16-11m terrorists

Document Sample
16-11m terrorists Powered By Docstoc
					              WORKING PAPERS EPP-LEA

                      Josep A. Rodríguez

                      José A. Rodríguez

             Grupo de Estudios de Poder y Privilegio
     Departament de Sociologia i Anàlisi de les Organitzacions
                    Universitat de Barcelona

Diciembre 2005
T h e M a r c h 1 1 tt h T e r r o r i s t N e t w o r k : In its weakness lies its strength

José A. Rodríguez**

          There is a growing recognition that not only individual but also collective action
is socially embedded. For a long time, research on social movements has highlighted
that networks play a fundamental role in structuring their strategies and activities, but
less is known so far about other collective actors, such as terrorist and insurgents
groups. The terrorist attacks of September 11th 2001 and March 11th 2004 provide a
clear illustration of the importance of network relational systems in shaping collective
action. One of the greatest difficulties for understanding the character of these new
collective actors is precisely to grasp their configuration and dynamics as networks.
Dominant theoretical and methodological approaches in Sociology (focusing on
individual actors and categorical groupings) are usually not well suited for this task.

          This paper aims at deepening our understanding of terrorist activities by taking
advantage of the opportunities that network analysis opens for studying and
visualizing the composition and dynamics of the terrorist networks responsible for the
March 11th bombing of several trains in Madrid. The tools of network analysis enable
us to highlight “relational qualities” of the main actors involved in the plot (such as
their centrality, capacity to serve as a bridge between different groups, channel
information, etc) and the broader characteristics of the network that facilitated the
commission of the action (internal cohesion, stability, flexibility). The main finding of
this paper is that loose ties are the key feature of this network. Loose connections
enable terrorist cells to maintain operative ties with a broader network from which they
can draw material support and receive ideological instigation. In addition, they present
three additional advantages over strong ties for clandestine terrorist groups: 1) they
give them stability in the face of arrests or mission failures, 2) they confer flexibility,
so that last minute adaptations can be made, 3) and security, in that they remain
largely invisible to police forces
          In addressing these issues, we use relational matrixes, covering a set of
relationships (disclosed in mass media reports) among the people incriminated so far,
and some of the frequently used statistical and visualization network programs (
Ucinet6, Netdraw, Key Player).

*   Department of Sociology and Analysis of Organizations. University of Barcelona. Avda
Diagonal 690, 08034 Barcelona (
       I am grateful for the assistance of Pau Marí Klose and the very valuable and
helpful comments provided by Charles Perrow, Wendell Bell, John Mohr, David
Ronfeldt, Ron Breiger and Joanne Vitello. Hugo García Andreu, Francisco José García
Vilchez, Ezequiel Moltó Seguí, Francisco Sempere Ruiz and Anna Ramon helped with
the data collection and cleaning.

       An earlier version of this paper was presented at the SUNBELT XXV International
Sunbelt Social Network Conference 16-21 February, 2005. Crowne Plaza Hotel,
Redondo Beach, California USA.

Research financed by a grant from the Spanish’s Ministry of Science and Technology
(SEC 2003/02353).

© JAR 2005
1. Terrorist Networks

       The September 11th 2001 and the March 11th 2004 tragic events have brought
to the spotlight a new form of organization and action: Terrorist Networks. Networks
are ideal for the new forms of terrorist acts and organization. Terrorist’s organizations
renew, transform and adapt themselves thanks to the networks which supply
information, build trust and facilitate their actions. (V. Krebs, 2001; Vos Fellman and R.
Wright, 2004). In order to study their configuration and dynamics, social researchers
interested in these phenomena are likely to benefit greatly from incorporating to their
analysis the tools of network analysis (Wasserman and Faust, 1994). Since its adoption
in social sciences, network analysis has broadened our understanding of numerous
phenomena, including organizations, economic transactions, migrations, or voting
patterns. Networks in which an individual is embedded have been shown to facilitate
individual action and success. At a different level of analysis, a large amount of
research highlights the role of social networks in furthering collective action through
social movements. One of the main findings in this literature is that resources
(information, economic support, opportunities to gather) generated within networks
unrelated to the social movement (church-based, in college, sport clubs or leisure
societies), but largely legal, can be easily used in other settings. As suggested by
James Coleman (1988: 99) in his seminal paper on social capital, clandestine activities
of political dissent in an authoritarian society (Coleman uses South Korean student
radical activists for the sake of illustration) benefit greatly from opportunities to meet
in “study circles”, groups of students who may come from the same high school or
hometown or church. These study circles are important forms of social capital in
serving as the basic organizational unit for demonstrations and other protests.

