals, scientific
       textbooks, law codices ... These areas fall clearly under embeddedness domains since they
       serve as fixed points for a tremendous amount of practices in economics, everyday life or in
       the scientific arena. Moreover, with standardized exchanges, measures, weights and units like
       ampere, ohm, calorie, british thermal unit (BTU), volt, newton, decibel, astronomical unit,
       carat, gallon, coulomb, becquerel, candela, kelvin ... one will be confronted with the peculiar
       phenomenon that only a very small part of the population will be able to reproduce the exact
       dimensions of say, a coulomb (6.23 x 1018 electrons) or an ampere (a unit of electric current
       equivalent to a flow of one coulomb per second).

       When one says that to imagine a language is to imagine a form of life ..., it is included in and implied by this statement that
       there are a number of forms of life and not just one. And just as surely this does not mean the cow-like, fish-like, dog-like
       and so on, but rather other human behaviour, other societies, real or imagined. (HALLER (1988:134).
    The most general indeterminacy “behind” these remarks may be referred to as “indeterminacy of attribution” whereby, inm principal, no
“fact of the matter” domains can be identified even for the most simple words like “table”, “blue” etc. Here, too, the totality of action
patterns may lead consistently to radically different forms of “attributions” ...

          Second, institutions form another important segment of embeddedness domains. Here,
          institutions can be loosely defined as spatio-temporal systems, having their spatio-temporal
          exits and entrances, with a common setting of shared and enforced encoded rules.123 Thus,
          schools count as a paradigmatic example for a societal institution in which common standards
          and goals, distributed throughout a national school-system, are to be achieved, independent
          and irrespective of the peculiar fact that each school is composed of different pupils, different
          teachers or a different administrative personnel ...
          Third, rituals, defined here as the combination of constant or fixed code-sequences or neuro-
          patterns (genotype) and specific action patterns (phenotype), can be named as another vital
          “fixed-point” ingredient in societal settings. Rituals are, under normal circumstances, linked
          to code-systems in manifold ways like in the case of greeting, entering a room (knocking from
          without, response from within, entering the room, greeting ...) or, to restrict the list of
          examples to just three, the extremely interesting code and action system of dancing. (For a
          short introduction on dance-codes, see TUFTE 1990:114p.) In this manner, rituals form an
          essential part of the embeddedness domains or, alternatively, of the “cement of society”.
          (ELSTER 1989)
          Fourth, another important component of embeddedness domains lies in the field of “implicit”
          or “tacit knowledge”, resulting from the slightly mysterious, though unquestionably very
          wide-spread configuration that, following Michael Polanyi, we know consistently more than
          we can say we know.124

Aside from the repertoire for the embeddednes-domains listed above, an extremely important addition has
to be made, namely the complexity of the description potential with respect to code-utilization contexts.
Take, as a prime example, “speech-acts” or “language games” in everyday-life configurations, then one
finds an astonishing rich field of investigations which have rarely been touched upon so far within the
social sciences. Take the following listing from Douglas R. Hofstadter on “surface-tests” in the speech

      For a similar definition, see SJÖSTRAND (1995:28) who defines institutions as -

          a social construct for a coherent system of shared and enforced norms. (IBID).
      In Polanyi´s words, the slightly paradoxical situation lies in the discrpancy -

          daß wir mehr wissen als wir zu sagen wissen. (POLANYI 1985:14)

In a more elaborate version, Polanyi gives a list of highly interesting experimental findings on the scope and the dimension of “implicit” or
“tacit knowledge”:

          Einige neuere psychologische Experimente haben unabhängig voneinander den grundlegenden Mechanismus aufgezeigt,
          mittels dessen Wissen 'implizit' erworben wird ... Die Autoren zeigten einer Versuchsperson eine große Zahl sinnloser
          Silben, wobei auf das Erscheinen einiger davon ein elektrischer Schlag erfolgte. Bald zeigte die Person Symptome der
          Antizipation des Stromstoßes beim Anblick der 'Schocksilben'; auf Befragen vermochte sie diese Silben gleichwohl nicht
          anzugeben ... Eine andere Variante dieses Phänomens wurde 1958 von Eriksen und Kuethe nachgewiesen. Sie setzten eine
          Versuchsperson einem Stromschlag aus, wann immer sie zufällig Assoziationen zu bestimmten Schockwörtern äußerte. Bald
          lernte die Person, die Äußerung solcher Assoziationen zu vermeiden, um dem Stromstoß zu entgehen, wußte jedoch nicht -
          wie sich auf Befragen herausstellte -, daß sie es tat. (IBID:16p.)

production alone and on the potential for detecting underlying “deep-structures” or “generative
mechanisms” -

       looking at word frequencies (e.g., ... is „the‟ the most common word? ...are some low-frequency words
       used unnaturally often? ...);
       observing sensitivity to tone (e.g., are formal and slang expressions in the input text understood? is
       humor based on improper mixtures of tone understood? ...);
       examining types of errors (e.g., misspellings, transposition errors, improperly used words or phrases,
       blends of all sorts ...);
       examining word flavors as a function of subtle details of the context (e.g., what contextual pressures
       lead to choosing „jock‟ over „athlete‟, or vice versa? to saying „lady‟ as opposed to „woman‟?...);
       examining level of abstraction of word choices (e.g., what pressures lead to choosing between „Fido‟,
       „the dog‟ and „some mammal‟? ...);
       looking at default assumptions regarding gender (e.g., what kinds of circumstances lead to generation
       of agent nouns with feminine endings, such as „heroine‟, „millionairess‟, or „farmette‟? ...);
       observing how throwaway analogies are understood and generated (e.g., is the abstraction hidden in
       remarks ... interpreted correctly and instantly? are such remarks produced in the standard contexts that
       would call for them?);
       observing how throwaway counterfactuals are understood and generated (e.g., is the subtle blend
       implicit in remarks such as „I wouldn´t have felt that way if I‟d been my father‟ ... interpreted
       correctly and instantly? ...);
       paying attention to timing data (the speed taken to generate the output can be used to make some
       inferences in the mechanisms behind the scenes) (HOFSTADTER 1995:489f.)125

   In this manner, another set for “surface features” can be identified in the area of body movements. Here, a variation of the Hofstadter-
quotation leads to research tasks like -

       looking at frequencies of movements (e.g., ... slight movements of the head towards another person or away from someone?
       ...are some unusual finger-movements recorded unnaturally often? ...);
       observing sensitivity to tone (e.g., are formal and slang expressions in the input text understood? is humor based on improper
       mixtures of tone understood? ...);
       examining types of divergencies between speech and body movements (e.g., agreeing on a proposal completely, while slightly
       shaking one´s head, turning away from one´s opponent, etc.);
       examining movement flavors as a function of subtle details of the context (e.g., what contextual pressures lead to choosing a
       specific change in one´s sitting position compared to the previous configuration? ...);
       examining level of complexity of movement choices (e.g., what pressures lead to choosing a complex movement, consisting of
       a coordinated effort between head, finger, arm and leg-movements instead of a simple change in the left hand only? ...);
       looking at specific movements regarding gender (e.g., what kinds of gender-specific circumstances lead to certain types of
       movement patterns in case of, say, a male or a female person entering a room? ...);
       observing how throwaway movements are understood and generated (e.g., is the “message” hidden in someone´s body
       movements ... interpreted correctly and instantly? are such movements produced in the standard contexts that would call for
       them? ...);
       paying attention to timing body movements (the speed taken in lifting one´s arm to point into a specific direction can be used
       to make some inferences on the interaction pattern between motoric and language performances behind the scenes)

More generally, small and even tiny deviations in the flow of speech, unspectacular, though characteristic
changes in word orders, small movements of a finger, a specific position while sitting in a chair ... qualify
as embeddedness domains too, since in all these instances an explanatory framework can be built up,
leading to speech processing programs which produce small and even tiny deviations in the flow of
speech, unspectacular, though characteristic changes in word orders ...
Fourth, with the inclusion of rarely explored external code-behavior patterns126, the main embeddedness-
domains for code-systems have been identified. Starting with the Wittgenstein-Kripke set of -

                              {Community specific consensus, forms of life, criteria} -

continuing with the second class, consisting of -

                           {Standardizations, institutions, rituals, “implicit knowledge”}

and finishing with the Hofstadter group of -

                         {observable surface features in the contexts of code-utilizations}

three principal embeddedness-domains have been found beyond the cognitive operations within persons.
At this point, it becomes tempting to ask whether these domains can be ordered in a more stringent and
coherent manner - and whether these three groups can be seen as a comprehensive enumeration of
embeddedness domains.

          First, starting with embeddedness domains like “criteria” and “tacit knowledge”, one
          arrives at a basic distinction between consensual code-based practices from consensual
          code free routines within spatio-temporal societal ensembles. Thus, practices of “implicit
          knowledge”can be clearly separated from the case of “criteria” for code-components where,
          by necessity, code-based routines are involved.
          A second dimension, under the heading of institiutionalization, refers to those societal
          fixed points where standards, weights, measures, constatnts have become common practice
          and where, additionally, institutions or, alternatively, organizations, distributed over a large
          spatio-temporal field, have been set in proper operation.

These two dimensions, combined with the necessary third dimension on the position of the observer, lead
to the subsequent table. (See Table 3.1, next page)
Table 3.1: A Morphological Space for Embeddedness Domains

   One of the rare examples in the social sciences lies in the field, called “objective hermenutics”. (OEVERMANN et al 1979, 1983a,

                                    OBSERVER/                            OBSERVER/
                                    OBSERVER                             OBSERVED
                                    (self-referential)                   (referential)

                                   DIMENSION3                           DIMENSION3
                                 G-LEVEL   P-LEVEL                    G-LEVEL   P-LEVEL
                                 (CODES)   (CODE-FREE)                (CODES)   (CODE-FREE)

      INSTITUTIONAL              Criteria,       “Implicit            Criteria,      “Implicit
                                 Rules,          Knowledge”,          Rules,         Knowledge”,
                                 Roles,          Rituals,             Roles,         Rituals,
                                 Goals,          Consensual           Goals,         Consensual
                                 Surface-        Routines,            Surface        Routines,
                                 Features,       etc.                 Features,         etc.
                                 Rituals,                             Rituals,
                                 Standardi-                           Standardi-
                                 zation, etc.                         zation, etc.


                                 Criteria,       “Implicit            Criteria,      “Implicit
                                 Rules,          Knowledge”,          Rules,         Knowledge”,
                                 Surface,        Rituals,             Surface        Rituals,
                                 Features,       Consensual           Features,      Consensual
      NON-INSTITUTIONAL          etc.            Routines, etc.       etc.           Routines, etc.

Fifth, the final problem consists in drawing the limits and the boundaries of embeddedness domains,
especially in the case of rule-systems, of intentional attributions, of emotional ascriptions and the like.
The reliance on societal “fixed points” and standardizations relative to specific communities leads, finally,
very directly to the question whether a solitary islander of the Robinson Crusoe variety -

     cannot be said to follow any rules, no matter what he does? I do not see that this follows. What does
     follow is that if we think of Crusoe as following rules, we are taking him into our community and
     applying our criteria for rule following to him. (KRIPKE 1985:110)

Whether attributions from an “attributional vocabulary” of the “internal variety” should be applied to
domains outside human communities, no a priori answers can be provided, neither with respect to the
Robinson Crusoe-problem nor with respect to animal behavior or in relation to machine activities
(NEUMAIER 1987). All that can be shown from the current {E,C,S} point of view, are the shifting
boundaries for “internal” ascriptions which, over the last decades, have seen a remarkable extension
towards higher organized animals and even plants. In these at times highly intriguing instances
(CHENEY/SEYFARTH 1990, GRIFFIN 1992, MORTON/PAGE 1992, SERPELL 1986), a separate
decision has to be made on the inner- or the outer limits of rule following, intentional attributions,
emotional ascriptions, or, more generally, of the adequacy or the inadequacy of the so-called “internal
stances” ...

                                 3.2. BUILDING BLOCKS

                                      FOR EXCHANGES

                                 WITHIN AND BETWEEN


With the next step, an integration between several well-known, but unfortunately separated domains will
turn out to produce far-reaching consequences. The prime interest will be focused on the basic
components and elements in the flows and exchanges between and within the dual levels of National
Innovation Systems. In the end, five different types of building blocks will be identified which make up
the “flows of knowledge and information”.

     The first domain for exchanges within or across levels lies, quite naturally, in the embedded
     code-systems themselves, including social rule systems. Thus, code-elements, strings, and,
     above all, programs and program sequences within specific genres can be exchanged and
     transmitted, at least in principle between and across levels.

     For the second area, a quotation from a widely acclaimed book on a radical version of the
     “genetic survival of the fittest” biology will serve as an introductory step -

        Wir (müssen) uns zum Verständnis der Evolution des modernen Menschen zunächst davon
        freimachen, das Gen als die einzige Grundlage unserer Vorstellungen von Evolution anzusehen
        ... Das neue Urmeer ist die „Suppe‟ der menschlichen Kultur. Wir brauchen einen Namen für
        den neuen Replikator, ein Substantiv, das die Assoziation einer Einheit der kulturellen
        Vererbung vermittelt, oder eine Einheit der Imitation ... das Mem ... Beispeiele eiens Mems
        sind Melodien, Gedanken, Schlagworte, Kleidermode, die Art, Töpfe zu machen oder Bögen zu
        bauen. So wie Gene sich im Genpool vermehren, indem sie sich mit Hilfe von Spermien oder
        Eiern von Körper zu Körper fortbewegen, so verbreiten sich Meme im Mempool, indem sie von
        Gehirn zu Gehirn überspringen mit Hilfe eines Prozesses, den man als Imitation bezeichnen
        kann.(DAWKINS 1978:225pp.)

     In a similar spirit, Douglas R. Hofstadter sees, consistent with the multiple agency-principle
     of the second chapter, an overall “war” or, less belligerent, a highly competitive game for
     resources at the meme-level -

        As a library is an organized collection of books, so a memory is an organized collection of
        memes. And the soup in which memes grow and flourish - the analogue to the „primordial soup‟
        out of which life first oozed - is the soup of human culture ... There need not be an exact copy
        of each meme, written in some universal memetic code, in each person´s brain. Memes, like
        genes, are susceptible to variation or distortion - the analogue to mutation. Various mutations of
        a meme will have to compete with each other, as well with other memes, for attention - which is
        to say, for brain resources in terms of both space and time devoted to that meme. Not only must
        memes compete for inner resources, but, since they are transmissible visually and aurally, they
        must compete for radio and television time, billboard space, newspaper and magazine column-
        inches, and library shelf-space. Furthermore, some memes will tend to discredit others, while
        some groups of memes will tend to be internally self-reinforcing. (HOFSTADTER 1985:51p.)

While the preceding discussion on evolution, code-systems, embeddedness domains and, finally,
embeddedness relations has established a far more comprehensive framework, the meme-concept refers to
an interesting level relation of the format -

                                             memes (high level)
                                            programs (basic level)

as well as to the diffusion potential of memes within a group, a community, a national or a global society.
Moreover, growth processes of the meme pool are, formally speaking, strictly independent from the
development patterns of the program ensembles.

       First, memes are, quite naturally, high level programs themselves which are loosely linked to
       programs within well-defined scientific genres. Take, as reference case, the social science
       programs which have been discussed in greater detail within the chapter on scientific
       creativity in Section II, then one is confronted with a peculiar phenomenon. Using Popper´s
       “Logic of Scientific Discovery” as prime example, one can find, at the level of “catch
       phrases”, memes of the format “hypothetical-deductive method”, “falsifiability”, “relativity of
       basic statements”, “degrees of testability” or “corroboration”. For a proper scientific discourse
       in disciplines outside the philosophy of science camps, one will need more comprehensive
       memes like {“falsifiability”, “A theory is falsified only if we have accepted basic statements
       which contradict it” (POPPER1968:86)} and the like. Inside the philosophy of science fields,
       it is still advisable to make use of the entire program, discussing and picking out also un-
       familiar aspects of Popper´s seminal book.127
       Second, memes and meme-diffusions may follow a “logic of their own” independent of and
       even contrary to the original program-contexts. Thus, one may find instances where the
       program part which is linked to a particular meme exhibits clearly a different context meaning
       than the subsequent meme-utilizations. Using a typical example form the philosophy of
       science, the prominent Feyerabend-meme of “Anything goes” appeared originally as an
       argument against the universal applicability of rationality rules in science -

           Es ist ... klar, daß der Gedanke einer festgelegten Methode oder einer feststehenden Theorie der
           Vernünftigkeit auf einer allzu naiven Anschauung vom Menschen und seinen sozialen
           Verhältnissen beruht. Wer sich dem reichen, von der Geschichte gelieferten Material zuwendet
           und es nicht darauf abgesehen hat, es zu verdünnen, um seine niedrigen Instinkte zu
           befriedigen, nämlich die Sucht nach geistiger Sicherheit in Form von Klarheit, Präzision,
           „Objektivität‟, „Wahrheit‟, der wird einsehen, daß es nur einen Grundsatz gibt, der sich unter
           allen Umständen und in allen Stadien der menschlichen Entwicklung vertreten läßt. Es ist der
           Grundsatz: „Anything goes.‟ (FEYERABEND 1993:31p.)

       Nowadays, the “Anything goes”-meme has been linked to postmodernist memes of games,
       recombinations and self-reflexivity, leading to an entirely different network of associative
       domains. In this manner, one will find an astonishingly large number of meme-exchanges

    It should be mentioned however, that comprehensive analyses on the reception patterns of literary or philosophical texts have revealed an
astonishing tendency towards “dominant memes” or, alternatively, “dominant program quotations” which are repeated ad nauseam whereas
most other parts and aspects of literary or philosophical texts remain outside the sphere of reproduction. For a similar focus, see e.g.
SCHMIDT (1987:63pp.), RUSCH 1987.

      which, in a strong sense of the word, have become independent of their original program
      contexts - and which are (re)produced within an entirely new semantic environment ...
      Third, “implicit knowledge” domains as practices free from the codification process, though
      they can be imitated and learned, should not be considered as memes or as components of the
      meme-pool simply because the do not correspond to the central “genotype logic” which has
      occupied, so far, the core position in the specification of dual level systems. Thus,
      (re)productive ensembles of the


      variety alone, while vital and essential in the development of National Innovation Systems,
      will not count as a proper subset in the meme-pools. They, the meme-pools, remain dependent
      on code-systems and programs generated within natural languages, numbers, symbols,
      musical notations and the like ...
      Fourth, since the exchanges and imitation processes at the level of non-codified routines -


      do play an extremely important funcion in the case of scientific (re)production as well, the
      genotype counterpart or, alternatively, the genotype basis lie, as ill be recalled from chapter
      2.4 within the present section, in “neural groups” or, borrowing Marvon Minsky´s term, in
      “agents” (MINSKY 1989).

         As a library is an organized collection of books, so societal practices, not encoded in forms of
         rules, are, at the genotype level, an organized collection of “neural groups” or “agents”. And
         the soup in which agents grow and flourish - the analogue to the „primordial soup‟ out of which
         life first oozed - is the “wet mind” of the human brain ... There need not be an exact prototype
         of each agent in each person´s senso-cognitive-motoric setting. Agents, like genes or memes,
         are susceptible to variation or distortion - the analogue to mutation. Various mutations of an
         agent will have to compete with each other, as well with other agents, for utilization - which is
         to say, for senso-cognitive-motoric resources in terms of both space and time devoted to that
         agent. Not only must agents compete for inner resources, but, since they are transmissible
         visually at the phenotype level, they must compete for sufficient phenotype diffusion.
         Furthermore, some agents will tend to discredit others, while some groups of agents will tend to
         be internally self-reinforcing.

So far, three different types of genotype-based building blocks for exchanges and transmission have been
identified. The first one, consisting of programs, is situated strictly at the level of embedded code-systems

and their different genres, the second one, composed of memes, is again code-based, but far better suited
and adapted for language games at the phenotype-level like discussions, story telling, lectures, chats and
the like. The third type of building blocks is encoded in the “language of the brain” and gets transmitted at
the phenotype level of imitations and learning only ...128
To complete the summary of essential building blocks, two different groups, this time, not surprisingly, at
the phenotype levels, will be identified. For the first set, a P-analogy to the neural building block of
“agents” will be set up and will be labeled, following in the line of genes and memes, as “acteme”, i.e. as
a phenotype action element, suited for transmission and exchange.

           As a library is an organized collection of books, so the obeservable behaviors, encoded in forms
           of rules or not, are, at the phenotype level, an organized collection of action sequences or
           “actemes”. And the soup in which agents grow and flourish - the analogue to the „primordial
           soup‟ out of which life first oozed - are the settings and environments of human practices ...
           There need not be an exact prototype of each acteme in each person´s action-repertoire.
           Actemes, like genes or memes, are susceptible to variation or distortion - the analogue to
           mutation. Various mutations of an acteme will have to compete with each other, as well with
           other actemes, for utilization - which is to say, for routines in terms of both space and time
           devoted to that acteme. Not only must actemes compete for resources within a person, but,
           since they are transmissible visually at the phenotype level, they must compete for sufficient
           phenotype diffusion. Furthermore, some actemes will tend to discredit others, while some
           groups of actemes will tend to be internally self-reinforcing.

Finally, a P-analogon to the G-level building block of “memes” will be introduced by referring to higher
order “chunks” of actemes - or “Eigen-behaviors” for short. Following Heinz von Foerster (1984), one
may define “eigen-behaviors” as a recursively organized pattern of activities which, dependent on initial
conditions, lead to a specific “attractor state” or, alternatively, to an “Eigen-wert”. Thus, (almost) all of
the language games are organized in a recursive manner, having initial or entrance stages, a set of
recursive intermediate steps and stabilized final states, “basins of attractions” or “Eigen-werte”. (For a list
of examples, see MÜLLER 1991)

       Was ich hier - um die Entsprechung mit dem Formalismus zu unterstreichen - 'Eigenverhalten' nenne,
       ist offensichtlich eine Verhaltenskompetenz ... Ich schlage vor, die Verhaltenskompetenzen mit dem
       Namen von Gegenständen zu bezeichnen, auf die sie sich beziehen. (FOERSTER 1984:17)

Thus, throwing a ball, searching for an object, discussing, shopping, dancing, writing a research report,
eating, swimming etc. are all recursively organized clusters of actemes which, as can be easily recognized,
are transmitted and exchanged en bloc.
   It should be added that the three different building blocks, specified so far - programs, memes, agents - do not exhaust the ralm of
possible elements which includes, inter alia, discourses and discourse components as well ....

       As a library is an organized collection of books, so societal practices, encoded in forms of rules
       or not, are, at the phenotype level, an organized collection of clusters of actemes or eigen-
       behaviors. And the soup in which eigen-behaviors grow and flourish - the analogue to the
       „primordial soup‟ out of which life first oozed - are the settings and environments of human
       practices ... There need not be an exact prototype of each eigen-behavior in each person´s
       routine-repertoire. Eigen-behaviors, like genes or memes, are susceptible to variation or
       distortion - the analogue to mutation. Various mutations of an eigen-behavior will have to
       compete with each other, as well with other eigen-behaviors, for utilization - which is to say,
       for routines in terms of both space and time devoted to that eigen-behavior. Not only must
       eigen-behaviors compete for resources within a person, but, since they are transmissible
       visually at the phenotype level, they must compete for sufficient phenotype diffusion.
       Furthermore, some eigen-behaviors will tend to discredit others, while some groups of eigen-
       behaviors will tend to be internally self-reinforcing.

With the inclusion of eigen-behaviors, the set of elementary building blocks for exchanges and
transmissions can be summarized within Table 3.2 (next page).

Table 3.2: Building Blocks for Exchanges and Transmissions

                                               Exchange and Transmission of Actemes
                                               and “Eigen-Behaviors”

                                 P                                                     P
       Exchange and                                                                            Exchange and
       Transmission of                                                                         Transmission of

       Memes,                                                                                Programs,
       Programs,                                                                               Memes

                                 G                                                     G

                                               Exchange and Transmission of Programs,
                                               Memes, Agents

In this manner, an important step toward a comprehensive analysis of the “embeddedness relations” has
been undertaken which offers both a high variety and potentially dense flows of exchanges and
transmissions across and within the dual levels of National Innovation Systems.

                                                  3.3. THE FOUR



To start with, embeddedness-relations, very generally speaking, fall into at least four different classes,
following different aspects of genotype-phenotype combinations. The first important point to emphasize
lies in the fact that all four exchange pathways play a vital role in the case of National Innovation
Systems. Four paradigmatic examples from the field of National Innovation Systems should make these
four directions more transparent.

       First, a National Innovation System can be studied with respect to the role and function,
       scientific programs play in the constitution and maintenance of production processes at the
       level of firms, the output innovations of economic sectors or the withinput innovations of
       economic clusters
       Second, the output or, alternatively, the programs of the scientific system can be analyzed also
       with respect to their intra- or inter-systemic diffusion which helps, then, to distinguish
       between programs with a high “distribution power” or alternatively, to use a term from Niklas
       Luhmann, a high “interpenetration power” from programs with low distribution power,
       restricted to very small circles of the scientific knowledge basis only.
       Third, an important part in the investigation of programs results from their encoding, or,
       alternatively, their production processes where scientists or technicians arrange and compose
       code-elements into strings and, at the end, into programs within established scientific
       Fourth, a National Innovation System can be studied with respect to the diffusion and the
       adaptations of its non-codified practices, its local adaptation and modification processes or its
       “informal” organizational changes alone without the support of encoded genotype programs.

   For a fascinating account of the many facets in the “construction” and “manufacturing” of a scientific article, see KNORR-CETINA

Each one of the four examples above exhibit one particular type of embeddedness relation. Thus, formally
the four main types of embeddedness-relations can be summarized, using, once again, the essential dual
level framework, as -


In other words, the four embeddedness relations in the {ECS} framework comprise, on the one hand, all
possible relations between the genotype and the phenotype levels and, on the other hand, all “building
blocks” for exchanges and transmissions, outlined in the previous chapter.

Table 3.3: The Four Embeddedness Relations

                                            Recombination of Actemes
                                            and “Eigen-Behaviors”
                                            (EMBEDDEDNESS RELATION IV)

                                P                                                P

        UTILIZATION of                                       ACTIVATION of              ENCODING of

        Memes,                                             Actemes and              Programs,
        Programs,                                            Agents                    Memes
        (EMBEDDEDNESS                                        (EMBEDDEDNESS              (EMBEDDEDNESS
        RELATION II)                                         RELATION IV)               RELATION III)
                                G                                                G

                                            Recombination of Programs,
                                            Memes, Agents
                                            (EMBEDDEDNESS RELATION I)

Some important additions and qualifications become immediately necessary in order to arrive at a fruitful
metatheoretical or, alternatively, transdisciplinary understanding of “embeddedness relations” in genetics
and in socio-economic ensembles.

G  G: This type of transfers comprises all those change in the genotype-ensembles which
are brought about by activities at the genotype levels or by disturbances from outside. While
predominantely closed recursive operations of the G  G type belong to the normal working
conditions in the case of the genetic code, it seems difficult at first to identify processes of this
type in societal embedded code systems, since most changes in the G-domain take the form of
P  G. Nevertheless, outside disturbances like the physical destruction of libraries,
mechanical defects in the storage of computers or computer networks, viruses in the case of
communication and information technologies and the like can exert a powerful influence on
the G  G transformation. In short, the natural “depreciation rate” for societal programs,
processes of physical destruction of programs and, more generally, the “physics of
information” (ZUREK 1990) must be considered as the most essential transformations for this
particular Genotype  Genotype-domain.
G  P: The second flow within the fourfold embeddedness-relations comes through
(re)production processes proper and comprises, in the case of National Innovation Systems,
the (re)productive “generative” impact of scientific programs for the science system and
especially impoprtant, for other societal systems, especially, but not exclusively, the economic
ensemble. This type of (re)productive relation has been developed in its most advanced forms
within the genetic code system where a zygote, the product of parental recombination,
produces, by rapid cell differentiations and morphogenesis, a new organism - the “wonder of

   If the swarm of stars in a spiral galaxy, clustered swirling in the high blackness of space,
   astonishes us with the wonder of the order generated by mutually gravitating masses, think with
   equal wonder at our own ontogeny. How in the world can a single cell, merely some tens of
   thousands of kinds of molecules locked in one another´s embrace, know how to create the
   intricacies of a human infant? No one knows. If Homo habilis wondered, if Cro-Magnon
   wondered how they came to be, so too must we. (KAUFFMAN 1995:93)

P  G: The third type of embeddedness-relations plays, following the “central dogma of
genetics”, virtually no role in the case of the genetic code. In societal systems however, this
area has been and will be of tantamount importance since it comprises both the creation of
new as well as the adaptation of already available programs. Moreover, it marks one of the
most dramatic differentia specifica between humans and non-humans -

   Andere Lebewesen können vieles besser als Menschen: Manche können besser sehen, hören
   und riechen, andere schneller laufen und kräftiger zubeißen ... Aber der Mensch verfügt über
   Wissen. Damit hat er die Erde erobert. Der Rest des Universums erwartet sein Kommen, so
   vermute ich, mit einiger Beklommenheit. (v. DOREN 1996:11)

     P  P: In biology, P  P enters the (re)production game especially in those instances where
     one can identify the reproduction of competencies and performances not explicitly coded for
     within the genotype-level. Thus, characteristic operations of a particular group of animals like
     specific songs of birds which are reproduced generation after generation fall clearly under the
     fourth type of embeddedness relations. In societal systems, one can observe a huge amount of
     exclusively P-centered transfers. In a first approximation, the domains of “implicit” or “tacit”
     operations, while hardly qualifiable as “knowledge” in one of its standard instantiations
     identified in Section II, fulfills the P  P requirement in a paradigmatic manner. Following
     the standard account of “implicit knowledge” like the one given by Michael Polanyi, one is
     confronted with -

        first, an action pattern, exhibiting competent and consistent selection procedures on the
        part of actors ...
        second, a corresponding gap at the genotype-level since the selection procedures do not
        possess codified rules or other genotype programs, reproducible by the competent actors
        or other persons ...
        third, an imitation and learning potential for the specific action pattern under
        consideration whereby new actors are able to acquire or, alternatively, to learn the
        specific action pattern.

     Thus, an interesting paradox will conclude the discussion of P  P embeddedness-relations.
     And this paradox states that any explicit example for “implicit knowledge”, since it must be
     undertaken at the G-level, destroys, by necessity, the


     character of the overall transfer process ....

So far, the embeddedness relations have been defined irrespective of specific systemic considerations.
What follows next, will be some further useful definitions and classifications for intra- as well as inter-
systemic embeddedness relations.

                                  3.4. INTER-SYSTEMIC


Following the building-block-constitution of exchange processes as well as the four directions for
embeddedness relations, the next step will present a very brief, but informative summary on exchanges
within complex socio-economic ensembles. Using one of the main results of Section IV already in the
present section, namely the multi-systemic configuration of societal systems, encompassing at least four
different types of sytemic groups, then one can build up at least four important intra- and inter-systemic
exchange relations. To simplify matters, these four types will be specified for one of the embeddedness
relations only, namely for the productive/reproductive case of G  P. In a similar manner, the four
intersystemic exchanges could be specified, then, for the remaining three cases, too.

     ONEiONEj-RELATION: Under this group, one may subsume (re)productive programs which
     have been produced within a specific socio-economic system i and which, moreover, play an
     indispensable role in the (re)production of a specific system in question, either of i or of
     ananother system j. In the case of National Innovation Systems, a scienific organization
     scheme for the restructuring of institutes may be considered as a classical case for a intra-
     systemic one-one relation, whereas a scientific blueprint for a particular machine may be
     qualified as a paradigmatic instance for a one-one-exchange between science and economics.
     ONEMANY-RELATIONS: Not surprisingly, this field encompasses the problem of program-
     diffusion among various systems. Restricted to a NIS-context only, one may think of new
     organizational schemes like “lean” arrangements, then one is immediately confronted with a
     wide diffusion potential, ranging from lean reconfigurations within the scientific system itself
     to “lean management-types” in the economic sphere or to “lean public administrations” ...
     Here, one finds a high diffusion potential of specific programs, from two different societal
     systems up to the totality of societal ensembles. After all, there is an open room in analogy
     formation and program extension to arrive at appropriate schemes for “lean household
     production” innthe age of postmodernity ...
     MANYONE-RELATIONS: Here, one finds, probably not unexpected, the integration of several
     programs from different societal systems for the (re)production of a special arrangement
     within a specific system. Referring, once again, to National Innovation Systems, one may
     think of a complex socio-technical ensemble in which programs from the scientific area, from
     the cultural and artistic domains as well as from the economic system have been combined to
     form a multi-systemic program-sequence, necessaqry in the (re)production of the socio-
     technical complex under consideration.

     MANYMANY-RELATIONS: Finally, the last inter-systemic exchange can be performed with
     respect to . Take, once again, a socio-technical ensemble like a personal computer in the area
     of information- and communication technologies, then the genotype constitution comes from a
     variety of different societal systems including the cultural and artistic ones and the diffusion
     potential goes well beyond the confines of the economic system into the scientific areas, the
     political-administrative spheres and the like ...

With this short list of four types of inter-systemic exchange relations, the basic conceptual framework for
describing flows within NIS-networks has been finished. What follows next are some additional useful
definitions and retrictions which should help to clarify the basic components and linkages within an
overall NIS-context.

                       3.5. EMBEDDEDNESS RELATIONS:
                               BASIC DEFINITIONS FOR

                           SOCIO-TECHNICAL SYSTEMS

Consequently, some basic concepts which have been used already in the course of the present discussion
will be introduced in a more rigorous fashion. More concretely, a basic NIS unit will be introduced at
some length, namely the component, refrerred to as “socio-technical systems”.

     Socio-technical systems - any dual level system , composed at the G-level of sets of
     (re)productive/non-(re)productive programs and at the P-level of machine components as well
     as practices. Thus, the social part is restricted to the domains of G-rule systems, code-free
     routines or “implicit knowledge”, whereas the technical part is linked to the G-design for
     machines as well as their P-building blocks ...
     Input of socio-technical systems - all ingredients (fuel, electricity, other forms of energy,
     backward linkages to other socio-technical systems, etc.) necessary for the continued
     opertation of a socio-technical system ...
     Output of socio-technical systems - a variety of different output areas, ranging from products
     or the manufacturing of other socio-technical systems to forward linkages to other socio-
     technical systemsproducts and, especially important, to waste and to a degradation of the
     human resources employed. Recalling the Second Law of thermodynamics -

        in the context of entropy, every action, of man or of an organism, nay, any process in nature,
        must result in a deficit for the entire system. (GEORGESCU-ROEGEN 1976:10)

     Reproductive Programs - any G-level program which, following the G  P relation, plays a
     necessary part in the maintenance of socio-technical systems.
     Productive Programs - any G-level program which, once again in the G  P direction, fulfills
     a necessary and indispensable function in the generation of new products or of new socio-
     technical systems.

Two additional comments should be made on the list of definitions above.

     On the one hand, the distinctions between products and socio-technical systems follows along
     the line of increasing complexity. Intuitively speaking, a socio-technical system producing
     food or textiles will fall paradigmatically under the scheme of -

                                    socio-technical system  product

     whereas an automatized socio-technical assembly hall for cars should be seen as a typical case
     of -

                             socio-technical system  socio-technical system

     The output in the form of cars requires, under normal circumstances, a complex of practices
     and competencies which are indispensable for their handling and maintenance. Moreover, the
     definition for socio-technical systems applies, quite naturally, to automobiles which, in any
     case, can be characterized as a hyper-complex dual level socio-technical system, composed at
     the G-level of sets of (re)productive/non-(re)productive programs from different code-systems
     (language, numbers, pictorial ..) and at the P-level of clusters of machine components as well
     as practices, necessary for its operations and (re)production.

At the present stage, an important intermediate level has been reached, namely the introduction of the
concept of “hyper-complex systems” ...

                                     4. THE EMERGENCE

                                 OF HYPER-COMPLEXITY

After the systemic exposition of codes and embeddedness-relations in evolutionary configurations in
general and in socio-economic ensembles in particular, the next concept to be introduced is the one on
hyper-complexity which can be defined, first, in a straightforward manner.

      An evolutionary system  is said to be hyper-complex if it is characterized by a multiplicity of
      code-systems and by a comprehensive set of embeddedness-relations.

Thus, Table 5.1 (next page) recapitulates and enlarges a summary, already shown in Section II, where one
can see, in addition, the comparatively high number of code-combinations and thus, of potential building
blocks present in any large scale societal system, from science to economics and beyond ...
At this point it becomes appropriate to introduce a thrid version of emergence, namely, not surprisingly,
emergence3 which is characterized by the following configuration.

      An evolutionary system  is said to be emergent3 if it is multi-layered of the type

                                         ,l  ,l´ (l´  l v l  l´) t

                                       (l,l´) =   () - (l) t
                                               l  l'

      and if the following two conditions hold -

                                             ,t  ,t (t´  t) t

                                       (t,t´) =   () - (t) t
                                               t  t'

      In other words, a system is said to be emergent3 if it is to be qualified simultaneously as
      emergent1 and as emergent2.

Thus, at the level of emergent3 systems, the potential for the “re-entry” of simplicity stays intact -

      Komplexität auf einer tieferen Ebene kann neue Einfachheiten auf einer höheren generieren.
      (COHEN/STEWART 1994:317).

Table 5.1: The Hyper-Complex Universe of National Innovation Systems

DOMAIN                      CODE-SYSTEM CODE-ELEMENTS                   COMBINATION WITH
                                                                          OTHER CODE-SYSTEMS

BIOLOGICAL                  Genetic               Four Bases: Adenin,     Biological Code  Language Code130
                            Code                  Cytosin, Guanin,        (especially the language of science)
                                                  Thymin                   Number Codes

HUMAN                       Natural               Letters of an           Natural Languages  Pictorial Codes
“LIFEWORLDS”                Languages             Alphabet                 Number Codes, etc.

HUMAN                       Number                Sets of Various         Number Codes  Language
“LIFEWORLDS”                Codes                 Numbers {}, {},       Codes, etc.

