Cognitive Science at UCLA: by HC12052114417


									    Cognitive Science at UCLA:
Opportunities and Recommendations

                A Report To:

         Brian Copenhaver, Provost
       College of Letters and Science
     University of California, Los Angeles

       By the Review Panel Members:

         Bennett Bertenthal, Ph. D.
        James L. McClelland, Ph. D.
          Paul Smolensky, Ph. D.

                May 23. 2001
Context of Review and Charge to the Review Panel

Our committee met at UCLA between May 8th and May 10th. We were charged with
assessing the current status of Cognitive Science at UCLA, offering insights into the
future directions of this discipline, and recommending strategies and structures for
elevating Cognitive Science to the level of an Organized Research Unit (ORU). During
our visit we met with Provost Brian Copenhaver, Dean Fred Eiserling, Dean Pauline Yu,
and Dean Scott Waugh, as well as with the Vice Chancellor for Research, Roberto
Peccei. We also met with representatives from several key departments related to
Cognitive Science: Computer Science, Linguistics, Philosophy, and Psychology.
Additional meetings were held with other groups that are relevant to a Cognitive Science
initiative and included faculty from the Brain Mapping Center, Computational Social
Science, the Culture and Evolution group, and the Brain Research Institute. We were
especially fortunate to meet with the past as well as current directors of the Cognitive
Science Research Program (Prof. Charles Taylor, Prof, Keith Holyoak, and Prof. Phil
Kellman). All of these meetings were highly informative, and provided us with a good
deal of insight into past accomplishments as well as future opportunities for a Cognitive
Science initiative at UCLA. Although we learned about some of the disappointments
that have accompanied the quest for establishing a more stable and more comprehensive
program, we were duly impressed by the energy and enthusiasm of the community as a
whole. Without exception, the comments we received were constructive and collegial,
and suggested that the faculty at UCLA are prepared to commit time and resources
toward an interdisciplinary program that they perceive as benefiting multiple departments
and research groups. The substance of our report is based on these interviews as well as
on supplementary written material about current departments and programs at UCLA and
strategic long-range plans written by the Deans and Provost of the College of Letters and
         Before turning to the substance of our report, we would like to express our
appreciation to our hosts for arranging a cordial and very informative visit. We found
everyone extremely cooperative and helpful, and we are especially grateful to Associate
Provost David Wilson for organizing our visit and demonstrating remarkable flexibility
in attending to our last minute requests for changes in the schedule. If this visit was
successful, then everyone involved deserves credit.

The Bottom Line

        UCLA has an outstanding opportunity to capitalize on its existing strengths by
creating a Cognitive Science Institute or Center combining features of an Organized
Research Unit and Interdisciplinary Degree Program. The Institute could foster the
synergistic exploitation of formal (including computational) and empirical approaches to
the study of cognitive functions, with key strengths in vision, language, and reasoning.
Such an Institute could become a major jewel in the crown of the University, if sufficient
resources and commitments are made to ensure its success. It is important for the
University to act quickly and decisively to capitalize on this outstanding opportunity.

        In what follows we begin with a fuller discussion of the background, including a
characterization of the challenges and opportunities for cognitive science, the value added
by a cognitive science program, and the potentials and imperatives for UCLA. We then
follow this by offering our view of the essential characteristics of cognitive science as a
discipline, an expression of the particular scientific framing that we think could serve
UCLA well in its efforts to differentiate its Cognitive Science effort from that of other
universities, and a discussion of the relationship between Cognitive Science and other
disciplines. The final section of the report provides details of our recommendations on
the specific structures, strategies, and resources that should be followed to ensure success
of the effort to create an outstanding Cognitive Science Institute at UCLA.