       Significantly less has been said about the underground networks that support
insurgent activities by terrorist groups. The main difference between these networks
and the ones studied in sociological research on social movements is that all their
activities need to remain largely secretive in order to survive. Unlike what happens with
social movements embedded within broader networks, counter-insurgency activity by
police and security forces (whatever the methods used: arrests of suspected members,
infiltration, co-optation, selective killings), threatens not only the success of any
particular mission, but may even put into risk the very existence of the broader
network that serves as the basis of organization of terrorist activism. Due to these
persistent threats to its capabilities, terrorist networks display distinctive features that
deserve specific attention.
       This paper identifies certain aspects of the social structure of the terrorist
networks that, while offering the means for sustaining cooperation over the period
required to prepare the terrorist action, minimizes the risks of detection. As was the
case of secret societies studied by Erikson (1981), the March 11th network also builds
upon knowledge and trust generated by prior relations. My first goal is to provide a
picture of the web of ties and relations (kinship, ongoing friendship, co-habitation,
participating in common meetings, contacts) among the individuals involved in the
terrorist plot that helped build trust among them. Beyond trust, the stability of terrorist
network benefits from ties that are less traceable by counter-terrorist services. On one
hand, the terrorist network responsible for the bombings in Madrid drew on strong ties
forged before their arrival to Spain (i.e. in training camps that the terrorists attended in
Afghanistan or Pakistan), which remained apparently inactive and hidden for long
intervals while the terrorist stayed in the country. On the other hand, terrorist networks
thrive on “reliability ties” forged among people that trust each other even when they
have never interacted before, on the basis of their participation in the same endeavors
(such as the war in Chechnya or previous terrorist attacks). These ties facilitate a
sense of belonging to a closed community, in some aspects resembling Ronfeldt’s
tribe idea (2005). Both trust and reliability limit the need of hierarchical oversight and,
therefore, the likelihood that the network could fall apart if beheaded. A central part of
this paper is devoted to showing one of the most outstanding feature of the terrorist
network: the dominance of loose ties. Actors that display low centrality when we
merely focus on direct relationships turn out to be key players in the network of
indirect ties. Indirect links are the hidden strength of the network by allowing for wide
network reach while reducing the visibility of the network to outsiders.

       The March 11th network, built upon prior relationships which provided
acquantinance and trust, supports, shelters and generates the smaller attack network
that directly perpetrated the action, also known as the March 11th
terrorist cell, and is also linked to, and a part of, Al Qaeda’s much larger and more
diffuse network.
2. The terrorist bombings of March 11

       The terrorist attack of March 11, 2004 was the first strike by Al-Qaeda on the
soil of a western democracy (and close ally of the U.S. in its “war” against terrorism)
after the tragic events of September 11, 2001. During the early morning of March 11,
amid the morning rush hour, several bombs planted in backpacks left in train wagons
exploded within few minutes of each other (between 7:39 and 7:42 am), in what
became the deadliest terrorist attack against civilians on European soil. The
consequences of the bombing (191 civilian casualties and more than 1500 people
injured) surpassed by far those of other terrorist actions that the Basque separatist
organization ETA had carried out in Spain over the past three and a half decades. Most
of the victims were low to middle-income passengers, who used those trains for daily
commuting between Madrid’s working class suburbs and downtown. Many of them
were foreign born, including some Muslim immigrants. The explosions came on three
days before general elections. This fact led several politicians (especially within the
ruling party, but also from PSOE, the main opposition party) to publicly attribute
responsibility to ETA early that same morning.

       However, only a few hours after the blast, police inquiry had already produced
some preliminary results which cast doubt on ETA’s authorship. The analysis of the
failing explosive devises placed in the train wagons revealed that both the methods of
detonation (through mobile phones) and the explosive substance found in the
backpacks (Goma 2) did not match ETA’s customary procedures. Even though the
government stubbornly stuck to the blaming of ETA for the following two days (for
political reasons related to the incoming elections), police soon shifted its main line of
investigation to Islamic groups.

       Along with the general shock created by its magnitude, the attacks sparked a
hot political debate over their authorship and the political responsibilities for them.
The first official statement issued shortly after the attacks pointed to ETA as the prime
subject. Shortly after, the opposition party (PSOE) and part of the mass media
questioned ETA’s authorship and pointed instead to the involvement of Islamic groups.
The final days before elections witnessed a major confrontation between the
conservative government maintaining ETA’s authorship, in spite of contradictory
evidence regarding explosives and modus operandi, and the opposition parties and
part of the mass media criticizing the government’s hiding of facts about the Islamic
implication. During the March 12th demonstrations against the bombings, which
gathered more than 12 million people throughout the whole country, anger against the
government paralleled the sadness for the dead and wounded. The government was
criticized for supposedly lying about the authors of the attacks and Spain’s
involvement in the Iraq war was seen as its cause. The first political consequence of
the attacks was the discredit of the conservative government and party and an
unexpected turn around in the March 14 elections. The conservative party (PP), which
won the 2000 elections by a large absolute majority, unexpectly lost the majority to
the opposition socialist party (PSOE) which during the campaign had promised a fast
withdrawal of Spanish troops from the Iraq war.