HUMAN                       Pictorial             Symbols from a          Pictorial Codes  Musical Codes
“LIFEWORLDS”                Codes                 Symbol-Library           Language Codes, etc.

HUMAN                       Musical               Musical Notes           Musical Codes  Language Codes
“LIFEWORLDS”                Codes                                          Number Codes, etc.

HUMAN                       Rule-                 Rules as Parts of       Rule Codes  Number Codes
“LIFEWORLDS”                Codes                 a “Rule-System”          Pictorial Codes, etc.

Moreover, there are strong reasons for assuming that upon a closer inspection of “building blocks”, their
arrangements and their recombinations across different societal domains, “One can hear in Bach´s music
Newton´s Principia Mathematica in the same manner than the new era of quantum-mechanics in the
music of Bela Bartok.” (Robert Colodny).

      Here, the symbol  denotes the combination of two code-systems.

                       5. EMBEDDEDNESS DOMAINS AND

                          EMBEDDEDNESS-RELATIONS -

                        A BACKGROUND-JUSTIFICATION

                            FROM COMPLEXITY-THEORY

It goes (almost) without saying that the newly established foundations offer a suitable basis for the
integration and the interconnectability of the social sciences, including economics, with the meanwhile
highly interlinked disciplinary complex of cognitive sciences, the neuro-sciences, “Artificial Intelligence”,
“Artificial Life” and the like. Not only that, the conceptualization of “embeddedness-relations”, “scientific
production”, “knowledge”, “information” or “transfers” under the integrative framework of “embedded
code systems” at the genotype and phenotype levels has paved the way for a direct and immediate
utilization of the rich repertoire of complex modeling and the stock of model-families developed under
the headings of learning, adaptation, self-modification, self-organization and the like. In other words, the
social science-foundations, built up in the course of finding an adequate schematization of National
Innovation Systems, provides one of the few successful research paths for a “big cognitive spurt” from a
position of relative backwardness with small or no comparative advantages to a status of a core-member
within the current scientific growth poles.
What is still missing however, is an additional justification which do not only point to the feasibility of an
externalized approach but which provide strong reasons for its unavoidability. Thus, before changing to
Section IV and to the multiple constitutions of families of National Innovation Systems, some remarks
from complexity theory will provide un-conventional reasons for the non-standard conceptualizations
offered within the previous chapters of Section III.

                                      5.1. A VIEW FROM

                               CODE-BASED NOWHERES

So far, the cognitive gains resulting from an externalization of the embeddedness domains and of
multiplying the embeddedness relations are entirely compensated by another consequence , namely the so-

called “observer-dependencies” and “relativisms”, resulting from an externalist view. Put very briefly, the
main results in the last chapters -

       die Ersetzung fiktiver notwendiger und hinreichender Bedingungen des Regelfolgens ... durch
       Behauptbarkeits- oder Rechtfertigungsbedingungen von Aussagen über das Regelfolgen unter
       geeigneten Umständen (STEGMÜLLER 1986:87) -

have the immediate consequence that attributions and conditions of rule-following and the like are
relative to specific “contexts”. The most important aspect in the sensitivity to spatio-temporal
environments lies in the observer-dependency which states, in the words of Humberto Maturana,
Francisco Varela and Heinz von Foerster, that attributions, including self-attributions, have to be
attributed by someone and to someone. And this “contextualization” leads in turn - under the heading of
“code-based nowheres” - to a far more human and even humanist approach to problems of knowledge,
justifications, revisions or adaptations. In the perspective of Thomas Nagel, one is confronted with the
following “edge of transcendence” configuration or, alternatively, the subsequent “condicio humana” -

       It is necessary to combine the recognition of our contingency, our finitude, and our containment in the
       world with an ambition of transcendence, however limited may be our success in achieving it. The
       right attitude ... is to accept aims that we can achieve only fractionally and imperfectly, and cannot be
       sure of achieving even to that extent. (NAGEL 1986:9)

The resulting consequences of a shift to external embeddedness domains and a multiplicity of genotype-
phenotype embeddedness relations are, within the current NIS-context, too far-reaching and clearly
transcend the scope of a theoretical introduction to the basic architecture of National Innovation
Systems.131 Suffice it to say that, among other areas, the favorite topics of French social analysis, namely
the ramifications of power-differentials132, acquire a prominent role since problems of the “definition
power”, the power of establishing standardizations, the accessibility to processes of standard-formation,
the participation in the selection of standardizations and “paradigmatic examples”, the resource-gaps in
the processes of reaching agreements, etc. turn out to be, within the externalized framework, of utmost

    Some consequences should be mentioned though, at least as a footnote. Observer-dependencies lead, for example, to an increased
importance of “response sets” and “context-effects” in the course of measuring social processes. Models and model-constructions, being
observer-dependent, are no longer conceived as “approximating reality”, but as a coherence-phenomenon between different types of data
based-descriptions and model-consequences. Moreover, observer-dependencies lead to a more cautious evaluation of failure- and success-
stories within the history of science and the like ...
    On this point, see esp. FOUCAULT 1979.

                                  5.2. THE FRAME-PROBLEM -


The remaining parts of Section III will be devoted to an unusual strengthening of the externalization-
perspective, developed so far. The basis for the subsequent explorations is marked by the simple and un-
controversial statement that rule-following, language or, more generally, code-games, allow for a high
degrees of variations - and errors.

         Das ist klar: daß wenn Einer sagt: 'Wenn du der Regel folgst, so muß es so sein', er keinen klaren
         Begriff von Erfahrungen hat, die dem Gegenteil entsprächen.
         Oder auch so: Er hat keinen klaren Begriff davon, wie es aussähe, wenn es anders wäre. Und das ist
         sehr wichtig. (WITTGENSTEIN 1984:238)

Since alternatives constitute, by necessity, a state space with specific trajectories, particular structures as
well as admissable domains and inadmissable areas, there must be ample room for potential errors, too. In
particular, correct rule following is centered, as has been demonstrated in the chapter on “embeddedness
domains”, on paradigmatic action-sequences in the public domain.

         Was ist 'eine Regel lernen'? - Das.
         Was ist 'einen Fehler in der Anwendung machen'? - Das.
         So rechnet man. Und Rechnen ist dies.
         Wie weiß ich, daß die Farbe die ich jetzt sehe 'grün' heißt? Nun, zur Bestätigung könnte ich andere
         Leute fragen; aber wenn sie mit mir nicht übereinstimmten, würde ich gänzlich verwirrt sein und sie
         vielleicht oder mich für verrückt halten. Das heißt: entweder mich nicht mehr zu urteilen trauen, oder
         auf das was sie sagen nicht mehr wie auf ein Urteil reagieren.
         Wenn ich ertrinke und 'Hilfe!' rufe, wie weiß ich, was das Wort Hilfe bedeutet? Nun, so reagiere ich
         in dieser Situation.
         Nun, so weiß ich auch, was 'grün' heißt und auch wie ich die Regel in dem besondern Fall zu befolgen
         habe 133 -

Thus, fuzziness of boundaries, indeterminacies of action-attributions, errors and the like become the
necessary bye-product of non-trivial, high-dimensional state-determined agencies, human and otherwise.
Moreover, doubts concerning attributions, drawing of boundaries or the content of messages are

      A compilation from WITTGENSTEIN (1971:ÜG 28 & ÜG 47) and WITTGENSTEIN (1984:BGM VI 36).

dependent on inconsistencies between a specific attribution, a special boundary drawing or a concrete
message and the space of alternatives.

      Ich sage Einem: 'Der und der war heute vormittag bei mir und hat mir das und das erzählt.' Wenn es
      erstaunlich ist, so fragt er mich vielleicht: 'Du kannst dich nicht darin irren?' Das mag heißen: 'Ist das
      auch gewiß heute vormittag geschehen?', oder aber: 'Hast du ihn auch gewiß recht verstanden?
      (WITTGENSTEIN 1971:ÜG 648)

The central point in the whole phenotype-chapter lies in the peculiar “fact” that there are no differences in
principle, just in degrees, between the operations in understanding a text, a symbol, a picture, a book, a
piece of music and the operations in understanding non-trivial machines, The criteria for understanding
are, in both instances, within the same public domain of paradigmatic instances and agreed upon
attribution strategies. Especially with respect to the understanding of “understanding”, the criteria are
marked by a highly counter-intuitive symmetry-condition which in its most puzzling form has been
formulated by Ludwig Wittgenstein.

      Wenn man aber sagt: 'Wie soll ich wissen, was er meint, ich sehe ja nur seine Zeichen', so sage ich:
      'Wie soll er wissen, was er meint, er hat ja auch nur seine Zeichen. (WITTGENSTEIN 1971:PU 504)

Moreover, an additional important point, emphasized within the Austrian philosophical tradition
especially by Otto Neurath or Ludwig Wittgenstein, lies in the relatively short chain of justifications for
attribution strategies. After only several steps of challenges and justifications, one will reach a state of
“dialectic equilibrium” where no more additional justifications can be provided.

      Die Gründe werden mir bald ausgehen. Und ich werde dann, ohne Gründe, handeln. (IBID:PU 211)

I.e., based on a necessary insufficient amount of “information”, one has to decide whether a person
understands a specific numeric sequence or not, whether a child has the ability to read, whether someone
“understands” a foreign language, an artistic style, etc.
In this spirit, the mysteries of the “frame-problem” which has been featured prominently in Section II
turns out to be superficial since they become either a “matter of triviality” or a “matter of history and
development” . Looking back at the previous sections, the “frame problem” has two different types of

      First, there can be the case of a varying number of active rules within a specific decision
      cycle. In the language of “classifier systems”, someone “knows” that P (“tomorrow is a
      holiday”) and Q (“I have to buy milk for the coffee”). In the decision situation whether to buy
      milk today or not (Qt v Qt+1), P and its customers (“On holidays, shops are closed in Austria”,

         etc.) do not enter into the bidding contest. The resulting action pattern Qt+1, postponing the
         purchase of milk, . The important point to stress lies inthe possibility for an insufficient
         number of ruels in the actual decision configuration. Occurences of this type simply cannot be
         ruled out, especially not in the case of non-trivial machines.
         Second, the architectonic problem behind the “frame problem” lies in finding salient answers
         to the number of relevant rules. The final arbiter for salience lies, however, in the long
         evolutionary history which, following Dawkins has achieved means and ways for “taming
         chance” -

             To „tame‟ chance means to break down the very improbable into less improbable small
             components arranged in series. No matter how improbable it is that an X could have arisen
             from a Y in a single step, it is always possible to conceive of a series of infinitesimally graded
             intermediaries between them. However improbable a large-scale change may be, smaller
             changes are less improbable. And provided we postulate a sufficiently large series of
             sufficiently finely graded intermediaries, we shall be able to derive anything from anything else
             ... (DAWKINS 1986:317)

In a generalized version of the trivial and historical solutions to the “frame problem”, any form of life
“has” its specific “basins of attractions”, its “fixed points”, its “social fitness landscapes” of local maxima
and minima ... Moreover, these “basins of attraction” exhibit all the essential dynamic characteristics like
a sensitivity to initial conditions, discontinuous changes and “jumps”, multiple local equilibria where the
following metaphor holds -

         So bin ich nun auf dem harten Felsen angelangt, und mein Spaten biegt sich zurück. Ich bin dann
         geneigt zu sagen: 'So handle ich eben.' (WITTGENSTEIN 1971:PU 217)

Rule-following and language games, they are neither fixed and pre-programmed action-sequences nor
self-contained and closed monads “without windows”. On the contrary, all our code-games have a high
potential for recombinations and for errors. Take a sentence like the following which has occupied a local
maximum for centuries -

                                               We are satisfied that the earth is flat.134

For millenia and for a large number of societal ensembles, explorations in cosmology and astronomical
reasoning, justifications in , everyday questions and answers on the shape of the earth have led, after a

      A variation to Wittgenstein´s

                                      We are satisfid that the earth is round (WITTGENSTEIN 1971:ÜG 299)

sequence of steps, into this particular basin of attraction where additional questions, challenges and
disturbances suffered the fate of returning into this particular region in cognitive space, again and again,
round and round ...
Moreover, the following quotation, written in the years around 1950, make the sensitivity to socio-
economic contexts as well as the drastic change in the shape of knowledge landscapes abundantly clear.

     Wir alle glauben, es sei unmöglich, auf den Mond zu kommen; aber es könnte Leute geben, die
     glauben, es sei möglich und geschehe manchmal. Wir sagen: diese wissen Vieles nicht, was wir
     wissen. Und sie mögen sich ihrer Sache noch so sicher sein - sie sind im Irrtum, und wir wissen es.
     Wenn wir unser System des Wissens mit ihrem vergleichen, so zeigt sich ihres als das weit ärmere.
     (WITTGENSTEIN 1971b: ÜG 286)

  Thus. like in the case of We are satisfied that the earth is flat, local maxima or, alternatively,

     poles which indicate the endpoints of possible gradation (HALLER 1988:134)

  have to be, in a very strong sense, the necessary endpoints in cognitive space. Phrased in a different
  perspective, these poles or local maxima are, for the space and time being, the “blind spots” in
  particular forms of life. Around local maxima, blindness becomes unavoidable. An alternative way of
  putting this important point is the following variation, once again from Wittgenstein II. -

     Aber schließt du eben nicht nur vor dem Zweifel die Augen, wenn du sicher bist? - Sie sind mir
     geschlossen. (WITTGENSTEIN 1971:PU, p. 261)

                             5.3. THE NP-CHARACTER OF

                                  INTERNAL SOLUTIONS

At this point, it should be fruitful to come up with a conjecture which, if corroborated, would provide a
final justification for the external approach to “embeddedness domains” and “embeddedness relations”.
This conjecture takes its starting point from a well-known philosophical “uncertainty principle”, namely
Quines Indeterminacy of translation which has been characterized by Quine, inter alia, in the following

      There can be no doubt that rival systems of analytical hypotheses can fit the totality of speech
      behavior to perfection, and can fit the totality of dispositions to speech behavior as well, and still
      specify mutually incompatible translations of countless sentences insusceptible of independent
      control.(QUINE 1975:72).

With respect to this particular task, one does not even need two different languages and an inconclusive
lexicography, with or without fact of the matter. Upon closer inspection, an identical problem arises with
respect to two native speakers within the same language community. Even more, for one and the same(?)
person at two different points in time “rival systems of analytical hypotheses can fit the totality of speech
behavior to perfection, and can fit the totality of dispositions to speech behavior as well”. Thus, the
generalized version of Quine´s Indeterminacy-relation becomes strikingly similar to another philosophical
riddle, namely to Goodman´s “new problems of induction” which result, by and large, from sufficiently
available “degrees of freedom” in attributing strongly contradictory hypotheses to a given set of
“observational data”. Add to these two paradoxes the Wittgensteinian conundrum of rule-following or
Hume´s paradox of the intertemporal non-transferability of causal relations, then one finds a striking
parallel between all four paradoxes which can be identified by using the following analogy. Take a typical
NP-problem, namely the “traveling salesman” (DEVLIN 1994, WAGNER 1994), then the following
complexity barrier comes into play.

      For just a few elements (C  5) and, consequently, a small number of combinations between
      them, the problem at hand remains trivial and can be solved in a very short time. Changing
      however, to large configurations (C  1000), consisting of more than thousand components
      and a vast number of possible combinations too, the problem spaces cannot be explored in a
      reasonable amount of time. In the case of the traveling salesman, no efficient algorithm has
      been found so far which would solve the problem at hand in polynomial time.

Since the salesman-configuration belongs to problems of the NP-class (non-deterministic polynomial), the
conjecture, emerging from the four rule-paradoxes, lies, first, in the NP-type character of these and similar
riddles. This, in turn, gives rise to the the following hypothesis of the transformation of NP-attribution
problems into P-ones:

      THE NP  P-CONJECTURE: The complexities of socio-economic ensembles, consisting of a
      large number of possible choices for attributing a particular rule and, moreover, of a long
      sequence of different actions, makes a comprehensive exploration into the required rule spaces
      impossible. Instead, in order to guarantee relatively quick real-time solutions, a large amount
      of intersubjectively easily accessible and, thus, of external restrictions has been imposed on
      attribution modes, placing rule-attributions and, more generally, the attribution of “internal

       stances” at the “edge of chaos”. Thus, the NP-character of attribution problems is permanently
       transformed into a P-class problem of “sufficient complexity” ...135

In other words, due to the NP-character of rule attributions, the only path for satisficing “real time
solutions” lies in the proliferation of a large number of external fixed points and restrictions in form of
criteria, standardizations, rituals, “implicit knowledge” and the like ... Only these temporarily fixed and
constant domains136 are able to guarantee that, on the one hand, appropriate attributions can be
accomplished in a small amount of time and that, on the other hand, processes of code-acquisitions and
finding the appropriate attributions can be learned successfuly in the life-course of a few years only. After
all, children in the age of five or six years exhibit a rich repertoire of attributing different types of
“internal stances” to humans, play toys, trees, flowers, animals ...
It would go well beyond the scope of the present Section to justify the “NP  P - Conjecture” in greater
detail. Likewise, the linkages between the “NP  P-Conjecture” and the “Frame-problem” can be given
only in its crudest form, stating that the imposition of a large amount of community specific restrictions is,
in all probability, the only satisficing way for a dissolution of the peculiar “frame problem” ...
At the end of the present section, it must be sufficient to point out to the overall goal-orientation which
has motivated the writing of the present volume.

       What is your aim in the alternative constitution of National Innovation Systems? - To show the social
       and economic science flies the exit from the conventional fly-glasses ... -

This ambitious goal, so it is expected, will be reached after Section IV which is devoted to the multiple
constitution of National Innovation Systems. After all, the overall goal for stressing so heavy importance
on the theoretical foundations in the current NIS-project, lies in the hope that, by doing so, a potentially
fruitful new transdisciplinary perspective can be established on evolutionary processes in hyper-complex
societal configurations like National Innovation Systems. Moreover, the construction of this meta-
theoretical framework is tied to another wish, this time with respect to the degree of ordering in cognitive
spaces and can be expressed in the following recombination -

                                                 Zeitweiser Friede in den Gedanken.
                        Das ist das ersehnte Ziel dessen, der sich komplexen Problemen widmet.137

    “Edge of chaos” and sufficient complexity” refer both to the same phenomenon, namely to a state, allowing “both stability and
flexibility”. (KAUFFMAN 1995:86)
    It should be added, once again, that these societal “fixed points” and “restrictions” exhibit a long-term history of changes and revulsion,
too. Nevertheless, following Wittgenstein, it seems useful to distinguish between the flow of water within a river from the movement of the
river-bed itself ...
    The “locus classicus” for the above variation comes, as so often, from Ludwig Wittgenstein -

                                                          Friede in den Gedanken.
                              Das ist das ersehnte Ziel dessen, der philosophiert. (WITTGENSTEIN 1978:87)

Following the discussion throughout Section III however, an important qualification with respect to
observer-dependencies must be added. Temporal peace within cognitive landscapes, this is the locally
best, we can hope for ...





Instead of presenting a detailed outlook of the main chapters of Section IV, a brief review as well as
preview will be given on the main targets which at the end of Section IV will have been readily

     First, a new epigenetic theory background will be completed in which the two conflicting
     views on National Innovation Systems, the first one as a knowledge based perspective, the
     second one as a learning-oriented approach, have been combined under partially new
     headings like “genotype-and phenotype-levels”, “embedded code systems”, “genotype-
     phenotype-interactions” and the like.
     Second, the “great evolutionary vision”, or, to quote Daniel C. Dennett, “Darwin´s dangerous
     idea” (DENNETT 1995), will be taken very seriously, furnishing a homogeneous evolutionary
     basis which will serve, inter alia, as an evolutionary “fitting” foundation for the subsequent
     empirical investigations and, more generally, for analyses in the area of evolutionary
     economics, evolutionary sociology or, of epigenetic social sciences for that matter.
     Third, the new conceptual framework will fulfill, moreover, an important additional function
     since it will contribute substantially in the construction and in the interpretation of the results
     of the Austrian Survey of Innovation and Transfer (ASIT). To mention just one example, the
     meta-theoretical innovation concept has led to a substantial widening with respect to
     innovation domains in ASIT, leading to the inclusion of “organizational innovations” as a new
     element alongside the traditional domains of product and process innovations.
     Fourth, the epigenetic perspective on code-based reproductions of embedded or dual level
     systems, i.e. on the “Two Great Chains of Becoming”, will open up radically new pathways for
     the comparative analysis of socio-technical and of biological evolution, beautifully
     summarized by a single sentence by Stuart Kauffman -

                Tissue and terracotta may evolve by deeply similar laws. (KAUFFMAN 1995:192)

     Fifth, the present evolutionary framework will allow, moreover, a transdisciplinary model-
     analysis on changing development patterns in National Innovation Systems. Why? Simply
     because a heavy emphasis has been placed on the construction of a meta-theoretical
     framework or, alternatively, of a core-apparatus which should be applicable, mutatis
     mutandis, for any type of evolutionary system, ranging from the biological areas to the
     domains of technical systems, scientific theories or National Innovation Systems, too. Since a
     variety of complex models have been successfully applied to the biological, the ecological or
     the neural domains already (See, e.g. GALE 1990, KOSSLYN/ANDERSEN 1992), the
     “transdisciplinary apparatus” should serve as an appropriate “bridging component” for the
     utilization of complex models in socio-technical systems or National Innovation Systems, too.

     Sixth, the new theoretical perspective will become of tantamount importance in the radical
     shift with respect to science and technology policies undertaken in Volume VI, where the dual
     level NIS-framework will open up entirely new ways for establishing high density genotype-
     ensembles, closely interlinked phenotype-networks and the like.

Thus, Section IV will bring the “quadriga” of essential building blocks for National Innovation Systems -

                    Evolutionary Systems and Innovations in General (Section I)
                 Embedded Code-Systems at the Level of NIS-Genotypes (Section II)
            Embeddedness-Domains and Embeddedness Relations between the NIS Genotype
                            and the NIS-Phenotype-Levels (Section III)
              Complex Epigenetic Systems at the Level of NIS-Phenotypes (Section IV)

to an end by focusing, within the first two chapters, on basic phenotype network arrangements and,
especially important, on intersystemic network connections. The following chapter three will then be
devoted to the multiple possibilities in the constitution of National Innovation System by introducing
different NIS-types of varying degrees of complexity, starting from a single systemic configuration like
the economic system, the scientific system, the cultural system, etc., to double ensembles like the duality
of the economic and the scientific system, to triple sets like the trias between science, economy and the
state and, finally, to societal NIS-configurations. Finally, chapters four and five will summarize the
preceding discussions and demarcations by offering, on the one hand, a new “phenomenology” of
National Innovation Systems both of the traditional and posttraditional, i.e. knowledge based variety and
by furnishing, on the other hand, a phenotype framework for the analysis of “scientific creativity” which
supports the genotype-schemes of “scientific creativity”, introduced in Section II.
It should be added right from the beginning that the elaborated and unusually intensive discussion on
evolutionary systems, on “embedded code-systems” or on “embeddedness relations” has yielded an
enormous “cash avlue” (Wilfried Sellars) which becomes apparent simply by looking at alternative
theoretical NIS-explorations. Take, for example, conventional definitions on National Innovation Systems
like the following -

     The network of institutions in the public and private sectors whose activities and interactions initiate,
     import and modify and diffuse new technologies may be described as the „national system of
     innovation‟ (FREEMAN 1987:3)
     A system of innovation is constituted by elements and relationships which interact in the production,
     diffusion and use of new and economically useful knowledge (LUNDVALL 1992:2) -

then a comparison with the current NIS-specifications laid out in the courses of Sections I - IV reveals an
astonishing number of discrepancies and differentiations. According to the current epigenetic theory
background -

     A National Innovation System is a class of different types or families of evolutionary, dual level
     systems, composed of -
         a multiplicity of genotype ensembles or, alternatively, of “helices” and a multiplicity of
         phenotype arrangments, interlinked via multiple “embeddedness relations” ...
         different types of generality (single NIS-ensembles (NIS1), dual NIS-ensembles (NIS2), triple
         NIS-ensembles (NIS3), n-ary NIS-ensembles (NISn)) ...
         different forms of building blocks for each of the genotype levels, comprising, above all,
         different groups of “embedded code systems”, “programs”, etc. ...
         different forms of building blocks for each of the phenotype levels on the micro-levels
         (individuals, departments, firms, institutes, socio-technical systems ...) and on the macro-levels
         (economic sectors, clusters, professions, scientific disciplines, production types, etc.) ...
         a universal mode of recombinations, i.e. of innovations and deovations, via “recombination
         oprators”, applicable both to the genotype and to the phenotype levels ...
         different forms of fitness landscapes, depending on the choices of building blocks and on the
         main focus of analysis (genotype-oriented, phenotype-centered or focused on genotype-
         phenotype interactions) ...
         different patterns of production, interaction and diffusion of genotype-innovations
         (denovations) (new programs) and phenotype-innovations (denovations) (new inputs, new
         inputwithinputoutput structures, new outputs), again depending on the selections of
         building blocks and on the principal level of investigation (genotype, phenotype,
         genotypephenotype) ...

The most surprising differences in the epigenetic NIS-perspective, introduced in the present report, lie, on
the one hand, in the complete dismissal of core NIS-concepts like “new technologies” or “new
knowledge” which have been integrated into a new {C-E-S}-framework of “embedded code systems”
(ECS), of “complex epigenetic systems” (CES) or of “self-contained ensembles” (SCE), all of them based
on “evolutionary stable classifications” (ESC). On the other hand, the new conceptual apparatus allows a
far more refined and sophisticated analysis of NIS-processes where new forms of knowledge diffusion
and new waves of technology propagation can be studied together with a large number of additional
processes in the area of scientific production, in the domain of information and entropy, in the
institutional settings or in the organizational changes which could be consistently integrated within the
new epigenetic conceptual “umbrella” ...
Thus, the next chapters will bring in a highly condensed form an overview of the phenomenology of
National Innovation Systems, first mostly at the phenotype level alone and second on both their genotype

and their phenotype-levels. This, in turn, will become the conceptual foundation both for the scope and
the content of the definitions, demarcations and boundaries which will be introduced in the empirical
study of the Austrian System of Innovations in Volume IV.

                             1. PHENOTYPE NETWORKS

The first important theoretical concept for the constitution of National Innovation Systems at the
phenotype level is the notion of “networks” which will be employed over and over in Volume IV as a
“description device” for the interaction patterns of scientific institutes, economic firms, bridging
institutions and the like. Diagram 1.1. presents, in its most general form, the basic arrangement for NIS-
relevant phenotype networks.

Diagram 1.1: The Basic Network-Configuration for the Phenotype Level

From Diagram 1.1, the network depicted has a series of general features which can be considered as
typical for NIS-phenotype networks.

     First, a network consists of a variable number of nodes, ranging from small numbers like the
     configuration in Diagram 1.1 with n = 6 to large numbers (n  1000) and beyond. In the case
     of the Austrian science system, for example, the total number of NIS-relevant institues, both
     from universities and the research segment, amounts to roughly 1200. In principle, networks
     may consist of millions of nodes or basic units, depending on the particular choices of
     building blocks, the topic under consideration and the like.
     Second, each of the network nodes, like simple input-output systems, possesses a certain
     number of input linkages from other network nodes and output linkages to other network
     nodes. These relations can be qualified as the internal linkage structure of a network. In the
     case of the science system, for example, the connections between network components, i.e.
     the linkages between science-institutes, may be qualified as the internal network structures of
     the overall science system.
     Third, in a similar manner, the network nodes, as becomes clear from Diagram 1.1, too, have
     a number of linkages with their extra-systemic environment. Again, the input linkages from
     other systems as well as the output linkages into other systems will be summarized as the
     external linkage structure of a network. Taking, once again, the science system as reference
     case, the connections between institutes on the one hand and R&D departments of firms, with
     the state bureaucracies, with the household sector or with the domain of arts and culture will
     be labeled as external network structures of the science compartment within a National
     Innovation System.
     Fourth, the linkages between network nodes must be, in principle, observable and measurable.
     This implies, in turn, that in many instances of applications and, especially important, in
     (almost) all cases of explanations, a specific segment of the overall interactions and
     connections has to be selected. For example, the science exchanges between institutes can be
     observed with respect to personnel mobility, with respect to financial flows, with respect to
     research cooperations, with respect to exchange of information, with respect to transfers and
     the like. While some of these linkages can be observed and measured simultaneously, it must
     be added immediately that many essential linkages, though obviously present in the science
     network, can be observed and measured only in a superficial manner. Take, as a prime
     example, the utilization of the network “knowledge basis”, i.e. the actual utilizations of
     programs, generated within a national science network, in the form of reading, adapting,
     recombining, etc., then one will be confronted with an unsolvable observation and
     measurement problem. On the one hand, easily available methods like citation analysis
     capture only a small fragment of the actual utilization patterns and, on the other hand,

       observations and measurements of a random sample only would require an extremely
       comprehensive design, which, in all probability, cannot be implemented at all. Moreover, only
       a descriptive analysis can be given on the persisting overall patterns of network structures
       since no explanatory framework can be built up which would be able to account
       simultaneously for the actual distributions of financial flows, mobility in personnel, transfer
       exchanges, research cooperations and the like within the overall science network. As will be
       shown in Volume V, only network models applicable to one or two particular core aspects like
       changes in production types or personnel mobility, can be transformed into complex, dynamic
       network models.
       Fifth, the same restrictions and “caveats” must be applied to the environment-network-
       relations which, once again, can be accounted for in some selected aspects and not even in the
       most essential ones. Taking once again the science network as reference point, then one can
       easily see that the relations between, say, the media, political parties, interest groups or non-
       government institutions on the one hand and scientific institutes or individual researchers on
       the other hand can only be dealt with by relying on the self-assessment of scientists
       themselves.138 Why? Because, once again, it would be completely impossible to develop a
       research design by using participant observation for a representative sample of institutes for a
       relatively long period of time ... 139

In this manner, some striking general features of phenotypes networks have been identified which has
been introduced, once again, as a metatheoretical concept, applicable to biological and socio-economic
domains - and, in a self-referential mode, to itself, too. What follows next, lies in the straightforward
extension of single networks to more general cases.


NIS-networks, due to their generality, can be identified for any type of large-scale systemic configuration,
ranging from the networks of institutes in the science arena, to the firm-network in economic ensembles
and to the household-network in the “life world” dimensions .... Moreover, for two phenotype networks,
one can define a single “exchange module”, for three societal networks a “triple exchange module” and,
more generally, for n societal systems a n(n-1) /2 set of modules. What remains to be shown at this
    In two recent seminars on the sociology of science at the University of Vienna, empirical designs have been developed and tested with
respect to both areas, namely with respect to the internal linkage strauctures and the external ones. Not surprisingly, scientists have turned
out to be highly insufficient reporters and commentators of their own research activities ...
    The problem at hand becomes structurally similar to the devlopment of most comprehensive road maps which, in the process of getting
closer to “reality”, become a 1:1 edition to the “real world” ...

point, is the specific linkage structure between societal networks. In essence, three probably unexpected
problems can be identified:

       First, it must be stressed that for the question of modeling the linkages between two or more
       large scale social systems a considerable problem arises, namely the question of attribution,
       or, in terms of the Austrian School of Economics, of Zurechnungen. Consider, for matters of
       simplicity, a technician in a laboratory of a medium size firm, writing an article for an
       international journal on new modeling tools, developed within a joint project with a research
       institute from abroad. While, on the individual level, no multiplicity of the self takes place, the
       attributions on the genotype levels as well as on some phenotype levels turn out to be, already
       in such a straightforward instance, highly complicated, since, on the one hand, a single person
       from the economic system has contributed, by publishing in a science journal, to the science
       helix. On the other hand, the person in question must be counted, due to her or his program
       capacities, both as a member of the national and the international science system and of the
       national economic system ...
       Second, in the presence of striking linkages in two large central NIS-domains like the
       economic and the scientific system, it must be viewed as the most promising research strategy,
       to define a third system, viz. a transition system, which takes its boundaries with respect to the
       exits and the entrances of the two NIS-systems under investigation. Thus, the internal and
       external connections between science and economy should be modeled as a dynamic process
       of three interacting large scale units.
       Third, any attempt to link a comparatively large number of NIS ensembles, though a viable
       and valuable research strategy in its own right, is faced, aside from genuine modeling
       difficulties, complexity barriers and lack of adequate data-bases, with an important
       methodological problem. The integration of different NIS-ensembles into an international
       GIS-complex (Global Innovation System) can neither be assumed the most fundamental level
       nor the final goal for IS-analyses. Contrary to stylized perspectives like the one raised by
       Immanuel Wallerstein140, a Global Innovation System approach remains just a single reserach
       path to be followed amidst a truly infinite phase space of possible research trajectories. Even
       worse, a GIS-framework cannot even be qualified as the most important or the most
       fundamental one either, since notions like relevance or importance remain intimately linked to
       the eyes of the observer. And, fortunately or unfortunately enough, the variation of
       perspectives seems to undergo, almost by necessity141, a process of persistent widening and
       enlargement whereby new like the introduction of Regional Innovation systems (RIS) open up

    See his recent, methodologically oriented book on the world systems approach, viz. WALLERSTEIN (1991).
    The main reason for this phenomenon is directly connected with the expansion and the differentiation in the scientific system where,
generally speaking, the number of different disciplines and conflicting schools within disciplinary confines exhibits an apparently secular

         and, as a genuinely non-intended side-effect, new areas and new frontiers for recombinations,
         permutations, and variations on the “thema” of Innovation Systems arise ...142

In this manner, the remarks on the phenotype-constitution of National Innovation Systems has been
finished. The two important points for the subsequent construction of NIS-families, namely the focus of
networks on the phenotype-level and the combinability of large scale networks into more comprehensive
ensembles, have been put forward in a condensed and, hopefully, plausible fashion.

                                         3. BASIC ARCHITECTURES

                         FOR NATIONAL INNOVATION SYSTEMS

What follows next is a short summary of basic NIS-ingredients for different families of National
Innovation Systems. The G-description for National Innovation Systems includes, quite obvious, different
helix-compositions whereas the P-part consists of various network-combinations. Moreover, the
subsequent NIS requirements can be understood, inter alia, as a “blending” of societal formations with
the evolutionary framework, proposed through Section I to IV.


At the beginning, a family of National Innovation Systems will be presented which hardly qualify as
“national” since here single societal systems like the scientific one, the economic one and the like as well
as their national or international environments become the core topic of analysis. The G and P
requirements for National Innovation Systems I include, quite obvious, the following ingredients.

      Following George Boole, one is led there to a simple metaphor of light and darkness which, according to George Boole,

         are not strictly conterminous, but are separated by a crepusular zone, through which the light of the one fades gradually off
         into the darkness of the other ... -

and where, by force of analogical reasoning,

         it may be said that every region of positive knowledge lies surrounded by a debatable and speculative territory, over which it
         in some degree extends its influence and its light. (BOOLE 1958:400)

     First, the helix pool for NIS I consists of all those programs, memes and neural groups which
     are an integral part of the G  P type of embeddedness-relations. Thus, for the scientific
     system, programs of the internal output as well as programs from other spheres with
     indispensable roles in the (re)production of the science system, become integral components
     of the helix pool. To point to somewhat unusual domains not mentioned already in the helix-
     chapter in Section II, organization schemes, developed and “tested” within pre-modern crafts
     and guilds, monasteries or state bureaucracies, have become part and parcel of many European
     university systems. Architectonic codes for buildings have been used in the construction of
     universities and research institutes. Symbolic or graphical codes, used and propagated in life-
     world domains like education or “visual information systems”, are gaining rapid ground as a
     way for transporting scientific messages as well. New genres at the G-level, initialized and
     standardized within the electronic or within the print media, enter as an important means of
     communicating and diffusing new scientific results. Thus, the helix pool for a NIS-I ensemble
     will be a mixture between helix-components from various societal systems.
     Second, the P-level for a NIS I-configuration consists, by and large, of a collection of specific
     network groups, depending on the choices of particular building blocks. In Part I of Volume
     V, the possibilities for generating a large number of different building block-compositions for
     the economic or for the scientific system, will be discussed at length. At this point, it must be
     sufficient to point to an irreducible multiplicity of different arrangements of building blocks
     which, in turn, require selection procedures for a specific choice of basic elements on part of
     the “scientist in action” ...
     Third, the internal structures, restrictions and flows for G-level and P-level processes
     comprise a large number of different specification pathways, ranging from theory-networks to
     scientific communication processes, financial flows or intra-scientific cooperation schemes.
     Fourth, the external relations and exchanges of NIS I are, once again, an “open land”, which
     leads from specific theory networks and their cognitive environments or from local and
     limited environmental linkages at the national level to dense relations within the global

The specific choices of boundaries, components and exchange relations determine, in effect, the type of
fitness landscapes, the innovation or denovation character of new network features, etc. Thus, the four
specification requirements for NIS I-ensembles have to leave most of the actual specification work to the
specific research operations and selections of NIS-researchers. This implies, e converso, that the concrete
specification choices are highly informative with respect to a particular NIS I-configuration - and with
respect to the “scientific observers” who have made these distinctions more or less deliberately ...


Despite the biological paradigm of the “double helix” in the DNA-arrangement, the NIS-double helix as
well as National Innovation Systems II exhibit a radically different configuration.