Challenges and Opportunities for Cognitive Science in the 21st Century

        We live in a world that is being transformed by advances in high-speed
computing, communications, and information technologies. These technologies connect
people to people, people to powerful databases and instruments, and people to institutions
around the world. This connectivity is enabling more people to become knowledge
workers, and it is fundamentally changing how we live as well as work. We are moving
rapidly toward a “knowledge society,” in which the creation, communication and
interpretation of information is one of biggest challenges facing individuals and
organizations alike.
        With each passing month, information is generated at ever increasing rates as new
technologies for delivering multimedia text and data via the World Wide Web continue to
proliferate. Consider, for example, how dependent we’ve become on web browsers and
search engines in such a relatively short period of time. Ironically, we often find
ourselves overwhelmed by the very sources of technology that were designed to make
our lives less stressful and more efficient.
        In order to better exploit this expansive growth of information, it is necessary to
develop techniques to rapidly and efficiently transform data into knowledge. The
challenge is to develop more powerful methods for the transmission, representation,
storage, and manipulation of information by human and artificial agents. Cognitive
Science offers one of the best opportunities for addressing these challenges through the
development of formal models that are designed to help us understand learning and the
acquisition of knowledge. Typically, the symbolic mental representations used by
cognitive scientists (e.g., connectionist models, formal rules of logic and grammar)
overlap, but are nevertheless different from those methods and models used by
neuroscientists and cognitive neuroscientists. The principal goal of this discipline is to
develop and test formal models for understanding the ways in which humans and
machines think and learn about all sorts of knowledge, ranging from common everyday
knowledge to the more abstract forms of knowledge that are necessary for scientific
reasoning. Cognitive scientists integrate ideas from many disciplines, such as nonlinear
dynamics, statistical learning, cognitive psychology, linguistics, and neural networks to
develop and test theories of mental representations and cognitive processes underlying
knowledge and learning in perceptual tasks, such as visual and auditory pattern
recognition, as well as in high-level tasks, such as language, mathematics, concept-
formation, reasoning and problem solving. This broad-based scientific inquiry has begun
to revolutionize our understanding of learning and intelligence, and the knowledge gained
is especially relevant to industrial and educational applications.

Value Added by a Cognitive Science Program

        Thus far, the interdisciplinary training of students has been the most tangible
outcome of Cognitive Science programs, but there is good reason to expect that
partnerships with private industry and education are likely to increase in the years to
come. We are already beginning to witness the creation of valued partnerships between
private industry and Cognitive Science Centers, such as the Institute for Research in
Cognitive Science at the University of Pennsylvania. Likewise, there are innovative
partnerships emerging between centers focusing on instructional technologies and
educational activities, such as those established at Vanderbilt University and the
University of California at Berkeley.
        Industrial applications. The applications of Cognitive Science research into how
machines, animals, and humans think and solve problems is wide ranging, and include
automated decision making, robotics, intelligent sensing devices, and smart instruments.
Even today, computers must still be told what to do on a step-by-step basis. In the future,
computers will be developed that are able to make adaptive decisions that are not entirely
controlled by a program. The development of these smart machines would significantly
improve productivity in science and engineering, manufacturing, and other areas. For
example, the new Center for Automated Learning and Discovery at Carnegie Mellon
University is developing new computer methods that use historical data to improve future
decisions, such as analyzing past medical records to identify future high-risk patients, or
analyzing past customer purchase records to predict future purchase behavior. This type
of problem is known by various names such as data mining, machine learning, advanced
data analysis and knowledge discovery from databases, but regardless of the name the
discipline addressing this problem is Cognitive Science.
        Additional examples of the relevance of Cognitive Science research are abundant.
The voluminous amounts of text and natural language that are digitally stored in
databases require efficient methods for processing meaning and abstracting content.
Cognitive scientists are developing highly efficient algorithms for accomplishing these
functions and these developments are leading to very profitable partnerships with private
industry. As one last example, consider the need for more advanced search and retrieval
mechanisms to take advantage of the rapidly expanding digitial libraries, networked
resources, multimedia archives, and electronic communication. Ideally, humans should
be able to call upon personalized search devices to help them learn. These devices would
take account of existing knowledge, search goals, and modes of learning. They would
help learners explore complex, unstructured, rapidly expanding, and multimedia
resources. Such devices would keep track of the search process, suggest next steps, and
develop expectations about the searcher. To create these tools, both advances in machine
learning and advances in human learning will be necessary. Cognitive scientists with
expertise in memory processes, human knowledge representation, knowledge
development, and spatial reasoning will play a crucial role in these developments.
         Education. The traditional concept of education as a process that takes place over
a set period of years in a highly structured group setting is rapidly giving way to a far
more multivariate view of education as a process of lifelong learning that is pursued in a
variety of ways and places to satisfy a wide range of individual needs. The education and
training demands of a technologically literate general population are enormous and will
continue to grow, especially as private industry requires new and additional training for
its workforce. Unions and companies are pouring more money into worker retraining
than ever before. Indeed, it is projected that the typical worker of the 21st century will
change jobs seven times during his or her career. Guidelines and models for retraining
and re-education are needed, as are new forms of technology and intelligent systems that
will play an increasingly important role in meeting these education needs. In recent
years, cognitive scientists have made considerable progress in developing automated
tutoring systems for the training and retraining of the workforce. The Department of
Defense and private industry are investing considerable resources in the development of
these systems. Automated tutoring systems that teach school children reading and math
are also receiving a good deal of support by the federal government. Researchers using
these systems are reporting improvements in performance that exceed gains produced
with traditional forms of teaching, especially among children from low-income families.
         It should also be mentioned that the pedagogy currently used in K-12 education is
a holdover from the Industrial Age model where teachers teach and students listen and
remember, and where the needed skills are definable and students move along more-or-
less in unison. Today these skills are less appropriate and have little connection with
problems and contexts outside the school proper. New theories and models are needed
to redirect and refocus our education system. Cognitive scientists possess the knowledge
and skills to contribute significantly to this initiative, and some of their projects have
been funded generously through the newly established Interagency Education Research