         Two days after the terrorist action, police arrested five suspects (three
Moroccans and two citizens from India), after untangling how the mobile phones used
in the terrorist action were obtained. The three Moroccans had lived in Madrid for
several years, where they ran a Telephone Calling Shop (Locutorio “Nuevo Siglo”) that
provided a convenient meeting space in which personal relations could foster. Three
weeks later, the police identified an apartment in Leganes (a working class suburb in
the south of Madrid) that harbored the perpetrators of the terrorist action on March
11.1 When the police stormed in, the terrorists blew up the premises, killing
themselves (seven people) and one of the special force officers participating in the

         Out of the seven dead terrorists, five were Moroccan, one Tunisian (#30) and
one Algerian (#50). S. B. Abdekmajid Fakhet (#30 and nicknamed “The Tunisian”) and
Jamal Ahnidar (#31 and nicknamed “The Chinese”) were considered by the police as
two of the leaders of the attacks. “The Tunesian” had arrived in Spain in 1996, worked
in real estate and was working towards a PhD in Economics. “The Chinese”, which had
arrived in Spain in 1990, ran a small store and, as the Oulad brothers (Mohamed #36
and Rachid #37) and Abddenabi Koujma #43, was involved in small drug smuggling.
Most of them belonged to the Abu Dahdah Spanish Al-Qaeda’s cell. In addition, “the
Tunisian” was also part of the Salafiya Jihadija Group (from Casablanca) and the Allah’s
Soldiers Group. The identity of #50 (first labeled as El Emir) was not known until
several months after the blast when his remains were identified as Allekema Lamari,
linked to the GIA (Algerian Islamic Group). With the death of these seven terrorists,
along with some previous arrests, police believed that the central structure behind the
attacks was dismantled.

         Two weeks after the terrorist action, nineteen suspects had been arrested.
Police work in the following months resulted in further arrests: fourteen in April and
six in May (five of them affiliated with an Al Qaeda cell). By the end of June, the police
had been able to track down most of the perpetrators and associates. More than 50
people had been arrested or interrogated, including twelve Spanish citizens that
supplied the explosives.

         Graph 1 illustrates the development of police work uncovering the terrorist
network over the period examined in this paper (March to June).
The first arrests followed the classical line of police work: linkages to material
elements used in the attacks. In this case, the first arrests were related to the
explosives and the mobile phones used to activate them and led directly to three

1   The members of the Leganes group were also involved in a failed attempt to blow up
the track of the high-speed train between Madrid and Seville on April 2, the day before
the assault took place.
Moroccans (Jamal Zougan #1, Mohamed Chaoui #3 and Mohamed Bekkali #2) and two
Indians (Suresh Kumar #5 and Vinay Kholy #4) accused of manipulating the phones
(Panel A). From then on, police focused on relations and contacts of the first arrested
(Panel B). The Telephone Calling Shop (Locutorio “Nuevo Siglo”), run by two of the first
detainees (Jamal Zougan #1, Mohamed Chaoui #3) became a space where relations
generated suspicions and led to the first link to Al Qaeda through the relation of the
these two people, also believed to be direct perpetrators of the attacks, with the Syrian
Imad Eddin Barakat #7, alias Abu Dahdah, the head of the Spanish Al Qaeda cell. Most
of those arrested or detained for interrogation during the rest of the month of March
had some previous relations with the first five or had relations through the Telephone
Calling Shop (Panel C). Following the explosive’s clues lead to Rafa Zuher # 23 , a
police confident, and to the first Spanish people responsible for providing of the
explosives (J. Emilio Suarez #21 and his brother in law Antonio Toro #35) (Panels D
and E). The first days of April were marked by the failed attempt to bomb the tracks of
the high speed train AVE Madrid-Seville and the suicide of the seven terrorists in
Leganes (Panel F). During the rest of April (Panel G) the international linkage of the
network was established as it grew with individuals and relations to Al Qaeda, other
terrorist attacks (Casablanca and September 11th ), participation in training camps
(Afghanistan and Pakistan) and/or wars (Chechnya). By the beginning of June police
were able to untangle the Spanish structure that obtained and supplied the explosives
used in the attacks (the diagonal shaped structure at the top of the network in Panel
H). By the end of June most of the March 11th terrorist network had been unveiled.
                                       GRAPH 1
                              Uncovering the Network

       A: March 13 th, 2004                             E: March 22-31

         B: March 14-17 th                                 F: April 3

           C: March 18 th                                G: April 4-30

           D: March 21                                 H: May-June 2004

In black the new actors at each time
Data and Methods

         I used press accounts in the two major Spanish daily newspapers (El País and El
Mundo) to reconstruct the terrorist network. Information gathering began on March 11
(2004) and extended over the next four months. By the end of April, extensive
information about the terrorist cell directly responsible for the terrorist attack had
already been disclosed. More evidence about the network appeared in May and June, as
new detainees provide further details about the broader network involved in the plot.
The information used in constructing relational matrices was completed and revised
during the following months using other media sources and reports of the Ministerio
del Interior2. The empirical analysis was carried out with UCINET 6 and Netdraw.

         The ties used in the analysis, and that appear in the graphs, cover several
substantive relationships that weave the terrorist network together:
       • Binding ties: kinship, friendship, personal contacts.
         •   Repeated interaction in particular settings: The Telephone Calling Shop,
             living together.
         •   Reliability ties: Al Qaida relations, training in Afghanistan and Pakistan,
             participation in the War in Chechnya, involvement in terrorist actions prior
             to March 11th (such as an action against the Spanish consulate in
             Casablanca in 2003, or the Sept 11th attacks), and membership in Al Qaeda.