     First, the helix pool for NIS II comprises, quite obvious, all those programs, memes and
     neural groups which exhibit the G  P type of embeddedness-relations, outlined in Section
     III. Thus, for the scientific-economic systems, programs of the internal output as well as
     programs from other spheres with indispensable roles in the (re)production of the science and
     economy system, become necessary components of the pool of helices. To point to somewhat
     unusual domains, organization schemes, developed and “tested” within state prisons have
     become ingredients in the work organization of firms. Architectonic codes for buildings have
     been used in the construction of firms, banks, universities and research institutes. Information
     systems whose design has been developed within the scientific arena, have been adapted to the
     needs and requirements of Management Information Systems (MIS). Thus, the helix pool for a
     NIS-II ensemble must be a mixture between helix-components from various societal systems.
     Second, the P-level for a NIS II-configuration consists, by and large, of a set of interlinked
     networks whose specification depends, however, crucially on the choices of the specific
     building blocks. At this point, it must be sufficient to point to a simple “law of increasing
     diversity”, since a large number of different specification choices in NIS I leads to a rapid
     increase in subsequent NIS-species like NISi II, NISj III, etc. Take, for matter of simplicity,
     hundred different building block approaches for each of three different NIS I-ensembles, then
     one immediately arrives at -

                                  NIS II:   3 x 102 . 102 = 3 x 104
                                   NIS III:   102 . 102 . 102 = 106

     different arrangements for NIS II- or NIS III-configurations.
     Third, the internal structures, restrictions and flows for G-level and P-level processes have to
     include both an intrasystemic as well as an intersystemic part. Once again, the “law of
     increasing diversity” operates on these exchanges as well, since large numbers both for the
     intra- and the intersystemic exchanges are available, in principle.
     Fourth, the external relations and exchanges of NIS II are, once again, dependent on the
     particular building block-specifications, leading from specific cognitive networks and their

       cognitive environments to global systems-environment networks with a dense linkage

The specific choices of boundaries, components and exchange relations determine, in effect, the type of
fitness landscapes, the innovation or denovation character of helix and network features and the like.
Thus, the four specification requirements for NIS II-ensembles have to leave most of the actual
specification work still open and undecided. This implies, e converso, that the concrete specification
choices are highly informative with respect to a particular NIS II-configuration - and with respect to the
role of the scientific observer who, folowing Spencer Brown, has made these NIS II distinctions ...143


Under the heading of an n-ary societal National Innovation System (N= 4, 5, 6, ...) the following set of
minimal requirements can be laid out. The introductory step will consist in more precise and empirically
accessible definitions of the scope and the extension of the term “societal” which will turn out to be far
from trivial or self-evident. Two essential requirements, one called “totality condition”, the other one
“spatio-temporal completeness-condition” will be postulated which a societal ensemble must meet

       First, a societal configuration must fulfill the “totality requirement” and must, thus, be
       composed in a way to include core processes in at least six different domains, namely
       production processes in the economic system, the state system, the scientific system, the
       cultural and artistic system, in the system of mass media and, finally, in the household system.
       Second, a societal ensemble must fulfill the demand for “spatio-temporal completeness”, that
       is, the requirement to account for the “agencies” of the “building blocks” selected. Thus,
       societal configurations, taking individuals as their basic element, must be defined in a way so
       that “action-sequences” or routines of individuals “do matter”. I.e., any action, especially
       everyday activities at the working place, in households, in public places etc., must be
       integrated into the specifications of the societal ensemble under investigation.

    In a mode of minor variations on the NIS-II thema, a very similar chapter could be written on NIS III-ensembles, comprising the
economy, the science system and the state and consisting, on the genotype levels, of a “triple helix” and, on the phenotype levels, of densely
interlinked networks.

Thus, a societal NIS-configuration must be able to account for innovation and denovation processes in at
least six core areas of contemporary societies in an inclusive manner. Consequently, the following
specification rules must be introduced.

     First, a societal NIS-S helix must be composed of at least four different helices, including,
     aside from the triple helix pool, also the genotype-realm of cultural, artistic, mass media or
     everyday activities.
     Second, likewise the phenotype-composition of a societal National Innovation System must
     take into account at least a fourfold division of systems, having, at its minimum, a partitioning
     of systems into an economic, scientific, political, and a life-world-ensemble in which arts,
     culture and mass media are integral components.
     Third, the most essential demand for a societal NIS lies in its successful “completeness
     perspective” and requires a partitioning of NIS-systems which accounts for the “agencies” of
     particular choices of building blocks. Thus, societal NIS-configurations must be defined in a
     way so that for any action and action sequence of a building block element, a subset relation
     can be established between these actions and action sequences and the set of NIS-systems.
     Fourth, appropriate fitness-landscapes have to be furnished for the systemic ensembles
     outside the NIS-III configurations which, especially for artistic and cultural domains, will turn
     out to become a very challenging and unexplored type of research both for the genotype and
     for the phenotype levels.

In this manner, the basic systemic specifications for the most comprehensive and as yet most unknown
NIS-ensemble, namely the societal National Innovation Systems (NIS-S) have been laid down.

                           4. THE PHENOMENOLOGY OF


After the description of different NIS-families, a short set of definitions, features and trends will
summarize the quintessential new perspectives which have been developed during the present theory part
for the basic architectures and devleopment patterns of National Innovation Systems, past and present.
More concretely, the following areas will be covered -

       Additional basic definitions for National Innovation Systems with respect to their spatio-
       temporal dimensions as well as to their helix and network constitutions ...
       Identification of four code-based “mega-stages” along a terrestrial time dimension, offering
       an unusual justification for a fundamental phase transition in current National Innovation
       Systems ...
       Predominant process types of “creative constructions” and “creative destructions” ...
       Phase transitions in National Innovation Systems from traditional ones to “knowledge based”
       ensembles or, alternatively, to “knowledge and information societies” ...
       Ten “mega-trends” as well as ten “mega-characteristics” for each of the genotype and
       phenotype levels of “Knowledge Based National Innovation Systems”

Thus, the present chapter must be viewed as one of the core-areas of the entire project since here a variety
of new definitions and development patterns will be suggested, which are directly based on the new dual
level epigenetic framework, introduced within the last two hundred pages ...

                                       4.1. BASIC DEFINITIONS

The first phenomenological step consists in integrating popular, yet largely undefined notions like
“knowledge societies”, “information societies” within the current epigenetic NIS-framework. Thus, the
notions of “knowledge societies” on the one hand and “information societies” on the other hand should
become different parts of the current NIS approach144. In doing so, two preliminary remarks must be given

       First, despite the synonymity of knowledge and information in the classical works on
       knowledge, production and information (MACHLUP 1962, 1980, MARSCHAK 1976,
       PORAT 1977), a new separation between “knowledge” and “information”is mainly justified
       by the different utilization contexts identified in Section II.
       Second, the proposed dual level differentiation for knowledge and information is only vaguely
       similar to a difference introduced first by Kenneth J. Boulding, namely to a focus on stock
       phenomena for knowledge and flow processes for information, since both stock and flow
       processes can be studied at the knowledge and information levels, too.
       Third, it must be added that the utilization contexts for knowledge and information proper are
       not restricted to a particular level alone, but are distributed across the genotype and across the
    For reasons of similarity with the concept of a “knowledge based economy”, the notion of “knowledge based NIS” has been chosen
instead of the concept of “Knowledge and Information Based National Innovation System”.

     phenotype levels. Thus, the seven contexts for knowledge identified in Section II are clearly
     located both at the genotype and at the phenotype level. Likewise, the four information
     domains can be studied in a phenotype and in a genotype setting, too.
     Fourth, the subsequent definitions for “knowledge and information domains” must be seen,
     consequently, as an additional suggestion independent of the established utilization contexts,
     specified so far.

In the spirit of the epigenetic framework, knowledge, information and innovations can be associated with
the genotype and phenotype levels in the following manner.

     The “NIS knowledge base” or, alternatively, the NIS-helix comprises structures, processes and
     histories of a specific NIS-ensemble (1) at the genotype levels of “embedded code systems”,
     “programs”, “memes”, etc. and (2) at the levels of genotype  phenotype relations.
     The “NIS information networks” integrate structures, processes and histories of a NIS-
     ensemble (1) at the phenotype levels of individuals, groups, institutions, economic sectors,
     etc. and (2) at the phenotype  genotype relations.
     “National Innovation Systems” encompass, finally, structures, processes and histories at the
     levels of NIS-knowledge bases (helices) and NIS-information networks.
     From its temporal boundaries, the term “National Innovation Systems” can be applied, in one
     of the family-types introduced in the previous chapters, to any large-scale formation after the
     emergence of a capitalist “world economy” and after the development of well-defined national
     boundaries, i.e. basically from the “long 16th century” onwards.
     Spatially, “National Innovation Systems” can be applied, within the temporal restrictions
     specified above, to highly developed regions or core areas of the “world-economy” as well as
     to peripheral or semiperipheral domains.
     Following the preceding chapter, National Innovation Systems can be studied as dual level -

                                   single systemic ensembles (NIS I)
                                     double configurations (NIS II)
                                         triple systems (NIS III)
                                      societal complexes (NIS-S)

     In addition, the present analysis of National Innovation Systems can be extended, mutatis
     mutandis, to lower spatial levels like Regional Innovation Systems (RIS) and to global
     investigations of GIS-development patterns ...

Table 4.1: The Multiple Regional Levels for Innovation Systems

                                            INNOVATION SYSTEMS

           WITH REGIONAL BOUNDARIES                                AT THE GLOBAL LEVEL

           single systemic ensembles (RIS I)                       single systemic ensembles (GIS I)
           double configurations (RIS II)                          double configurations (GIS II)
           triple systems (RIS III)                                triple systems (GIS III)
                          ...                                                ...
           societal complexes (RIS-S)                              societal complexes (GIS-S)

Fianlly, the basic architecture for National Innovation Systems, including both the knowledge and the
information concept, can be represented in the format of Table 4.2.

Table 4.2: The Architecture of National Innovation Systems

                                      INFORMATION NETWORKS

                                      P                   P

                BOTTOM-UP                                       TOP-DOWN
     EMBEDDEDNESS RELATIONS                                   EMBEDDEDNESS RELATIONS

                                      G                   G

                                      KNOWLEDGE HELICES (KNOWLEDGE BASES)

In this manner, an extremely important phenomenological aspect of National Innovation Systems, namely
their spatio-temporal boundaries as well as their systemic confines, has been completed.

                           4.2. FOUR MEGA - STAGES OF

                          CODE-BASED DEVELOPMENTS

One of the implicit justifications for the plausibility of the framework introduced so far lies in the “non-
intended side-effect” of arriving at a highly ordered temporal sequence of truly cosmic proportions which
becomes clearly visible in the stage-sequences of Table 4.3. Here, a severely modified adaptation from
BENINGER (1986:63) allows the identification of a spiral pattern in the unfolding of code-systems -

     Life is code-making .. There is no original. Everything melts into something else. Everything melts.
     There are no prototypes, only casts (TZONIS 1990:85)

Two additional remarks must be made with respect to the arrangement of Table 4.3.

     First, the stages are to be understood as additive or accumulative, implying that stage II has
     two essential sources for recombinations, namely a G  G base as well as a P  P
     mechanism of learning and adaptations, stage III, in addition, a P  G relation ...
     Second, in Table 4.3 one can see immediately a classical “dialectical” spiral of returning, after
     a number of intermediate steps, to an initial configuration, but at a “higher” state. Thus the
     stylized pattern of -

                               G (genetic code system)  P (organisms)
                           G (genetic recombinations)  P (new organisms)
                           P(organisms)  P (organisms, including humans)
           P (humans)  G (human code systems, including the biotechnological code-system)
                         G (human code systems)  G (genetic code-system)
          G (recombinations in biotech-code & recombinative operations)  P (new organisms)

     shows very clearly that in the “Meta Stage IV”, recombinations at the G-level occur within the
     rhythms and confines of human development clocks and are no longer bound to the biological
     clockworks which, for billions of years, have been the dominant “engines of change”.
     Moreover, due to this dialectical return at a different vertical state, the most important
     justification for postulating an essential phase transition within the present decades has been

Table 4.3: Time-Line of Four Basic Stages of Code-Based Developments

(logarithmic)                                         STAGE I:
                                                      Genotype  Genotype Interaction (Genetic Code)
                                                      G  P Generation (Genetic Code)
                                                      G  G Recombinations (Genetic Code)
1 billion                                      
                                                      “Cambrian Explosion”
                                                      STAGE II:
100 million                                           Phenotype  Phenotype Interaction
                                                      Learning by Imitations (“Implicit Knowledge”)
                                                      P  P Recombinations
10 million                                     
1 million                                          
                                                      STAGE III:
100.000                                               Phenotype  Genotype Interactions
                                                      Learning by Encoding (Constructions of Human Code-Systems,
                                                      especially Natural Languages and Pictorial Codes)
10.000                                         
                                                      Non-Pictorial Scriptures
                                                      P  G Recombinations
1000                                           
100 years                                      
                                                      STAGE IV:
10 years                                              Genotype (Human Code Systems)  Genotype Interactions
                                                      (Genetic Code (GC))
                                                      Genotype (GC)  Genoptype (Scrientific Language) 145
                                                      Genotype (Re)production (GC)  Operations in Biotechnology
                                                      Genotype Recombination (GC)  Recombinative Biotechnology
                                                      G  G Recombinations (Scientific Languages, Other Human
                                                      Code Systems)

    stands for a “transcription relation”, implying that a specific code-system has been transcribed or, alternatively, translated into another

code-system. “Transcription relations” occur quite frequently like in the case of “morse code  language code”, etc.

After two intermediate chapters on essential patterns of changes, a preliminary phenomenology of the
current phase transition from traditional NIS-ensembles to “Knowledge Based National Innovation
Systems” or, alternatively, to modern “knowledge and information societies” will be provided.

                       4.3. “CREATIVE CONSTRUCTIONS”

                      AND “CREATIVE DESTRUCTIONS” -


Before entering into the core dimensions of “Knowledge Based Innovation Systems”, a challenging
research task will be underatken by specifying, in a formal mode of description, those processes and
operators which, in sum, generate the trends and basic features of current National Innovation Systems,
including the current phase transition to “knowledge and information societies”. The reference point for
the subsequent explorations into the evolutionary dynamics of National Innovation Systems is, at least for
an Austrian contribution, an (almost) natural one, namely Joseph A. Schumpeter´s view of processes of
“creative destructions”.

     Der ... evolutionäre Charakter des kapitalistischen Prozesses ist nicht einfach der Tatsache
     zuzuschreiben, daß das Wirtschaftsleben in einem gesellschaftlichen und natürlichen Milieu vor sich
     geht, das sich verändert und durch seine Veränderung die Daten der wirtschaftlichen Tätigkeit ändert;
     diese Tatsache ist zwar wichtig, und diese Veränderungen (Kriege, Revolutionen usw.) bedingen oft
     auch eine Veränderung der Industrie; sie sind aber nicht ihre primäre Triebkraft. Auch ist dieser
     evolutionäre Charakter nicht einer quasi-automatischen Bevölkerungs- und Kapitalzunahme oder den
     Launen des Geldsystems zuzuschreiben, von denen genau das gleiche gilt. Der fundamentale Antrieb,
     der die kapitalistische Maschine in Bewegung setzt und hält, kommt von den neuen Konsumgütern,
     den neuen Produktions- und Transportmethoden, den neuen Märkten, den neuen Formen der
     industriellen Organisation, welche die kapitalistische Unternehmung schafft ... Die Eröffnung neuer,
     fremder oder einheimischer Märkte und die organisatorische Entwicklung vom Handwerksbetrieb und
     der Fabrik zu solchen Konzernen wie dem U.S.-Steel illustrieren den gleichen Prozeß einer
     industriellen Mutation - wenn ich diesen biologischen Ausdruck verwenden darf-, der unaufhörlich
     die Wirtschaftsstruktur von innen heraus revolutioniert, unaufhörlich die alte Struktur zerstört und
     unaufhörlich eine neue schafft. Dieser Prozeß der „schöpferischen Zerstörung‟ ist das für den
     Kapitalismus wesentliche Faktum. (SCHUMPETER 1975:137p.)

With the above quote from Schumpeter, one is confronted with notions like “mutations”, the
“evolutionary character of capitalist development”, “new markets”, “new products” and the like which
have the distinctive advantage of having been systematized within the new epigenetic framework laid out
so far. Following the Schumpeterian great vision, processes of creations and destructions or, alternatively,
of innovations and denovations in hyper-complex systems comprise a variety of generative features which
can be recapitulated in six main points.

         First, innovations as well as denovations must be seen as an endemic attribute of “complex
         epigenetic systems” (CES), consisting, at the genotype level, of recombinations within an
         embedded code-system fulfilling the five main code-requirements (see also Section II, chapter
         1) and, at the phenotype level, of action changes in a large number of distributed, autonomous
         units with a definite set of criteria for average fitness and comparative advantages. Thus,
         permanent processes of innovations and denovations at dual levels must be considered as the
         normal or reference case - and stationary reproductions as a special configuration only. This,
         in turn, implies, following Stephen Toulmin, a complete reversal with respect to “core” and
         “peripheral” concepts in the analysis of change and development. As a prominent example
         from the history of science, one may refer to the emergence of the Newtonian synthesis where,
         contrary to the previous research traditions, inspired mainly by Aristotle, “force” and “inertia”
         had become the core concepts - and “state of rest” a special, “derivative” term only. (See, e.g.
         COHEN 1977)
         Second, an essential NIS-area for creations comes through the “codification processes” P 
         G, i.e. through the encoding of actions, of socio-technical machines, of experiments within
         laboratories, the encodings of encodings and the like -

             Codification is a (generally) fuzzy program which is compiled in a suitable processor (a brain, a
             CIT-machine, etc.) and is undergoing actual or potential execution. 146

         Moreover, the encodings in the context of economic, scientific, artistic or other productions
         may turn out to be not only a tedious “task”, but also a transformation with erroneous and
         misleading solutions at the G-level. Not only that, generalizing Quine´s “indeterminacy of
         translation” into a more general encoding perspective, the resulting “tasks” are, by necessity,
         fuzzy and open to more than one “correct” or “justified” solution only.
         Third, at the G  G level, recombinations in programs lead, as has been pointed out in
         Section II, to a permanent reconfiguration of cognitive spaces and its fitness functions. As has
      The above quotation is a slight modification from Gordon Pask´s definition of concepts -

         A concept is a (generally) fuzzy program which is compiled in a suitable processor (usually a brain) and is undergoing
         actual or potential execution. (PASK 1995:244)

been demonstrated in the chapter on scientific creativity, these G-level “acts of creation” or,
alternatively, of recombinations can be compressed into a small class of “creativity operators”
through which existing programs can be re-arranged and re-configured into innovative or
denovative forms, depending on the overall contexts. Thus, the helices for National
Innovation Systems undergo a permanent process of “creation”, increasing the available G-
pool practically on an hourly basis.
Fourth, the other vertical route G  P, leading from the level of genotype programs to the
(re)production of action-sequences, machine components, products and, more generally, of
socio-technical systems, offers, once again, a wide array for “creations” and, subsequently, for
“destructions”. The transformation of machine designs into prototypes, the incorporation of
new machine designs into existing socio-technical configurations, the “execution” of a new
marketing strategy, etc. fall under this special type of creation processes, absolutely vital in the
case of National Innovation Systems. Moreover, the G  P-transition must be seen as a
process, which, like its P  G counterpart, may generate erroneous action sequences,
machine components, products and, more generally, socio-technical systems, too -

   Execution is a (generally) fuzzy operation which is performed in a suitable setting (organisms,
   wet machines, etc.) and is undergoing actual or potential action-sequences.

Fifth, at the P  P level, numerous changes and, thus “creations” take place, once again, like
in the G  G case, at a permanent basis. Upon closer inspection, any type of learning or
adaptation which resides within “the hands, the hearts and the brains” of individuals must be
qualified, according to the definitions for embeddedness relations in Section III, as an act of
“creation” and, consequently, of a “destruction” of an established practice. Moreover, heavy
emphasis must be placed on the “coupled nature” of many P  P alterations and
modifications, leading from an initial reconfiguration to a cascade of coupled modifications.
In this manner, the Schumpeterian notion of “new products” should be seen as a highly
complex phenomenon, surrounded by a dense array of P  P learning or, alternatively, of
“implicit knowledge”.
Sixth, processes of destructions can be described and analyzed as phenomena sui generis at
two different levels, too. Taking the science system as reference case, a lot of fascinating self-
organized destruction processes can be identified, culminating in the observation that science,
generally, has established very short decay-intervals since almost all of its productions forty
years ago, while full of interesting and still poorly explored research areas, are completely out
of actual utilization contexts. (For a detailed summary, see MÜLLER 1994) While still
physically present at the G-level, these programs have been reduced to a very small niche,
open for historians of science, library science and the like. (For an interesting analysis in this
area, see esp. LEM 1983) Likewise, “destructions” at the phenotype-level could be analyzed

       in principle, along a highly revealing dimension, namely along a reversible-irreversible axis.
       Here, probably the most fascinating research problem would consist in the identification of
       criteria, differentiating between reversible changes in action-sequences or in new product
       designs and irreversible modifications in socio-technical systems like the replacement of
       standardized “dominant designs” by entirely new alternatives or the phase transition from
       traditional NIS or, alternatively, labor societies to knowledge and information societies.

This short summary makes it abundantly clear that Schumpeter´s “creative destructions” occur not only in
spectacular P-events like the “opening of markets” or “new products” but in numerous activities and
changes at a duality of genotype and phenotype levels. In this manner, the genotype and phenotype-based
summary of the perennial creative and destructive changes leads to the next chapters which will present
important “mega-trends” of the très longue durée which capture some well-known processes as well as
some poorly understood development patterns.

                                    4.4. MEGA - TRENDS IN THE

                                                   EVOLUTION OF

                            NATIONAL INNOVATION SYSTEMS

In the present chapter, a material mode of description will lead to the identification of ten mega-trends
which characterize important aspects in the evolution of National Innovation Systems over the last
centuries. Following the spatio-temporal boundaries of National Innovation Systems specified in chapter
4.1, the decisive break in the evolution of societal formations has been accomplished through the
emergence of a self-propagated global economic system of integrated local markets, consisting of a
sufficiently large number of distributed units and of monetary exchanges and flows beyond the
“reachability” of political control mechanisms.147 In other words, market systems have surpassed, in an
irreversible manner, the barriers and thresholds of the “requisite varieties” of political steering and control
systems. Via this phase transition from many local societal systems to a global world economy, a large
number of well-defined regional, national or global fitness landscapes for economic systems and other
large scale ensembles has emerged which, in addition, are coupled in character. Within this evolution of
NIS evolutions, the ten historical core-processes or, alternatively, “mega-trends” can be summarized in
the following way.
   On this point, see especially Part I in Volume II which is centered around historical development patterns from an evolutionary, though
mostly phenotype-centered point of view.

First, at the end of the 18th century the scientific system as a whole has become
institutionalized in its modern version of an “autonomous unit”, recombining education and
research. (STICHWEH 1991, SWOBODA 1978) Moreover, the modern university system has
been linked to a far more significant degree with the state apparatus than with the economic
spheres (WAGNER 1990). Consequently, NIS II-ensembles, integrating the linkages between
science and economy only, miss a comparatively large proportion of societal innovation
dynamics since a core-actor both at the genotype and at the phenotype levels has not been
included properly.
Second, during the spectacular scientific growth processes especially after 1945 from “little
science” to “big science” systems (de SOLLA PRICE 1974), a “take-off phase” has occured
which has led to a period of “self-sustained growth” and to a global ensemble in which
processes of cognitive and spatial segmentations occupy an ever-increasing importance.
Moreover, the fitness landscapes within the scientific system, both at the genotype levels of
scientific programs and at the phenotype levels of institutes, have experienced a marked
process of disciplinary and regional differentiations, leading to well defined basic-applied
domains. to center-periphery relations or to advanced, intermediate and lagging types of
scientific production.
Third, the scientific system has reacted to the fast and vastly growing number of scientific
producers by a methodological as well as an epistemological device of “planned
obsolescence”, beautifully summarized by Robert M. Pirsig.

  If scientists had simply said Copernicus was right and Ptolemy was wrong without any
  willingness to further investigate the subject, then science would have simply become another
  minor religious creed. But scientific truth has always contained an overwhelming difference
  from theological truth: it is provisional. Science always contains an eraser, a mechanism
  whereby new Dynamic insight could wipe out old static patterns without destroying science
  itself. Thus science, unlike orthodox theology, has been capable of continuous, evolutionary
  growth ... The pencil is mightier than the pen. (PIRSIG 1991:222)

Following Table 4.4 below, the “scientific eraser” or, alternatively, the process of program
destructions within science must be considered as an inverse function of the number of
participating scientists.

       Table 4.4: The Self-Organization of Knowledge Destruction

       Number of Knowledge Producers                                      Knowledge Life Cycle148
              (in Persons)                                                  (in Years)

                      100                                                      103
                      104                                                      102
                      108                                                      101
                      1012                                                     100

       Due to the inverse relation - the higher the number of scientists at the global level, the shorter
       the average time span for replacing (almost) the entire “knowledge basis” - a variation to the
       Rousseau function on the dependence of important innovations (II) on the number of
       contributions C

                                                              II  C

       leads to the “self-organized destruction” within the science system -

                                                           LCt  K
                                                                      4    N

       where LC denotes the life-cycle of the knowledge bases, N the number of scientists and K a
       societal parameter, dependent on the long-term “sustainability” of the entire NIS-ensemble.
       Thus, assuming K to be in the range of 103 years, the resulting “knowledge turnover” or
       “replacement phase” assumes the following values.

       Table 4.5: The Self-Organization of Knowledge Destruction

       Number of Knowledge Producers                                      Knowledge Life Cycle
                  (in Persons)                                              (in Years)

                      104                                                      100.0
                      105                                                       56.2
                      106                                                       31.6
                      107                                                       17.8
                      108                                                       10.0
                      109                                                        5.6

   It should be added that the concepts of “knowledge” or “knowledge destruction” refer to the genotype levels of program production and
program destruction as they have been introduced in Section II.

       In this manner, scientific niches are permanently re-created and being filled anew, sometimes
       with very minor modifications only.149
       Fourth, following the classical studies in the sociology of science (MERTON 1985), a strong
       argument can be made that “universalism” or “communism” are to be considered as central
       long-term characteristics in the normative orientation of the modern science system and that,
       as a corollary, dense international networks can be identified from the early stages of the
       modern science system onwards. Moreover, switching to the genotype level, an interesting
       program feature can be detected throughout the past two centuries, namely the presence of
       program components which are considered of applicative relevance not to a single disciplinary
       domain only but for many different scientific disciplines. Thus, the evolution of scientific
       programs has brought about a comparatively large number of transdisciplinary helix
       components like the early Newtonian program, classical mechanics, the conception of
       “Unified Science” by the Vienna Circle, cybernetics or General Systems Theory, which, at
       least during a short interval of time, have been considered as an appropriate transdisciplinary
       basis for the scientific area as a whole.
       Fifth, the evolution of scientific disciplines has been accompanied by a comparatively small
       growth within the science of science fields, having led, in the long run, to the emergence of a
       notoriously weak spot in scientific analyses, namely the science system itself. Moreover, the
       institutionalization of philosophy of science departments especially after 1945 has contributed
       significantly to a neglect of the “dark empirical sides” of scientific production processes and
       to “stylized heroism” with respect to goal orientation and in-built rationality of the overall
       scientific system.

           The worst offender in this regard is possibly the philosopher of history, who is distinguished by
           having virtually no contact with the field of knowledge he is philosophizing about ... The basic
           premises with which they approach their subject are therefore laughably naive and simplistic.
           On the basis of such assumptions, however, a formidable logical machinery is set in motion,
           producing conclusions that cannot fail to reflect the quality of the underlying assumptions.
           (SWOBODA 1978:65)

       Sixth, a highly interesting secular shift can be observed with respct to the history of a central
       knoweldge utilization context, namely with respect to the development of encyclopedias or,
       alternatively, of transdisciplinary designs. This particular phase transition may be labeled as

    One of the most interesting research tasks especially in the social sciences lies in concrete analyses of scientific journals from sixty,
seventy or eighty years ago which, on the one hand, are composed of totally “forgotten” and “out-dated” articles and which, on the other
hand, contain materials, insights and discussions which, with very small adaptations and contemporary references, could be re-used within
current debates, too ...

change from accumulative types of encylopedic designs to structural ones. The “dominant
metaphor” in this area has been provided by Heinz von Foerster -

  Let me confess that I am a man who is weak in properly carrying out multiplications. It takes
  me a long time to multiply a two or three digit number, and, moreover, when I do the same
  multiplication over and over again in most of the time I get a different result. This is very
  annoying, and I wanted to settle this question once and for all by making a record of all correct
  results. Hence, I decided to make myself y multiplication table with two entries, one on the left
  (X) and one at the top (Y) for the two numbers to be multiplied and with the product (XY)
  being recorded at the intersection of the appropriate rows and columns ... In preparing this table
  I wanted to know how much paper I need to accomodate factors X, Y up to a magnitude of, say,
  n decimal digits. Using regular-size type for the numbers, on double-bond sheets of 81/2 x 11
  inch, the thickness D of the book containing my multiplication table for numbers up to n
  decimal digits turns out to be approximately

                                            D = n . 102n-6 cm

  For example, a 100 x 100 multiplication table ( 100 = 102; n = 2) fills a „book‟ with thickness

                               D = 2. 104-6 = 2. 10-2 = 0.02 cm = 0.2 mm

  In other words, this table can be printed on a single sheet of paper ... Now I propose to extend
  my table to multiplications of ten digit numbers. This is a very modest request, and such a table
  may be handy when preparing one´s federal Income Tax. With our formula for D, we obtain for
  n = 10:

                                          D = 10.1020-6 = 1015

  In other words, this multiplication table must be accomodated on a book-shelf which is 1015 cm
  long, that is, about 100 times the distance between the sun and the earth, or about one light day
  long. A librarian, moving with the velocity of light, will, on the average, require 1/2 day to look
  up a single entry in the body of this table.
  This appeared to me not to be a very practical way to store the information of the results of all
  ten-digit multiplications. But, since I needed this information very dearly, I had to look for
  another way of doing this. I hit upon a gadget which is about 5 x 5 x 12 inch in size, contains 20
  little wheels, each with numbers from zero to nine printed on them. These wheels are sitting on
  an axle and are coupled to each other by teeth and pegs in an ingenious way so that, when a
  crank is turned an appropriate number of times, the desired result of a multiplication can be

   read off the wheels through a window. The whole gadget is very cheap indeed and, on the
   average, it will require only 50 turns of the crank to reach all desired results of a multiplication
   involving two ten-digit numbers.
   The answer to the question of whether I should „store‟ the information of a 1010 x 1010
   multiplication table in the form of a 81/2 x 11 inch book 6 billion miles thick or in the form of a
   small manual desk computer, is quite obvious, I think. However, it may be argued that the
   computer does not „store‟ this information but calculates each problem in a separate set of
   operations. My turning of the crank does nothing but give the computer the „adress‟ of the
   result, which I retrieve at once - without the computer doing anything - by reading off the final
   position of the wheels. If I can retrieve this information, it must have been put into the system
   before. But how? Quite obviously, the information is stored in the computer in a structural
   fashion. In the way in which the wheels interact, in cutting notches and attaching pegs, all the
   information for reaching the right number has been laid down in its construction code, or, to put
   it biologically, in its genetic code. (FOERSTER 19995:314p)

By now, the main differences between accumulative and structural ways of “knowledge
storage” should be fairly obvious. The interesting long term trend to be observed lies in the
fact that the history of transdisciplinarity from the French Enlightenment to the present day
can be conceptualized as a phase transition from

   a period of high accumulation, best exemplified in the French Encyclopedie which can
   be characterized by a maximum degree of accumulative arrangements and a minimal
   degree of structural storages

to the present day stock of complex or, alternatively, transdisciplinary models which contain

   a maximum degree of structural “encodings” and a minimum degree of accumulation,
   namely a relatively small number of successful applications. (For more details, see
   MÜLLER 1994)

The phase transition in transdisciplinary designs may be considered, finally, as a corollary to
the third megatrend since accumulative forms of transdisciplinary composition are linked to
small scientific communities and structural types to a comparatively large number of
scientific producers.
Seventh, a significant acceleration in the speed of economic and, more generally, societal
evolution has been accomplished by the successful recombination of distributed forms of
market organization with equally distributed socio-technical machinery, giving rise to a
distributed system of low, medium and high technology firms, incorporating socio-technical

       systems of increasing complexity. Through this recombination at the start of the first of the
       subsequent five industrial revolutions150, an energy and power barrier has been overcome
       which, in previous societal formations, has become an essential impediment for a long-term
       period of growth and expansion. Moreover, the distributed firm structures have brought about,
       in turn, an overwhelming importance of and reliance on infrastructural segments - railroads,
       automobiles, communication and information technologies, etc. - which, so far, have played
       the role of leading sectors in periods of “High Innovations”.
       The most decisive consequence of the recombination between firm organization and scientific
       units was clearly the proliferation of new “investment outlets” or, alternatively, economic
       niches. Thus, the explorations in program spaces under the guidance and goals of concrete
       firm-objectives has brought about a relation of the following form -

                                                          dPI/dt  AI/dt

       I.e., science-induced potential innovations (PI) have been constantly higher than the actual
       utilization (AI) of innovations within the economic system.
       Eighth, within the emergence of a technologically advanced distributed firm system, one of
       the far-reaching phenotype  genotype-related processes has been achieved through the
       “encoding” of implicit knowledge in production processes. One of the most interesting
       descriptions of this “encoding strategy” can be found in one of the early modern codification
       experiments within the industrial-scientific sphere, namely in the French Enlightenment.

           Wir wandten uns an die tüchtigsten Handwerker in Paris und unserem Königsreich. Wir
           machten uns die Mühe, sie in ihren Werkstätten aufzusuchen, sie auszuforschen, nach ihrem
           Diktat Aufzeichnungen zu machen, ihre Gedanken zu entwickeln, aus diesen Gedanken die
           jeweils eigentümlichen Fachausdrücke zutage zu fördern, Verzeichnisse derselben anzufertigen,
           und sie zu erklären ... Es gibt Handwerkmeister, die gleichzeitig Schriftsteller sind und wir
           können sie hier nennen. Aber ihre Zahl ist sehr klein. Die meisten unter denen, die mechanische
           Künste ausüben, haben sie nur aus Not ergriffen und arbeiten nur unter der Leitung ihres
           Instinkts. Unter tausend findet man kaum ein Dutzend, die sich einigermaßen klar ausdrücken
           können, sei es in bezug auf die Werkzeuge, die sie benutzen, sei es in bezug auf die
   On this point, see once again Part I of Volume II, where, following Schumpeter and others (FREEMAN 1983, 1986,
FREEMAN/SOETE 1995, KLEINKNECHT 1987), the following five industrial revolutions have been distinguished for the last two-
hundred years. Turning to the “leading sectors”for each of these revolutions (ROSTOW 1978), one may add the following economic
segments which, moreoevr, exerted powerful forward and backward linkages and required, finally, a large number of adaptations in the
monetary apparatus, in the state-organization and the like (for the case of the German railroads, see, e.g. SPREE 1978):

                                   First Industrial Revolution (1780/90 - 1820): Textile Industry et al.
                                     Second Industrial Revolution (1850 - 1870/73): Railroads et al.
                     Third Industrial Revolution (1896 - 1913/20): Chemical Industry, Electric Steam Engines et al.
                         Fourth Industrial Revolution (1948 - 1966/73): Car Industries, Aircraft Industies, et al.
                     Fifth Industrial Revolution (1993 - ???): Communication- and Information Technologies et al.

  Werkstücke, die sie herstellen ... Es gibt so eigenartige Handwerke und so feine Verfahren, daß
  man über sie wohl nur dann treffend sprechen kann, wenn man selbst in ihnen tätig ist, eine
  Maschine eigenhändig bedient und sieht, wie das Werkstück unter den eigenen Augen entsteht.
  Wir mußten uns deshalb öfters Maschinen verschaffen, sie aufstellen, selbst Hand anlegen,
  sozusagen Lehrlinge werden und schlechte Werkstücke machen, um die anderen lehren zu
  können, wie man gute macht. So überzeugten wir uns von der Unkenntnis, in der man sich den
  meisten Gegenständen des Lebens gegenüber befindet und von der Notwendigkeit, aus dieser
  Unkenntnis herauszukommen. (DIDEROT 1969:48p.)

It goes without further comment that a successful encoding like the one undertaken in the
edition of the French “Encyclopédie” increases both the diffusion rate for advanced socio-
technical systems and the potential for local adaptations and reproductions considerably.
Moreover, a fascinating history could be written on the power struggle for “encoding
relations” within firms, starting from the early experiments of Frederick Winslow Taylor at
“Bethlehem Steel” -

  The managers assume, for instance, the burden of gathering together all of the traditional knowledge
  which in the past has been possessed by the workman and the task of classifying, tabulating and reducing
  this knowledge to rules, laws and formulae which are immensely helpful to the workmen in doing their
  daily work ... Almost every act of the workman should be preceeded by one or more preparatory acts of
  the management which enables him to do his work better and quicker than he otherwise could (TAYLOR
  1971:26pp.) -.

up to the present day arrangements of “quality circles” and “fractal firms” ...(See also
Nineth, roughly one hundred years after the successful “blending” of distributed firm
structures with socio-technical systems, another highly consequential recombination has taken
place via the integration of research laboratories on the one hand and firms on the other hand,
leading to intra-firm R&D activities which, by now, occupy, on the OECD average, more than
half of the total national R&D expenditures. The merging of R&D with industries and, at least
partially, services has been undergoing two different stages -

  The traditional approach, referred to ... as first generation R&D management, has been largely
  intuitive. Research and development is treated as an overhead item, and budgets are set in
  relation to some business measure (for example, sales) ... Within this budget framework,
  decisions about areas of concentration and project continuation may be left largely to R&D
  management ...In second generation, or systematic, R&D management, managers outside the
  R&D area participate in suggesting or reviewing projects ... Arguing that research and

           devlelopment projects are investments - as in a sense they are - corporate management seeks to
           base their justification on rate of return or payout. (MAGEE 1991:XIp.)

       Tenth, seen in a long term historical perspective, the “prosperity phases” of the “long swings”
       have generated a pattern of increasing “scientification” which can be demonstrated via the
       developmental course from textiles (first wave, low science contribution), railroads (second
       wave), chemical and electrical engineering (third wave), traffic systems (fourth wave) to the
       current wave of communication and information technologies and the future wave of
       biotechnologies which will continue to be almost exclusively dependent on (re)productive
       science and technology programs.