Potentials and Imperatives

        UCLA is geographically and intellectually positioned to become a national center
in Cognitive Science. The intellectual infrastructure for such a center is already quite
strong, and with a few strategic appointments in the College and the School of
Engineering, it is highly likely that a critical mass of talent will have been assembled to
establish an extremely visible and productive Institute. It is our opinion that a
considerable investment in resources and space is still necessary to make this center a
reality. In the remainder of this report, we will try to flesh out in more detail the nature
of these investments.
        In considering these investments, we urge the administration to remain cognizant
of the erratic history of Cognitive Science at UCLA. A Cognitive Science group began
with informal weekly meetings in 1985-1986. In 1989 the Cognitive Science Research
Program was formally recognized as an Organized Research Program at UCLA. By
1991, this program had obtained $4M in extramural funds and was responsible for
establishing COGNET, a fiber optic ring network of over 40 advanced Apollo computer
workstation. At that time a Development Plan was submitted by this Program to
establish a more permanent facility and to improve infrastructure support.
Unfortunately, this plan was not approved and progress toward establishing an
independent Center or Institute has been discouragingly slow ever since. This lethargic
response to the dynamic efforts of the Program and especially its directors has led some
of the most distinguished faculty members involved with this program to withdraw or
worse yet to move to other institutions with well established Cognitive Science Centers.
        The recent history of Cognitive Science at UCLA poses a serious impediment
toward achieving the goal of developing a world-class Institute. Such an Institute
requires the appointment of an internationally recognized scholar and the appointment of
additional faculty with expertise in relevant areas. Currently, UCLA is experiencing
problems in the recruitment and retention of relevant faculty. If the University is truly
committed to the development of a Cognitive Science Center or Institute, then it is
imperative for the administration to move rapidly toward the creation of such an Institute
in order to harness the momentum created by recent efforts. It is in the spirit of
encouraging the University to act swiftly that we have acted quickly to provide this

                             Cognitive Science at UCLA

What is cognitive science?