         The biographic information on the 70 people involved in the network indicates
its international character. Only six are Spaniards and most of the others are Arabian
(with only the exceptions of two Indians and one person form Indonesia). Almost half
(32 out of 70) are Moroccan, ten are from Syria, and the rest come from Algeria,
Tunisia, Lebanon, Palestine or Egypt. The Moroccan origin is even more dominant in
the so-called terrorist cell (those directly involved in the attack): ten out of thirteen are
from Morocco.
         Although biographical information is scarce, we have obtained substantial
relational and affiliation information. Twenty three people had kinship ties within the
network, seven went to training camps either in Afghanistan or Pakistan, three were
involved in the war in Chechnya, fifteen had some links to Al Qaeda or Bin Laden, and
fourteen were involved in previous terrorist attacks (11S and Casablanca).

         A note of caution is warranted here. The data obtained might be incomplete,
possibly biased and perhaps politically manipulated. As noted by other researchers of
criminal and terrorist networks (Sparrow 1991, Krebs 2004), the application of network
analysis to criminal activity is likely to encounter several problems. Even after a
thorough investigation, security forces may not be able to uncover all the nodes and
links among them. Networks are also likely to have fuzzy or ever-changing
boundaries, which make it difficult to decide who to include and not to include in the
analysis. And networks are also dynamic, that is to say they are in a continuous

process of change and transformation. Perhaps this last aspect is the very key to the
network as an organizational form. It is this state of change which facilitates, at a
moment’s notice, the optimal flow of communication or the best manner to organize a
terrorist act. The network changes continually, strengthening and intensifying certain
relationships and letting others fall away.

       To these concerns, I could add those derived from the information that
government or the security forces choose strategically to disclose or keep secret for
partisan purposes. Most of the information that has been made public depends on
police and judicial investigations which are still in progress. Obviously, we must be
cautious of any possible political influence in the investigation and /or in the
surrounding publicity of the investigations. In fact, during the days that followed the
attack, every bit of information or rumor entered the political and media battle
surrounding the electoral campaign. The information was, and is part, of a large-scale
ongoing political and media battle which resulted in the unexpected change of the
governing party in the elections of March 14th (three days after the attack). It has been
said that the attacks had great political repercussions. However it is not only the
attacks themselves but also the very definition, vision and public reconstruction of the
facts that had political reverberations. However, we are confident that, by gathering
data over a longer-term period, our research study has minimized some of these

       Using the data from the press we can construct a map of a part of the network
involved in the March 11th attacks in Madrid that provides us with a certain view of its
structure, however incomplete. Even though much has been said and written about the
March 11th attack and its impact (García-Adallio, 2004; F. T. Reinares and A. Elorza,
2004), it has never before been visualized or analyzed as a network. In this paper we
intend to do precisely that: visualize, display, and uncover the network.

We will start our analysis by focusing on the smaller network directly involved in the
attack (the perpetrator network) and continue with the complete network and the
resulting network after extracting the perpetrator cell. Our analysis attempts to explain
how these networks get built and to assess the impact of trust and reliability ties
(based on affectivity, prior knowledge, prior international relations, participation in
other attacks, and association with Al Qaeda) on their dynamics. Kinship and friendship
relationships, together with the prior relationships, are clearly relevant as the builders
of a network that favors, engenders, and makes possible the emergence of the terrorist
attack of March 11th . It appears that weak relations and loose ties are responsible for
the birth of the terrorist attack network. In addition it is also evident that, as occurred
in the attacks of September 11th (V. Krebs, 2001 and 2002), sporadic contact also had
great importance in the consolidation of the network. In this particular case those
contacts took place at the “Locutorio Nuevo Siglo” (Telephone Calling Service Shop).
3 . T h e M a r c h 1 1 tt h N e t w o r k

           The lion’s share of the available information on the March 11th attack (and
probably of the police investigation as well) deals with the group of thirteen direct
perpetrators of the terrorist action, which we will call Field Operative Group (FOG). FOG
consists of people who placed the bombs on the train wagons (some of them detained
shortly after the attack) and those who died a few days later during the assault of
security forces to their premises in the outskirts of Madrid (in the district of Leganes).
With the exception of S.B. Abdelmajid Fakhet (#30) and El Emir (#50),3 the rest of the
deceased in Leganes were directly involved in the terrorist action on March 11th .

                                                 GRAPH 2
                               Field Operacional Group (FOG) Network*

*In darker green: Dead in the Leganés explosion.

           The Graph is the result of aggregating three relational matrixes based on the
alternative substantive ties that cemented the terrorist network: b i n d i n g t i e s (kinship,
friendship and acquaintance), r e p e a t e d e n c o u n t e r s (through the Telephone Calling
Service Shop “Locutorio Nuevo Siglo”), and r e l i a b i l i t y (based on links to the
international terrorist network (relation with Al Qaeda and/or O. Bin Laden),
participation in previous terrorist attacks (September 11, Casablanca - Morocco) and
the relationships based on their participation in wars or at training camps in Pakistan,

    El Emir is also known by another name in some accounts: Allekeme Lamari.
Afghanistan or Chechnya. Relational intensity among the members of FOG is indicated
by the thickness of connecting lines. Intensity ranges from 0 to 3 depending on the
extent of overlap of different kinds of ties. For example, Jamal Zougam (#1), who is the
most central actor in this network, maintains the strongest relationship possible with
Mohamed Chaoui (#3) as they are brothers, operate together the Telephone Calling
Shop and also share similar links to Al Qaeda terrorist networks. His relation with El
Emir (Allekema Lamari) is, however, weaker as they only have in common relations with
the Al Qaeda network.