Thus, a long term megatrend of an increasing (re)productive relevance of the science generated
knowledge bases offers a direct link to the subsequent chapter of the knowledge and information shapes
to come. Moreover, these ten megatrends151 sum up, in a partially new perspective, the preceding
theoretical discussions on evolutions, creative constructions or creative destructions. What follows after a
long-term view on the history of traditional National Innovation Systems, is, by a method of inversion, an
equally long term perspective on the future of “modern times”.

                        4.5. CURRENT PHASE TRANSITIONS OF

                             NATIONAL INNOVATION SYSTEMS

Leaving aside the historical trend patterns, a short summary of twenty essential future-oriented features
will be presented which justify the recombination of “knowledge”, “information” or “innovation system”
   It should be noted that the enumeration of “megatrends” has been confined to the NIS II ensemble only, leaving aside many vital
“megatrends” within the NIS-S configuration. To mention just two.

       First, the last five hundred years have experienced, aside from the growth and diffusion patterns of the economy - science
       interrelationships, an international as well as a national differentiation of other large-scale systems of the “societal protective
       belt” like insurance schemes, labor organizations, national state legislations or international arrangements which, in sum,
       fulfilled the indispensable double-function of protecting individuals and the environment from destruction through the
       development of the distributed firm system on the one hand and of safe-guarding the future expansion of the market domains
       on the other hand ...151
       Second, the development of National Innovation Systems so far has been accompanied by a clear separation between
       household work and production processes especially within the economic system. Moreover, household work proper, though
       still the basis for societal (re)production, has been undergoing a proces of marginalization and gender discrimination, being
       restricted to an unpaid service area, mostly filled out by female activities only. Due to the centrality of both paid as well as
       unpaid forms of labor for the (re)production of the entire NIS-complex, both nationally and internationally, an alternative
       characterization of National Innovation Systems can be provided by the label of “labor societies” where labor is to be
       associated with human labor only.

with generalizations like “society” and “nation”. Moreover, the twenty “mega-characteristics” are, from a
long-term point of view, sufficiently specific to differentiate between contemporary forms of “knowledge
and information societies” or, alternatively, of “Knowledge Based National Innovation Systems” on the
one hand and previous societal architectures and constitutions on the other hand. Seen in a very long term
perspective, knowledge and information societies are the result of a current phase transition which, so far,
has been given various names, namely as a change from

                                       modern societies  postmodern societies
                                    traditional societies  posttraditional societies
                                          modernization I  modernization II
                                       modernization  reflexive modernization
                                    capital-labor  knowledge-information, etc.152

Each of these new states, as they are specified within current research programs, is seen, however, as an
unsatisfactory attempt at catching the “core engines” of epigenetic developments in contemporary
societies. Consequently, the phase transition within modernity will be postulated, first, in a formal mode
by distributing the information and knowledge parts in the following way:

        The “Knowledge Society” comprises a series of mostly new features (1) at the genotype levels
        of helices, i.e. of “embedded code systems”, “programs”, “memes” etc. and (2) at the levels of
        genotype  phenotype relations.
        The “Information Society” encompasses a set of novel characteristics (1) at the phenotype
        levels of networks between individuals, groups, institutions, economic sectors, etc. and (2) at
        the phenotype  genotype relations.
        “Knowledge Based National Innovation Systems” or, alternatively, “Knoweldge and
        Information Societies” integrate, finally, new elements at the levels of “Knowledge Societies”
        and at the levels of “Information Societies”.

Thus, the phase transition of -

      Traditional National Innovation Systems               Knowledge Based National Innovation Systems153

    Here, only some recent literature like BECK (1992, 1993), BECK/GIDDENS/LASH (1994), LYOTARD (1982), DRUCKER (1993),
LASH/SZERSZYNSKI/WYNNE (1996) or ZAPF (1994) has been included. It should be added that a huge number of similar phase
transitions could have been used instead like “postcapitalist society” (DAHRENDORF), “postmaturity economy” (ROSTOW), “computer
revolution” (HAWKES), “knowledge economy” (MACHLUP), “global village” (McLUHAN), “postcivilized area (BOULDING),
“technological society” (ELLUL), “posteconomic society” (KAHN), “technetronic era” (BRZEZINSKI), “age of information” (HELVEY),
“information economy” (PORAT), “telematic society” (NORA and MINC), “communications revolution” (WILLIAMS), “second industrial
divide” (PIORE and SABEL), “third wave” (TOFFLER) - and many others ...
    Following the specifications of a complex economic macro-model in Volume V, the phase-transition might be characterized in the
following manner, too, namely as a change from

                                       labor societies  knowledge and information societies

or. alternatively, focusing on the differences in the (re)production and maintenance of embeddedness
domains between a traditional mode, centered on human labor and a new mode, based on (re)productive
helices and interlinked networks, a change from -

                                  Labor Societies              Knowledge and Information Societies

has been chosen as the new epigenetic transition perspective. It should be added that this phase transition
occurs within an ongoing global evolution of a capitalist world economy. Moreover, in the historical part
of Volume II, it will be shown that societies prior to the second half of the 20th century, while equipped
with knowledge bases and with inter-systemic information networks, especially in the form of science-
economy networks, lack most of the vital elements for “knowledge and information societies” which have
come into existence in the last fifty years only and which become, consequently, the “differentia
specifica” of Knowledge Based National Innovation Systems.

                          4.5.1. MEGA - CHARACTERISTICS OF

                                      “KNOWLEDGE SOCIETIES”

The “principium divisionis” for modern “knowledge societies” lies in ten core-features at the genotype
levels of Knowledge Based National Innovation Systems and can be summarized as follows.154

       First, the number of scientific programs ready for exchange and transfer within the scientific
       system itself has reached, by now, a comparatively high level and will increase heavily in the
       future. Referring, once again, to the new mode of knowledge production (NOWOTNY 1995),
       “transdisciplinarity”, expressed in terms of the utilization of programs from a specific
       scientific discipline SDi within a different scientific domain SDj, will turn out to be, unlike the
       previous two centuries of disciplinary reclusions, a common routine practice. The most
       important cognitive basis for a strongly increased program exchange must be seen in the fact
       that, as will be shown in closer detail Part II of Volume II, a core-set of non-trivial models for,
       generally speaking, pattern formation and pattern modifications has become available whose
       applicability extends to all domains where “structure formation matters”, i.e. to practically any

    Due to the NIS II-character of the whole project, the main characteristics will be focused on the science-economy complex only. Thus,
the following set of “mega-features” should not be interpreted as a NIS-S picture.

type of “empirically” oriented scientific discipline both in the natural and in the social
sciences.This, in turn, will exert strong influences on the qualification structure of scientists
where “knowledge navigators”, traveling in the cognitive spaces of existing disciplines and
adapting available programs to the needs of a particular scientific field, will become of
paramount importance for this particular diffusion and exchange process.
Second, the proliferation of new science programs will run more and more in a self-referential
mode of “second order programing”. Program features, including topoi like science of science,
modeling of modeling, control of control, communications of communications, functions of
functions, evolution of evolution, understanding of understanding or, with a self-referential
turn, innovations of innovations and creative explorations of creative explorations, will
become established “normal science domains” without invoking new hierarchies at the
program level. Moreover, the growing number of approaches, perspectives and the increasing
potential for recombinations within the scientific program spaces will make it almost
imperative to change from “first order explorations” to “second order investigations” on
common principles and rules inhererent in different “first order analyses”.

Table 4.6: Ordering “Second Order”-Programs

AUTHOR                  FIRST ORDER                         SECOND ORDER
                        PROGRAMS                            PROGRAMS

von Foerster            Observed Systems                    Observing Systems
Müller                  Self-referentiality Excluded        Self-referentiality Included
Pask                    Purpose of the Model                Purpose of the Modeller
Umpleby                 Interaction among Variables         Interaction between Observer
                        in System                           and System Observed
Varela                  Controlled Systems                  Autonomous Systems

For an interesting ordering of the second order concept one may refer to a scheme developed
by Stuart Umpleby (1991) which, with some self-referential modifications and adaptations,
has been reproduced as Table 4.6. Accordingly, second order explorations are characterized,
by and large, by two distinct and independent requirements.

   On the one hand, scientific programs on a specific issuei like { ..., communication,
   control, evolution, knowledge, information, innovation, modeling, systems ... } must be
   devoted to an analysis of a large number of scientific approaches to problemsi in the

          area of { ..., communication, control, evolution, knowledge, information, innovation,
          modeling, systems ... }. (Second Level- Requirement)
          On the other hand, the investigation must exhibit in the course of analyzing the main
          chatacteristics of the topic under consideration the main characteristics itself. (Self-
          Referentiality Requirement)

       More generally, second order brain research, second order biology or second order social
       science on a specific domaini should be characterized by the following heterarchic
       configuration -

                              Second Order Programi  {Programsi}
                                          

       Consequently, an epigenetic theory preserves in its second order format the main features of
       epigenetic development just like a model of different approaches to modeling fulfills the
       criteria specified for adequate modeling and so forth ...155
       Third, the structure of scientific tasks, both of the internally and of the externally induced
       variety, will advance in a fashion which requires a permanent recombination of competencies
       and problem solutions across disciplines and which, therefore, must be qualified as
       transdisciplinary in nature, too. In other words, the growing network interlinkages at the level
       of societal Innovation Systems require, as their genotype corollary, densely interconnected
       programs from different disciplines as well. Moreover, scientific task-integration will play an
       ever increasing role in the (re)production of other systemic domains, especially in the
       (re)production processes of the economic system and, due to the growing intersystemic
       linkages, of the science system itself. The current configuration exhibits already, like in the
       case of the Austrian Innovation System described in Volume IV, the building up of dense
       program transfer flows both as a source for innovations and as quality control for their
       successful implementations. Thus, the local or regional program basis, i.e. “the power of the
       local helices”, will exert an increasing role for the local or regional innovation potential.
       Fourth, growing inter-systemic diffusions can be identified at the program level, making, on
       the one hand, societal “needs” or “objectives” more clearly visible in the specification of
       scientific tasks and, on the other hand, scientific programs more recognizable at the
       formulation of societal “needs” and “objectives”. Restricted to the G-level only-

          communication between research and society increasingly takes the form of diffusion processes
          that carry scientific and technological knowledge into society while social norms and

   For a concrete specification, see Volume V where the diffusion process of a special class of complex models, namely models using
master-equation, is analyzed within the specific framework of master equations ...

  expectations held by different institutions and communities are brought home more forcefully
  to the research communities. (GIBBONS et al. 1994:38)

Moreover, the new forms of integrating the actions and reactions of extra-scientific ensembles
more directly into the intra-scientific production of programs as well as into the genres of
distribution, will increase significantly since they are supported by a powerful evolutionary
pattern, namely by the irreversible increase in the self-organizing capacities of large scale
societal systems which, in a seemingly paradoxical manner, can only be accomplished in a
sustainable fashion by integrating more and more aspects of the environment into the internal
operations of the system itself.
Fifth, within science, very rapid “catching up” processes will be recorded especially in the
domains of the social sciences and in the humanities which, so far, have been characterized
either by “comparative disadvantages of backwardness” or by a “low” or “soft profile”. Thus,
due to the growing potential for program exchanges and for program adaptations from a
transdisciplinary model core, the current “malaise“ or “crisis” can be overcome in a rapid
manner. A few examples should be sufficient to demonstrate the enormous “catching up-
potential” which is currently being built up.

  Historiography will be able to use most of the complex model stock for the analysis and
  the simulation of long-term development patterns or structural changes. Moreover, the
  large simulation potential inherent in the complex model repertoire will offer an
  opportunity for new “experimental settings” which, so far, have been largely missing
  within the historiographic routines.
  Anthroplogy, recombined with the cognitive sciences, will find a vast and unexplored
  intellectual space in which, moreover, the enormous mass of empirical material can be
  successfully re-analyzed. By concentrating on learning models or on processes of
  recombinations and “analogy formation”, challenging insights will be extracted, based
  on the permanent laboratory called “society”.
  Sociology, or political science for that matter, may use any of the available
  transdisciplinary models for the analysis of structural changes both at the micro- and at
  the macro-level. It may very well be argued that the quick adaptation of complex
  models to the available “disciplinary matrices” will lead to a new “epistemic culture”
  (Karin Knorr-Cetina) within the areas of sociology or political science which will be
  characterized by attributes like “closed laboratory settings”, “possible world research”
  and the like. (For more details, see MÜLLER 1995b)
  Even the “hardest” of the social sciences, namely economics, will profit enormously by
  integrating, on the one hand, the cognitive science domain and, on the other hand, the
  evolutionary complex. The resulting areas of cognitive based “experimental economics”

           and of “evolutionary economics” may well lead to a profound paradigm change,
           substituting the established core and the periphery relations.

       Sixth, the integration of various embedded code-systems will proceed, once again due to the
       support of communication and information technologies, in the future at an even accelerated
       rate, giving rise to families of new hyper-programs which consist of pictorial code-systems,
       language codes, number codes and which, following Vilem Flusser (1988, 1989), are not
       arranged in a one-dimensional manner.

           On the one hand, “the society of text” (Edward Barrett) will change into the “society of
           hyper genres” where recombined code systems, predominantly pictorial in charcter, will
           replace the traditional code systems and genres.
           On the other hand, the interfaces between two and three dimensional pictorial codes,
           complex models and simulations will pave the way for a “science of hyper-genres”, in
           which various code-systems will act in a complementary manner, overcoming, thus, the
           in built limitations of the clusters of traditional scientific languages. (For a summary,
           see BRODLIE et al. 1992)

       Seventh, the helices of a knowledge society have aquired, over the last decade, an embodied
       technological format, namely the global internet which, in an astonishingly short amount of
       time, has evolved into a highly interesting genotype support system for the storage and
       retrieval of essential parts of societal helices, including the science helix. To give a self-
       referential example, the internet contained, by January 1996, a huge number of relevant
       programs on topics like knowledge, science policy or innovation systems.156 Moreover, the
       materials in the internet reflect, in a much more successful way than the on-going literary
       production, the current state of the art within scientific fields since programs can become part
       of the internet helix in the very moment of being “finished”, without the delays and the “gate-
       keeping processes” inherent in the publication of books, articles in journals, etc. Finally, the
       internet leads, totally unlike the traditional libraries, to a fusion of vast numbers of regional
       programs into a single, globally accessible program ...

    As a very small sample of the available materials, the following list of small and, at times, very large articles and surveys has been
compiled and ordered in a simple alphabetical manner - Bardige A., The Invention of Knowledge. The Unique Artifacts Theory; Colorado‟s
Model State Standards for Science Education; Education Human Resources for the Information and Library Professions of the 21st
Century; GNOSIS: Knowledge Systematization: Configuration System for Design and Manufacturing; Gaines, B. R. and M.L.G. Shaw, A
Learning Model for Forecasting the Future of Information Technology; Gaines, B.R. and M.L.G. Shaw, Eliciting Knowledge and
Transferring it Effectively to a Knowledge-Based System; Gaines, B.R. and M.L.G. Shaw, Knowledge Acquisition Tools based on Personal
Construct Psychology; Gaines, B. R. and M. L.G. Shaw,WebMap: Concept Mapping on the Web; Gaines, B. R., An Ounce of Knowledge is
Worth a Ton of Data: Quantitative Studies of the Trade-Off between Expertise and Data based on Statistically Well-Founded Empirical
Induction; Haywood T. - Info Rich/Info Poor; International Society for Knowledge Organization (ISKO) - Brochure No.3 (30.11.94);
Knowledge Science Institute (KSI) - Self-description; Martin, D. C., C. S. Ang,. Univ. of California, San Franc., Libraries in the Information
Age; RITE/LINCS: Research Instiute for Technologies in Education and Learning through Interactive Networked Collaboration and
Simulation - New Information; Schomaker L. - Navigation in Hyperspace and Cognitive Representation/Literature (12.12.94); T. Nicolas,
E. Cosyn et al. - Knowledge Space Assessment via the Web.

Eighth, the internet helix offers an immediate basis for recombinations, since a large number
of distinct search processes in space and time like the identification of relevant literature,
reading it, making notes, modifying specific points, etc. collapse into a single internet work
process. In a bewildering fashion, the recombinative potential of the internet has been has
been described by a fifty year old engineer, interviewed by Sherry Turkle.

  It´s like a brain, self-organizing, nobody controlling it, just growing up out of the connections
  that an infant makes, sights to sounds, ... people to experiences ... Sometimes I‟ll be away from
  the Web for a week and a bunch of places that I know very well will have „found‟ each other.
  This is not an engineering problem. It´s a new kind of organism. Or a parallel world. No point
  to analyze it. No way you could have built it by planning it. (TURKLE 1995:45)

Ninth, one of the most path-breaking characteristics of “knowledge societies” lies in the
subsequent relation -

                            genetic code  scientific language codes

which implies that “knoweldge societies”, in contrast to previous societal knoweldge
formations, are in command of changing not only their phenotype environments, but also their
genotype outfits in hitherto untouched environmental domains of plants, organisms and the
like. Thus, in knowledge societies, totally unlike previous societal knowledge bases, genetic
programs have become subjects and substance of science programs.
Tenth, one of the elementary new characteristcs of knowledge societies refers to the
phenomenon that special helix domains will take a similar path to the one which has been
followed in the long history of the genetic code. More concretely, the vertical (re)production
relations -


will become, in selected areas of the economy, the state or even the household domains, a self-
organized process of the type -

     program  program reproduction  self assembly (HOFSTADTER 1982:504pp.)

Thus, the simultaneous diffusion of communication and information technologies on the one
hand and biotechnologies along different dimensions from the nanoworlds
(CRANDALL/LEWIS 1992) to the “extended phenotypes” on the other hand will serve as one
of the essential “missing links” in the generation of self-assemblies which, especially in

       combination with the cognitive sciences, may very well become the dominant
       transdisciplinary research designs throughout the next century.

Seen in the light of the ten genotype features, most of the new characteristics cannot be identified in
societies of the first half of the 20th century or, a fortiori, in any other societal formation prior to the 20th
century. In this sense, the emergence of a “knowledge society” must be viewed as a phase transition which
has occured over the last five decades only - and which, as an irreversible trait of modern societies, is
highly likely to remain in the decades ahead.

                          4.5.2. MEGA - CHARACTERISTICS OF

                                    “INFORMATION SOCIETIES”

In a similar manner, the main-features of “information societies” or, alternatively, of the phenotypes of
Knowledge Based National Innovation Systems, can be described in the subsequent way. 157

       First, intrasystemic exchanges and cooperations within the science system become, to a
       significantly larger extent than during the period of “High Disciplines”, transdisciplinary in
       nature, leading to the institutionalization of research institutes with the explicit goal of
       pursuing transdisciplinary projects and to spin-offs from universities which, once again,
       exhibit a transdisciplinary composition of work-teams. Couched in terms of the innovation
       and creativity framework, a process of and reconfiguring              traditional disciplinary
       competencies has set in in order to arrive at transdisciplinary problem solutions in advanced
       areas like -

           theoretical neurophysics; the modeling of evolution, including the evolution of behavior;
           strategies to troublesome states of minds and associated higher brain functions; nonlinear
           systems dynamics, pattern recognition and human thought; fundamental physics, astronomy,
           and mathematics; archaeology, archaeometry, and forces leading to extinction of flourishing
           cultures; an integrated approach to information science; (or) the heterogeneity of genetic
           inventories of individuals. (COWAN 1988:236)

    It must be added as a “Caveat lector” that, once again due to the NIS II-character of the whole project, the mega characteristics of
“information societies” will be centered on the science-economy domain only ...

Second, under the heavy impact of communication and information technologies, the
scientific output and its utilization contexts will change in the future from the dominant types

                                       Pt  Gt and Gt+k  Pt+k

into a coupled form -

                                 Gt, Pt  Gt+l, Pt+l       with (l < k)

Thus, the starting point will be given by a specific “task” at a particular point in time t under
the availability of socio-technical systems at the P-level and of programs at the G-level.
Subsequently, the final product will consist in a new socio-technical system and, at least
partially, in new programs at the genotype level. This process which, under the heading of
“unity between the context of discovery (G-level) and application (P-level)”, has been found
as the characteristic trait for the Mode II manner of scientific production (GIBBONS et al.
1994), will become a wide-spread and standardized manner of scientific production.
Third, the transdisciplinary task structures in many technologically advanced fields and in
(almost) any important societal domain will and must be accompanied by a growing
heterogeneity in the intra-scientific network composition, by the temporality of intra-scientific
network relations and, finally, by “multiple cognitive identities” -

    The capacity to cooperate with experts from other fields, to come to see the world and its
    problems in a complementary way and to emphatise with different presuppositions, involves the
    capacity to assume multiple cognitive and socieal identities ... Biologists working in
    environmental science, computer scientists in the analysis of gene sequences and
    mathematicians in ecological modelling can equally gain reputation on both their native and
    new grounds. (GIBBONS 1994:149)

Fourth, an increased “resonance” (Niklas Luhmann) can be identified with respect to the
communication networks at the intersections of science, public, media, interest groups.

    The previous one-way communication process from scientific experts to the lay public
    perceived to be scientifically illiterate and in need of education by experts has been supplanted
    by politically backed demands for accountability of science and technology and new public
    discussions in which experts have to communicate a more „vernacular‟ science than ever
    before. (GIBBONS et al. 1994:36)

Moreover, the new forms of dense outside communication networks will be likely to increase
since they are supported via a secular development pattern, namely via a growing number of
highly educated and scientifically trained persons in advanced societies.

   The new demands for accountability and for more communication between the community of
   scientific and technical experts and the „attentive‟ public are interconnected and emanate from
   the spread of higher education through society. (IBID.)

In this manner, processes of “science meets the public”or “science meets the media” will
become part of the day to day routines of scientists in the same manner as the “scramble for
public funds” or the games of persuasion have occupied a central stage even in the periods of
“Undisturbed Theory”.
Fifth, another highly interesting phenomenon of “information societies” can be seen in the fact
that the physical spaces of scientific production will become more heterogeneous, too. What
one can observe here, are at least two different processes.

   On the one hand, the “settings” for scientific production will become more widely
   distributed, ranging from the traditional sites of universities, research institutes or R&D
   laboratories in industries to different areas like the media, the service sector, especially
   to the firm related services like managment consulting, insurance and banking, to
   special interest groups, to political parties or to hybrid communities “at the boundaries
   and in the spaces between systems and subsystems”. (GIBBONS et al:37)
   On the other hand, each of these settings produces a highly specialized know-how or, as
   will be shown in the last chapter of Volume V, a specific “know-do” on specific societal
   domains which, if organized in a recursive manner, may offer interesting “cognitive
   disturbances” for the established institutions of program proliferation. The booming
   literature on new management strategies which, to a large extent, originates outside the
   traditional places of science, offers a highly interesting example. Here, the first step in
   the “power of recursion” has set in already, since many of these “postmodern”
   management approaches have adopted the terminology of the “sciences of complexity”,
   by introducing concepts like “chaos”, “fractals”, “self-organization”, “evolution” and
   recombining them with the core-notions of the domain like “firms”, “management”,
   “strategies” ... (For a tiny sample, see BURRUS/GITTINES 1994,

Sixth, the inter-systemic networks, especially the linkages between science and economy, are
advancing at a distinctly higher and more densely interlinked regional manner. Above all, the
interlinked networks in the area of conducting economic innovations, be it in form of new

products, new processes or new organizations, become necessary pre-requirements for their
successful implementation and maintenance. Thus, a clear sign of “information societies” can
be seen in the markedly increased interdependencies between economic production processes
and scientific networks or, more functionally defined, scientific support infrastructures. This
process may be labeled as “third generation R&D” -

   to distinguish it from the primitive hands-off „strategy of hope‟ or the somewhat more
   systematic but incomplete project-management approach. Third generation research and
   devleopment management is a continuous interactive process. It demands active dialogue and a
   sense of partnership in technology among the leadership of R&D and other key managers
   focused on business strategy ... This style of R&D management requires regular review of the
   R&D project portfolio in relation to product and market strategy. (MAGEE 1991:XII)

Thus, “third generation R&D” implies that research and development on the one hand and
goals or strategies at the level of firms or other institutions on the other hand become
entangled in a recursive fashion.
Seventh, the “implicit knowledge” relation of -


will be supplemented by another embeddedness relation of the same basic format which can
be written as -

                                       P [SCIT]  P [SCIT]

and which refers to the support systems of communication and information technologies.
While this specific source for learning processes has emerged with the distribution of
telephone networks, radio and television throughout the 20th century already, the interactions
and relations in the age of information societies will be distributed, to a far more significant
degree than in previous societal formations, between “synchronous time” person to person
relations on the one hand and relations in “asynchronous time” and CIT mediated on the
otrher hand.

   Virtual communities ranging from MUDs (Multi User Domains, K.H.M.) to computer bulletin
   boards allow people to generate experiences, relationships, identities, and living spaces that
   arise only through interaction with technology. (TURKLE 1995:21)

Eighth, again at the -


embededness relation of “information societies”, a significant increase, at least in principle,
can be recorded with respect to the accessibility of the “knowledge bases”, regional, national
and global. More generally, single work places or households in “information societies”, once
again due to the support of CIT, have acquired, compared to previous decades and centuries, a
significantly higher direct access to the available “helices” in scientific, artistic, cultural
sytemic domains. The runaway growth of the Internet as well as its current prices may even be
considered as a “compensational device”, running counter to established reachability barriers
of helix components in scientific and economic highly advanced areas. Moreover, under the
vision of an “Open Data Network” two major building blocks, aside from the entrance and
exit options for service and network providers, have been specified which should guarantee
the “catching ip” potential of the Internet in the near and distant future. Thus, the “Open Data
Network” -

   does not force users into closed groups or deny access to any sectors of society, but permits
   universal connectivity, as does the telephone system ... It permits the introduction of new
   applications and services over time. (NRENAISSANCE COMMITTEE 1994:44)

Ninth, processes of codifications or, to be more precise, the embeddedness relations of the P
 G format, will be intensified in the manner of -

                                      P [SSIM]  G [SSIM]

I.e., codification will experience an astonishing and unprecedented rate of progress, especially
through the use of semi-intelligent machines (SIM), based on self-adapting learning programs
like neural networks, genetic algorithms, fuzzy systems and the like which can act as “feature
detectors” and as “observers” in the epistemological sense of the word. Moreover, along the

                                      P [SSIM]  G [SSIM]

relations, a new reconfiguration of the cognitive side of production processes, both within
science and outside, is set in motion which -

   is based on machines that extend, multiply and leverage our mental abilities. (KURZWEIL

     Tenth, it goes almost without specifying a specific mega-trend that the CIT-support will and
     must lead to a new dimension with respect to “the power of the global” irrespective of the
     growing importance of the local networks. Thus, production processes in all major societal
     ensembles, economy, science, culture and others - will be undergoing a rapid process of
     globalization. With respect to the economic system exclusively, globalization manifests itself
     along two different dimensions.

        On the one hand, due to low transport costs, production processes can be separated,
        distributed globally and re-assembled at some specific sites, service work, including
        quality control, consulting or testing, can be distributed across different physical spaces.
        Along the first dimension, a new set of global and globally interacting acteurs has
        emerged which, contrary to the “multinationals” of the preceding decades may be
        referred to as “transnational corporate networks”. (See e.g., SPYBEY 1996:75pp.,
        RUIGROK/TULDER 1995)
        On the other hand, globalization in the age of CIT-support implies a change which has
        been beautifully summarized in the slogan “bits for atoms” -

           In the same ways that hypertext removes the limitations of the printed page, the post-
           information age will remove the limitations of geography ... If instead of going to work
           by driving my atoms into town, I log into my office and do my work electronically,
           exactly where is my workplace? (NEGROPONTE 1996:165)

With the above features of “information societies”, a basic “transformation sketch” has been completed.
Combined with the ten mega-characteristics for “knowledge societies”, a comprehensive picture on the
basic architectures of contemporary OECD-societies has emerged. Moreover, this new fabric of genotype-
phenotype interactions has, by now, at least an additional advantage since it exhibits two significant U-
turns in the core of societal architectures.

     First, the heterogeneity in the composition of scientific networks far away from the
     disciplinary boundaries and the multiplicity of the settings for science production outside the
     conventional scientific arena marks a significant U-turn of a long term development pattern of
     nearly two hundred years.
     Second, the return of households as places for market production and,.moreover, their re-
     integration into the global production networks may be considered as an even more profound
     U-turn since the gradual differentiation of household work from market oriented production
     has been one of the characteristic traits in the emergence of the new world system.

Finally, it would be relatively easy to transform the twenty “mega-characteristics” of contemporary
knowledge and information societies into societal trends by postulating, in general, positively sloped
trajectories for each of these features. In this manner, a comprehensive long-term outlook on future
explorations and standardizations of fitness landscapes both at the G- and P-levels as well as, finally, in
the creation or destruction of P-level and G-level niches can be obtained which, due to the empirical bases
offered in Volume IV, differ markedly from the current trend for “trends” (For a critical summary, see
RUST 1995)

                                                 5. PROCESSES OF

                                      SCIENTIFIC INNOVATIONS

                                    AT THE PHENOTYPE-LEVEL:
                                          AN EXEMPLAR-BASED


The final chapter in the theoretical part is aimed at bringing the components, accumulated in the course of
Section I to Section IV together for a single cause so that the “comparative advantages” of the entire
epigenetic program becomes more transparent. Moreover, the problem to be analyzed, namely the one of
scientific creativity at the phenotype level, has four peculiar features.

       First, it offers a view on the micro- and macro-constitution of scientific systems as well as on
       one of its characteristic development patterns.
       Second, it belongs to the category of very poorly understood processes158, due to its “shoelace
       structure”.159 Moreover, it exhibits, as an unintended side effect, that, due to the magic of self-
       referentiality, the science system has become the scientifically least comprehended large scale

   For an evolutionary justification on the diffulties in understanding recombinative processes, see e.g. RIEDL 1987.
   An interesting description of “bootstrapping” or “shoelace processes” can be found, surprisingly, in SOROS (1994:42pp). Here, the
“shoelace structure” is depicted via a pair of recursive finctions -

                                                                y = (x)
                                                                x = (y)

which “do not produce an equilibrium but a never-ending process of change”. (IBID.)

      Thirdt, it turns out to be of utmost importance for the dynamics and the well-functioning of
      National Innovation Systems and for their distribution power.
      Fourth, its mastering at the cognitive level should offer a new basis for innovation-oriented
      science- and technology policies.

Thus, Chapter 5.2. will bring, after a final comparison of the present dual level approach to National
Innovation Sysytems with another well-known “actor network program”, combining genotype- and
phenotype levels too, the explorations into “theory spaces” to a halt ...

                     5.1. AGENCIES, THE SCIENCE-HELIX,

                                  AND MACRO-CHANGES
                              AT THE PHENOTYPE-LEVEL

Before entering into a detailed discussion of the Austrian social science networks prior to 1933/38, a short
comparison seems appropriate between the present meta-evolutionary or, alternatively, transdisciplinary
framework and another controversial network approach, namely Bruno Latour´s “actor networks”. The
first part will be concentrated on some striking similarities between these two programs, the second block,
not surprisingly, on prevalent differences.

      First, an interesting identity can be detected on the level of agencies, since in both cases
      “agency” refers to a multitude of components, ranging from practices of scientists to
      “agency”-operations at the level of plants, microbes, amoebae, the animal kingdom, the
      explosions of a “supernova” and the like. The embeddedness relations in Section III have
      introduced a very definite position, highly similar to the following quote from Bruno Latour -

         In order to map the development of a scientific controversy or of technical innovation, the STS
         field has learned to doubt the dichotomy between nature on the one hand, and society on the
         other ... Alternative narratives have been developed under the heading „actor-network theory‟
         that stress the heterogeneity and variability of associations of humans and non-humans.
         (LATOUR 1992b:33)

Second, the internal-external distinction has lost its epistemically privileged position since the
scientific system has been built up ab ovo as an interlinked and relational ensemble in which
most interesting processes of innovations and diffusion occur in an interplay of internal and
external factors, leading to interlinked relationships of the format -

     Internal Processes or Structures =  (Internal and External Processes or Structures)
     External Processes or Structures =  (External and Internal Processes or Structures)

Thus, the innovation and diffusion process of research programs needs its local and global
scientific alliances for the promotion of research programs. Likewise, outside linkages and
coalitions to the political-administrative sphere, to political parties, to firms, to media, to
minority groups and the like become essential and co-decisive instances determining the
future trajectory of research programs -

   What we call society and what we call science were indissolubly combined in the work of
   numerous scientists, politicians and soldiers. (LATOUR 1994:611)

Third, a decisive importance has been attached in both approaches to the realm of scientific
texts and to their socio-economic contexts. In the present approach, this combination is
brought about by a dual level distinction between code-systems and their embedded
environments and by including natural languages as a specific type of code-system. In
Latour´s case, one finds observations and general statements on the unity between texts and
their environment like the following -

   A scientific text is not only a more or less transparent medium to convey information to the
   author´s scientific colleagues, nor is it only a document to help historinas, psychologists, or
   sociologists retrieve the state of the mind of its author or the context it has been written. As
   many decades of literary theory have helped us see, texts are a little bit less and a good deal
   more tha information and document. They build a world of their own that can be studied as
   such in relative and provisional isolation from the other aspects. They are localized events, with
   their own matter and their own practice. (LATOUR 1992a:129)

Fourth, a radical symmetry condition is imposed in the case of the present dynamic network
approach, too. Taking the self-referential models for science dynamics in Volume V as prime
example, one is confronted with the possibility of modeling four different configurations with
practically the same set of equations, namely the wide-spread diffusion of -

                                Strongly corroborated research programs
                                Weakly corroborated research programs

     as well as the non-diffusion of -

                                Strongly corroborated research programs
                                Weakly corroborated research programs

     Thus, the present evolutionary approach is definitely not confined to the area of success
     histories alone, but is capable of explaining all possible configurations in an, traditionally
     formulated, “externalist manner”. In this sense, Latour´s symmetry condition can be upheld in
     the case of the present evolutionary framework as well.

Despite clear “family resemblances”, one can detect very easily some deep-going differences between
Latour´s “actor-networks” and the present evolutionary network-approach.

     First, an important divergence can be detected at the level of research designs and data
     generation. In Latour´s case, one is confronted with a qualitative approach, stressing the
     history of single cases or debates in the history of science. The current NIS approach, on the
     other hand, attempts a quantitative micro and macro picture of ongoing research and transfer
     processes as well as a genotype description of NIS-helices. Here, one is confronted with both
     a massive data and indicator generation, arriving at empirical patterns of National or Regional
     Innovation Systems en bloc.
     Second, the present evolutionary approach has been built up as a meta-theoretical or, less
     hierarchy oriented, as a transdisciplinary framework, using concepts, relations and
     perspectives characteristic for any type of evolutionary system, biological, social and
     otherwise. In Latour´s case, an integrated and very comprehensive approach for the history of
     “society-in-science” has been identified which, however, sticks to the disciplinary domain of
     the historiography of science. Within the current epigenetic perspective, a conceptual
     apparatus has been built up which can be employed, in principle, in biology, ecology,
     sociology, historiography, anthropology and the like ...
     Third, an important ingredient in the present evolutionary approach comes, as can be seen
     especially in Sections II and III, from the contemporary discussions in “cognitive neuro-
     sciences” which, hopefully, will shed new and empirically guided light on processes of
     scientific creativity, on the cognitive structure of paradigm changes and “Gestaltwechsel” or
     on the shape of intellectual fitness landscapes. In Latour´s case, a largely descriptive, though
     highly illustrative account on problem spaces and problem solutions can be found which
     fulfills essentially the indispensable function of a “true story” (Arthur Danto), highly

     characteristic for historical accounts. Nevertheless, “true stories” like the “intentional stance”
     will, in due course, become part of the “folk social science” repertoire and, in all probability,
     will undergo a process of sharply decreasing scientific utilization in the near future...
     Fourth, an essential difference can be seen in the model-utilizations compatible with these
     two “actor-network”-approaches. While it has become the explicit intention of the present
     evolutionary outlook to have a comfortable intersection with the domain of currently available
     complex models, no similar ambition, despite Latour´s incorporation of “socio-technical
     graphs” (LATOUR et al. 1992) can be recorded. Moreover, another significant difference is
     marked by the utilization contexts, since simulations play, quite naturally, an essential part in
     the evolutionary approach, while they are of little or no relevance in Latour´s program.

Thus, the final part in the section on NIS-families will be devoted to the problem of macro-scientific
innovations or, alternatively, of “creativity” for which all the necessary ingredients have been introduced
and defined. Thus, a recombination on recombinations will yield, hopefully, a new explanation scheme
for the generation of an, namely of a large number of cognitive innovations within a small spatio-
temporal region.

                                 5.2. MACRO-PATTERNS

                            OF SCIENTIFIC CREATIVITY

Hence, the final part in the theoretical section will be devoted to a phenotype-framework for processes of
scientific creativity and innovation which is based on, which corresponds to and which supplements the
genotype-version introduced in Section II. More specifically, four network conditions will be formulated
which can be seen as necessary conditions for a big and creative spurt within a small region in a relatively
short period of time of thirty to forty years. Thus, the phenomena under investigations should be
concentrated - aside from the Viennese case - on examples like the Scottish Enlightenment centered
around Edinburgh, the French Enlightenment in Paris or Berlin after the First World War up to 1933. In
all these cases one will find the following five features which, in combination, have generated a
“bootstrapping-effect” towards extraordinary levels of scientific achievements.

     Sufficiently Large Number of Network-Components with „a Keen Sense of What Is
     Interesting”: A creative or innovative scientific network at the phenotype-level requires, first,
     a “critical mass” of distributed creative or, alternatively, innovative phenotype units like

        institutes, research groups, individual scientists and the like. At first sight, this condition
        carries all the charms of an outright tautology. However, a clear separation can be made
        between a highly entropic ensemble of irregular and isolated creative problem solutions - like
        “bubbles” at the surface of boiling water - and a configuration in which the same number of
        creative network components leads to a highly ordered creative “big spurt” throughout the
        scientific network. Thus, the quintessential point for scientific creativity as mass-phenomenon
        at the phenotype level must lie in a very special list of attributes which should be present aside
        from a critical group of creative units. This additional list is located in an area which,
        following Heinz von Foerster, acts as a powerful order-generator, namely in the domain of
        “order by disturbances”.