        Cognitive Science, the scientific study of the mind, is distinct from and
complementary to neuroscience, the study of the brain. As with any biological organ, a
key question concerning the brain is how it achieves its functions. For other organs, say
the heart, the function is relatively clear, and can be characterized precisely via other
sciences such as physics. Not so the brain. The function of the brain is to enable
cognition, including perception, memory, thought, reasoning, and language. We cannot
turn to physics or any other traditional branch of science for a precise characterization of
these functions. A new science is needed: cognitive science.
        Mental functions like thought and language might seem inherently vague, and not
subject to scientific study. Indeed, they are inherently abstract --- but not inherently
vague, given a precise language for stating theories of abstract functions. Such a language
developed in the mid-20th century: computer science. Cognitive science, the modern
science of mental function, was born at this time, founded on the principle that cognition
is computation --- that even abstract functions like reasoning and language can be
characterized with formal precision using the language of computation, and that with
such precision, exact theoretical predictions can be made and put to the empirical test.
        On the theoretical side, then, cognitive science is the construction of precise,
computational theories of mental function (both human and artificial). These theories
may be formally articulated through computer simulation, or through abstract formalisms
such as logic and formal grammar. Cognitive theories frequently also employ more
traditional mathematical and statistical analysis.
        Empirical evaluation of theories in cognitive science employs data of many types.
Controlled experiments in the psychological laboratory yield precise data concerning
human behavior. Corpora of naturally-occurring spoken and written language, and
examples of language elicited by linguists working in the field around the globe, provide
an enormous empirical testing ground for formal theories of language. And studies
employing individuals with disorders of cognitive function and/or recent neuroimaging
techniques produce important data which become relevant to cognitive science given
additional hypotheses concerning the way cognitive theories are realized in the brain.
        How theories of cognition are realized in the brain is the central question of
cognitive neuroscience. This emerging discipline is anchored on one end by
neurobiology, which provides detailed knowledge of the physical brain; it is anchored on
the other end by cognitive science, which provides detailed knowledge of mental
function, and of the precise computation required to carry out these functions.
Neurobiology, cognitive neuroscience, and cognitive science all have crucial --- and quite
distinct --- roles to play. Such a multi-level view is at odds with the reductionist urge to
believe that we need only study neurons, since ultimately all mental functions are
subserved by neurons. But this is just as mistaken as to believe that we only need to study
quarks, since ultimately all neural functions are subserved by quarks. Strong science is
needed not only at the level of subatomic particles, but also at the higher levels of
molecular and cellular neuroscience and at the still higher level of cognitive function.
Understanding of cognitive function is provided by cognitive science, and only by
cognitive science. Only cognitive science provides the knowledge needed to design better
educational methods, better human-computer interfaces, and better systems for automatic
understanding of spoken and written language.

Defining Cognitive Science at UCLA

         A striking and distinctive attribute of UCLA is the university’s excellence in both
the formal and the empirical approaches to understanding cognitive functions. We believe
this unusual strength and balance affords the basis for defining Cognitive Science at
UCLA as the synergistic exploitation of both formal and empirical approaches. These
two approaches are the essential engines of scientific progress, in all branches of
scientific inquiry. They are the vehicles that turn informal observation and intuition into
new knowledge. Formal approaches give to cognitive science precisely what Euclid gave
to geometry. For example, an interesting idea about the representation and processing of
knowledge can only be assessed for its sufficiency by casting it in a formal framework
that allows the encoding of specific content and the derivation of ensuing implications.
Correspondingly, the observation of empirical facts, particularly ones predicted or not
predicted by a proposed formalization, provides the essential proving ground that allows
the discrimination between what cognition might be like and how it actually is. When
these two approaches are combined, as they are in the best work in the field of cognitive
science, they afford the essential ingredients necessary to make the study of cognition a
         Most major universities have cognitive science units of some sort, and most of
these would agree that synergy of theoretical and empirical methods is key. But the
degree of excellence already in place at UCLA, combined with the strong motivation
within these areas of excellence to build greater synergy between them, is unique to
UCLA. Furthermore, UCLA’s strength in highly formal areas of cognitive theory is a true
mark of distinction among universities seeking to build in cognitive science.
         It is for these reasons that we propose the synergistic combination of formal and
empirical approaches to be the fundamental defining characteristic of Cognitive Science
at UCLA. The potential for synergy characterizes UCLA’s existing strength in three
important subdomains of cognitive science: vision, language, and reasoning. Between
them, these subdomains stake out what might be called the corners of the tent of
cognitive science. Vision begins at the interface between physics and biology --- where
the photon meets the neuron --- and extends upward into perception, spatial cognition,
imagery, and representation for action. Reasoning occupies the highest ground, but one
where we now understand that explicit thought and deduction coexist with implicit,
intuitive processes that are equally important for insight and discovery. Language begins
at the interface between the physical and the mental --- where motor programs driving
oral muscles generate pressure waves and where these waves generate perceptions of
speech sounds --- and extends upward to the patterning of mental sounds and words that
constitute grammar, to the expression of thought, and to the interface between
individuals, affording a basis for the sharing and dissemination of knowledge and for the
construction of culture and society. Building further synergy in these three areas --- as
well as others that are ripe for fuller development, such as memory, learning and
development --- will provide an exciting target for UCLA’s Cognitive Science initiative.