        One of the most striking features of the web of relations among the members
of FOG is its apparently lack of cohesiveness. What was often branded by press
accounts as a terrorist “cell” appears as a rather sparse one, in which many members
are loosely tied to the rest or even disconnected. Four of the terrorists killed in the
explosion in Leganes are not connected by any means to the central structure.
According to the available published information, these four direct perpetrators had no
previous direct relations with the other perpetrators (with exception of the relation
between the Oulad brothers).

        The core of the comet-shaped “cell” consists of six out of thirteen members.
Unlike the rest, they are closely connected to each other. Three of them, Jamal Zougam
(#1), Mohamed Chaoui (#3) and Said Berrak (#41), stand out as the most central
figures of the network and also as having a very intense relationship through
overlapping ties forged through friendship and informal contacts, encounters at the
Calling Shop, and relationships to Al Qaeda. The other three actors, who form the tail
of the comet, hang from the relationship with Jamal Zougan (1), who plays a key role in
intermediation. Fouad el Morebit is only indirectly connected (through Mohamed
Chedali) while Allekema Lamari (#50) and Mohamed Chedadi (#6) maintain only one
direct tie.

        The first question that we want to answer is: What is the origin of the relations
between these people? That is to say… How can we explain their relationships? And
how was the network formed?

        Panel A to D of Graph 2 provide a closer picture of the web of alternative sets of
relationships that tie members of FOG together. The so-called b i n d i n g t i e s (based on
kinship, friendship and contacts), shown in Panel A, account for a very small part of the
relations and produce a sparse network. The majority of the relationships were
engendered at the Locutorio Nuevo Siglo (the Telephone Calling Shop). This network of
r e p e a t e d e n c o u n t e r s (Panel B) brings together the central actors of FOG. The
aggregation of binding ties and repeated encounters (Panel C) connects almost all the
actors in the comet-shaped cell shown in Graph 2. R e l i a b i l i t y t i e s (Panel D), based on
the active participation in the terrorist network Al Qaeda (relation with Al Qaeda and/or
O. Bin Laden, participation in previous terrorist attacks, participation in training camps
in Pakistan or Afghanistan, and participation in the war in Chechnya) form the basis of
the strong and key relationship between Zaugan, Chaoui, and Berrak, who make up the
cohesive center, but play a minor role in weaving the “cell” together and leave the
majority of actors disconnected. Only A. Lamari is incorporated into the group
through reliability ties. No doubt, the strongest connective linkages are created
through relations that involve time spent together and face-to-face interaction,
engendering what we might call here i n t e r a c t i o n - b a s e d t r u s t . However, interaction-
based trust still leaves five active terrorist operatives unconnected.

How then can we explain the expected relationships among the active terrorist
operatives (the terrorist cell)?

        One way of doing so is to think that these actors were not alone, but rather
working within a much larger network. It would be this larger network (which is a
complete network including collaborators, informers and others) that would give rise
to the relations among all the active actors. The active attacking network would
therefore be the product of this larger and more complete network.

        The loosely structured “cell” takes on a new character when each set of
relationships is mapped out against the backdrop of the larger configuration of actors
involved through the whole period. The complete network is made up of 70 people
mentioned in police reports and the press in relation to the attacks (contacts,
information sources, logistic collaborators, material and explosive suppliers) and that
obviously played an important part in the attack.

                                   GRAPH 2A FOG: Binding Ties*

* Based on kinship, friendship and contacts
                        GRAPH 2B     FOG: Repeated Encounters*

* Repeated Encounters through the Telephone Calling Shop (Locutorio “El Nuevo Siglo”)

                       GRAPH 2C     FOG: Interaction Based Trust*

* Aggregation of Binding Ties and Repeated Encounters
                             GRAPH 2D      FOG: Reliability Ties*

* Based on membership in Al Qaida networks, Al Qaida relations, training in Afganistan and
Pakistan, participation in the War in Chechenya, involvement in terrorist actions prior to March
11 th

       The Complete Network is the relational structure resulting from the
incorporation of binding ties (kinship, friendship and contacts), encounters at the
Locutorio Nuevo Siglo (the Telephone Calling Shop) and relations with the international
Al Qaeda network. The relational intensity, shown in the graphs as the thickness of the
connecting lines, is the result of adding the value 0 or 1 from each of the individual
relations. The Complete Network is a very complex structure although it is less dense
than the FOG network (see G r a p h 3 ). The proportion of direct existing relationships is
only 9% compared with 30% in the FOG network. It is a network where there is fluid
communication among almost all the actors with little effort and social cost, with the
exception of the 6 actors who are not directly connected. The relations among all the
connected members are, on the average, possible through fewer than two
                                         GRAPH 3
                                   COMPLETE NETWORK