            Ich möchte ... zwei Mechanismen als wichtige Schlüssel zum Verstehen selbstorganisierender
            systeme nennen: den einen können wir nach Schrödingers Vorschlag das Prinzip „Ordnung aus
            Ordnung‟ nennen, den anderen das Prinzip „Ordnung durch Störung ... In meinem Gasthaus
            ernähren sich daher selbstorganisierende Systeme nicht nur von Ordnung, für sie stehen auch
            Störungen auf der Speisekarte. (FOERSTER 1985:125p.)

        Thus, the following four points will be concentrated on internal as well as on external
        disturbance processes which, combined with the requirement of a sufficiently large number of
        distributed innovative units, may lead to discontinuous upward movements in creative or,
        alternatively innovative scientific outputs ...160
        “Phase Transitions in Epistemic Regimes and Great Visions”: A typical disturbance factor
        lies, second, in the transition from a long-established “epistemic regime” (Björn Wittrock) to
        a new one. Creative mass-movements within small regions are to be expected with a much
        higher probability during transition intervals since the intensity of search processes for new
        theoretical schemes will be greatly enhanced. Moreover, the phenomenon of “epistemic
        regimes” does not only refer to the cognitive code-spaces alone, but comprises the
        technological, institutional and societal settings as well. Thus, scientific creativity within
        small regions should reach their peaks, to modify a phrase from Christopher Langton, “at the
        edge of cognitive chaos” or, to be more precise, between a global phase transition from a long
        established to a new “epistemic culture” (Karin Knorr-Cetina). During these intervals, search
        processes for radically new schemes of ordering should, in all probability, increase and,
        moreover, the acceptance of new perspectives far away from the existing “dominant research
        designs” should reach its maximum. Another important point is worth being mentioned.
        Global revulsions and restructurings in cognitive space can be generated by to two broad

    For a comprehensive survey of the Austrian networks in the First and in the Second Republic see especially MÜLLER 1988, 1994a and
1995. The simple criterion of having a keen sense of what is interesting at the level of scientific disciplines is sufficient to distinguish clearly
between research groups in the social sciences prior to 1933/38 and ensembles after 1945. To summarize the main difference in a single
sentence, Vienna was a highly interesting place for scientists from outside whereas after 1945 one was virtually forced to leave Austria in
order to get a better undestanding of contemporary advances and discussions within the social sciences ...

       domains, namely by cognitive recombinations on the one hand and by socio-economic
       revolutions and drastic reconfigurations on the other hand. 161 Though not impossible a priori,
       scientific creativity at a massive scale throughout a variety of disciplines will be extremely
       difficult to achieve amidst the standardization phase of dominant research programs.162
       Network-Flows - Local Solutions and Heterogeneity: Third, the most decisive cognitive
       determinant for a period of dense creative activities within a small region consists, on the one
       hand, in the existence of local solutions for the reconfigurations taking place in the global
       intellectual space and, on the other hand, in a sufficient degree of heterogeneity, allowing for
       different perspectives, approaches and research programs. Take a typical note of a participant
       in the highly creative Viennese network up to 1934, namely of Friedrich A. Hayek -

           Haberler sagte (mir), es sei gerade ein sehr gutes, kritisches Buch von „einem von ihnen” (den
           Logischen Positivisten, K.H.M.) erschienen, von Karl Popper ... Ich las das Buch und fand dort
           alles, was mir wichtig war. Ich war nie ein geschulter Philosoph. aber ich hatte das, was man
           eine hypothetisch-deduktive Methode nennt, für meinen eignenen Zweck entwickelt und fand
           bei Popper meine eigene Einstellung wissenschaftlich begründet. Ich habe die Logik der
           Forschung über Nacht gelesen. Es war ein ähnliches Erlebnis wie das mit England. Poppers
           Denken war mir so kongenial wie das englische Denken (KREUZER 1983:17p.) -

       then one sees the “disturbance character” as well as the micro-mechanisms of these
       disturbances. A local solution within the scientific network to a global problem, namely the
       reconfiguration of science and demarcation criteria between “science, good, bad and bogus”,
       leads to an adaptation process on part of another network-participant, in this case the austro-
       liberal group. In turn, the new austro-liberal outlook on the status of economic science leads to
       a discussion with austromarxist economists who, in turn, react to the austro-liberal position ...
       Thus, local solutions create an “echo-effect” and can induce, under an appropriate ensemble,
       an internally propagated dynamic which, decades later, has turned into an explicable “creative
       singularity” ...163
       The necessity for heterogeneity is in most instances fulfilled in a trivial sense since phase
       transitions and the emergence of new epistemic regimes are coupled at least with the
       interaction process between a “traditional” ensemble and its “modern” counterparts. However,
       more heterogeneous instances can be found in those networks where, like in the case of
       Vienna, the modernist camp within the social sciences can be separated into two different
       highly innovative groups, namely into austro-liberals and austro-marxists ... (On this point,
       see MÜLLER 1987, 1988)
    One can observe an interesting coincidence between broad emancipatory societal movements on the one hand and the construction of
new forms of encyclopedias on the other hand, reflecting the new state of development. For a short summary, see MÜLLER 1994b.
    In the Austrian case, one can clearly observe that both the internal developments in the global scientific arena and the external socio-
economic environments exhibited radically different patterns.
    A typical “singularity” description can be found, for example, in HALLER 1986:108.

Network-Features - High Degrees of Interlinkages and “Criticality”: Fourth, even a network
of very innovative units, furnishing a heterogeneous set of local solutions to a reconfiguration
process of the global intellectual space may be insufficient to generate a period of “High Local
Theory”. In order to do so, the distributed units must be interlinked in a dense manner, leading
to an overall local network-configuration with heavy local flows between the distributed units.
It goes almost without saying that a high local density has been observed in the case of the
scientific “fin de siecle”, and “fin de republique” networks both from scholars outside
(SCHORSKE 1981) and from inside.
Dense local interconnections become especially important since a hypothesis from systems-
dynamics can be proposed, stressing the importance of “critical thresholds”. These “critical
barriers separate between dense exchange flows and rare exchange occurrences in a
discontinuous manner. Above the critical thresholds, creative reconfigurations and adaptations
can take place in a rapid way and “bootstrapping-processes” of scientific innovations can
emerge. Below the critical barriers, in the subcritical domain, the erratic pattern of isolated
“bubbles” prevails. Thus, the phenomenon of “self-organized criticality”, if corroborated by
empirical investigations, would point to another interesting dynamical feature, namely to the
sensitivity to initial conditions which are able to differentiate, despite similar distributions in
scientific output, between a “big jump” and a continuous slow and sparse exchange
Network-Environments - Turbulences: Fifth, reference must be made to a final disturbance
factor, namely to turbubulences outside the scientific system as well. The two main reasons
for demanding external turbulences as additional elements can be summarized in the
following way.

   On the one hand, external turbulences may lead to an increase in the utilization of the
   science system and, moreover, to a change in the linkages between science and other
   societal systems.
   On the other hand, heavy environmental turbulences affect the cognitive space
   significantly, especially, but not exclusively, in the social sciences.

It would go well beyond the scope of the present section to work out this point in close detail,
but some additional hints will be provided in Volume VI on science and technology policies
which will be devoted, inter alia, to the problem of stimulating genotype as well as phenotype
creativity, too.

                         6. NATIONAL INNOVATION SYSTEMS -

                                      TENTATIVE CONCLUSIONS

It would be relatively easy to continue from here to a comprehensive and genuinely evolutionary program
on NIS II, NIS III or NIS-S architectures, stressing the importance of the institutional settings, the
epigenetic structures of different innovation types or the recombinative character of new technologies or
new knowledge. Due to “real time” restrictions and due the focus on the empirical features of National
Innovation Systems however, the theoretical section will and must come to an end at this particular point
of having established a set of macro-requirements for regional innovation processes at an unusually high
scale. After all, the explanation sketch offered here has brought to light an unconventional class of
conditions whose presence and absence will be studied in Volume IV on the empirical helix
characteristics and network features of the Austrian Innovation System.164 At this particular juncture in
cognitive space, some tentative conclusions will be re-presented which, by and large, have been specified
as the overall goals of the theory volumes already. Thus, in the eyes of the observing author, the following
points have been reached and fulfilled.

       First, a new epigenetic theory background has been built up in which the two conflicting
       views on National Innovation Systems, the first one as a knowledge based perspective, the
       second one as a learning-oriented approach, have been combined under partially new
       headings like “genotype-and phenotype-levels”, “embedded code systems”, “genotype-
       phenotype-interactions” and the like.
       Second, the “great evolutionary vision”, or, to quote Daniel C. Dennett, “Darwin´s dangerous
       idea” (DENNETT 1995), has been taken very seriously, furnishing a homogeneous
       evolutionary basis which serves, inter alia, as an evolutionary “fitting” foundation for the
       subsequent empirical investigations and, more generally, for analyses in the area of
       evolutionary economics, evolutionary sociology or, most generally, of epigenetic social
       sciences for that matter.
       Third, the new conceptual framework fulfills moreover, an important additional function since
       it contributes substantially in the construction and in the interpretation of the results of the
       Austrian Survey of Innovation and Transfer (ASIT). To mention just one example, the meta-
       theoretical innovation concept has led to a substantial widening with respect to innovation
       domains in ASIT, leading to the inclusion of “organizational innovations” as a new element
       alongside the traditional domains of product and process innovations.

    On the point of the overall relevance of the new meta-theoretical or, alternatively, transdisciplinary framework, see also Volume V of the
current project on complex modeling.

     Fourth, the epigenetic perspective on code-based reproductions of embedded or dual level
     systems, i.e. on the “Two Great Chains of Becoming”, has opened up radically new pathways
     for the comparative analysis of socio-technical and of biological evolution, beautifully
     summarized in two sentences by Stuart Kauffman -

        Organisms arise from the crafting of natural order and natural selection, artifacts from the
        crafting of Homo sapiens. Organisms and artifacts sodifferent in scale, complexity and
        grandeur, so different in time scales over which they evolved, yet it is difficult not to see
        parallels. (KAUFFMAN 1995:191)

     Fifth, the present evolutionary framework allows, moreover, a transdisciplinary model-
     analysis on changing development patterns in National Innovation Systems. Why? Simply
     because a heavy emphasis has been placed on the construction of a meta-theoretical
     framework or, alternatively, of a core-apparatus which should be applicable, mutatis
     mutandis, for any type of evolutionary system, ranging from the biological areas to the
     domains of technical systems, scientific theories or National Innovation Systems, too. Since a
     variety of complex models have been successfully applied to the biological, the ecological or
     the neural domains already (See, e.g. GALE 1990, KOSSLYN/ANDERSEN 1992), the
     “transdisciplinary apparatus” should serve as an appropriate “bridging component” for the
     utilization of complex models in socio-technical systems or National Innovation Systems, too.
     Sixth, the new theoretical perspective becomes of tantamount importance in the radical shift
     with respect to science and technology policies undertaken in Volume VI, where the dual level
     NIS-framework will open up entirely new ways for establishing high density genotype-
     networks, closely interlinked phenotype-ensembles and the like.

Thus, the self-proclaimed goal areas, introduced at the beginning of Section IV, have been reached and,
moreover, implemented in a {C, E, S}-framework where complexity and codes, epigenesis and evolution,
and, finally, systems and structures have been re-combined into a new transdisciplinary ensemble which
has left the traditional confines of social science analyses and which, by doing so, has offered new and
highly promising trajectories within the “spaces of possible NIS-investigations”.


AARON, H.J., C.L. SCHULTZE (1992)(eds.), Setting Domestic Priorities. What Can Government Do? Washington:The Brookings Institution.
ABERNATHY, W.J., K.B. CLARK, A.M. KANTROW (1983), Industrial Renaissance. New York:Basic Books.
ABRAMOVITZ, M. (1989), "Notes on Postwar Productivity Growth. The Play of Potential and Realization", in: International Seminar on Science,
       Technology and Economic Growth. Paris:OECD.
AGAZZI, E. (1991), The Problem of Reductionism in Science. Amsterdam:Kluwer.
AHRWEILER, G., P. DÖGE, R. RILLING, R. (eds.) (1994/5). Memorandum Forschungs- und Technologiepolitik 1994/5 - Gestaltung statt
       Standortverwaltung, in: Forum Wissenschaft, Studien Bd. 26. Marburg: BdWi-Verlag.
AICHHOLZER, G., R. MARTINSEN, J. MELCHIOR (1994), “Technology Policy under Conditions of Social Partnership: Development and Problems of an
       Integrated Strategy in Austria”, in: G. AICHHOLZER, G. SCHIENSTOCK (eds.) (1994). Technology Policy. Towards an Integration of Social and
       Ecological Concerns.Berlin:de Gruyter, 375-403
AICHHOLZER, G., R. MARTINSEN, J. MELCHIOR (1994). Österreichische Technologiepolitik auf dem Prüfstand. Vienna:IHS Series Political Science.
AICHHOLZER, G. SCHIENSTOCK (eds.) (1994). Technology Policy. Towards an Integration of Social and Ecological Concerns.Berlin:de Gruyter
AIGINGER, K. et al. (1992), Industriepolitik 2000. Studie des Österreichischen Instituts für Wirtschaftsforschung. Vienna:WIFO.
AINSLIE, G. (1992), Picoeconomics. The Strategic Interaction of Successive Motivational States within the Person. Cambridge:Cambridge University Press.
ALEMANN, U.v. et al.. (1992), Leitbilder sozialverträglicher Technikgestaltung. Ergebnisbericht des Projektträgers zum NRW-Landesprogramm "Mensch
       und Technik – Sozialverträgliche Technikgestaltung". Opladen:Westdeutscher Verlag.
ALEMANN, U.v., P. JANSEN, H. KILPER, H., L. KISSLER (1988) (eds.), Technologiepolitik. Grundlagen und Perspektiven in der Bundesrepublik
       Deutschland und in Frankreich. Frankfurt:Campus.
ALEXANDER, J. C. et al. (1987)(eds.), The Micro-Macro Link. Berkeley:University of California Press.
ALEXANDER, J. (1994), “Modern, Anti, Post, and Neo: How Social Theories Have Tried to Understand the „New World‟ of „Our Time‟”, in: Zeitschrift für
       Soziologie 3, 165 - 197.
ALEXANDER, J.C. (1995), Fin de Siècle Social Theory. Relativism, Reduction, and the Problem of Reason. London:Verso.
ALVESSION, M. (1995), Management of Knowledge-Intensive Companies. Berlin:de Gruyter.
AMENDOLA, M., J.L. GAFFARD (1989), Macro-Dynamic Analysis of the Innovation Process. Issues and Methodological Approaches, in: International
       Seminar on Science, Technology and Economic Growth. Paris:OECD.
ANDERSON, A.E., J. MANTSINEN (1980), “Mobility of Resources, Accessibility of Knowledge, and Economic Growth”, in: Behavioural Science 25, 5.
ANDERSON, A. E. (1981), “Structural Change and Technological Development”, in: Regional Science and Urban Economics 11, 351 - 361.
ANDERSON, J.R., R. THOMPSON (1989), “Use of Analogy in a Production System Architecture”, in: VOSNIADOU, ORTONY, 267 - 297.
ANDERSON, P.W., K.J. ARROW, D. PINES (1988)(eds.), The Economy as an Evolving Complex System. The Procewedings of the Evolutionary Paths of
       the Global Economy Workshop, Held September, 1987 in Santa Fe, New Mexico. Redwood City:Addison-Wesley.
ANDREASEN, L.E., B. CORIAT, F.d. HERTOG, R. KAPLINSKY (1995)(eds.), Europe´s Next Step: Organisational Innovation, Competition and
       Employment. Ilford:Frank Cass.
ARIETI, S. (1976), Creativity. The Magic Synthesis,.New York:Basic Books.
ARONOWITZ, S., W. DIFAZIO (1994), The Jobless Future. Sci-Tech and the Dogma of Work. Minneapolis: The University of Minnesota Press
ARROW, K.J., S. HONKAPOHJA (1985)(eds.), Frontiers of Economics. Oxford:Basil Blackwell.
ARROWSMITH, D.K., C.M. PLACE (1994), Dynamische Systeme. Mathematische Grundlagen. Heidelberg:Spektrum Akademischer Verlag.
ARTHUR, W.B. (1989), The Economy and Complexity, in: D.L. STEIN (1989), 713 - 740.
ASHBY, R. (1981), Mechanisms of Intelligence: Ross Ashby´s Writings on Cybernetics, ed. by R. Conant. Seaside:Intersystems Publications.
ASHENFELTER, O., R. LAYARD (1986)(eds.), Handbook of Labor Economics, Bd.2. Amsterdam.
ATTENBOROUGH, D. (1995), Das geheime Leben der Pflanzen. Wie Pflanzen sich orientieren, verständigen, fortbewegen, ums Überleben kämpfen - eine
       neue Sicht der Pflanzenwelt. Bern:Scherz Verlag.
AVENI, A.F. (1989), Empires of Time. Calendars, Clocks and Cultures. New York:Basic Books.
AVERILL, J.R., E.P. NUNLEY (1992), Die Entdeckung der Gefühle. Ursprung und Entwicklung unserer Emotionen. Hamburg:Kabel Verlag.
AYRES, R. (1984), The Next Industrial Revolution. Reviving Industry through Innovation. Cambridge.
AYRES, R. (1989), "Information, Computers, CIM and Productivity", in: International Seminar on Science, Technology and Economic Growth.
AYRES, R.U., U.E. SIMONIS (1994)(eds.), Industrial Metabolism: Restructuring for Sustainable Development. Tokyo:United Nations University Press.
BACHARACH, M., S. HURLEY (1994)(eds.), Foundations of Decision Theory. Issues and Advances. Oxford:Blackwell Publishers.
BACKHOUSE, R.E. (1994)(ed.), New Directions in Economic Methodology. London:Routledge.
BAIER, B. (1996), “Cultural Tourism - the Next Generation”. Gibt es einen Kulturtourismus im virtuellen Raum? Vienna:IHS Research Paper.
BAILY, M.N., A.K. CHAKRABARTI (1988), Innovation and the Productivity Crisis. Washington.
BALZER W., C.U. MOULINES, J.D. SNEED (1987), An Architectonic for Science. The Structuralist Program. Dordrecht:Reidel.
BALZER W., D.A. PEARCE, H.J. SCHMIDT (1984), Reduction in Science. Structure, Examples, Philosophical Problems. Dordrecht:Reidel.
BARKOW, J.H., L. COSMIDES, J. TOOBY (1992)(eds.), The Adapted Mind. Evolutionary Psychology and the Generation of Culture. Oxford:Oxford
       University Press.
BARLOW, C. (31993)(ed.), From Gaia to Selfish Genes. Selected Writings in the Life Sciences. Cambridge:The MIT Press.
BARLOW, H.B. (1972), „Single Units and Sensation: a Neuron Doctrine for Perceptual Psychology?“, in: Perception 1, 371 - 394.

    The publications in the bibliography contain, on the one hand, bibliographical references throughout the seven final volumes of the NIS-
project and, on the other hand, interesting publications which have been relevant in the shaping and structuring of the final reports.

BARLOW, H.B. (1994), „The Neuron Doctrine in Perception“, in: M.S. GAZZANIGA (1995)(ed.), The Cognitive Neurosciences. Cambridge:MIT Press. 415
       - 435.
BARNSLEY, M. (1988), Fractals Everywhere. Boston:Academic Press.
BARNSLEY, M.., S.G. DEMKO (1986)(eds.), Chaotic Dynamics and Fractals. San Diego:Academic Press.
BARRETT, E. (1989)(ed.), The Society of Text. Hypertext, Hypermedia, and the Social, Construction of Information. Cambridge:The MIT Press.
BARROW, J.D. (1996), Die Natur der Natur. Wissen an den Grenzen von Raum und Zeit. Reinbek bei Hamburg:Rowohlt.
BARTON, G.E., R.C. BERWICK, E.S. RISTAD (1987), Computational Complexity and Natural Language. Cambridge:The MIT Press.
BASIEUX, P. (1995), Die Welt als Roulette. Denken in Erwartungen. Reinbek bei Hamburg:Rowohlt.
BATTEN, D., J.L. CASTI, B. JOHANSSON (1987)(eds.), Economic Evolution and Structural Adjustment. Berlin:Springer.
BAUER, F.L., G. GOOS (21973/74), Informatik. Eine einführende Übersicht, 2 vol. Berlin:Springer.
BAUM, E.B. (1988), "Neural Nets for Economists", in: P.W. ANDERSON, K.J. ARROW, D. PINES (1988), 33 - 48.
BAYER, K., M. PENEDER, M., F. OHLER, W. POLT (1993), Zwischen Rohstoff und Finalprodukt. Die wirtschaftliche Wettbewerbsfähigkeit des
       Wirtschaftsbereiches Holz-Papier. Vienna:TIP.
BAYER, K. et al. (1994), Technologiepolitisches Konzept 1994 der Bundesregierung. Vienna:BMWF.
BECHER, G. (1993), “Industrielle Forschungs- und Technologiepolitik in der Bundesrepublik. Instrumente, Wirkungen, Meßmethoden - am Beispiel von
       Fördermaßnahmen zugunsten von kleinen und mittleren Unternehmen”, in: KÖNIG, K., N. DOSE (eds.), Instrumente und Formen staatlichen
       Handelns. Verwaltungswissenschaftliche Abhandlungen, Vol. 2, Köln:Karl Heymanns Verlag, .453-493.
BECHMANN, G., T. PETERMANN (eds.) (1994), Interdisziplinäre Technikforschung. Genese, Folgen, Diskurs. Franfurt:Campus.
BECK, U. (1986), Risikogesellschaft. Auf dem Weg in eine andere Moderne. Frankfurt:Sihrkamp.
BECKER, G. (1981), A Treatise on the Family. Cambridge:Harvard University Press.
BECKER, J., B. WAGNER, K. WEINER, (1988)(eds.), EUREKA. Westeuropäische Technologiepolitik im Spannungsfeld wirtschafts- und
       sicherheitspolitischer Interessen. Protokoll des Marburger Symposiums am 4.-5.12.1987. Marburg:Universitätsdruckerei.
BEENSTOCK, M. (21984), The World Economy in Transition. London:George Allen&Unwin.
BEER, S. (1994a), Decision and Control. The Meaning of Operational Research and Management Cybernetics. Chichester:John Wiley&Sons.
BEER, S. (1994b), The Heart of Enterprse. Companion Volume to „Brain of the Firm‟. Chichester:John Wiley&Sons.
BEER, S. (21994c), Brain of the Firm. Companion Volume to „The Heart of Enterprise‟. Chichester:John Wiley&Sons.
BELL, D. (1979a), Die nachindustrielle Gesellschaft. Reinbek:Rowohlt.
BELL, D. (1979b), Die Zukunft der westlichen Welt. Kultur und Technologie im Widerstreit. Frankfurt:Fischer.
BENDALL, D.S. (1983)(ed.), Evolution from Molecules to Men. Cambridge:Cambridge University Press.
BENIGER, J.R (1986), The Control Revolution. Technological and Economic Origins of the Information Society. Cambrisge:The MIT Press.
BENNETT, C.H. (1988), "Dissipation, Information, Computational Complexity and the Definition of Organization", in: D. PINES (1988), 215 - 233.
BERGER, J. (1986)(Ed.), Die Moderne - Kontinuitäten und Zäsuren. Soziale Welt, Sonderband 4. Göttingen.
BERLINSKI, D. (1986), Black Mischief. The Mechanics of Modern Science. New York:William Morrow and Company.
BERTALANFFY, L. (1968), General System Theory. Foundations, Development, Applications. New York:Harper&Row.
BEYME, K.v. (1995), “Steuerung und Selbstregelung. Zur Entwicklung zweier Paradigmen”, in: Journal für Sozialforschung 3/4, 197 - 217.
BIEHL, W. (1981), Bestimmungsgründe der Innovationsbereitschaft und des Innovationserfplges. Eine empirische Untersuchung von
       Investitionsentscheidungen mittelständischer Maschinenbauunternehmen, Berlin.
BIERFELDER, W.H. (1989): Innovationsmanagement. München:DTV.
BIJKER, W.E., J. LAW (1992)(eds.), Shaping Technology/Building Society. Studies in Sociotechnical Change. Cambridge:The MIT Press.
BILGRAMI, A. (1994), Belief and Meaning. The Unity and Locality of Mental Content. Cambridge:Basil Blackwell.
BINGHAM, J.E., G.W.P. DAVIES (21978), A Handbook of Systems Analysis. London:Mc Millan.
BINSWANGER, H.C. et al.(1983)(Ed.), Arbeit ohne Umweltzerstörung. Strategien für eine neue Wirtschaftspolitik. Frankfurt am Main.
BIRKE, L., R. HUBBARD (1995), Reinventing Biology. Respect for Life and the Creation of Knowledge. Bloomington:Indiana University Press.
BLÜMLE, G. (1975), Theorie der Einkommensverteilung. Eine Einführung. Berlin:Springer.
BMWF (1993), Forschungsförderungen und Forschungsaufträge 1992, Faktendokumentation der Bundesdienststellen für 1992 . Vienna:BMWF.
BMWF/BMUK (eds.) (1992), Diversification of Higher Education in Austria. Background Report to the OECD. Vienna:BMWF/BMUK.
BODEN, M.A. (1990), The Creative Mind. Myths and Mechanisms. London:Cardinal.
BODMER, W., R. McKie (1994), The Book of Man. The Human Genome Project and the Quest to Discover Our genetic Heritage. New York:Scribner.
BOLDRIN M. (1988), Persistent Oscillations and Chaos in Dynamic Economic Models: Notes for a Survey, in: P.W. ANDERSON, K.J. ARROW, D.
       PINES (1988), 49 - 75.
BOOLE, G. (1958), An Investigation of the Laws of Thought on Which Are Founded the Mathematical Theories of Logic and Probability . New York:Dover
BORGES, J.L. (1974), Die Bibliothek von Babel. Stuttgart:Reclam Verlag.
BORTZ, J., N. DÖRING (21995), Forschungsmethoden und Evaluation. Berlin:Springer.
BÖS, D., H.D. STOLPER (1984)(eds.), Schumpeter oder Keynes? Zur Wirtschaftspolitik der neunziger Jahre. Berlin:Springer Verlag.
BOUDON, R. (1979), Widersprüche sozialen Handelns. Neuwied:Luchterhand.
BOUDON, R. (1980), Die Logik gesellschaftlichen Handelns. Eine Einführung in die soziologische Denk- und Arbeitsweise. Neuwied:Luchterhand.
BOULDING, K. (1981), Ecodynamics. A New Theory of Societal Evolution. Beverly Hills-London.
BOURDIEU, P. (1982), Die feinen Unterschiede. Kritik der gesellschaftlichen Urteilskraft. Frankfurt:Suhrkamp.
BOURDIEU, P. (1983), "Ökonomisches Kapital, kulturelles Kapital, soziales Kapital", in: R. KRECKEL (1983), 183 - 198.
BOURDIEU, P. (1985), Sozialer Raum und 'Klassen'. Leçon sur la leçon. Zwei Vorlesungen. Frankfurt:Suhrkamp.
BOURDIEU, P. (1987), Sozialer Sinn. Kritik der theoretischen Vernunft. Frankfurt:Suhrkamp.
BOURDIEU, P. (1988), Homo academicus. Frankfurt:Suhrkamp.
BOURDIEU, P. (1991), Language and Symbolic Power. Cambridge:Cambridge University Press.
BOYD, R., P.J. RICHERSON (1985), Culture and the Evolutionary Process. Chicago:The University of Chicago Press.
BOZEMAN, B., J. MELKERS (eds.) (1993), Evaluating R&D Impacts: Methods and Practise. Boston:Kluwer Academic Publishers.
BRAUDEL, F. (1982), Civilization and Capitalism 15 th - 18th Century, Vol. 2 The Wheels of Commerce. New York: Harper&Row.
BRAUDEL, F. (1986), Civilization and Capitalism 15 th - 18th Century, Vol. 3 The Perspective of the World. New York: Harper&Row.
BRAUN, C.-F. v. (1994): Der Innovationskrieg: Ziele und Grenzen der industriellen Forschung und Entwicklung, München:Hanser.
BREUER, H. (1995), dtv-Atlas zur Informatik. Tafeln und Texte. München:dtv.
BROCKHOFF, K. (1979): Delphi-Prognosen im Computer-Dialog. Experimentelle Erprobung und Auswertung kurzfristiger Prognosen. Tübingen:J.C.B.

BRODLIE, K.W. et al (1992)(eds.), Scientific Visualization. Techniques and Applications. Berlin:Springer.
BROOKS, D.R., E.O. Wiley (21988), Evolution as Entropy. Toward a Unified Theory of Biology. The University of Chicago Press.
BROOKS, R.A. (1992), “Artificial Life and Real Robots” in: F.J. VARELA, P. BOURGINE, op. cit., 3 - 20.
BROSE, P. (1982), Planung, Bewertung und Kontrolle technologischer Innovationen. Berlin.
BROTCHIE, J.F., P. HALL, P.W. NEWTON (1987)(eds.), The Spatial Impact of Technological Change. London-New York:Croom Helm.
BROWNING, H.L., J. SINGELMANN (1978), "The Transformation of the U.S. Labor Force: The Interaction of Industry and Occupation", in: Politics and
        Society 3-4, 481 - 509.
BRUCKMANN, G. (1977)(ed.), Langfristige Prognosen. Möglichkeiten und Methoden der Langfristprognostik komplexer Systeme. Würzburg:Physica
BUCHANAN, J.M. (1986), Liberty, Market and the State. Political Economy in the 1980s. Brighton:Harvester Press.
BÜHL, W. L.(1992), "Vergebliche Liebe zum Chaos", in: Soziale Welt 1.
BÜHL, W.L.(1995), Wissenschaft und Technologie. An der Schwelle zur Informationsgesellschaft. Göttingen:Verlag Otto Schwartz&Co.
BUND, MISEREOR (1996)(eds.), Zukunftsfähiges Deutschland. Ein Beitrag zu einer global nachhaltigen Entwicklung. Basel:Birkhäuser Verlag.
BUNDESKAMMER DER GEWERBLICHEN WIRTSCHAFT (ed.) (1981). Wissenschafter für die Wirtschaft. Wien:BWK.
BUNDESWIRTSCHAFTSKAMMER (1987): Kooperative Forschungsinstitute. Leistungen, Zielsetzungen und Probleme. Vienna:BWK.
BUNGE, M. (1977), Treatise on Basic Philosophy. Ontology I: The Furniture of the World. Dordrecht:Reidel.
BUNGE, M. (1979), Treatise on Basic Philosophy. Ontology II: A World of Systems. Dordrecht:Reidel.
BUNGE, M. (1983a), Treatise on Basic Philosophy. Epistemology and Methodology I: Exploring the World. Dordrecht:Reidel.
BUNGE, M. (1983b), Treatise on Basic Philosophy. Epistemology and Methodology II: Understanding the World. Dordrecht:Reidel.
BURKS, A.W. (1970), Essays on Cellular Automata. Urbana:University of Chicago Press.
BURRUS, D., R. GITTINES (1994), Technotrends. How to Use Technology to Go Beyond Your Competition. New York:HarperBusiness.
BURTSCHER, K. (1994), Innovation als Interaktion von Wirtschaft, Wissenschaft und Politik. Vienna:TIP.
CALLON, M. et al. (1988), Mapping the Dynamics of Science and Technology.
CALVIN W.H. (1990), The Cerebral Symphony. Seashore Reflections on the Structure of Consciousness. New York:Bantam Books.
CALVIN, W.H., G.A. OJEMANN (1995), Einsicht ins Gehirn. Wie Denken und Sprache entstehen. München:Carl Hanser Verlag.
CALVIN, W.H. (1996), Wie der Schamane den Mond stahl. Auf der Suche nach dem Wissen der Steinzeit. München:Carl Hanser Verlag.
CAMPBELL, J. (1984), Grammatical Man. Information, Entropy, Language, and Life. Harmondsworth:Penguin.
CAPRA, F. (1996), Lebensnetz. Ein neues Verständnis der lebendigen Welt. Bern:Scherz Verlag.
CARNAP, R. (21961), Logical Foundations of Probability. Chicago:University of Chicago Press.
CARNAP, R., R. JEFFREY (1971)(eds.), Studies in Inductive Logic and Probability, vol. 1. Berkeley:University of California Press.
CARR, M. (1985)(ed.), The AT Reader. Theory and Practice in Appropriate Technology. London:Intermediate Technology Publications.
CARTWRIGHT, N. (1989), Nature´s Capacities and Their Measurement. Oxford:Clarendon Press.
CARVALLO ; M.E. (1988)(ed.), Nature, Cognition and System I. Current Systems-Scientific Research on Natural and Cognitive Systems. Dordrecht:Reidel.
CASDAGLI, M., S. EUBANK (1992)(eds.), Nonlinear Modeling and Forecasting. Redwood City:Addison-Wesley.
CASTI, J.L. (1983), „Emergent Novelty and the Modeling of Spatial Processes“, in: Kybernetes, 167 - 175.
CASTI, J.L. (1985), Nonlinear System Theory. Orlando.
CASTI, J.L. (1986), "Metaphors for Manufacturing: What Could it be Like to be a Manufacturing System?", in: Technological Forecasting and Social
        Change 29, 241 - 270.
CASTI, J.L. (1988), "Linear Metabolism-Repair Systems", in: International Journal of General Systems 14, 143 - 167
CASTI, J.L. (1989a), "Newton, Aristotle and the Modelling of Living Systems" in:. CASTI J.L., A. KARLQVIST (1989d), 47 - 89
CASTI, J.L. (1989b), "(M,R) Systems as a Framework for Modelling Structural Change in a Global Industry, in: Journal of Social and Biological Structures
        12, 17 - 31.
CASTI, J.L. (1989c), Alternate Realities. Mathematical Models of Nature and Man. New York et al.
CASTI J.L., A. KARLQVIST (1989d)(eds.), From Newton to Aristotle. New York:Basic Books.
CASTI, J.L. (1992), Reality Rules, 2 Bde. New York et al.
CASTI, J.L. (1994), Complexification. Explaining a Paradoxical World through the Science of Surprise. New York:Harper Collins Publishers.
CASTI, J.L., A. KARLQVIST (1995)(eds.), Cooperation and Conflict in General Evlutionary Processes. New York:John Wiley-Interscience Publication.
CASTI, J.L. (1996), Five Golden Rules. Great Theories of 20th Century Mathematics - and Why They Matter. New York:John Wiley&Sons.
CAUDILL, M. (1992), In Our Own Image. Building an Artificial Person. New York:Oxford University Press.
CAUSEY, R. (1977), Unity of Science. Dordrecht:Reidel.
CHENEY, D.L., R.M. SEYFARTH (1990), How Monkeys See the World. Inside the Mind of Another Species. Chicago:University of Chicago Press.
CHECKLAND, P. (1987/88), „Images of Systems and the Systems Image. Excerpts from the Presidiential Adress to ISGSR Budapest June 1987“, in: IFSR
        Newsletter 16, 1 - 3.
CHURCHLAND, P.M. (1995), The Engine of Reason, the Seat of the Soul. A Philosophical Journey into the Brain. Cambridge:The MIT Press.
CLARK, C. (1960), The Conditions of Economic Progress. London:Francis Pinter.
CLEEREMANS, A. (1993), Mechanisms of Implicit Learning. Connectionist Models of Sequence Processing. Cambridge:The MIT Press.
CLEMENT, W., P.F. AHAMMER, A. KALUZA (1980), Bildungsexpansion und Arbeitsmarkt. Befunde zur Entwicklung in Österreich. Vienna:WUV
CLIFFORD, J. (1988), The Predicament of Culture. Twentieth-Century Ethnography, Literature, and Art. Cambridge:Harvard University Press.
COHEN, I.B. (1977), “History and the Philosopher of Science”, in: P. SUPPES ( 21977)(ed.), The Structure of Scientific Theories. Urbana:University of
        Illinois Press, 308 - 349.
COHEN, J., I. STEWART (1994), Chaos-Anti-Chaos. Ein Ausblick auf die Wissenschaft des 21. Jahrhunderts. Berlin:Byblos-Verlag.
COLE, S. (21995), Making Science. Between Nature and Society. Cambridge:Harvard University Press.
COLEMAN, J.S. (1986), „Social Theory, Social Research, and a Theory of Action“, in: American Journal of Sociology 91, 1309 - 1335
COLEMAN, J.S. (1987), „Microfoundations and Macrosocial Behaviour“, in: J.C. ALEXANDER et al., op. cit., 153 - 173.
COLEMAN, J.S. (1990), Foundations of Social Theory. Cambridge:Harvard University Press.
COLLINS, H.M. (1990), Artificial Experts. Social Knowledge and Intelligent Machines. Cambridge:The MIT Press.
COLLINS, H.M. T. Pinch (1993), The Golem. What Everyone Should Know about Science. Cambridge:Cambridge University Press.
CORNING, P.A. (1983), The Synergism Hypothesis. A Theory of Progressive Evolution. New York:McGraw-Hill.
CORNWALL, J. (1983), The Conditions for Economic Recovery. A Post-Keynesian Analysis. Oxford.
CORNWALL,J. (1977), Modern Capitalism. Its Growth and Transformation. Oxford.
COVENEY, P., R. HIGHFIELD (1995), Frontiers of Complexity. The Search for Order in a Chaotic World. New York:Fawcett Columbine.
COWAN, G.A. (1988), "Plans fo the Future", in: D. PINES (1988)(ed.), 235 - 237.