Existing Cognitive Science at UCLA

         The four disciplines traditionally seen as providing the supporting pillars of
cognitive science are Computer Science, Linguistics, Philosophy, and Cognitive
Psychology. UCLA has notable strength in all these areas. The core of UCLA’s strength
lies in Linguistics and in Cognitive Psychology. Both are world-class departments, and
furthermore, the strengths of each connect solidly to cognitive science. Both departments
have unfortunately lost key senior faculty in recent years. But we are confident that the
existing strength of these departments, if combined with a major initiative in Cognitive
Science, will make it possible to attract other first-rate scientists, and to increase the
synergy that both departments see as a high priority. Philosophy too is a highly ranked
department; like the rest of this department, the faculty relevant to cognitive science are
few but strong. Computer Science, the weakest of these units, nonetheless has a number
of excellent faculty in both vision and reasoning, and (together with Electrical
Engineering’s speech lab), a base for building strength in computational linguistics.
         There are several universities with strong programs in some but not all of these
contributing disciplines, and there may even be one or two where all of these disciplines
are well represented. But in several of the universities where the contributing disciplines
are strong, the spirit of collaboration needed to bring them together in a truly integrated
fashion is lacking. The crucial element favoring UCLA is its cooperative culture, and the
deep (and long-standing) desire of faculty to bring the campus together to build an
integrated cognitive science community. (For comparisons of UCLA with other cognitive
science units in the country, see Appendix A.)
        Outside the core disciplines, UCLA has strength in a number of allied disciplines.
Strong formal research on vision is underway or planned in both Mathematics and
Statistics. The Computation in the Social Sciences group spanning anthropology,
sociology and political science provides a unique social-level counterpart of the
computational modeling at the individual level conducted in cognitive science, and an
intriguing connection to research on the evolution of mind, culture, and language. Finally,
a range of strengths within neurobiology, both in Letters and Science and in Medicine,
provide a solid interface between cognitive science and the allied disciplines within
neuroscience. A resource of particularly important potential value to cognitive science at
UCLA is the Brain Mapping Center.