Size of nodes proportional to degree
Line thikness according to intensity of relations
Square: FOG network members

       In Graph 3 we can point out the central core of the network, with more contacts
and a greater presence of intense relationships, around which simpler sub-structures
and less implicated actors revolve. The three central actors in the previous FOG
network (J. Zougam (#1), M. Chaoui (#3) and S. Berrak (#41)) are also part of the
central core and most cohesive substructure in this Complete Network. Together with
Abu Dahdah (7), A. Azizi (11) and S. Gaby Eid (63), who plays an important role in
intermediation and inter-connection among various parts of the network, they form
the nerve center of the network. Not only FOG members stand at the center of a more
cohesive set of actors but FOG actors previously unconnected to its central structure
now exhibit multiple indirect ties back into the network. A key player in cementing the
network is Abderrahim Zbakh (The “Chemist” #19). The Chemist links each of the
unconnected members of FOG (Abddenabi Koujma, Anuar Asri Rifaat, Rachid Oulad
and Mohamed Oulad) with the central structure, through ties based on friendship and
personal contact. The FOG action network becomes a connected structure, with all
actors linked (although indirectly). The Complete Network is the social space which
produces the FOG action network and assures relations among all of its members.
           Just as in the previous FOG network, here we also want to learn about the role
played by different types of relationships (whether they are kinship or friendship,
connections to international groups, or repeated encounters). We want to specifically
assess their contribution to the establishment of ties among the perpetrators (FOG

           Kinship ties, although important, only connect a very small part of the network
(G r a p h 3 A ). Friendship and contacts produce the largest set of connected actors ( G r a p h
3 B ), link the most central actors and also bind together all the FOG people (thanks to
the intermediation role played by “The Chemist” # 19). He is a key boundary spanner
with access to information flows generated in the cluster of active participants in the
repeated encounters at the Telephone Calling Shop (Locutorio Nuevo Siglo) and direct
contact with the “peripheral“ perpetrators of the terrorist action. He is, along with
Jamal Zougam #1, the most central actor of the interaction-based trust network.
Relationships created in the Telephone Calling Shop (Locutorio Nuevo Siglo) are once
again important because they connect one of the central parts of the Complete
Network as well as many of the perpetrators of the March 11th attack (G r a p h 3 C ).
These relationships form the very cohesive center of the interaction-based trust
network (also formed by kinship, friendship and contact) which now indeed unites
(directly and indirectly) the great majority (89%) of actors in the Complete Network
(G r a p h 3 D ) . Trust plays a key role in the formation of the network and facilitates the
action of the FOG sub-network.

           “The Chemist” is not the only actor surrounding, but not pertaining to FOG, that
plays a decisive role in the interaction-based network (G r a p h 3 D ). Nine out of the
thirteen most central actors in the network –those having more than ten direct
connections (degree) to other nodes– are not “field operatives” participating in the
terrorist actions. In addition to Abderrahim Zbakh, other actors like Nauma Oulad
Akcha (19 connections), Hamid Amidal (13 connections), Vinay Kholy (11 connections),
Suresh Kumar (11 connections) and Seeman Gaby Eid (11 connections) are central
figures in the network which allow for a wide network reach with minimal direct
involvement of FOG members. According to the publicly released information, all of
them except Semaan Gaby Eid, are directly or indirectly connected among themselves
through encounters at the Telephone Calling Shop (G r a p h 3 C ).4 This suggests that
many of the coordination tasks and activities required to prepare the terrorist actions
may have been carried out beyond the boundaries of FOG, possibly within a “second
circle” of actors surrounding FOG. This may have helped minimize the risks of
exposition of FOG members to the surveillance of security forces.

4   Seeman Gaby Eid is also an interesting figure. He mantains only one direct link to a
group which attends meetings at the Telephone Calling Shop, but plays a crucial
bridging role with the cluster of Spanish citizens who supply the explosives used in the
terrorist action.
     GRAPH 3A
Binding Ties: Kinship
                      GRAPH 3B
         Binding Ties: Friendship and contacts

                      GRAPH 3C
Repeated Encounters through the Telephone Calling Shop
                            GRAPH 3D : Interaction Based Trust*

* Aggregation of Binding Ties and Repeated Encounters

                                 GRAPH 3E : Reliability Ties*

* Based on membership in Al Qaida networks, Al Qaida relations, training in Afganistan and
Pakistan, participation in the War in Chechenya, involvement in terrorist actions prior to March
11 th
         Linkages to the international terrorist network produce the other cohesive
center of the Complete Network. Interestingly, none of the members of the second
circle belong to the cluster of actors with resilient ties to Al- Qaeda (based on relations
to Al Qaeda, training in Afghanistan or Pakistan, involvement in prior terrorist attacks
or participation in the War in Chechnya) (G r a p h 3 E ). What is striking is that only three
FOG members (Jamal Zougam, Mohamed Chaoui and Said Berrak) are embedded in this
smaller group of 22 actors 5. These three people who were key figures in the FOG
network are also central in this cluster. There are reasons to believe that the need for
security dictates that the interconnection between the Al-Qaeda international structure
and the local operatives is kept at low levels of observability. In fact, the Al-Qaeda
pedigree of boundary spanners at both clusters facilitates personal trust regardless of
the density of the common relations in which they are embedded. Reputation within
the terrorist network is likely to make direct ties unnecessary. The key figure which
brokers the relationship between the Al-Qaeda international network and the field
operatives in Madrid is Imad Eddin Barakat [alias Abu Dahdah, head of Al Qaeda’s cell
in Spain]. These ties linking the March 11th attack network to the international terrorist
network Al Qaeda point to the international character of the Madrid attack. Even
though the Complete Network is largely based on trust relations, it is apparently
activated by links with the Al Qaeda international network.