CRAMER, F. (31989), Chaos und Ordnung. Die komplexe Struktur des Lebendigen. Stuttgart:DVA.
CRANDALL, BC., J. LEWIS (1992)(eds.), Nanotechnology. Research and Perspectives. Papers from the First Foresight Conference on Nanotechnology.
       Cambridge:The MIT Press.
CUHLS, K., T. KUWAHARA (1994), Outlook for Japanese and German Future Technology. Heidelberg:Physica-Verlag.
CUMMINS, R. (1985), The Nature of Psychological Explanation. Cambridge:The MIT Press.
CYERT, R.M., D.C. MOWERY (1987)(eds.), Technology and Employment. Innovation and Growth in the U.S. Economy. Washington.
DAMASIO, R.A. (1994), Descartes´Error. Emotion, Reason and the Human Brain. New York:Grosset/Putnam Book.
DAMASIO, R.A. (1995), Descartes Irrtum. Fühlen, Denken und das menschliche Gehirn. München:Paul List Verlag.
DANIEL, H.-D, R. FISCH, R. (1988)(eds.) (1988), Evaluation von Forschung, Methoden, Ergebnisse, Stellungnahmen. Konstanz:Universitätsverlag.
DARNTON, R. (1979), The Business of Enlightenment. A Publishing History of the Encyclopédie 1775 - 1800. Cambridge:Harvard University Press.
DAVID, P.A. (1993), „Clio and the Economics of QWERTY“ in: U. WITT (ed.), 267 - 272.
DAVID, P.A., D. FORAY, OECD (1994), Accessing and Expanding the Science and Technology Knowledge Base. (DSTI/STP/TIP(94)4) Paris:OECD.
DAVIES, M., T. STONE (1995)(eds.), Folk Psychology. The Theory of Mind Debate. Oxford:Blackwell Publishers.
DAVIES, P. (1995), About Time. Einstein´s Unfinished Revolution. Harmondsworth:Penguin Books.
DAWKINS, R. (1978), Das egoistische Gen. Berlin:Springer.
DAWKINS, R. (1982), The Extended Phenotype. Oxford:Oxford University Press.
DAWKINS, R. (1986), The Blind Watchmaker. Harlow:Longman Scientific&Technical.
DAWKINS, R. (1995), River out of Eden. A Darwinian View of Life. London:Weidenfeld&Nicolson.
DAY, P., R. CATLOW (1995)(eds.), Bicycling to Utopia. Essays on Science and Technology. Oxford:Oxford University Press.
DELL´MOUR, R., F. LANDLER, W. RABITSCH (1985), Bildungswesen und Qualifikationsstruktur. Einige Simulationsrechnungen. Wien:BMWFK.
DENNETT (1986a), Content and Consciousness. London:Routledge&Kegan Paul.
DENNETT, D.C. (1986b), “Cognitive Wheels: the Frame Problem of AI”, in: C. HOOKWAY (1986)(ed.), Minds, Machines and Evolution. Philosophical
       Studies. Cambridge:Cambridge University Press, 129 - 151.
DENNETT, D.C. (1991), Consciousness Explained. Boston:Little, Brown and Company.
DENNETT, D.C. (1995), Darwin´s Dangerous Idea. Evolution and the Meanings of Life. New York:Simon &Schuster.
DESAI, M. (1981), Testing Monetarism. London:Frances Pinter.
DESCHAMPS, J.P., P. R. NAYAK (1995), Product Juggernauts. How Companies Mobilize to Generate a Stream of Market Winners. Boston:Harvard
       Business School Press.
DEUTSCH, K.W., B. FRITSCH (1980), Zur Theorie der Vereinfachung: Reduktion von Komplexität in der Datenverarbeitung für Weltmodelle. Königstein
       im Taunus:Athenäum.
DEWDNEY, A.K. (1995), Der Turing Omnibis. Eine Reise durch die Informatik in 66 Stationen. Berlin:Springer.
DIDEROT, D. (1969), Enzyklopädie. Philosophische und politische Texte aus der „Encyclopèdie‟ sowie Prospekt und Ankündigung der letzten Bände.
DEVLIN, K. (31994), Sternstunden der modernen Mathematik. Berühmte Probleme und neue Lösungen. München:dtv.
DIMITZ, E., B. HARTMANN (1991), “Soziale Auswirkungen/Qualifikation”, in: G. HUTSCHENREITER et al. (1991), Evaluierung der
       Technologieförderungsprogramme der Bundesregierung 1985/1987. Research-Report. Vienna:BMÖWV, 351-384.
DONKERSLOOT, H. (1995), “The Dutch Foresight Steering Committee: A Process-Oriented Approach”, Paper Presented at the Six Countries Program
       Stockholm Conference.
DORAN, J., N.G. GILBERT (1994)(eds.), Simulating Societies: the Computer Simulation of Social Phenomena. London:University of London Press.
DOREN, C.v. (1996), Geschichte des Wissens. Basel:Birkhäuser Verlag.
DOSE, N., A. DREXLER (eds.) (1988), Technologieparks. Voraussetzungen, Bestandsaufnahme und Kritik. Opladen:Westdeutscher Verlag.
DOSI, G. et al (eds.) (1988). Technical Change and Economic Theory, London:Pinter.
DOUGLAS, M. (1987), How Institutions Think. London:Routledge and Kegan Paul.
DOUGLAS, M. (1992), Risk & Blame. Essays in Cultural Theory. London:Routledge.
DRETSKE, F.I. (1981), Knowledge and the Flow of Information. Oxford:Basil Blackwell.
DRUCKER, P.F. (1993), Die postkapitalistische Gesellschaft. Düsseldorf:Econ-Verlag.
DUBACH, P. (1977), „Morphologie als kreative Methode in der Langfristplanung“, in: G. BRUCKMANN (ed.), op.cit., 112 - 125.
DUPRIEZ, L.H. (1947), Des Mouvements Economiques Generaux. Louvain.
DÜRR, H.P. (1995), Die Zukunft ist ein unbetretener Pfad. Bedeutung und Gestaltung eines ökologischen Lebensstils. Freiburg:Herder.
ECO, U. (1972), Einführung in die Semiotik. München:dtv Verlag.
ECO, U. (21981), Zeichen. Einführung in einen Begriff und seine Geschichte. Frankfurt am Main:Suhrkamp.
ECO, U. (1992), Die Grenzen der Interpretation. München:Carl Hanser Verlag.
ECO, U. (1993), Die Suche nach der vollkommenen Sprache. München:Verlag C.H. Beck.
EDELMAN, G.M. (1987), Neural Darwinism. New York:Basic Books.
EDELMAN, G.M. (1990), The Remembered Present. A Biological Theory of Consciousness. New York:Basic Books.
EDELMAN, GH.M. (1992), Bright Air, Brilliant Fire. On the Matter of the Mind. New York:Basic Books.
EGGBAUER, H, u.a. (1991), Möglichkeiten des Technologietransfers. Vienna:Technical University.
EIGEN, M. (1987), Stufen zum Leben. Die frühe Evolution im Visier der Molekularbiologie. München-Zürich.
EIGEN, M., P. SCHUSTER (1979), The Hypercycle: A Principle of Natural Self-Organization. Berlin:Springer.
EISENHARDT, P., D. KURTH, H. STIEHL (1995), Wie Neues entsteht. Die Wissenschaften des Komplexen und Fraktalen. Reinbek:Rowohlt.
EKINS, P. (1986)(ed.), The Living Economy. A New Economics in the Making. London:Routledge&Kegan Paul.
ELIAS, N. (21971), Was ist Soziologie? München:Piper.
ELIAS, N. (1988), Die Gesellschaft der Individuen. Frankfurt:Suhrkamp.
ELLSWORTH, P.T., J.C. LEITH (61984), The International Economy. New York:Basic Books.
ELSTER, J. (1983), Explaining Technical Change. A Case Study in the Philosophy of Science. Cambridge:Cambridge University Press.
ELSTER, J. (1989), The Cement of Society. A Study of Social Order. Cambridge:Cambridge University Press.
EMERY, F.E. (1978), Systems Thinking. Selected Readings, vol. 1. Harmondsworth:Penguin.
ERD, R., O. JACOBI, W. SCHUMM (1986)(eds.), Strukturwandel in der Industriegesellschaft. Frankfurt:Campus.
ERDMANN, G., B.FRITSCH (1989), Synergismen in sozialen Systemen, ein Anwendungsbeispiel. Zürich: Working Paper Nr.94/89 des Instituts für
       Wirtschaftsforschung der ETH Zürich.
ERESHEFSKY, M. (1992)(ed.), The Units of Evolution. Essays on the Nature of Species. Cambridge:The MIT Press.
ERICSON, R.V., N. STEHR (1992)(eds.), The Culture and Power of Knowledge. Inquiries into Contemporary Societies. Berlin:de Gruyter.

ESSER, H. (1991), Alltagshandeln und Verstehen. Zum Verhältnis von erklärender und verstehender Soziologie am Beispiel von Alfred Schütz und
       „Rational Choice‟. Tübingen:J.C.B. Mohr (Paul Siebeck).
ETZIONI, A. (1994), Jenseits des Egoismus-Prinzips. Ein neues Bild von Wirtschaft, Politik und Gesellschaft. Stuttgart:Schäffer-Poeschel.
ETZIONI, A. (1995), The Spirit of Community. Rights, Responsibilities and the Communitarian Agenda. New York:Fontana Press.
ETZKOWITZ, H., L.LEYDESDORFF (1995)(eds.), Universities and the Global Knowledge Economy. A Triple Helix of University-Industry-Government
       Relations. Amsterdam:Science&Technology Dynamics.
EUROPEAN CENTRE (1993), Welfare in a Civil Society. Report for the Conference of European Ministers Responsible for Social Affairs.
       Bratislava:European Centre.
EUROPEAN COMMISSION (1994a), The Community Innovation Survey: Status and Objectives, Eurostat and DG-XIII. Brussels:EU.
EUROPEAN COMMISSION (1994b), The European Report on Science and Technology Indicators 1994. Report. Brussels: EUR 15879 EN.
EVERED, D. (1989)(ed.), The Evaluation of Scientific Research. Ciba Foundation Conference.Chichester:John Wiley&Sons.
FARKAS, V., P. KRASSA (1985), Lasset uns Menschen machen. Schöpfungsmythen beim Wort genommen. Berlin:edition meyster.
FEATHERSTONE, M., S. LASH, R. ROBERTSON (1995)(eds.), Global Modernities. London:Sage Publications.
FEIGENBAUM, E.A., J. FELDMAN (1995)(eds.), Computers and Thought. Menlo Park:The AAAI Press.
FELDERER, B. (1993)(ed.), Wirtschafts und Sozialwissenschaften zwischen Theorie und Praxis. 30 Jahre Institut für Höhere Studien in Wien .
       Vienna:Physica Verlag.
FELDERER, B., D. CAMPBELL (1994), Forschungsfinanzierung in Europa. Trends- Modelle, Empfehlungen für Österreich. Wien:Manz Verlag.
FELDERER, B., D. CAMPBELL (1995), Die Evaluation der akademischen Forschung im internationalen Vergleich: Strukturen, Trends und Modelle. First
       Interim Report. Vienna:IHS.
FELDERER, B. (1996), “The Importance of R&D for Future Industries and the Wealth of Nations”, in: IHS Newsletter 4, 1 - 3.
FELDMAN, M.W. (1988), “Evolutionary Theory of Genotypes and Phenotypes: Towards a Mathematical Synthesis”, in: D. PINES (1988)(ed.), 43 - 52.
FELT, U., H. NOWOTNY (1995)(eds.), Social Studies of Science in an International Perspective. Vienna:Institute for Theory and Social Studies of Science.
FERRY, L. (1995), The New Ecological order. Chicago:The University of Chicago Press.
FESSEL+GFK (1993): Forschung und Entwicklung in österreichischen Betrieben. Vienna:FESSEL.
FINKE, R.A., T.B. WARD, S. M. SMITH (1992), Creative Cognition. Theory, Research, and Applications. Cambndge:The MIT Press.
FISCHER, H. (1985) (ed.), Forschungspoltik für die 90er Jahre. Wien:Springer-Verlag.
FLECKER, J. (1986), “Soziale Kriterien in der staatlichen Technologieförderung - Wenn Ja, warum nicht?”, in: M. SCHERB, I MORAWETZ (1986)(eds.),
       Stahl und Eisen bricht. Industrie und staatliche Politik in Österreich.Wien:WUV, 47-80.
FLEISSNER, P. (1987)(ed.), Technologie und Arbeitswelt in Österreich. Trends bis zur Jahrtausendwende, 4 Vol. Wien:Akademie der Wissenschaften.
FLUSSER, V. (1988), Krise der Linearität. Bern:Benteli Verlag.
FLUSSER, V. (21989), Die Schrift. Hat Schreiben Zukunft. Göttingen:Immatrix Publications.
FLORA, P. (1974), Modernisierungsforschung. Zur empirischen Analyse der gesellschaftlichen Entwicklung. Opladen:Westdeutscher Verlag.
FLORA, P., A.J. HEIDENHEIMER (21984)(eds.), The Development of Welfare States in Europe and America. New Brunswick-London.
FOERSTER, H.v. (1982), Observing Systems. Seaside:Intersystems Publications.
FOERSTER, H.v. (1984), Erkenntnistheorien und Selbstorganisation, in: DELFIN IV, 6 - 19.
FOERSTER, H.v. (1985), Sicht und Einsicht. Versuche zu einer operativen Erkenntnistheorie. Braunschweig:Vieweg.
FOERSTER, H.v. (1986), Sketch on Recursions. A Primer. Pescadero:HVF.
FOERSTER, H.v. (1990), “Kausalität, Unordnung, Selbstorganisation” in: K.W. KRATKY, F. WALLNER (1990)(eds.), Grundprinzipien der
       Selbstorganisation. Darmstadt:Wissenschaftliche Buchgesellschaft, 77 -95.
FOERSTER, H.v. (1993), Wissen und Gewissen. Versuch einer Brücke. Frankfurt:Suhrkamp.
FOERSTER, H.v. (21995), Cybernetics of Cybernetics. Minneapolis:Future Systems.
FONTANA, W. L.W. BUSS (1995), A Methodology for Generating and Modeling Self-Maintaining Systems. A Research Agenda. Laxenburg:IIASA.
FORESTER, T. (1985)(ed.), The Information Technology Revolution. Oxford:Basil Blackwell.
FORAY, D., E. CONESA (1993), The Economics and Organization of „Remote‟ Research Programes: Beyond the Frontier of Knowledge. Mimeo.
FORREST, S. (1991)(ed.), Emergent Computation. Self-Organizing, Collective, and Cooperative Phenomena in Natural and Artificial Computing
       Networks. Cambridge: The MIT Press.
FOSTER, R.N. (1986), Innovation. The Attacker's Advantage. New York:Summit Books.
FOUCAULT, M. (31979), Überwachen und Strafen. Die Geburt des Gefängnisses. Frankfurt:Suhrkamp.
FOURASTIE, J. (1954), Die große Hoffnung des 20. Jahrhunderts. Köln.
FRANK, H. (1992), Modellversuch Wissenschaftler gründen Firmen. Vienna:BMWF.
FREEMAN, C., J. CLARK, L. SOETE (1982), Unemployment and Technical Innovation. A Study of Long Waves and Economic Development.
       London:Frances Pinter.
FREEMAN, C. (1983)(ed.), Long Waves in the World Economy. London:Butterworths.
FREEMAN, C. (1986)(ed.), Design, Innovation and Long Cycles in Economic Development. London:Frances Pinter.
FREEMAN, C. (1987), Technology and Economic Performance. Lesons from Japan. London:Pinter Publishers.
FREEMAN, C. (1988) (ed.), Output Measurement in Science and Technology. Dordrecht:Klouwer..
FREEMAN, C. (1989), The Nature of Innovation and the Evolution of the Productive System, in: International Seminar on Science, Technology and
       Economic Growth. Paris:OECD.
FREEMAN, C., L. SOETE (1994), Work for All or Mass Unemployment. Computerised Technical Change into the 21st Century. London: Frances Pinter.
FREEMAN, J.A. (1991), Neural Networks. Algorithms, Applications, and Programming Techniques. Reading et al.
FRICKE, W. (ed.) (1992), Jahrbuch Arbeit und Technik 1992. Bonn:Dietz-Verlag.
FRIEDBERG, E. (1995), Ordnung und Macht. Dynamiken organisierten Handelns. Frankfurt:Campus Verlag.
FRIEDEBURG, L., J. HABERMAS (1983)(eds.), Adorno-Konferenz 1983. Frankfurt am Main:Suhrkamp.
FRIEDRICH, J., T. HERRMANN, M. PESCHEK, A. ROLF (1995)(eds.), Informatik und Gesellschaft. Heidelberg:Spektrum.
FRISCH, H. (1981)(ed.), Schumpeterian Economics. Eastbourne:Praeger.
FRITSCH, B. (1974), Wachstumsbegrenzung als Machtinstrument. Stuttgart:Deutsche Verlags-Anstalt.
FUKUYAMA (1991), Trust. The Social Virtues and the Creation of Prosperity. New York:The Free Press.
FURGER, F. (1994), Ökologische Krise und Marktmechanismen. Umweltökonomie in evolutionärer Perspektive. Opladen:Westdeutscher Verlag.
GALABURDA, A.M. (1989)(ed.), From Reading to Neurons. Cambridge:The MIT Press.
GALE, J.S. (1990), Theoretical Population Genetics. London:Unwin Hyman.
GALL, J. (31990), Systemantics. The Underground Text of Systems Lore. How Systems Really Work and Especially How They Fail. Ann Arbor:The General
       Systemantics Press.
GARDNER, H. (1985), The Mind´s New Science. A History of the Cognitive Revolution. New York:Basic Books.

GARDNER, H. (1993), Creating Minds. An Anatomy of Creativity Seen through the Lives of Freud, Einstein, Picasso, Stravinsky, Eliot, Graham, and
Ghandi. New York:Basic Books.
GARZ, D., K. KRAIMER (1983)(eds.), Brauchen wir andere Forschungsmethoden? Beiträge zur Diskussion interpretativer Verfahren. Frankfurt:Scriptor.
GATES, B. (1995), Der Weg nach vorn. Die Zukunft der Informationsgesellschaft. Hamburg:Hoffmann und Campe.
GAUDIN, T. (1995), 2100. Spiece´s Odyssey. Montiers:Foundation 2100.
GAZZANIGA, M.S. (1985), The Social Brain. Discovering the Networks of the Mind. New York:Basic Books.
GAZZANIGA, M.S. (1995)(ed.), The Cognitive Neurosciences. Cambridge:MIT Press
GEERTZ, C. (1983), Local Knowledge. Further Essays in Interpretative Anthropology. New York:Basic Books.
GELL-MANN, M. (1988), “The Concept of the Institute” in: D. PINES (1988)(ed.), 1 - 20.
GELL-MANN, M. (1994), Das Quark und der Jaguar. Vom Einfachen zum Komplexen - die Suche nach einer neuen Erklärung der Welt. München:Piper.
GERGEN, K.J. (21994), Toward Transformation in Social Knowledge. London:Sage Publications.
GEORGESCU-ROEGEN, N. (1971), The Entropy Law and the Economic Process. Cambridge:Harvard University Press.
GEORGESCU-ROEGEN, N. (1976), Energy and Economic Myths. Institutional and Analytical Economic Essays. New York:Pergamon Press.
GERKEN, G. (1995), Wild Future. Abschied von den kalten Strategien. Düsseldorf:Econ.
GERKEN, G. (1996), Multimedia. Das Ende der Information. Wie Multimedia die Welt des managements verändert. Düsseldorf:Metropolitan Verlag.
GERSHUNY, J. (1981), Die Ökonomie der nachindustriellen Gesellschaft. Produktion und Verbrauch von Dienstleistungen. Frankfurt:Campus.
GERSHUNY, J. (1983), Social Innovation and the Division of Labour. Oxford:Oxford University Press.
GESCHKA, H. (1977): Delphi, in: BRUCKMANN, G. (ed): Langfristige Prognosen. Möglichkeiten und Methoden der Langfristprognostik komplexer
        Systeme. Würzburg:Physica-Verlag.
GIBBONS, M. et al. (1994), The New Production of Knowledge. The Dynamics of Science and Research in Contemporary Societies. London:Sage.
GIBBS, W.W. (1995), “Lost Science in the Third World”, in: Scientific American 8, 76-83.
GIBSON, K.R., T. INGOLD (1994)(eds.), Tools, Language and Cognition in Human Evolution. Cambridge:Cambridge University Press.
GIDDENS, A. (1989), Sociology. Cambridge:Polity Press.
GIDDENS, A. (1991), Modernity and Self-Identity. Self and Society in the Late Modern Age. Cambridge:Polity Press.
GILBERT, S.F. (41994), Developmental Biology. Sunderland:Sinauer Associates.
GLANVILLE, R. (1988), Objekte. Berlin:Merve-Verlag.
GLASERSFELD, E.v. (1987), Wissen, Sprache und Wirklichkeit. Arbeiten zum radikalen Konstruktivismus. Braunschweig:Friedr.Vieweg&Sohn.
GLATZ, H. (1991), Kleinstaaten im Wirtschaftlichen Strukturwandel. Industrie- und technologiepolitische Strategien ausgewählter Industrieländer.
        Vienna:Bundesministerium für Öffentliche Wirtschaft und Verkehr.
GLATZ, H. (1992), “Die Industrie- und Technologiepolitik kleiner europäischer Länder im Vergleich”, in: Wirtschaft und Gesellschaft 1: 47-74.
GOEL, V. (1995), Sketches of Thought. Cambridge:The MIT Press.
GOETHE, J.W.v. (1977), Sämtliche Werke, vol. 3. Epen, West-östlicher Divan, Theatergedichte. Zürich:Artemis-Verlag.
GOLD, B., G. ROSEGGER, M.G. BOYLAN (1994), Evaluating Technological Innovations. Methods, Expectations and Findings. Toronto:Lexington
GOLDMANN, W. (1985), “Forschung, Innovation und Technologie in Österreich”, in: H. Fischer (ed.)(1985), Forschungspolitik für die 90er Jahre.
        Wien:Springer, 187-208
GOLDMANN, W. (1990), “Industriepolitik in Österreich”, in: Wirtschaft und Gesellschaft 1, 43-64
GOLDMANN, W. (1992), “Österreichische Industrie- und Technologiepolitik”, in: Wirtschaft und Gesellschaft 4, 461-468.
GOLEMAN, D. (1995), Emotional Intelligence. Why It Can Matter More Than IQ. New York:Bantam Books.
GOODMAN, N. (1973), Sprachen der Kunst. Ein Ansatz zu einer Symboltheorie. Frankfurt am Main:Suhrkamp.
GOODWIN, B. (1995), How the Leopard Changed Its Spots. The Evolution of Complexity. New York:Charles Scribner´s Sons.
GOODWIN, R.M., L.F.PUNZO (1987), The Dynamics of a Capitalist Economy. A Multi-Sectoral Approach. Cambridge:Polity Press.
GOTTINGER, H.W. (1983), Coping with Complexity. Perspectives for Economics, Management and Social Sciences. Dordrecht:Reidel.
GOTTWEIS, H., M. LATZER (1991), “Technologiepolitik”, in: H. Dachs, P Gerlich et al. (eds.) (1991), Handbuch des politischen Systems Österreichs.
        Wien:Manz-Verlag, 601-612.
GOUILLART, F.J., J.N. KELLY (1995), Business Transformation. Reframing-Restructuring-Revitalizing-Renewing. Wien:Ueberreuter.
GOULD, S.J. (1982), „Is a New and General Theory of Evolution Emerging“, in: J. MAYNARD SMITH (ed.), 129 - 145.
GOULD, S.J. (1989), Wonderful Life. The Burgess Shale and the Nature of History. New York:W.W. Norton&Company.
GRANDE, E., J. HÄUSLER (1994), Industrieforschung und Forschungspolitik. Staatliche Steuerungspotentiale in der Informationstechnik
GRANOVETTER, M. (1985), “Economic Action and Social Structure: The Problem of Embeddedness”, in: American Journal of Sociology 91, 479 - 495.
GRAY, J., A. REUTER (1993), Transaction Processing. Concepts and Techniques. San Mateo:Morgan Kaufmann.
GREWENDORF, G. (1995), Sprache als Organ - Sprache als Lebensform. Frankfurt:Suhrkamp.
GRIBBIN, M&J. (1993), Being Human. Putting People in an Evolutionary Perspective. London:Phoenix.
GRIFFIN, D.R. (1992), Animal Minds. Chicago:The University of Chicago Press.
GRUPP, H. (1992)(ed.), Dynamics of Science Based Innovation. Heidelberg:Springer-Verlag.
GRUPP, H. (1995), Der Delphi-Report. Innovationen für unsere Zukunft. Stuttgart:Deutsche Verlags-Anstalt.
GRÜTZMANN, K. (1993), Simulation und Analyse eines dynamischen Entscheidungsmodells mit Gedächtniseffekten, Diplomarbeit am Institut für
        Theoretische Physik. Stuttgart:University of Stuttgart.
GRÜTZMANN, K., G. HAAG (1994), “Gedächtniseffekte und die Entstehung des Neuen”, in: WISDOM 3/4, 128 - 137.
GUSTAFSSON, B. (1979)(ed.), Post-Industrial Society. London:Frances Pinter.
HAAG, G., W. WEIDLICH (1984), “A Stochastic Theory of Interregional Migration”, in: Geographical Analysis 16, 331 - 357.
HAAG, G. (1989), Dynamic Decision Theory: Applications to Urban and Regional Topics. Dordrecht:Kluwer.
HAAG, G. et al. (1992)(eds.), Economic Evolution and Demographic Change. Formal Models in the Social Sciences. Berlin:Springer.
HAAG, G., K.H. MÜLLER (1992), „Employment and Education as Non-Linear Population Networks I&II“, in: G. HAAG et al. (eds.), 349 - 407.
HAAG, G., K. GRÜTZMANN (1993), “A New Theory of Nested Decision Processes with Memory Effects”, in: Papers of Regional Science 72, 311 - 325.
HABICH, R., H.H. NOLL (1993), Soziale Indikatoren und Sozialberichterstattung. Internationale Erfahrungen und gegenwärtiger Forschungsstand.
HAARMANN, H. (1990), Universalgeschichte der Schrift. Frankfurt:Campus-Verlag.
HABERLER, G. (1948), Prosperität und Depression. Eine theoretische Untersuchung der Konjunkturbewegungen. Bern.
HABERMAS, J. (1981), Theorie kommunikativen Handelns, 2 vol. Frankfurt:Suhrkamp.
HABERMAS, J. (1988), Nachmetaphysisces Denken. Philosophische Aufsätze. Frankfurt:Suhrkamp.

HABICH, R., H.H. NOLL (1993), Soziale Indikatoren und Sozialberichterstattung. Internationale Erfahrungen und gegenwärtiger Forschungsstand.
HAKEN, H. (1977), Synergetics. An Introduction - Nonequilibrium Phase Transitions and Self-Organization in Physics, Chemistry and Biology. Berlin:
HAKEN, H. (1980)(ed.), Dynamics of Synergetic Systems. Berlin:Springer.
HAKEN, H. (1982a), Synergetik. Eine Einführung. Berlin:Springer.
HAKEN, H. (1982b(ed.), Evolution of Order and Chaos in Physics, Chemistry, and Biology. Berlin:Springer.
HAKEN, H. (1983), Advanced Synergetics. Instability Hierarchies of Self-Organizing Systems and Devices. Berlin:Springer.
HAKEN, H. (1991), Synergetic Computers and Cognition. A Top-Down Approach to Neural Nets. Berlin:Springer.
HALLER, M. (1996), “Einstellungen zur sozialen Ungleichheit im internationalen Vergleich” in: M. HALLER, K. HOLM, K.H. MÜLLER, W. SCHULZ, E.
       CYBA (1996)(eds.), Österreich im Wandel. Werte, Lebensformen und Lebensqualität 1986 bis 1993. Wien:Verlag für Geschichte und Politik & R.
       Oldenbourg Verlag, 188 - 220.
HALLER, R. (1986), Fragen zu Wittgenstein und Aufsätze zur österreichischen Philosophie. Amsterdam:Rodopi.
HALLER, R. (1988), Questions on Wittgenstein. Lincoln:University of Nebraska Press.
HAMILLTON, F.E.I. (1987)(eds.), Industrial Change in Advanced Economies. London:Croom Helm.
HANAPPI, G. (1994), Evolutionary Economics. The Evolutionary Revolution in the Social Sciences. Aldershot:Avebury.
HANISCH, E. (1994), Der lange Schatten des Staates. Österreichische Gesellschaftsgeschichte im 20. Jahrhundert. Wien:Ueberreuter.
HANSEN, H.R. (1996), Klare Sicht am Info-Highway. Geschäfte via Internet&Co. Wien:Verlag Orac.
HARBORDT, , S. (1974), Computersimulation in den Sozialwissenschaften, 2 vol. Reinbek:Rowohlt.
HARTMANN, F. (1993)(ed.), Standort und Perspektiven der außeruniversitären Sozialforschung. Vienna:Forum Sozialforschung.
HARTWICH, H.H. (ed.)(1986), Politik und Macht der Technik. Tagungsbericht: 16. wissenschaftlicher Kongreß der DVPW. Opladen:Westdeutscher Verlag.
HAUBERT, R. (1994), Nutzung der Informationsfunktion von Patenten für kleinere und mittlere Unternehmen. Seibersdorf:Forschungszentrum Seibersdorf.
HAWKING, S.W. (1991), Anfang oder Ende? Inauguralvorlesung. Paderborn:Junfermann Verlag.
HAWKING, S., R. PENROSE (1996), The Nature of Space and Time. Princeton:Princeton University Press.
HAWTHORN, G. (1994), Die Welt ist alles, was möglich ist. Über das Verstehen der Vergangenheit. Stuttgart:Klett-Cotta.
HAYEK, F. v. (1949), Individualism and Economic Order. London:Routledge.
HAYEK, F. v. (1967), Sttudies in Philosophy, Politics and Economics. London:Routledge.
HAYKIN, S. (1994), Neural Networks. A Comprehensive Foundation. New York:Macmillan College Publishing Company.
HEGSELMANN, R., U. MUELLER, K.G. TROITZSCH (1996)(eds.), Modelling and Simulation in the Social Sciences from the Philosophy of Science point
       of View. Dordrecht:Kluwer Academic Publishers.
HEIMS, S.J. (1991), The Cybernetics Group. Cambridge:MIT Press.
HEINZ, W.R. (1992)(ed.), Institutions and Gatekeeping in the Life Course. Weinheim:Deutscher Studien Verlag.
HELBING, D. (1993), Stochastische Methoden, nichtlineare Dynamik und quantitative Modelle sozialer Prozesse. Aachen: Shaker.
HELLER, A. (1995), Ist die Moderne lebensfähig? Frankfurt:Campus.
HELLER, A., S. PUNTSCHER RIEKMANN (1996)(eds.), Biopolitics. The Politics of the Body, Race and Nature. Aldershot:Avebury.
HEMPEL, C.G. (1942), “The Function of General Laws in History”, in: The Journal of Philosophy 39, 35 - 48.
HEMPEL, C.G., P. OPPENHEIM (1948), “Studies in the Logic of Explanation”, in: Philosophy of Science 15, 135 - 175.
HEMPEL, C.G. (1965), Aspects of Scientific Explanation and Other Essays in the Philosophy of Science. New York:Free Press.
HENDERSON, J. (1991), The Globalisation of High Technology Production. Society, Space and Semiconductors in the Restructuring of the Modern World .
HENDRY, D.E. (1987), “Econometric Methodology: a Personal Perspective”, in: T.F. BEWLEY (1987)(ed.), Advances in Econometrics. Fifth World
       Congress. Cambridge:Cambridge University Press, 29 - 48.
HENNIG, W. (1995), Genetik. Berlin:Springer.
HENSEL, M. (1990), Die Informationsgesellschaft. Neuere Ansätze zur Analyse eines Schlagwortes. München:Verlag Reinhard Fischer.
HERFEL, W., W. KRAJEWSKI, I. NIINILUOTO, R. WOJCICKI (1995)(eds.), Theories and Models in Scientific Processes. Amsterdam:Rodopi.
HERITIER, A. (1993)(ed.), Policy-Analyse. Kritik und Neuorientierung. PVS-special edition 24. Opladen:Westdeutscher Verlag.
HERKEN, R. (21995), The Universal Turing Machine. A Half-Century Survey. Wien:Springer.
HERMAN, A., STUBNYA, G. (1994), Information System in Transition - The Hungarian Science, mimeo
HESKETT, J.L. (1986), Managing in the Service Economy. Harvard Business School Press.
HESS, D.J. (1995), Science and Technology in a Multicultural World. The Cultural Politics of Facts and Artifacts. New York:Columbia University Press.
HILDEBRANDT, E, R. SELTZ (1985), “Gewerkschaftliche Technologiepolitik zwischen Statussicherung und Arbeitsgestaltung”, in: LUTZ, B (Ed.),.
       Soziologie und gesellschaftliche Entwicklung. Verhandlungen des 22. Deutschen Soziologentags in Dortmund. Frankfurt:Campus, S. 434-447.
HILLEBRAND, G., W. ERNST (1993), Wirtschaftsbezogene Forschung in Europa. Analyse und Vergleich außeruniversitärer, wirtschaftbezogener
       Forschungseinrichtungen. Seibersdorf:Forschungszentrum Seibersdorf.
HILPERT, U. (1989), Staatliche Forschungs- und Technologiepolitik und offizielle Wissenschaft. Wissenschaftlich-technischer Fortschritt als Instrument
       politisch vermittelter technologisch-industrieller Innovation. Opladen:Westdeutscher Verlag.
HIPPEL, E. von (1994), The Sources of Innovation. Oxford:Oxford University Press.
HIRSCHMAN, A.O. (1992), Denken gegen die Zukunft. Die Rhetorik der Reaktion. München:Carl Hanser Verlag.
HOBSBAWM, E. (1995), Age of Extremes. The Short Twentieth Century 1914 - 1991. London:Abacus.
HOCHGERNER, J. (1988), “Zwischen Zielen und Zwängen. Überlegungen zur Positionsbestimmung der Gewerkschaften in der Technologiepolitik”, in:
       Österreichische Zeitschrift für Politikwissenschaft 3, 263-274.
HOCHGERNER, J. (1990)(ed.), Soziale Grenzen des technischen Fortschritts. Vergleiche quer durch Europa. Wien:Faltzer Verlag.
HOCHMUTH, U., J. WAGNER (1994)(eds.), Firmelpanelstudien in Deutschland. Konzeptionelle Überlegungen und empirische Analysen. Tübingen-Basel.
HODGSON, G.M. (1993), Economics and Evolution. Bringing Life Back into Economics. Cambridge:Polity Press.
HOFBAUER, J., K. SIGMUND (1984), Evolutionstheorie und dynamische Systeme. Mathematische Aspekte der Selektion. Berlin:Paul Parey.
HOFBAUER, J., G. PRABITZ, J. WALLMANNSBERGER (1995)(eds.), Bilder - Symbole - Metaphern. Visualisierung und Informierung in der Moderne.
       Wien:Passagen Verlag.
HOFINGER, C. (1994), Am Beispiel Politik: Daten-Akquisition für ein Mastergleichungsmodell”, in: WISDOM 3/4, 110 - 127.
HOFINGER, C., K. GRÜTZMANN (1994), “Das Politik-Modell: Attraktivitäten als Determinanten von Wählerbewegungen in Österreich 1970 - 1990”, in:
       WISDOM 3/4, 79 - 89.
HOFESTÄDT, R. (1994), Theorie der regelbasierten Modellierung des Zellstoffwechsels. Habilitationsschrift. Koblenz:Universität Koblenz-Landau.
HOFFMANN, W.G. (1965), Das Wachstum der deutschen Wirtschaft seit der Mitte des 19. Jahrhunderts. Berlin:Springer.