Locating Cognitive Science within UCLA

        The departments housing the core sub-disciplines of cognitive science span the
Division of Humanities (Linguistics, Philosophy), the Division of Life Sciences
(Psychology), and the College of Engineering (Computer Science). The allied units
involve in addition the Division of Physical Sciences (Statistics, Mathematics), the
Division of Social Sciences (Anthropology, Sociology, Political Science), and the
College of Medicine (Neurology et al.). Cognitive Science thus provides a tremendous
opportunity for campus-wide synergy among a truly impressive constellation of units. For
the same reason, it also clearly poses a significant institutional challenge: finding a way
of ensuring that a Cognitive Science unit has sufficient identity, strength, and resources in
and of itself to ensure that its success does not depend crucially on the prolonged,
consistent donation of resources by multiple Departments, Divisions and even Colleges.
This institutional challenge has rarely been successfully met by universities seeking to
develop cognitive science, but we are extremely optimistic that UCLA will succeed
because of the remarkable culture of cooperation that we have consistently perceived.
        A related challenge for Cognitive Science at UCLA will be to maintain sufficient
focus amidst this broad array of related strengths. The focus must be to fully engage the
existing strengths within Linguistics and Cognitive Psychology, to increase their synergy,
and to significantly expand the strength in computation, either within Computer Science,
within other Departments, or within a Cognitive Science unit itself. If possible, bridging
to Philosophy would also be desirable. Increasing synergy between linguistics, cognitive
psychology, computation, and possibly philosophy will require hiring new bridging
faculty, in areas including, but not limited to, psycholinguistics, speech, computational
linguistics, and several areas of computational cognitive modeling. Increasing synergy
between the core departments will also require constructing new infrastructure to serve
all Cognitive Science faculty, regardless of their departmental affiliation (e.g.,
experimental and computer laboratories and personnel required to support them) . Other
important resources for increasing synergy at UCLA are discussed below in the context
of our recommended course of action.
        Relation to neuroscience. We have stressed that Cognitive Science represents a
distinct and important level of analysis, and it might be said that it exists as a science
placed at the crucial interface between the neural and biological sciences on the one hand
and the social sciences on the other. It will be very important for the success of
Cognitive Science at UCLA to adopt exactly the right stance with respect to the linkages
with research at these other levels. It is particularly important to be clear about the
relation between Cognitive Science and Neuroscience, because of the growing strength
and importance of Neuroscience in general and Cognitive Neuroscience specifically at
UCLA and throughout the worldwide scientific community. In our view, it is crucial to
consider Cognitive Science and Neuroscience as distinct and complementary disciplines,
ones that use neuroscience research methods for complementary purposes. It is likely
that researchers who study individuals with cognitive impairments and/or who rely on
functional imaging or other non-invasive brain monitoring methods will be important
participants in Cognitive Science. At the same time it is crucial to understand that the
effort to use these methods to address questions at the cognitive level is not the same as
the effort to address the biological underpinnings of cognition per se. The most
appropriate candidates for faculty positions in Cognitive Science will be those whose
work clearly addresses the functional questions using neuroscience research methods.
Faculty with such an orientation would be good candidates for appointments in Cognitive
Science. Such appointments might be joint appointments with psychology or linguistics,
and/or between Cognitive Science and units such as the Brain Mapping Center, the Brain
Research Institute, the department of Neurology, and the Neuropsychiatric Institute.

                       Structures, Strategies, and Resources

Formative Considerations

        Four considerations have played a key role in shaping our thinking about the
structures, strategies, and resources necessary to create a world class initiative in
Cognitive Science at UCLA. First, the scale of the commitment must be sufficient to
create a self-sustaining nucleus of scientific activity, yet not so large as to prevent a
strong overall sense of integration. Second, the deployment of these resources must be
make in a way that maintains and strengthens the key allied units at the University.
Third, the form of the commitment must be sufficiently explicit and credible to allay the
doubts of those who have been around long enough to see their hopes raised with general
promises only to see them dashed by the disappearance of key players and anticipated
future funds. Fourth, the structure created must be sufficiently solid [robust?] to ensure
its viability in the face of the fact that Cognitive Science necessarily spans important
administrative boundaries within the university.