         Besides the two clear clusters rising from the interaction at the Telephone
Calling Shop and the relations to Al Qaeda, we should also point to the diamond
shaped structure, located on the superior-right of the graph, responsible for the
supply of the explosives used in the attacks. S. Gaby Eid (#63) is a key person
connecting this structure of six Spaniards to the FOG network, as well as to the
“second circle” and to the international Al Qaeda cluster.

         Overall, the network displays a highly segmentary structure, with several
centers of activity with different competencies, assignments and even leaders. At the
local level, key FOG members maintain intense relationships with members of the
“second circle”, forged through kinship, friendship and direct interaction at the
Telephone Calling Shop, as well as with the cluster of actors linked to the international
network of Al Qaeda. However, several members of FOG remain loosely connected to
the core group and can not easily be tracked down in the event that part of the
organization was compromised. Beyond that, the principle of protecting the network as
a whole is achieved by minimizing inter-cluster direct interactions, such as those
between local and Al Qaeda international operatives.

    While those directly and materially involved in the March 11th attack (FOG actors) were
mostly Moroccan (10 out 13), the most cohesive and strongest space (the hub) of the
international network is dominated (two thirds) by other Arabian nationalities
(Lebanese, Syrian, Egyptian, Jordanian, Palestinian).
4. The power of loose ties: operational benefits of sparseness

           One of the most striking features of the new terrorist networks is that
cohesiveness and stability can be achieved within a weak relational structure. Unlike
conventional armed groups which are often hierarchical and centralized, large terrorist
organizations use dispersed forms of organization, which can balance the need for
covertness with the need for broader operational support and resources (money,
explosives) (Tsvetovat and Carley: 2005). Weak and looser networks are also easier to
reconstruct since they do not depend as much on strong relationships (sometimes
redundant) requiring a higher social effort to maintain.

           As seen above, the terrorist network responsible for the terrorist attack in
Madrid, as was the case in the September 11th attack, is not an exception to this
general pattern. The configuration that emerges exhibits very few –but very
significant- instances of intense connectedness among nodes and is frequently based
on loosely interconnected relationships (with distant and in some cases disconnected

           To study the power of loose relations we use two approaches. We compare the
network of high intensity relations (relational intensity higher than 1) with the network
of non-strong ties (relational intensity of one) in order to asses the relevance of weak
ties. To better understand loose ties we analyze indirect relations (trough one
intermediary) to see the relational potential of the existing network. As we previously
did, we also analyze and compare the power of interaction-base trust and reliability
generating indirect relations.

           The March 11th network is fundamentally based on weak (non-strong) relations
(G r a p h 4 A ). They connect all the actors and account for 88% of the total number of
relations, providing the form and dynamics to the network. In this type of system, the
weak rather than the intense relations become more important by defining and
creating the network. (This provides some alternative evidence regarding the strength
of weak ties (Granoveter, 1973, 1982).)
            GRAPH 4A
 (Relational intensity equal to 1)

            GRAPH 4B
          STRONG TIES
(Relational intensty higher than 1)
        Panel B of Graph 4 shows all the ties based on more than one substantive
relationship (i.e., binding ties, repeated encounters, and reliability ties). Strong and
intense relationships (resulting from redundancy) are very uncommon, they represent
only a very small part of the network, although in this case they form the nerve center,
only linking 26% of the actors and accounting for 12% of the relations. Most redundant
ties link members of the Al-Qaeda international cluster. According to the information
released in press accounts, many members of this cluster maintain strong
interpersonal bonds that have often been forged training together or sharing common
experiences as members of insurgent groups in the Middle East or fighting alongside
the Mujahaddin in Chechnya or Bosnia. Within FOG, Jamal Zougan and Mohamed
Chaoui work together at the Telephone Calling Shop and are stepbrothers. Both, along
with Said Berrak, have “Al-Qaeda ties”.

        Even though both structures revolve around the same core actors (#1 Jamal
Zougan, #3 Mohamed Chaoui and #7 I. Eddin Barakat), they differ in the sets of central
and key people surrounding them. Amer Azizi (#11) and M. Belfatmi (#15) share the
centrality in the strong ties network, while in the weak network #24 N. Oulad, #19 A.
Zbakh and #41 Said Berrak are also important and central.