HOFINGER, C., K. GRÜTZMANN (1994), „Das Politik-Modell: Attraktivitäten als Determinanten von Wählerbewegungen in Österreich 1970 - 1990“, in:
       WISDOM 3/4, 79 - 89.
HOFSTADTER, D.R (41982), Gödel-Escher-Bach. An Eternal Golden Braid. Harmondsworth:Penguin.
HOFSTADTER, D.R., D.C. DENNETT (1982) (eds.), The Mind´s I. Phantasies and Reflections on Self and Soul. Harmondsworth:Penguin.
HOFSTADTER, D.R. (1985a), Metamagical Themas. Questing for the Essence of Mind and Matter. New York:Basic Books.
HOFSTADTER, D.R. (1985b), Gödel, Escher, Bach - Ein Endloses Geflochtenes Band. Stuttgart:Klett-Cotta.
HOFSTADTER, D.R., FLUID ANALOGIES RESEARCH GROUP (1995), Fluid Concepts and Creative Analogies. Computer Models of the Fundamental
       Mechanisms of Thought. New York:Basic Books.
HÖLL, O. (1989), “Technologieentwicklung, wirtschaftliche Umgestaltung und die europäische Integration”, in: Österreichisches Jahrbuch für
       internationale Politik 6, 50-75
HOLLAND, J. (1986), „Escaping Brittleness: The Possibilities of General-Purpose Learning Algorithms Applied to Parallel Rule-Based Systems“, in: R.S.
       Michalski et al. (1986)(eds.), Machine Learning. An Artificial Intelligence Approach, Vol. II. Los Altos:Morgan Kaufmann Publishers, 593:623.
HOLLAND, J.H. (1988), "The Global Economy as an Adaptive Process", in: P.W. ANDERSON, K.J. ARROW, D. PINES (1988), 117 - 124.
HOLLAND, J.H., K.J. HOLYOAK, R.E. NISBETT, P.R. THAGARD (1989), Induction. Processes of Inference, Learning, and Discovery. The MIT Press.
HOLLAND, J.H. (1992), Adaptation in Natural and Artificial Systems. An Introductory Analysis with Applications to Biology, Control, and Artificial
       Intelligence. Cambridge:MIT Press.
HOLLAND, J.H. (1995), Hidden Order. How Adaptation Builds Complexity. Reading:Addison-Wesley.
HÖLLINGER, S., E. HACKL, C. BRÜNNER (1994)(eds.), Fachhochschulstudien - unbürokratisch, brauchbar und kurz. Vienna:Passagen Verlag.
HOLYOAK, K.J., P. R. THAGARD (1989), “A Computational Model of Analogical Problem Solving”, in: VOSNIADOU, ORTONY, 242 - 266,
HOLZINGER, E., B. FRÜHSTÜCK-PFNEISZL, A.S. LABURDA (1991), Der regionale Versorgungsbedarf an Bildungseinrichtungen. Expertengutachten
       des Österreichischen Instituts für Raumplanung. Vienna:ÖROK.
HOPCROFT, J.E.., J.D. ULLMAN (1990), Einführung in die Automatentheorie, Formale Sprachen und Komplexitätstheorie. Bonn:Addison-Wesley.
HORVAT, M., C. LAURER (1992), Die Außeninstitute als Technologietransfereinrichtungen: Intensivierung des Technologietransfers zwischen
       Wissenschaft und Wirtschaft. Vienna:Technical University.
HORX, M. (1995), 12 neue Trends. Megatrends für die späten neunziger Jahre. Düsseldorf:Econ.
HOYNINGEN-HUENE, P., G. HIRSCH (1988)(eds.), Wozu Wissenschaftsphilosophie? Positionen und Fragen zur gegenwärtigen
       Wissenschaftsphilosophie. Berlin:de Gruyter.
HUBER, J. (1979)(ed.), Anders arbeiten - anders wirtschaften. Dual-Wirtschaft: Nicht jede Arbeit muß ein Job sein. Frankfurt am Main:Fischer.
HUBER, J. (1982), Die verlorene Unschuld der Ökologie. Neue Technologien und superindustrielle Entwicklung. Frankfurt am Main:Fischer.
HUBER, J. (1991)(ed.), Macro-Micro Linkages in Sociology. Newbury Park:Sage Publications.
HUCKE, J., H. WOLLMANN (1991) (eds.), Dezentrale Technologiepolitik?. Technikförderung durch Bundesländer und Kommunen. Basel:Birkhäuser
HUDSON, R. (1994): "Techno-Trophies", in: The Wall Street Journal's Central European Economic Review, Vol. 2, summer 1994, p. 18.
HULL, D.L. (1988), Science as a Process. An Evolutionary Account of the Social and Conceptual Development of Science . Chicago:The University of
       Chicago Press.
HUMPHREY, N. (1995), Die Naturgeschichte des Ich. Hamburg:Hoffmann und Campe.
HURST, D.K. (1995), Crisis and Renewal. Meeting the Challenge of Organizational Change. Boston:Harvard Busieness School Press.
HUTSCHENREITER, G. et al. (1991). Evaluierung der Technologieförderungsprogramme der Bundesregierung 1985/1987. Vienna:Wifo-Research Report.
HUTSCHENREITER, G., H. LEO (1992), “Künftige Aufgaben österreichischer Technologiepolitik”, in: Wirtschaftspolitische Blätter 4, 453-461.
HUTSCHENREITER, G. (1993), Industriepolitik der EG: Grundlagen und neue Initiativen. Vienna:TIP.
HUTSCHENREITER, G. (1994a), Cluster innovativer Aktivitäten in der österreichischen Industrie. Vienna:TIP.
HUTSCHENREITER, G. (1994b), Innovation und Produktivitätsentwicklung der österreichischen Industrie. Vienna:TIP.
HUTSCHENREITER, G. et al. (1996), Technologiepolitisches Konzept 1996 der Bundesregierung. Expertenentwurf. Vienna:Ministry of Science, Transport
       and Arts and Ministry for Economic Affairs.
IFRAH, G. (1989), Universalgeschichte der Zahlen. Frankfurt:Campus-Verlag.
INDEPENDENT COMMISSION ON POPULATION AND QUALITY OF LIFE (1996), Caring for the Future. Making the Next Decades Provide a Life
       Worth Living. Oxford:Oxford University Press.
INDUSTRIEWISSENSCHAFTLICHES INSTITUT (1993), Die internationale Wettbewerbsposition der österreichischen Industrie aus ökonomischer und
       technologischer Sicht. Vienna:Wirtschaftsuniversität
       Unternehmen. Second Interim Report. Vienna:Economic University Report.
IRVINE, J., B.R. MARTIN (1984), Foresight in Science. Picking the Winners. London:Pinter Publishers.
IRVINE, J., B.R. MARTIN, P.A. ISARD (1991), Investing in the Future. An International Comparison of Government Funding of Academic and Related
       Research. Aldershot:Edward Elgar.
JACOBS, D. (1989), “Small Countries‟ Opportunities for Participating in Science-Based Development”, in: R. van TULDER (ed.), Small Industrial
       Countries and the Economic and Technological Development. University of Amsterdam:Working Document 9, 43-51.
JAMMER, M. (1974), The Philosophy of Quantum Mechanics. The Interpretations of Quantum Mechanics in Historical Perspective. New York:John
JANICH, P. (1996), Konstruktivismus und Naturerkenntnis. Auf dem Weg zum Kulturalismus. Frankfurt:Suhrkamp.
JANIK, A., “Work, Technology, Language: The Role of „Tacit Knowledge‟ in Experimental Physics”, in: U. FELT, H. NOWOTNY (1995)(eds.), Social
       Studies, op.cit., 135 - 142.
JANSHEN, D., O. KECK, W.D. WEBLER (1981)(eds.), Technischer und sozialer Wandel. Eine Herausforderung an die Sozialwissenschaften.
       Königstein:Verlag Anton Hain.
JANTSCH, E. (1972), Technological Planing and Social Futures. London:Francis Pinter.
JANTSCH, E. (1982), Die Selbstorganisation des Universums. Vom Urknall zum menschlichen Geist. München:DTV.
JASANOFF, S. (1990), The Fifth Branch. Science Advisers as Policy Makers. Cambridge:Harvard University Press.
JOHNSON-LAIRD, P. (1996), Der Computer im Kopf. Formen und Verfahren der Erkenntnis. München:DTV.
JONES, S. (1993), The Language of Genes. Solving the Mysteries of Our Genetic Past, Present and Future. New York:Anchor Books.
JONES, S., R. MARTIN, D. PILBEAM (31995)(eds.), The Cambridge Encyclopedia of Human Evolution. Cambridge:Cambridge University Press.
KAELBLING, L.P. (1993), Learning in Embedded Systems. Cambridge: The MIT Press.
KAGEL, J.H., A.E. ROTH (1995)(eds.), The Handbook of Expermental Economics. Princeton:Princeton University Press.
KAHNEMANN, D., A. TVERSKY (1973), "On the Psychology of Prediction", in: Psychological Review 80, 237:251.
KAHNEMANN, D., P. SLOVIC, A. TVERSKY (1982), Judgement under Uncertainty: Heuristics and Biases. Cambridge:Cambridge University Press.

KAMANN, D.J.F. (1985), Innovation, Industrial Organization, Networks and Employment. Groningen:Research Paper.
KAPFERER, J.N. (1996), Gerüchte. Das älteste Massenmedium der Welt. Leipzig:Gustav Kiepenheuer Verlag.
KAST, V. (1996), Die Dynamik der Symbole. Grund lagen der Jungschen Psychotherapie. München:Deutscher Taschenbuch Verlag.
KASVIO, A. (1995), Reinventing the Nordic Model. Can the Nordic Countries Succeed in 21st Century Global Competition . Tampere:Working Paper 2
       (Department of Sociology and Socialpsychology)
KATZ, R. (1988)(ed.), Managing Professionals in Innovative Organisations. New York: Harper Business.
KATZENSTEIN, P. (1985), Small States in World Markets. Industrial Policy in Europe Ithaka:Longman.
KAUFFMAN, S.A. (1990), "Requirements for Evolvability in Complex Systems", in: W.H. ZUREK (1990), 151 - 192.
KAUFFMAN, S.A. (1993), The Origins of Order. Self-Organization and Selection in Evolution. Oxford University Press.
KAUFFMAN, S.A. (1995), At Home in the Universe. The Search for the Laws of Self-Organization and Complexity. New York:Oxford University Press.
KAYE, B. (1993), Chaos and Complexity. Discovering the Surprising Patterns of Science and Technology. Weinheim:VCH Verlagsgesellschaft.
KENIS, P. (1992), The Social Construction of an Induistry. A World of Chemical Fibres. Frankfurt:Campus Verlag&Westview Press.
KERN, L. (1984)(ed.), Probleme der postindustriellen Gesellschaft. Königstein:Athenäum.
KILPER, H., G. SIMONIS (1992) “Arbeitsorientierte Technologiepolitik - Vergleichende Analyse staatlicher Programme von Arbeit und Technik”, in: K.
       GRIMMER et al. (eds.) (1992), Politische Techniksteuerung. Opladen:Westdeutscher Verlag, 203-226
KINDLEBERGER, C. P.,B.HERRICK (31981), Economic Development. Auckland:McGraw Hill.
KINGDON, J. (1993), Self-made Man. Human Evolution from Eden to Extinction. New York:John Wiley&Sons.
KITCHER, P. (1993), The Advancement of Science. Science without Legend, Objectivity without Illusions. New York:Oxford University Press.
KLAUS, G., H. LIEBSCHER (1979), Wörterbuch der Kybernetik. Frankfurt:Fischer Taschenbuch Verlag.
KLEINKNECHT, A. (1987), Innovation Patterns in Crisis and Prosperity. Schumpeter's Long Cycle Reconsidered. London:Macmillan Press.
KNEUCKER, R. F. (1985), “Förderungsentscheidungen in Österreich”, in: H. FISCHER (ed.) (1985). Forschungspolitik für die 90er Jahre. Wien:Springer
       Verlag., 211-237
KNEUCKER, R. F. (1993), “Wissenschaft, Forschung Technologie - Auswirkungen des EWR-Vertrages”, in: M. GEHLER, R STEININGER (eds.),
       Österreich und die Europäische Integration 1945-1993. Aspekte einer wechselvollen Entwicklung. Wien:Böhlau, 477-500.
KNEUCKER, R. F. (1995), “Eine neue Internationalität für Forschung und Forschungspolitik?”. in: R. MARTINSEN, G. SIMONIS (eds.) (1995),
       Paradigmenwechsel in der Technologiepolitik? Opladen:Leske + Budrich, 47-56.
KNIGHT, R., K. LEITNER (1993), The Potentials of Vienna´s Knowledge Base for City Development. Vienna:Magistratsabteilung 18.
KNOFLACHER, H., E. BUCHINGER, E. SCHIBEL (1994): Interaktionsmodell als Beitrag zur ganzheitlichen Kulturlandschaftsforschung. Seibersdorf:
       Forschungszentrum Seibersdorf.
KNORR-CETINA, K. (1984), Die Fabrikation von Erkenntnis. Zur Anthropologie der Naturwissenschaft. Frankfurt:Suhrkamp.
KNORR-CETINA, K. (1988), The Micro-Social Order. Towards a Reconception, in: N.G. FIELDING (1988)(ed.), Actions and Structure. Research Methods
       and Social Theory. London:Frances Pinter, 20 - 53.
KNORR-CETINA, K. (1995), “What Scientists Do”, in: U. FELT, H. NOWOTNY (1995)(eds.), Social Studies op.cit., 117 -133.
KNORR-CETINA, K. (1996), Epistemic Cultures. How Scientists Make Sense. (to be published).
KOCH, W.A. (1986), Genes vs. Memes. Modes of Integration for Natural and Cultural Evolution in a Holistic Model („ELPIS‟). Bochum:Studienverlag Dr.
       Norbert Brockmeyer.
KOCKELMANS, J. (1978) (ed.), Interdisciplinarity. Reflections on Historical, Epistemological, Educational and Administrative Issues. University of
       Pennsylvania Press.
KOEPKE, E. (1986), Das künstlerische Schaffen der Menschheit im Zusammenhang mit ihrer Bewußtseinsentwicklung. Schaffhausen:Novalis Verlag.
KOESTLER, A. (1993), Der Mensch - Irrläufer der Evolution. Die Kluft zwischen Denken und Handeln. Eine Anatomie menschlicher Vernunft und
       Unvernunft. Frankfurt:Fischer Taschenbuch Verlag.
KOHN, A. (1989), Fortune or Failure. Missed Opportunities and Chance Discoveries. Oxford:Basil Blackwell.
KOHONEN, T. (1995), Self-Organizing Maps. Berlin:Springer.
KÖNIG, H. (21970)(eds), Wachstum und Entwicklung der Wirtschaft. Köln:Kiepenheuer&Witsch.
KORNWACHS, K. (1993), Information und Kommunikation. Zur menschengerechten Technikgestaltung. Berlin:Springer.
KOSKO, B. (1995), Fuzzy-Logisch. Eine neue Art des Denkens. Düsseldorf:Econ-Verlag.
KOSSLYN, S.M., R.A. ANDERSEN (1992)(eds.), Frontiers in Cognitive Neuroscience. Cambridge:The MIT Press.
KOZA, J.R. (1992), Genetic Programming. On the Programming of Computers by Means of Natural Selection. Cambridge:The MIT Press.
KREIBICH, R. (1986), Die Wissenschaftsgesellschaft. Von Galilei zur High-Tech-Revolution. Frankfurt:Suhrkamp.
KREUZER, F. (1983)(ed.), Markt, Plan, Freiheit. Franz Kreuzer im Gespräch mit Friedrich von Hayek und Ralf Dahrendorf. Wien:Deuticke.
KRIPKE, S.A. (1985), Wittgenstein on Rules and Private Language. An Elementary Exposition. Oxford:Oxford University Press.
KROHN, E., G. KÜPPERS, H. NOWOTNY (1990)(eds.), Selforganization. Portrait of a Scientific Revolution. Dordrecht:Kluwer..
KROHN, W., G. KRÜCKEN (1993)(eds.), Riskante Technologien: Reflexion und Regulation. Einführung in die sozialwissenschaftliche Risikoforschung.
KRUPP, E. (ed.), Technikpolitik angesichts der Umweltkatastrophe. Heidelberg:Physica-Verlag.
KUHLMANN, S. (1994), “TA-Zentren an Universitäten - Ausländische Erfahrungen”, in: G. BECHMANN, T. PETERMANN (eds.)(1994),
       Technikforschung, op.cit., 347 - 378.
KUHLMANN, S., D. HOLLAND (1995), Evaluation von Technologiepolitik in Deutschland - Konzepte, Anwendung und Perspektiven. Heidelberg:Physika..
KUHN, T.S. (1973), Die Struktur wissenschaftlicher Revolutionen. Frankfurt:Suhrkamp.
KÜHNE, K. (1982), Evolutionsökonomie. Grundlagen der Nationalökonomie und Realtheorie der Geldwirtschaft. Stuttgart:Gustav Fischer Verlag.
KÜPPERS, B.O. (1987)(eds.), Ordnung aus dem Chaos. Prinzipien der Selbstorganisation und Evolution des Lebens. München:Piper.
KURZWEIL, R. (21992), The Age of Intelligent Machines. Cambridge:MIT Press.
KUTTRUFF, Silvia (1994), Wissenstransfer zwischen Universität und Wirtschaft: Modellgestützte Analyse der Kooperation und regionale Strukturieru ng -
       dargestellt am Beispiel der Stadt Erlangen. Erlangen:Diss.
LANDES, D.S. (81979), The Unbound Prometheus. Technological Change and Industrial Development in Western Europe from 1750 to the Present .
       Cambridge University Press.
LANGER, J. (1988)(ed.), Geschichte der österreichischen Soziologie. Konstituierung, Entwicklung und europäische Bezüge. Wien:Verlag für
LANGLEY, P., H. A. SIMON, G.L. BRADSHAW, J. M. ZYTKOW (1987), Scientific Discovery. Computational Explorations of the Creative Processes.
       Cambridge:The MIT Press.
LANGLOIS, R.N. (1989)(ed.), Economics as a Process. Essays in the New Institutional Economics. Cambridge:Cambridge University Press.
LANGTON, C.G. (1989)(ed.), Artificial Life. Redwood:Addison Wesley.
LANGTON, C.G., C. TAYLOR, J.D. FARMER, S. RASMUSSEN (1992)(eds.), Artificial Life II. Redwood:Addison Wesley.

LANGTON, C.G. (1994)(ed.), Artificial Life III. Redwood:Addison Wesley.
LASH, S., B. SZERSZYNSKI, B. WYNNE (1996)(eds.), Risk, Environment and Modernity. Towards a New Ecology. London:Sage Publications.
LASSNIGG, L. (1989), Ausbildung und Berufe in Österreich. Problemorientierte Beschreibung und Analyse des Systems beruflicher Erstausbildung .
LASSNIGG, L. (1993), “Connoisseurship und Selbstreferenz -- Bemerkungen zur Hochschulforschung in Österreich”, in: L. LASSNIGG (ed.),
        Hochschulreformen in Europa - Autonomisierung, Diversifizierung, Selbstorganisation. Vienna:IHS Series in Sociology, Vol. 2.
LASSNIGG, L. (1994), Changing Pathways and Participation in VOTEC - Country report: Austria. Vienna:IHS Research Report.
LASSNIGG, L. (1995), “Higher Education and Employment: Is There an Idiosyncratic Austrian Situation?”, in: European Journal of Higher Education 1,
LASTRES, H. M. (1994), The Advanced Materials Revolution and the Japanese System of Innovation. New York :St. Martin's Press.
LASZLO (1972), The Systems View of the World. The Natural Philosophy of the New Developments in the Sciences. New York:Free Press.
LASZLO, E. (1995), Kosmische Kreativität. Neue Grundlagen einer einheitlichen Wissenschaft von Materie, Geist und Leben. Frankfurt:Insel Verlag.
LATOUR, B. (1987), Science in Action: How to Follow Scientists and Engineers through Society. Cambridge:Harvard University Press.
LATOUR, B. (1988), The Pasteurization of France. Cambridge:Harvard University Press.
LATOUR, B. (1992a), We Have Never Been Modern. Cambridge:Harvard University Press.
LATOUR, B. (1992b), “Pasteur on Lactid Acid Yeast: A Partial Semiotic Analysis”, in: Configurations 1, 129 - 145.
LATOUR, B., P. MAUGUIN, G. TEIL (1992), “A Note on Socio-Technical Graphs”, in: Social Studies of Science 22, 33 - 57.
LATOUR, B. (1994), “Joliot: History and Physics Mixed Together”, in: M. SERRES (1994)(ed.), A History of Scientific Thought. Oxford:Blackwell, 611 -
LATZER, M., J. BAUER (eds.) (1994), Cash Lines. The Development and Regulation of Audiotex in Europe and the USA. Amsterdam.
LAUDAN, L. (1977), Progress and Its Problems. Toward a Theory of Scientific Growth. Berkely:University of California Press.
LAUDAN, L. (1981), Science and Hypothesis. Historical Essays on Scientific Methodology. Dordrecht:Reidel Publishing Company.
LEAKY, R., R. LEWIN (1995), The Sixth Extinction. Patterns of Life and the Future of Humankind. New York:Doubleday.
LEEBAERT, D. (1991)(ed.), Technology 2001 - The Future of Computing and Communications. Cambridge:The MIT Press.
LEINFELLNER, W. (1985), A Cyclic Model of Innovations, in: Rivista Internationale di Scienze Economiche e Commerciali 9, 849 - 863.
LEM, S. (1983), Philosophie des Zufalls. Zu einer empirischen Theorie der Literatur., 2 vol. Frankfurt am Main:Insel Verlag.
LEO, H. (1993a), Technischer Fortschritt, Arbeitslosigkeit und Technologiepolitik. Vienna:TIP.
LEO, H. (1993b), TIP Workshop Proceedings: Managing Transfer Sciences. Vienna:TIP.
LEO, H., G. Palme, E. Volk (1992), Die Innovationstätigkeit der österreichischen Industrie. Technologie- und Innovationstest 1990. Vienna:TIP.
LEO, H.. M. PENEDER, N. KNOLL, F. OHLER, M. LATZER (1994), Telekommunikation im Umbruch. Innovation - Regulierung - Wettbewerb.
LEYDESDORFF, L., P.v.d. BESSELAAR (1994)(eds.), Evolutionary Economics and Chaos Theory. New Directions in Technology Studies. London:Pinter
LIBERATORE, M. J. (1990) (ed), Selection and Evaluation of Advanced Manufacturing Technologies. Berlin:Springer.
LIGHTFOOT, D. (41986), The Language Lottery: Toward a Biology of Grammars. Cambridge:The MIT Press.
LIGHTFOOT, D. (1991), How to Set Parameters. Arguments from Language Change. Cambridge:The MIT Pres.
LINDBLOM, C. (1977), Politics and Markets. New York:Basic Books.
LINDBLOM, C. (1990), Inquiry and Change. The Troubled Attempt to Understand and Shape Society. New Haven:Yale Universioty Press.
LINESTONE, H.A., TUROFF, M. (1975)(eds): The Delphi Method, London:Francis Pinter.
LINSTONE, H.A., I. MITROFF (1994), The Challenge of the 21st Century. Managing technology and Ourselves in a Shrinking World. Albany:State
        University of New York.
LITTLER, D. (1988), Technological Development. Oxford:Basil Blackwell.
LLOYD, E.A. (1994), The Structure and Confirmation of Evolutionary Theory. Princeton:Princeton University Press.
LORENZ, K. (1973), Die Rückseite des Spiegels. Versuch einer Naturgeschichte menschlichen Erkennens. München:Piper.
LORENZEN, H.-P. (1985), Effektive Forschungs- und Technologiepolitik. Abschätzung und Reformvorschläge. Frankfurt:Campus.
LUDWIG, G. (21990), Die Grundstrukturen einer physikalischen Theorie. Berlin:Springer.
LUHMANN, N. (1984), Soziale Systeme. Grundriß einer allgemeinen Theorie. Frankfurt:Suhrkamp.
LUHMANN, N. (1986), Ökologische Kommunikation. Kann die moderne Gesellschaft sich auf ökologische Gefährdungen einstellen ?
        Opladen:Westdeutscher Verlag.
LUHMANN, N. (1988), Die Wirtschaft der Gesellschaft. Frankfurt:Suhrkamp.
LUHMANN, N. (1990), Die Wissenschaft der Gesellschaft. Frankfurt:Suhrkamp.
LUMSDEN, C.J., E.O. WILSON (1981), Genes, Mind and Culture - The Coevolutionary Process. Cambridge:Harvard University Press.
LUMSDEN, C.J., E.O. WILSON (1983), Promethean Fire. Reflections on the Origin of Mind. Cambridge:Harvard University Press.
LUNDGREEN, P., B. HORN, W. KROHN, G. KÜPPERS, R. PASLACK (1986), Staatliche Forschung in Deutschland 1870-1980. Frankfurt:Campus.
LUNDVALL, B.A. (ed.) (1992), National Systems of Innovation - Towards a Theory of Innovation and Interactive Learning.London:Pinter.
LUTZ, B. (1984), Der kurze Traum immerwährender Prosperität. Eine Neuinterpretation der industriell-kapitalistischen Entwicklung im Europa des
        20.Jahrhunderts. Frankfurt:Campus-Verlag.
LÜTZ, S. (1993). Die Steuerung industrieller Forschungskooperation. Funktionsweise und Erfolgsbedingungen des staatlichen Förderungsinstrument s
        Verbundforschung. Frankfurt:Campus-Verlag.
LYOTARD, J.F. (1982), Das postmoderne Wissen. Ein Bericht. Wien:Theatrum Machinarum.
LYOTARD, J.F. (1988), “Ob man ohne Körper denken kann”, in: H.U. GUMBRECHT, K.L. PFEIFFER (1988)(eds.), Materialität der Kommunikation.
        Frankfurt:Suhrkamp, 813 - 829.
MAASEN, S., E. MENDELSOHN, P. WEINGART (1995)(eds.), Biology as Society, Society as Biology: Metaphors. Dordrecht:Kluwer Academic
McCULLOCH, W.S. (1988), Embodiments of Mind. Cambridge:The MIT Press.
MACH, E. (1988), "Auszüge aus den Notizbüchern", in: R. HALLERr, F. STADLER (1988)(eds.), Ernst Mach - Werk und Wirkung. Wien:Hölder-Pichler-
MAGEE, J.F. (1991), “Foreword”, in: P.A. ROUSSEL, K.N. SAAD, T.J. ERICKSON (1991), IX - XIII.
MALECKI, E.J. (1981), "Product Cycles, Innovation Cycles, and Regional Economic Change", in: Technological Forecasting and Social Change 19, 291 -
MALIK, F. (1993), Systemisches Management, Evolution, Selbstorganisation. Grundprobleme, Funktionsmechanismen und Lösungsansätze für komplexe
        Systeme. Bern:Verlag Paul Haupt.
MANDELBROT, B.B. (31983), The Fractal Geometry of Nature. New York.

MARCHETTI, C. (1981), Society as a Learning System: Discovery, Invention, and Innovation Cycles Revisited. Laxenburg 1981.
MARIN, B. (1993), “Sozialforschung als Bereitstellungsgewerbe?”, in: F. HARTMANN (1993), op.cit., 113 - 125.
MARR, D. (1981), Vision. A Computational Investigation into the Human Representation and Processing of Visual Information . New York:W.H. Freeman.
MARTINDALE, C. (1990), The Clockwork Muse. The Predictability of Artistic Change. New York:Basic Books.
MARTINSEN, R. (1991), “Technologiepolitische Strategien von Kleinstaaten im europäischen Kontext - das Beispiel Österreich”, in: SWS-Rundschau 4,
MARTINSEN, R. (1994a), “Der lange Weg in die „Europäische Technologiegemeinschaft“ - europäische Integration und technologische Entwicklung zu
       Beginn der 90er Jahre”, in: Wirtschaftspolitische Blätter 5/6, 468-481.
MARTINSEN, R. (1994b), “Gentechnologie als Politikum. Über Herausforderungen an die Politik im Umgang mit Unsicherheit”, in: Was ist der Mensch?
       Menschenbilder im Wandel. Europäisches Forum Alpbach 1993, ed. by H. PFUSTERSCHMID-HARDTENSTEIN, Vienna:Ibera, 258-266.
MARTINSEN, R. (1994c), “Moderne Modernisierung. Zur Herausbildung der Integration des Sozialfaktors in technologiepolitische Konzepte”, in: Journal
       für Sozialforschung, 1/1994, 3-19.
MARTINSEN, R., J. MELCHIOR (1994), Innovative Technologiepolitik. Optionen sozialverträglicher Technikgestaltung mit einer Fallstudie über
       Österreich. Pfaffenweiler:Centaurus-Verlag.
MARTINSEN, R. (1995), “„Der lernende Staat‟ als neues Paradigma der politischen Techniksteuerung”, in: R. MARTINSEN, G. SIMONIS (eds.),
       Paradigmenwechsel in der Technologiepolitik? Opladen:Leske + Budrich, 13-29.
MARTINSEN, R., G. SIMONIS (eds.) (1995), Paradigmenwechsel in der Technologiepolitik? Opladen:Leske + Budrich.
MARTINSEN, R. (1996)(ed.), Politik und Biotechnologie - zum Verhältnis von machtbasierten Entscheidungen und technologischen Entwicklungspfaden.
       Frankfurt am Main:Campus.
MATTHES, J. (1983)(eds.), Krise der Arbeitsgesellschaft? Verhandlungen des 21. Deutschen Soziologentages in Bamberg 1982. Frankfurt:Campus Verlag.
MATURANA, H.R. (1985), Erkennen: Die Organisation und Verkörperung von Wirklichkeit. Ausgewählte Arbeiten zur biologischen Epistemologie.
MATURANA, H.R., F.J. VARELA (1987), Der Baum der Erkenntnis. Die biologischen Wurzeln des menschlichen Erkennens. Bern:Scherz-Verlag.
MATURANA, H. R. (1991), „Reality: The Search for Objectivity or the Quest for a Compelling Argument“, in: N. LESER et al. (1991)(eds.), Die
       Gedankenwelt Sir Karl Poppers. Kritischer Rationalismus im Dialog. Heidelberg:Carl Winter Universitätsverlag, 282 - 368.
MATZNER, E., J. KREGEL, A. RONCAGLIA (1987)(eds.), Arbeit für alle ist möglich. Über ökonomische und institutionelle Bedingungen erfolgreicher
       Beschäftigungs- und Arbeitsmarktpolitik. Berlin:edition sigma.
MAYNARD SMITH, J. (1974), Models in Ecology. Cambridge:Cambridge University Press.
MAYNARD SMITH, J. (1982)(ed.), Evolution Now. A Century after Darwin. London:Macmillan.
MAYNARD SMITH, J. (31985), Evolution and the Theory of Games. Cambridge:Cambridge University Press.
MAYNARD SMITH, J. (1989), Evolutionary Genetics. Oxford:Oxford University Press.
MAYNTZ, R., T.P. HUGHES (1988)(eds.), The Development of Large Technical Systems. Frankfurt:Campus.
MAYNTZ, R. (1993), “Policy Netzwerke und die Logik von Verhandlungssystemen”, in: Policy Analyse: Kritik und Neuorientierung (PVS Sonderheft 24),
       39 - 56.
MAYNTZ, R., F.W. SCHARPF (1995)(eds.), Gesellschaftliche Selbstregelung und politische Steuerung. Frankfurt:Campus Verlag.
MAZEY, S., J. RICHARDSON (1993)(eds.), Lobbying in the European Community. Oxford:Basil Blackwell.
MEIER, G.M. (41984)(eds.), Leading Issues in Economic Development. New York:Oxford University Press.
MELCHIOR, J. (1990), “Zur österreichischen Forschungs- und Technologiepolitik: Entwicklungen und Probleme im Kontext internationaler Diskussionen”,
       in: Österreichische Zeitschrift für Politikwissenschaft 3, 245-265
MENSCH, G. (1977), Das technologische Patt. Innovationen überwinden die Depression. Frankfurt am Main:Fischer.
MENSCH, G., W. WEIDLICH, G. HAAG (1991), “The Schumpeter Clock. A Micro-Macro-Model of Economic Change, Including Innovation, Strategic
       Investment, Dynamic Competition and Short and Long Swings in Industrial Transformation - Applied to United States and German Data, in: OECD
       (1991)(ed.), Technology and Productivity. The Challenge for Economic Policy. Paris:OECD.
MERTON, R.K. (1985), Entwicklung und Wandel von Forschungsinteressen. Aufsätze zur Wissenschaftssoziologie. Frankfurt:Suhrkamp.
MESAROVIC, MACKO, TAKAHARA (1970), Theory of Hierarchical Multi-Level Systems. New York:Free Press.
METZINGER, T. (1993), Subjekt und Selbstmodell. Die Perspektivität phänomenalen Bewußtseins vor dem Hintergrund einer naturalistischen Theorie
       mentaler Repräsentation. Paderborn:Ferdinand Schöningh.
METZINGER, T. (21996), Bewußtsein. Beiträge aus der Gegenwartsphilosophie. Paderborn:Ferdinand Schöningh.
MICHALSKI, R.S., J.G. CARBONELL, T.M. MITCHELL (1986)(eds.), Machine Learning. An Artificial Intelligence Approach. Los Altos:Morgan
MILLIKAN, R.G. (1984), Language, Thought, and Other Biological Categories. New Foundations for Realism. Cambridge: Cambridge:The MIT Press.
MILLER, J. (1978), Living Systems. New York:Basic Books.
MINSKY, M. (1990), Mentopolis. Stuttgart:Klett-Cotta.
MOLIERE, J.B. (o.J.), Komödien. Stuttgart:Deutscher Bücherbund.
MORRIS, D. (1981), Der Mensch, mit dem wir leben. Ein Handbuch unseres Verhaltens. München:Knaur.
MORRIS, D. (1996), Warum klappert der Storch? Körpersprache und Verhaltensformen der Tiere. München:Wilhelm Heyne Verlag.
MORTON, E.S., J. PAGE (1992), Animal Talk. Science and the Voices of Nature. New York:Random House.
MOWERY, D. C.(1994), Science and Technology Policy in Interdependent Economies. Boston:Kluwer.
MUEHLMANN, H. (1996), Die Natur der Kulturen. Entwurf einer kulturgenetischen Theorie. Wien:Springer
MUELLER, D.C. (1983)(ed.), The Political Economy of Growth. New Haven:Yale University Press.
MUELLER, U. (1996), “Evolutionary Explanations from a Philosophy of Science Point of View”, in: R. HEGSELMANN, U. MUELLER, K.G.
       TROITZSCH, op. cit., 101 - 122.
MÜLLER, F. (1991), Diffusionsprozesse von Innovationen. Vienna:BMWF.
MÜLLER, K.H. (1984a), Industrialization and Scoial Development. Investigating the Simultaneous Interactions. Vienna:UNIDO.
MÜLLER, K.H. (1984b), „Von Schichten und Geschichten“, in: K.H. MÜLLER., R. HINTEREGGER, E. STAUDINGER (1984)(eds.), Auf dem Weg in die
       Freiheit. Anstöße zu einer steirischen Zeitgeschichte. Graz:Leykam, 11 - 33.
MÜLLER, K.H. (1986), “Grundbedürfnisse”, in: M. BENEDIKT, R. BURGER (1986)(eds.), Kritische Methode und Zukunft der Anthropologie.
       Vienna:Wilhelm Braumüller, 121 - 130.
MÜLLER, K. H. (1987), “Die Idealwelten der österreichischen Nationalökonomen”, in: F. STADLER (1987)(ed.), Vertriebene Vernunft. Emigration und Exil
       österreichischer Wissenschaft. Wien:Jugend und Volk, 238 - 275.
MÜLLER, K.H. (1988a), “Lange Wellen und Beschäftigung”, in: H. SEIDEL, G. SCHIENSTOCK (1988), 31 - 45.
MÜLLER, K.H. (1988b), Hochzeit der Sozialwissenschaften 1871 - 1938, in: J. LANGER (1988), 51 - 69.
MÜLLER, K.H. (1988c), Weltwirtschaft und nationale Wissenschaftsentwicklung. Ein Erklärungssketch, in: F. STADLER (1988A), 341 - 399.