Specific Recommendations

        Institute. We advocate the creation of a scientific and educational unit that
combines characteristics of an ORU and an IDP, and has sufficient independent status to
be perceived as a distinct and viable focal point for active programs of research and
teaching. Such a unit could be called either a Center or an Institute. Here we call it the
Cognitive Science Institute, without presuming that this would necessarily be its actual,
formal name.
        Director and advisory committee. The Institute will need a Director of recognized
international stature, and it is clear to us that an international search is needed to identify
the best possible candidate. That said, we do not by any means intend to imply that the
right candidate could not emerge from UCLA's existing faculty. The Director should be
recruited and supported by an internal advisory committee of 5-8 leading scientists from
the key allied disciplines to provide both wisdom and energy as well as necessary
interfacing to the key allied units. It is anticipated that some members of this committee
might end up as central participants in the Institute, while others will fade into the
background as new faculty are hired into the Center. We feel it would be wrong to
constrain the specific area or home disciplinary affiliation of the Director, but rather urge
that the committee seek an individual who respects and is respected by both the empirical
and formal segments of the existing UCLA community.
         Faculty. We propose that 10 additional Faculty FTE's be placed under the
Director's control. It is expected that almost all of these faculty will be placed in joint
appointments with existing departments, and that for the most part they will be viewed as
serving as participants in and bridges between the Center and the home department. As
such the actual formulation and implementation of each search will require a dialog
between the Director and the Head of the relevant allied department, and the actual
search committee will necessarily include individuals from both units. Correspondingly
we expect that the evaluation of joint appointees will occur in a single integrated process
grounded within the home department and incorporating institute as well as departmental
         It is also expected that the Institute will need to make a small number of key
appointments that do not fit within the boundaries of any existing Center or Department.
We cannot presume to specify the exact institutional mechanisms that would be used for
such appointments, but we can suggest how their quality and suitability might be
evaluated. This could be done by a committee consisting of the Director and those
Faculty Members of the Institute and of the Advisory Committee who hold appropriate
rank, who would then report the result of their deliberations to the dean presiding over the
Institute. Precedent exists for such a procedure at least at some other Universities.
Over and above the core faculty who would be recruited by and for the Cognitive Science
Institute, there are several departments in the college and elsewhere in the university with
agendas that could be strengthened by the presence of a strong Cognitive Science
Institute. We believe the university would be well-served to offer some FTE for
recruitments of additional faculty in these units, with the understanding that such
recruitments would be undertaken without tapping the resources of the Institute, but with
its cooperation to help ensure the success of the recruitment effort.
         Faculty time for administration and interfacing. An interdisciplinary institute of
the sort we envision here requires real time and energy over and above that required of a
free-standing academic unit. In recognition of this, it is crucial that release time for
administration and interfacing be built into the expected activities of the Director and
other Institute Faculty. The equivalent of 200% of a faculty FTE should be specifically
earmarked for such activities. It is further understood that the teaching activities of the
faculty recruited jointly would serve the needs of the department and the institute in way
that is proportionate and where possible synergistic.
         Educational programs. Graduate and undergraduate education are both key
missions of the university and key activities that help to create a sense of shared mission
for any academic unit. It is our expectation that the Institute will sponsor an
Undergraduate Major in Cognitive Science, building on and superceding the existing
major of that name administered by the Department of Psychology. Regarding the
graduate program, a range of alternative approaches might be considered. One of the
possibilities appears to resonate well with all parties with whom we have discussed the
matter. In this approach, graduate students would be recruited jointly by the Institute and
allied Ph.D. granting programs. Students so recruited would enter both the allied Ph. D.
program and the Cognitive Science Program, and would receive their Ph. D. from the
home program with an added certificate from Cognitive Science. Such students would
function at their level in a way that parallels the role of the faculty, serving as participants
in and links between the Institute and the home department. The content of the Cognitive
Science Program would be determined by the faculty of the Institute, and would likely
involve requirements that would correspond to a year's worth of formal education. In
support of this year's worth of effort, the Center would be expected to contribute a year's
worth of funding. Such an approach has several advantages, including the maintenance
of links with allied units and the conferring of a degree in a recognized scientific
discipline. Use of this approach initially would not preclude the eventual development of
a Ph. D. in Cognitive Science, should the evolution of the Institute warrant it.
        Space as a focal point. An institute that will meet the needs of the new recruits
and of existing faculty who are anxious to participate and contribute will need its own
physical space. 16,000 net square feet (exclusive of corridors etc) is just sufficient to
allow the Institute to provide a sufficient venue for its scientific and educational
programs. This space should be based as close as possible to Psychology and Linguistics,
so as to maximize the benefit from and to these important allied units.
        Start up funding. Significant block funding to meet the needs for faculty startup
and program support must be set aside up front to ensure the success of the Center, and
the Director must be given discretionary control over the use of these funds, subject to
oversight at the Dean level. Nothing less than $7.5 Million would be sufficient to meet
programmatic and faculty startup needs for an initial five-year period, and to provide the
Director with fungible resources to pursue exciting research opportunities.
Supplementation of this figure for housing assistance will be necessary to make it
possible to succeed in faculty recruitment.
        Ongoing support. We expect that ongoing support from the university would
take the form of funding of the FTE's of the Director and the 10 Institute faculty, plus an
additional irreducible minimum of $275,000 per year for support staff and operating
expenses. It should be understood that the actual cost of supporting the Center's activities
would be far higher than this, but would be covered for the most part by grants from
extramural sources, especially NSF, NIH, DOD, and other government and private
sources. Finally we note that the university should expect a typical level of faculty turn-
over in the Institute, and so should expect to incur additional startup costs for the hiring
of replacements.

To top