MÜLLER, K.H. (1990); "Langfristige Systemanalyse des österreichischen Beschäftigungssystems", in: K.H. MÜLLER, K. PICHELMANN (1990), 49 - 169.
MÜLLER, K.H., K. PICHELMANN (1990)(eds.), Modell zur Analyse des österreichischen Beschäftigungssystems. Vienna:IHS.
MÜLLER, K.H. (1991a), "Langfristige Entwicklungen im österreichischen Beschäftigungssystem", in: Arbeitsmarkt 6, 4 - 9.
MÜLLER, K.H. (1991b), Symbole-Statistik-Computer-Design. Wien:Hölder-Pichler-Tempsky.
MÜLLER, K.H. (1991c), “Otto Neurath´s Pictorial Statisticis” , in: T.E. Uebel (1991)(ed.), Rediscovering the Forgotten Vienna Circle. Dordrecht:Reidel.
MÜLLER, K.H. , L. LASSNIGG (1992)(eds.), Langfristige Szenarienanalyse des österreichischen Bildungssystems. Vienna:IHS.
MÜLLER, K.H. (1993), Zement und Gesellschaft. Modernisierungsskizzen aus dem Geist Karl Polanyis. Vienna:IHS Research Paper.
MÜLLER, K.H (1994a) "Sozioökonomische Erklärungsskizzen für gesamtgesellschaftliche Transformationen", in: WISDOM 3/4, 138 - 154.
MÜLLER, K. H. (1994b), Technology Audit for Hungary: The Sector of Agricultural Machine Production. Vienna:IHS.
MÜLLER, K.H. (1995a), “Zement und Gesellschaft. Modernisierungsskizzen aus dem Geist Karl Polanyis”, in: K. R. LEUBE, A. PRIBERSKI (1995)(eds.),
      Krise und Exodus. Österreichische Sozialwissenschaften in Mitteleuropa. Wien:WUV-Universitätsverlag, 146 - 190
MÜLLER, K.H. (1995b), Epistemic Cultures in the Social Sciences. The Modeling Dilemma Dissolved. Vienna:IHS Research Paper.
MÜLLER, K.H. (1996), Die Sozialwissenschaften im Zeitalter ihrer transdisziplinären Reproduzierbarkeit. Vienna:IHS.
MÜLLER, K.H., G. HAAG (1994)(eds.), Komplexe Modelle in den Sozialwissenschaften. Special Edition of WISDOM 3/4.
MÜLLER, K.H. (1996), “Sozialwissenschaftliche Kreativität in der Ersten und in der Zweiten Republik”, in: Österreichische Zeitschrift für
      Geschichtswissenschaften 1, 9 - 43.
MÜLLER, K.H. et al. (1996), The Austrian Innovation System, 7 vol. Vienna:IHS.
MÜNCH, R., N.J. SMELSER (1987), „Relating the Micro and Macro“, in: J.C. ALEXANDER, B. GIESEN, R. MÜNCH, N.J. SMELSER (1987)(eds.), The
      Micro-Macro Link. Berkeley:University of California Press.
MÜNCH, R. (1988), Theorie des Handelns. Zur Rekonstruktion der Beiträge von Talcott Parsons, Emile Durkheim und Max Weber. Frankfurt:Suhrkamp.
MYERS, N. (1990), The Gaia Atlas of Future Worlds. Challenge and Opportunity in a World of Change. New York:Anchor Books.
NADEL, L., D.L. STEIN (1991)(eds.), 1990 Lectures in Complex Systems. Redwood City:Addison-Wesley.
NAGEL, T. (1986), A View from Nowhere. Oxford:Oxford University Press.
NAISBITT, J. (1982), Megatrends. 10 Perspektiven, die unser Leben verändern werden. München:Piper.
NAISBITT, J. (1995), Megatrends Asien. Acht Megatrends, die unsere Welt verändern. Wien:Signum.
NASCHOLD, F. (1985), “Zum Zusammenhang von Arbeit, sozialer Sicherung und Politik. Einführende Anmerkungen zur Arbeitspolitik”, in: F.
      NASCHOLD (ed.) (1985), Arbeit und Politik. Gesellschaftliche Regulierung der Arbeit und der sozialen Sicherung. Frankfurt am Main:Campus, 9-
NASCHOLD, F. (1991), “Internationale Konkurrenz, sektorale Produktionsregimes und nationalstaatliche Arbeitspolitik”, in: B. BLANKE, H. WOLLMANN
      (eds.) (1991), Die alte Bundesrepublik. Kontinuität und Wandel. Opladen:Westdeutscher Verlag,, 106-129.
NEFIODOW, L.A. (21991), Der fünfte Kondratieff. Strategien zum Strukturwandel in Wirtschaft und Gesellschaft. Wiesbaden:Gabler-Verlag.
NEGROPONTE, N. (1996), Being Digital. New York:Vintage Books.
NELSON, R.J. (1982), The Logic of Mind. Dordrecht:Reidel..
NELSON, R.R., S.G. WINTER (1982), An Evolutionary Theory of Economic Change. Cambridge:Harvard University Press.
NELSON, R. R. (1993)(ed.), National Innovation Systems: A Comparative Analysis. New York:Oxford University Press.
NEUMAIER, O. (1987), “A Wittgensteinian View of Artificial Intelligence” in: R. BORN (1987)(ed.), Artificial Intelligence. The Case Against.
      London:Croom Helm, 132 - 173.
NEUSER, W., K. NEUSER-OETTINGEN (1996)(eds.), Quantenphilosophie. Mit einem Nachwort von C.F. v. Weizsäcker. Heidelberg:Spektrum.
NEURATH, O. (1970), “Foundations of the Social Sciences”, in: O. NEURATH, R. CARNAP, C. MORRIS (1970)(eds.), Foundations of the Unity of
      Science. Toward an International Encyclopedia of Unified Science. Chicago:University of Chicago Press, 1 - 51.
NEURATH, O. (1981), Gesammelte philosophische und methodologische Schriften.(ed. by R. Haller and H. Rutte). Vienna:Hölder Pichler Tempsky.
NEWELL, A. (1990), Unified Theories of Cognition. Harvard University Press.
NICOLIS, G., I. PRIGOGINE (1977), Self-Organization in Nonequilibrium Systems. From Dissipative Structures to Order through Fluctuations. New
      York:John Wiley&Sons.
NICOLIS, G., I. PRIGOGINE (31982), Vom Sein zum Werden. Zeit und Komplexität in den Naturwissenschaften. München:Piper.
NIJKAMP, P. (1986)(Ed.), Handbook of Regional and Urban Economics, Bd.1. Amsterdam.
NOWAK, A., M. LEWENSTEIN (1994), “Dynamical Systems: A Tool for Social Psychology?” in: VALLACHER, NOWAK, 17 - 52.
NOWOTNY, E. (1985), “Technologiepolitik als wirtschaftspolitische Herausforderung”, in: H. FISCHER (ed.), Forschungspolitik für die 90er Jahre.
      Wien:Springer-Verlag,, 417-428.
NOWOTNY, H. (1990), In Search of Usable Knowledge. Utilization Contexts and the Application of Knowledge. Frankfurt:Campus Verlag&Westview
NOWOTNY, H. (1995), The Dynamics of Innovation. On the Multiplicity of the New. Budapest:Collegium Hungaricum.
NOWOTNY, H. (1996), “Mechanismen und Bedingungen der Wissensproduktion. Zur gegenwärtigen Umstrukturierung des Wissenschaftssystems”, in: Neue
      Züricher Zeitung, 6.7. Januar, 39.
NRENAISSANCE COMMITTEE et al. (1994), Realizing the Information Future. The Internet and Beyond. Washington:National Academy Press.
NUSSBAUM, M.C., A. SEN (1993)(eds.), The Quality of Life. Oxford:Clarendon Press.
OECD (1988a), New Technologies in the 1990s. A Socio-economic Strategy. Paris:OECD.
OECD (1988b), Reviews of National Science and Technology Policy (Austria). Paris:OECD.
OECD (1991a), Technology in a Changing World. Paris:OECD.
OECD (1991b), International Conference Cycle. Paris:OECD.
OECD (1992a) Innovation Manual: Proposed Guidelines for Collecting and Interpreting Innovation Data (Oslo Manual). Paris:OECD.
OECD (1992b), Science and Technology Policy. Review and Outlook 1991. Paris:OECD.
OECD (1992c), Wissenschafts- und Technologiepolitik. Bilanz und Ausblick 1991. Paris:OECD.
OECD (1992d), Technology and the Economy. The Key Relationships. Paris:OECD.
OECD (1993), National Systems for Financing Innovation. DSTI/STP/TIP(93)3/REV. Paris:OECD.
OECD (1994), Science and Technology Policy. Review and Outlook 1994. Paris:OECD.
OECD (1994/1)(eds.), Short-Term Economic Indicators. Transition Economies, Paris
OECD (1994a), Interaction in Knowledge Systems. Foundation, Policy Implications and Empirical Methods. DSTI/STP/TIP(94)15. Paris:OECD.
OECD (1994b), The Measurement of Scientific and Technical Activities. (Frascati Manual 1993) Paris:OECD.
OECD (1994c), National Innovation Systems. Work Plan for Pilot Case Studies. DSTI/STP/TIP(94)16/REV1.Paris:OECD.
OECD (1994d), Technology Flows in National Systems of Innovation. DSTI/STP/TIP(94)3. Paris:OECD.
OECD (1995a), Interim Report on Technology, Productivity and Job Creation. DSTI/IND/STP/ICCP(95)1. Paris:OECD.
OECD (1995b), Technology, Productivity and Job Creation. DSTI/IND/STP/ICCP(95)14/Part 1 - Part 7. Paris:OECD

OECD (1995c), Education at a Glance. OECD Indicators. Paris:OECD.
OECD (1995d), Main Science and Technology Indicators. Paris:OECD.
OEVERMANN, U., T. ALBERT, E. KONAU, J. KRAMBECK (1979), Die Methodologie einer 'objektiven Hermeneutik' und ihre allgemeine
                 forschungslogische Bedeutung in den Sozialwissenschaften, in: H.G. SOEFFNER (1979), 352 - 434.
OEVERMANN, U. (1983a), Hermeneutische Sinnkonstruktion: Als Therapie und Pädagogik mißverstanden, oder: Das notorische strukturtheoretische
                 Defizit pädagogischer Wissenschaft, in: D. GARZ, K. KRAIMER (1983), 113 - 155.
OEVERMANN, U. (1983b), Zur Sache. Die Bedeutung von Adornos methodologischem Selbstverständnis für die Begründung einer materialen
                 soziologischen Strukturanalyse, in: L. FRIEDEBURG, J. HABERMAS (1983), 234 - 289.
OFFE, C. (1984), "Arbeitsgesellschaft". Strukturprobleme und Zukunftsperspektiven. Frankfurt:Campus.
OGDEN, F. (1993), The Last Book You´ll Ever Read And Other Lessons from the Future. Toronto:Macfarlane Walter&Ross.
OLSON, M. (1982), The Rise and Decline of Nations. Economic Growth, Stagflation and Social Rigidities. Yale:Yale University Press.
ÖSTERREICHISCHES INSTITUT FÜR RAUMPLANUNG (1992), Innovationszentren und Technologietransfereinrichtungen in Österreich. Vienna:ÖIR.
OTTO, P., P. SONNTAG (1985), Wege in die Informationsgesellschaft. Steuerungsprobleme in Wirtschaft und Politik. München:DTV.
OVERBURY, R.E. (1969), „Technological Forecasting. A Critisim of the Delphi-Technique“, in: Long-Range-Planning, vol 1, No 4: 76-77
PACEY, A. (21992), The Maze of Ingenuity. Ideas and Idealism in the Development of Technology. Cambridge:The MIT Press.
PAHL, R. (1995), After Success. Fin de Siècle Anxiety and Identity. Cambridge:Polity Press.
PALLA, R. (1994), Verschwundene Arbeit. Ein Thesaurus der untergegangenen Berufe. Frankfurt:Eichborn Verlag.
PARAYIL; G: (1991): "Technological Change as a Problem-Solving Activity", in: Technological and Social Change, 40:235-247
PARSONS, T. (21961), The Structure of Social Action. A Study in Social theory with Special Reference to a Group of Recent European Writers . New
        York:Basic Books.
PARSONS, T. (1964), "Evolutionary Universals in Society", in: American Sociological Review 19, 339 - 357.
PARSONS, T. (1994), Aktor, Situation und normative Muster. Ein Essay zur Theorie sozialen Handelns. Frankfurt:Suhrkamp.
PASK, G. (1995), “Concept”, in: H. v. FOERSTER (1995), 244.
PASLACK, R. (1991), Urgeschichte der Selbstorganisation. Zur Archäologie eines wissenschaftlichen Paradigmas. Braunschweig-Wiesbaden.
PAVITT, K:L.R. (1987), “The Objectives of Technology Policy”, in: Science and Public Policy 14, 182 - 188.
PAVITT, K.L.R. (1991), “What Makes Basic Research Economically Useful”, in: Research Policy 20, 109 - 119.
PAULA, M. (o.J.), Ecodesign. Rahmenkonzept für einen F&E-Schwerpunkt umweltbewußte Produktgestaltung. Vienna:BMWF.
PEAK, D., M. FRAME (1995), Komplexität - das gezähmte Chaos. Basel:Birkhäuser Verlag.
PENEDER, M. (1994), Clusteranalyse und sektorale Wettbewerbsfähigkeit der österreichischen Industrie. Vienna:TIP.
PENROSE, R. (1995), Shadows of the Mind. A Search for the Missing Science of Consciousness. London:Vintage.
PERROUX, F. (1983), A New Concept of Development. Basic Tenets. Paris:OECD.
PESCHEL, M., W. MENDE (1986), The Predator-Prey Model. Do We Live in a Volterra World. Wien:Springer-Verlag.
PESCHL, M.F. (1994), Repräsentation und Konstruktion. Kognitions- und neuroinformatische Konzepte als Grundlage einer naturalisierten Epistemologie
        und Wissenschaftstheorie. Braunschweig:Vieweg.
PETSCHE, T., S.J. HANSON, J. SHAVLIK (1995)(eds.), Computational Learning Theory and Natural Learning Systems. Volume III: Selecting Good
        Models. Cambridge:The MIT Press.
PHILIPS, D., Y. BERMAN (1995), Human Services in the Age of New Technology. Harmonising Social Work and Computerisation. Aldershot:Avebury.
PIAGET, J. (1973), Einführung in die genetische Erkenntnistheorie. Frankfurt:Suhrkamp.
PIAGET, J. (1983), Biologie und Erkenntnis. Über die Beziehungen zwischen organischen Regulationen und kognitiven Prozessen . Frankfurt:Fischer.
PIATELLI-PALMARINI, M. (1980)(ed.), Language and Learning. The Debate between Jean Piaget and Noam Chomsky. London:Routledge&Kegan Paul.
PICHLER, F (1975), Mathematische Systemtheorie. Dynamische Konstruktionen. Berlin:de Gruyter.
PICHLER, F. (1990). “Österreichs Weg in die Europäische Technologiegemeinschaft”, in: Österreichische Zeitschrift für Politikwissenschaft 3, 317-327.
PICKERING, A. (1992)(ed.), Science as Practice and Culture. Chicago:The University of Chicago Press.
PINES, D. (1988)(ed.), Emerging Syntheses in Science. Redwood City:Addison Wesley.
PINKER, S. (1994), The Language Instinct. New York:William Morrow and Company.
PIORE, M.J., C.F. SABEL (1984), The Second Industrial Divide. Possibilities for Prosperity. New York:Basic Books.
PIRSIG, R.M. (1991), Lila. An Inquiry into Morals. New York:Bantam Books.
PLOTKIN, H. (1994), Darwin Machines and the Nature of Knowledge. Cambridge:Harvard University Press.
POLANYI, K. (1978), The Great Transformation. Politische und ökonomische Ursprünge von Gesellschaften und Wirtschaftssystemen. Frankfurt:Suhrkamp.
POLANYI, K. (1979), Ökonomie und Gesellschaft. Frankfurt:Suhrkamp.
POLANYI, M. (1985), Implizites Wissen. Frankfurt:Suhrkamp.
POLLARD, S. (1981), Peaceful Conquest. The Industrialization of Europe 1760 - 1970. Oxford:Oxford University Press.
POLLOCK, J.S. (1989), How to Build a Person: A Prolegomenon. Cambridge:The MIT Press
POLT, W. (1993), Information Technology Diffusion Policies for Small- and Medium-Sized Enterprises. Background Report.
        Seibersdorf:Forschungszentrum Seibersdorf.
POPCORN, F., L. MARIGOLD (1996), „Clicking‟. Der neue Popcorn Report. Trends für unsere Zukunft. München:Wilhelm Heyne Verlag.
POPPER, K.R (1965), Conjectures and Refutations. The Growth of Scientific Knowledge. New York:Harper&Row.
POPPER, K.R.(31971), Das Elend des Historizismus. Tübingen:J.C.B. Mohr.
POPPER, K.R. (1974), “Autobiography”, in: P. A. SCHILPP (1974)(ed.), The Philosophy of Karl Popper, Bd. 1, La Salle:Open Court, 1 - 181.
POPPER, K.R. (31975), Objective Knowledge. An Evolutionary Approach. Oxford:Oxford University Press.
POPPER, K.R., J.C. ECCLES (1982), Das Ich und sein Gehirn. München:Piper.
POPPER, K.R.. (1982a), Quantum Theory and the Schism in Physics. From the 'Postscript to the Logic of Scientific Discovery'. London:Totowa Press.
POPPER, K.R. (1982b), The Open Universe. An Argument for Indeterminism. From the „Postscript to the Logic of Scientific Discovery‟. Totowa:Rowman
        and Littlefield.
PORTER, M.E. (1985), Competitive Advantage. Creating and Sustaining Superior Performance. New York:The Fre Press.
PORTER, M.E. (1990), The Competitive Advantage of Nations. New York:The Free Press.
POSNER, M.I. (1989)(ed.), Foundations of Cognitive Science. Cmbridge:The MIT Press.
PRIGOGINE, I., I. STENGERS (1984), Order out of Chaos. Man´s New Dialogue with Nature. Toronto et al.
PRIGOGINE, I., I. STENGERS (1993), Das Paradox der Zeit. Zeit, Chaos und Quanten. München-Zürich.
PRISCHING, M. (1993), “Sozialpartnerschaft auf neuen Wegen? Ein Kommentar zur Vereinbarung vom November 1992”, in: Wirtschaftspolitische Blätter
        40, 114 - 121.
PRISCHING, M. (1996), Die Sozialpartnerschaft. Modell der Vergangenheit oder Modell für Europa? Eine kritische Analyse mit Vorschlägen für
        zukunftgerechte Reformen. Wien:Manzsche Verlags- und Universitätsbuchhandlung.

PSACHAROPOULOS (1987), "The International Comparison Model", in:PSACHAROPOULOPS (1987)(ed.), Economics of Education. Research and
       Studies. Oxford et al., 341.
PUTNAM, H. (51984), Philosophical Papers, vol.2: Mind, Language and Reality. Cambridge:Cambridge University Press.
PUTNAM, H. (1985), "What Theories Are Not", in: H. PUTNAM ( 51985), Philosophical Papers, vol.1: Mathematics, Matter, and Method.
       Cambridge:Cambridge University Press, 215 - 227.
QUINN, J. (1992), Intelligent Enterprise.A Knowledge and Service Based Paradigm for Industry. New York:Free Press.
RAMMERT, W. (1988): Das Innovationsdilemma. Technikentwicklung im Unternehmen. Opladen:Westdeutscher Verlag.
RAWLS, J. (81977), A Theory of Justice. Cambridge:Harvard University Press.
RAWLS, J. (1996), Political Liberalism. New York:Columbia University Press.
RAY, G.F. (1980), Innovation and Long-Term Economic Growth. Laxenburg:IIASA.
RESCHER, N. (1979), Cognitive Systematization. A Systems-theoretic Approach to a Coherentist Theory of Knowledge. Oxford:Oxford University Press.
RESCHER, N. (1982), Wissenschaftlicher Fortschritt. Eine Studie über die Ökonomie der Forschung. Berlin:de Gruyter.
RESCHER, N. (1996), Glück. Die Chancen des Zufalls. Berlin:Berlin-Verlag.
RIEDL, R. (1975), Die Ordnung des Lebendigen. Systembedingungen der Evolution. Hamburg:Paul Parey.
RIEDL, R. (21980), Strategie der Genesis. Naturgeschichte der realen Welt. München:Piper.
RIEDL, R. (1987), Begriff und Welt. Biologische Grundlagen des Erkennens und Begreifens. Berlin:Paul Parey.
RIFKIN, J. (1995), The End of Work. The Decline of the Global Labor Force and the Dawn of the Post-Market Era. New York:Jeremy P. Tarcher/Putnam
RITTER, H., T. MARTINEZ, K. SCHULTEN (21991), Neuronale Netze. Eine Einführung in die Neuroinformatik selbstorganisierender Netzwerke.
RITZER, G. (21988), Contemporary Social Theory. New York:Basic Books.
RITZER, G. (1993), The McDonaldization of Society. An Investigation Into the Changing Character of Contemporary Social Life. Thousand Oaks:Pine
       Forge Press.
ROBERTSON, G. et al. (1996)(eds.), FutureNatural. Nature, Science, Culture. London:Routledge.
RODE, R. (1993), High Tech Wettstreit 2000. Strategische Handels- und Industriepolitik. Frankfurt:Campus.
ROJAS, R. (1993), Theorie der neuronalen Netze. Eine systematische Einführung. Berlin:Springer.
ROOBEEK, A. (1990), Beyond the Technology Race. An Analysis of Technology Policy in Seven Industrial Countries. Amsterdam:Kluwer.
ROOT BERNSTEIN, R.S. (1989), Discovering. Inventing and Solving Problems at the Frontiers of Scientific Knowledge. Cambridge:Harvard University
ROPOHL, G. (1996), Ethik und Technikbewertung. Frankfurt:Suhrkamp.
ROSE, H. (ed.) (1995), Nutzerorientierung im Innovationsmanagement. Neue Ergebnisse der Sozialforschung über Technikbedarf und Technikentwicklung .
ROSEN, M.R. (1991), Life Itself. New York:Columbia University Press.
ROSENBERG, N. (1982), Inside the Black Box: Technology in Economics. New York:Cambridge University Press.
ROSENBERG, N. (1991), “Why Do Firms Do Basic Research (with Their Own Money)?”, in: Research Policy 19, 165 - 174.
ROSENBERG, N., W.E. STEINMÜLLER (1989), “Why Are Americans Such Poor Imitators?”, in: American Economic Review 23, 229 - 234.
ROSENBERG, N., R. LANDAU, D.C. MOWERY (1992)(eds.), Technology and the Wealth of Nations. Stanford:Stanford University Press.
ROSTOW, W.W. (21971), The Stages of Economic Growth. A Non-Communist Manifesto. Cambridge:Cambridge University Press.
ROSTOW, W.W. (1978), The World Economy. History & Prospect. Austin:University of Texas Press.
ROTH, G., H. SCHWEGLER (1981)(eds.), Self-Organizing Systems. An Interdisciplinary Approach. Frankfurt:Campus.
ROTHENBERG, D. (1995), Hand´s End. Technology and the Limits of Nature. Berkely:University of California Press.
ROTHWELL, R. (1980), Technology, Structural Change and Manufacturing Employment. IIASA.
ROTHWELL, R., W. ZEGVELD (1985), Reindustrialization and Technology. Harlow.
ROUSSEL, P.A., K.N. SAAD, T.J. ERICKSON (1991), Third Generation R&D. Managing the Link to Corporate Strategy. Boston:Harvard Business School
ROWE, G:, WRIGHT, G., BOPLGER, F:, (1991): "Delphi. A Reevaluation of Research and Theory", in: Technological Forecasting and Social Change, 39:
RUIGROK, W., R.v. TULDER (1995), The Logic of International Restructuring. London:Routledge.
RUMBERGER, R.W. (1984), High Technology and Job Loss. Stanford:Stanford University Press.
RUMELHART, D.E., J.L. McCLELLAND (1986), Parallel Distributed Processing. Explorations in the Microstructure of Cognition, 2 vol. Cambridge:The
       MIT Press.
RUMELHART, D.E (1989), “Toward a Microstructural Account of Human Reasoning”, in: VOSNIADOU, ORTONY, 298 - 312.
RUNCO M.A., R. S. ALBERT (1990)(eds.), Theories of Creativity. Newbury Park:Sage Publications.
RUSCH, G. (1987), Erkenntnis, Wissenschaft, Geschichte von einem konstruktivistischen Standpunkt. Frankfurt:Suhrkamp.
RUSCH, G., S.J. SCHMIDT (1993)(eds.), Konstruktivismus und Sozialtheorie. Frankfurt:Suhrkamp.
RUSSELL, P. (1995), The Global Brain Awakens. Our Next Evolutionary Leap. Palo Alto:Global Brain Inc.
RUST, H. (1995), Trends. Das Geschäft mit der Zukunft. Wien:Kremayr&Scherlau.
RYLE, G. (1969), Der Begriff des Geistes. Stuttgart:Reclam.
SACKS, O. (1995), An Anthropologist on Mars. Seven Paradoxical Tales. New York:Picador.
SAHAL, D. (1981), Patterns of Technological Innovation. London.
SALMON, W.W. (41975), The Foundations of Scientific Inference. Pittsburgh:The University of Pittsburgh Press
SATO, R., M.J. BECKMANN (1983)(eds.), Technology, Organization and Economic Structure. Berlin:Springer.
SAVIGNY (1970), Analytische Philosophie. Freiburg:Alber Kolleg.
SCHABEDOTH, H. J. (1994), Zukunft ohne Arbeit? Neue Wege aus der Strukturkrise, München.
SCHARPF, F.W. (1985), Strukturen der post-industriellen Gesellschaft oder: Verschwindet die Massenarbeitslosigkeit in der Dienstleistungs- und
       Informationsökonomie? WZB.
SCHARPF, F.W. (21987), Sozialdemokratische Krisenpolitik in Europa. Frankfurt:Campus.
SCHARPF, F.W. (1993), “Positive und negative Koordination in Verhandlungssystemen”, in: Policy Analyse: Kritik und Neuorientierung (PVS Sonderheft
       24), 57 - 83.
SCHERER, F.M. (1986), Innovation and Growth. Schumpeterian Perspectives. Cambridge:MIT Press.
SCHETTKAT, R., M. WAGNER (1989)(eds.), Technologischer Wandel und Beschäftigung. Fakten, Analysen, Trends. Berlin:de Gruyter.
SCHETTKAT, R., M. WAGNER. (1989) (eds), Technologischer Wandel und Beschäftigung. Fakten, Analysen, Trends. Berlin:de Gruyter
SCHIEBINGER, L. (1993), Schöne Geister? Stuttgart:Klett-Cotta.

SCHIEBINGER, L. (1995), Am Busen der Natur. Stuttgart:Klett-Cotta.
SCHIENSTOCK, G. (1994), “Technology Policy in the Process of Change: Changing Paradigms in Research and Technology Policy?”, in: G.
        AICHHOLZER, G. SCHIENSTOCK (eds.) (1994), Technology Policy Towards an Integration of Social and Ecological Concerns. Berlin:de
SCHIENSTOCK, G., B. STEFFENSEN (1995), “Lean Production als Leitbild der Restrukturierung einer Region. Die Wirtschaft Baden-Württembergs im
        Wandel”, in: J. FISCHER, S. GENSIOR (eds.)(1995), Soziale und technische Vernetzung von Arbeit und Arbeitsorganisation. Berlin:Sigma.
SCHIMANK, U., A. STUCKE (1994)(eds.), Coping with Trouble. How Science Reacts to Political Disturbances of Research Conditions. Frankfurt:Campus
        Verlag&St. Martin´s Press.
SCHLICHT, E. (1985), Isolation and Aggregation in Economics. Berlin:Springer.
SCHMEIKAL, B. (1996), “The Generative Process of Space-Time and Strong Interaction Quantum Numbers of Orientation”, in: R. ABLAMOWICZ, P.
        LOUNESTO, J.M. PARRA (1996)(eds.), Clifford Algebnras with Numeric and Symbolic Computations. Basel:Birkhäuser-Verlag, 83 - 100.
SCHMIDJELL, R. / HOBENDANNER, A. (o.J.), Bedingungen, Arbeitsweise bzw. Erfolge von Technologie- und Technologietransferzentren in Österreich -
        unter besonderer Berücksichtigung der Erfahrungen des Salzburger Technologiezentrums Techno-Z. Vienna:BMWF.
SCHMIDT, S.J. (1987)(ed.), Der Diskurs des radikalen Konstruktivismus. Frankfurt:Suhrkamp.
SCHMIDT, S.J. (1991)(ed.), Gedächtnis. Probleme und Perspektiven der interdisziplinären Gedächtnisforschung. Frankfurt:Suhrkamp.
SCHMITZ, C., B. ZUCKER (1996), Wissen gewinnt. Knowledge Flow Management. Düsseldorf:Metropolitan Verlag.
SCHMOOKLER,J. (1966), Invention and Economic Growth. Harvard University Press.
SCHÖNEBURG, E. (1993)(ed.), Industrielle Anwendung Neuronaler Netze. Fallbeispiele und Anwendungskonzepte. Bonn:Addison-Wesley.
SCHORSKE, C.E. (1981), Fin de siècle Vienna. Politics and Culture. New York:Basic Books.
SCHROEDER, K. (1990), “Struktur und Profil staatlicher Technologiepolitik”, in:TSCHIEDEL, R. (ed.): Die technische Konstruktion der Wirklichkeit.
        Gestaltungsperspektiven der Techniksoziologie. München:Profil-Verlag, 209 - 222.
SCHRÖDER, W.H., R. SPREE (1980)(eds.), Historische Konjunkturforschung. Stuttgart:Klett Cotta.
SCHUMPETER, J.A. (1961), Konjunkturzyklen. Eine theoretische, histori-sche und statistische Analyse des kapitalistischen Prozesses, 2 Bde. Göttingen.
SCHUMPETER, J.A. (41975), Kapitalismus, Sozialismus und Demokratie. München:Francke Verlag.
SCHUMPETER, J.A. (1989), Essays. On Entrepreneurs, Innovations, Business Cycles, and the Evolution of Capitalism. New Brunswick:Transaction
SCHUSTER, P. (1984)(eds.), Stochastic Phenomena and Chaotic Behaviour in Complex Systems. Berlin:Springer.
SCITOVSKY, T. (21992), The Joyless Economy. The Psychology of Human Satisfaction. New York:Oxford University Press.
SCOTT, A. (1995), Stairway to the Mind. The Controversial New Science of Consciousness. New York:Copernicus.
SCOTT-MORGAN, P., A.D. LITTLE (31995), Die heimlichen Spielregeln. Die Macht der ungeschriebenen Gesetze im Unternehmen. Frankfurt:Campus.
SERPELL, J. (1986), In the Company of Animals. A Study of Human-Animal Relationships. Oxford:Basil Blackwell.
SEARLE, J.S. (21984), Intentionality. An Essay in the Philosophy of Mind. Cambridge:Cambridge University Press.
SEGAL, L. (1988), Das 18. Kamel oder Die Welt als Erfindung. Zum Konstruktivismus Heinz von Foersters. München:Piper.
SENGE, P. (1990), The Fifth Discipline: Mastering the Five Practices of the Learning Organisation. New York: Doubleday.
SENGHAAS, D. (1982), Von Europa lernen. Entwicklungsgeschichtliche Betrachtungen. Frankfurt:Suhrkamp.
SHAPIN, S., S. SCHAFFER (1985), Leviathan and the Air-Pump. Hobbes, Boyle and the Experimental Life. Princeton:Princeton University Press.
SHARP, M., K. PAVITT (1993), “Technology Policy in the 1990s: Old Trends and New Realities”, in: Journal of Common Market Studies 2, 129-151.
SHNEIDERMAN, B. (1993), Sparks of Innovation in Human - Computer Interaction. Norwood:Ablex.
SIEDER, R., H. STEINERT, E. TALOS (1995)(eds.), Österreich 1945 - 1995. Gesellschaft - Politik - Kultur. Wien:Verlag für GHesellschaftskritik.
SIGMUND, K. (1995), Games of Life. Explorations in Ecology, Evolution and Behaviour. Harmondsworth:Penguin.
SILVERS, R.B. (1996)(ed.), Verborgene Geschichten der Wissenschaft. Berlin:Berlin-Verlag.
SIMON, H.A. (1977), Models of Discovery and Other Topics in the Methods of Science. Dordrecht:Reidel.
SIMON, H. A. (1993), Homo rationalis. Die Vernunft im menschlichen Leben. Frankfurt:Campus Verlag.
SIMON, W. (1996), Die neue Qualität der Qualität. Grundlagen für den TQM- und KAIZEN-Erfolg. Offenbach:Gabal-Verlag.
SIMONIS, G. (1989), “Technikinnovation im ökonomischen Konkurrenzsystem”, in: U.v. ALEMANN, H. SCHATZ, G. SIMONIS (1989)(eds.), Gesellschaft
        - Technik - Politik. Perspektiven der Technikgesellschaft. Opladen:Leske+Budrich, 37-73.
SINGH, J. (1972), Great Ideas of Opwerations Research. New York:Dover Publications.
SJÖSTRAND, S.E. (1995), “Towards a Theory of Institutional Change” in: J. GROENEWEGEN, C. PITELIS, S.E. SJÖSTRAND (1995)(eds.), On Economic
        Institutions. Theory and Applications. Aldershot:Edward Elgar, 19 - 44.
SKELTON, P. (1993)(ed.), Evolution. A Biological and Palaeontological Approach. Wokingham:Addison Wesley Publishing Company.
SMITH, K., E. DIETRICHS, S.O. NÅS (1995), The Norwegian National Innovation System. A Pilot Study of Knowledge Creation, Distribution and Use.
        Paper for the TIP-Jahreskonferenz. Oslo:STEP Group.
SMITH, N. (1984), Uneven Development. Nature, Capital and the Production of Space. Oxford.
SNEED, J.D. (1984), „Reduction, Interpretation and Invariance“, in: W. BALZER et al. (eds.), op. cit., 95 - 129.
SOBER, E. (31986)(ed.), Conceptual Issues in Evolutionary Biology. An Anthology. Cambridge:The MIT Press.
SOEFFNER, H.G. (1979)(ed.), Interpretative Verfahren in den Sozial- und Textwissenschaften. Stuttgart:Klett-Cotta.
SOETE, L, A. ARUNDEL. (1993)(eds.), An Integrated Approach to European Innovation and Technology Diffusion Policy. Maastricht:European
SOLA POOL, I.d. (1990), Technologies without Boundaries. On Telecommunications in a Global Age. Cambridge:Harvard University Press.
SOLLA PRICE, D.J. (1974), Little Science, Big Science. Von der Studierstube zur Großforschung. Frankfurt:Suhrkamp.
SOMMERHOFF, G. (1978), „The Abstract Characteristics of Living Systems“, in: F.E. EMERY (ed.), op. cit., 147 - 202.
SOROS, G. (1994), The Alchemy of Finance. Reading the Mind of the Market. New York:John Wiley&Sons.
SPEKTRUM DER WISSENSCHAFT (1995)(ed.), Schlüsseltechnologien. Special Issue, Vol. 4.
SPELLERBERG, A. (1994), Lebensstile in West- und Ostdeutschland. Verteilung und Differenzierung nach sozialstrukturellen Merkmalen. Berlin:WZB
SPELLERBERG, A. (1995), Lebensstile und Lebensqualität - West- und Ostdeutschland im Vergleich. PhD. Thesis. Berlin:FU Berlin.
SPIEGEL,I., D.d.SOLLA PRICE (1977)(eds.), Science, Technology and Society. A Cross-Disciplinary Perspective. London-Beverly Hills.
SPIES, M. (1993), Unischeres Wissen. Wahrscheinlichkeit, Fuzzy-Logik, neuronale Netze und menschliches Denken. Heidelberg:Spektrum.
SPIETHOFF, A. (1955), Die wirtschaftlichen Wechsellagen. Aufschwung, Krise, Stockung. Tübingen:Mohr.
SPREE, R. (1978), Wachstumstrends und Konjunkturzyklen in der deutschen Wirtschaft von 1820 bis 1913. Göttingen:Vandenhoeck und Ruprecht.
SPRÜNGLI, R.K. (1981), Evolution und Management. Ansätze zu einer evolutionistischen Betrachtung sozialer Systeme. Bern-Stuttgart.
SPYBEY, T. (1996), Globalization and World Society. Cambridge:Polity Press.
STACEY, R.D. (1991), The Chaos Frontier. Creative Strategic Control for Business. Oxford.

STADLER, F. (1988)(ed.), Kontinuität und Bruch 1938 - 1945 - 1955. Beiträge zur österreichischen Kultur- und Wissenschaftsgeschichte. Wien:Verlag
        Jugend und Volk.
STALK, G. Jr., T.M. HOUT (1990), Competing against Time. How Time Based Competition is Reshaping Global Markets. New York:The Free Press
STARBATTY, J., U. VETTERLEIN (1990), Die Technologiepolitik der Europäischen Gemeinschaft. Entstehung, Praxis und ordnungspolitische
        Konformität. Baden-Baden:Nomos.
STEGMÜLLER, W. (1986), Kripkes Deutung der Spätphilosophie Wittgensteins. Kommentarversuch über einen versuchten Kommentar. Stuttgart: Alfred
        Kröner Verlag.
STEHR, N. (1994), Arbeit, Eigentum und Wissen. Zur Theorie von Wissensgesellschaften. Frankfurt am Main:Suhrkamp.
STEIN, D.L. (1989)(ed.), Lectures in the Sciences of Complexity. The Proceedings of the 1988 Complex Systems Summer School. Redwood City:Addison
STEINFIELD, C., J. BAUER, L. CABY (1994)(eds.), Telecommunications in Transition. Policies, Services and Technologies in the European Community.
        London:Sage Publications.
STEINHÖFLER, K. H. (1992), “Zur Diskussion Technologiepolitik und Wettbewerbsfähigkeit”, in: Wirtschaftspolitische Blätter 4, 483-497
STEINHÖFLER, K. H. (1995), “Technologiepolitische Leitideeen für einen Kleinstaat wie Österreich”, in: R. MARTINSEN, G. SIMONIS (eds.),
        Paradigmenwechsel in der Technologiepolitik? Opladen:Leske + Budrich, 73-95.
STERNBERG, R.J., P.A. FRENSCH (1991)(eds.), Complex Problem Solving: Principles and Mechanisms. Hillsdale:Lawrence Erlbaum Associates.
STERNBERG, R.J., C.A. BERG (1992)(eds.), Intellectual Development. Cambridge:Cambridge University Press.
STERNBERG, R.J., R. K. WAGNER (1994)(eds.), Mind in Context. Interactionist Perspectives on Human Intelligence. Cambridge:Cambridge University
STICHWEH, R. (1991), Der frühmoderne Staat und die europäische Universität. Zur Interaktion von Politik und Erziehungssystem im Prozeß ihrer
        Ausdifferenzierung. Frankfurt:Suhrkamp.
STONEMAN, P. (1983), The Economic Analysis of Technological Change. Oxford University Press.
STONIER, T. (1990), Information and the Internal Structure of the Universe. An Exploration into Information Physics. London:Springer.
STRASSER, H., J.H. GOLDTHORPE (1985)(eds.) Die Analyse sozialer Ungleichheit. Kontinuität, Erneuerung, Innovation. Opladen:Westdeutscher Verlag.
STREIT, B. (1995)(ed.), Evolution des Menschen. Heidelberg:Spektrum.
SÜSS, W., G. BECHER (eds.) (1993), Politik und Technologieentwicklung in Europa. Analysen ökonomisch-technischer und politischer Vermittlungen im
        Prozeß der europäischen Integration. Berlin:Duncker&Humblot.
SWOBODA, W.W. (1978), Disciplines and Interdisciplinarity. A Historical Perspective, in: J. KOCKELMANS (1978), 49 - 92.
SWAAN, A.d. (1994), Der sorgende Staat. Wohlfahrt, Gesundheit und Bildung in Europa und den USA der Neuzeit. Frankfurt:Campus Verlag.
SWEDBERG, R. (1993)(ed.), Explorations in Economic Sociology. New York:Russell Sage Foundation.
SWEDBERG, R. (1994), Joseph A. Schumpeter. Eine Biographie. Stuttgart:Klett-Cotta.
SZTOMPKA, P. (1991), Society in Action. The Theory of Social Becoming. Cambridge:Polity Press.
TARDIF, T.Z., R. J. Sternberg, (1988), “What Do We Know about Creativity?”, in: R.J. STERNBERG (1988)(ed.), The Nature of Creativity. Contemporary
Psychological Perspectives. Cambridge:Cambridge University Press, 429 - 440.
TATSUNO, S. (1986), The Technopolis Strategy. Japan, High Technology, and the Control of the Twenty-first Century. New York:Prentice Hall Press.
TAYLOR, F.W. (1971), The Principles of Scientific Management. New York:Basic Books.
TEICH, A. H., S.D. NELSON, C. McENANEY (eds.) (1994), AAAS Science and Technology Science Yearbook 1993. Washington:American Association for
        the Advancement of Science, Washington.
THAGARD, P. (1988), Computational Philosophy of Science. Cambridge:The MIT Press.
THAGARD, P. (1992), Conceptual Revolutions. Princeton:Princeton University Press.
THERBORN, G. (1995a), European Modernity and Beyond. The Trajectory of European Societies 1945 - 2000. London:Sage Publications.
THERBORN, G. (1995b), “Routes to/through Modernity”, in: M. FEATHERSTONE et al. (1995)(eds.), Global Modernities. London:Sage Publications, 124
        - 139.
THOM, R. (1989), Structural Stability and Morphogenesis. An Outline of a General Theory of Models. Redwood City:Addison Wesley.
THRIFT, N. (1996), Spatial Formations. London:Sage Publications.
THUROW, L.C. (1996), The Future of Capitalism. How Today´s Economic Forces Shape Tomorrow´s World. New York:William Morrow and Company.
TICHY, G. (1990), “F&E Politik: Volkswirtschaftliche Bedeutung und Umsetzungsschwierigkeiten”, in: Österreichische Zeitschrift für Politikwissens