        By treating indirect relations among actors (linkages through one intermediate
actor) as another type of weak relations we uncover another outstanding feature: the
“strength of the weak network”. Indirect ties add considerable relational potential to
the network of direct ties, providing a compact under-layer of relationships, a
rearguard relational system that makes the network more cohesive while helping
ensure secrecy (G r a p h 5 ). In fact, the density of the network of indirect ties (27%) is
much higher than the network of direct ties (10%). Direct and indirect ties together
account for more than one third (37%) of all possible relations within the network. This
suggests that, despite sparseness of the terrorist network, there is a high degree of
interconnection among nodes that allows for a rapid flow of information and resources
when needed. The addition of indirect ties triples the relational capability of the
network with a very low social cost (one step) and confirms its enormous relational

        Relationships through one intermediary greatly raise the relational power of
certain actors. Thus, Jamal Zougam, Mohamed Chaoui and Said Berrak (central figures
in the normal Complete Network) double their relational capabilities. While their direct
ties provided access to 42, 39 and 32% of all the nodes, the addition of indirect ties
allow them to reach out to 78, 78 and 72% of them. In addition, Abdeluahid Berrak
(#40), Jamal Ahmidan (#31), S.B. Abdelmajid Fakhet (#30) and Basel Ghayoun (#28)
increase their number of ties until they are able to contact, either directly or indirectly,
nearly two thirds of the members of the network.

        More interestingly, the network of indirect ties reveals the “hidden centrality” of
actors that appear to be peripheral in the network of direct ties, such as Abdeluahid
Berrak, S.B. Abdelmajid Fakhet or Allekeme Lamari. Specifically, the latter has been
identified in police reports as the key operative designer of the attacks. By being the
most central actors of the indirect ties network, thus becoming potential central actors,
they are therefore key players in the articulation of the entire terrorist network.

                                            GRAPH 5
                             LOOSE NETWORK OF INDIRECT TIES
                                        (only two steps)

Size of nodes proportional to their degree (number of contacts)

        The system of indirect (weak and loose) relationships is extremely important as
it provides life to the terrorist network. Central actors are able to connect in only two
steps to most of the members of the network, pointing to the importance of the
relationships established by those with whom they are in direct contact. Actors who
most increase their networking capacity (those who are able to enlarge their network
through an intermediary) were not central actors in the normal network. In fact, those
only duplicate their relational capacity while previously non-central actors enlarge their
networking capacity much more. For instance, Allekeme Lamari (actor 50) has his
relational capacity increased tenfold, Abdeluahid Berrak (actor 40) multiplied by five,
S.B. Abdelmajid Fakhet and Basel Ghayoun (actors 30 and 28) by four, and Jamal
Ahmidan (actor 31) by three.

        Interaction-based trust relations generate, alone, 58% of the total amount of
indirect relations (G r a p h 5 A ) while reliability ties based on international links with Al
Qaeda generate only 18% of those relations (G r a p h 5 B ). Trust (interaction based) is the
most important element in the creation of the network (of both direct and indirect
relations) and links most of its members (89% in the direct network and 86% in the
indirect one). Although less inclusive, reliability is key in the articulation of a strong
cluster of relations among one third of the members of the network. The transitivity
dimension of trust is definitely more important than in reliability. As a result, trust
creates a more open and larger system of relations while reliability produces a more
limited and closed cluster.

                                          GRAPH 5 A

*Based on binding ties and repeated encounters
                                       GRAPH 5B

*Based on relations to the international Al Qaeda network
5. The Network after the March 11th Attack

         Terrorist organizations are often characterized in the literature as fluid and
dynamic. Many studies in traditional organization theory demonstrate that
decentralized non-hierarchical structures are often highly adaptative to challenges in
their environment. This general observation seems to apply to our still limited
knowledge on terrorist structures. Terrorist groups, and in particular Al-Qaeda-related
organizations, have shown a remarkable capacity to rebuild themselves as they face
attacks on their infrastructure aimed at disrupting its ability to affect harm. This makes
the longitudinal study of terrorist networks especially relevant.

         The fluidity and dynamism of terrorist networks is facilitated by the sparseness
of its relational structure. Loose structures and weak ties make terrorist networks very
invulnerable to discovery. Random arrests, even when numerous, may have little effect
on the capacity of the network to accomplish its goals if they do not disrupt key
information pathways and command and control structures.

         In this part of the paper we are first locking into what happened with the
network once the perpetrators (the FOG network) were identified and out of
commission (either detained or dead at Leganés) and, later, we are exploring the
actions that could be taken to further weaken it and prevent its reconstruction, in the
hypothetical case it has not rebuilt itself yet.

         G r a p h 6 shows the March 11th terrorist network when the perpetrators of the
attack and the terrorists who died in Leganes a few days later (FOG) have been
removed. Although the relational density is somewhat reduced and the distance among
actors increased, the network as such continues to exist. It basically consists of two
clusters, connected by a single linkage. The cluster at the top of the Graph includes
members of the “second circle” of actors surrounding FOG (but not directly involved in
the attacks) as well as the group of Spanish citizens who supplied the explosives. After
the first police raids, this cluster was severely enfeebled. Ten out of twenty nodes
within it were arrested, and it can be argued that the main communication pathways
were disrupted. Given his important bridging role, removal of actor 63 (S. Gaby Eid) is
highly consequential.

         In contrast, the cluster at the bottom of the Graph remains largely intact and
still with a high level of cohesion. It includes the members of the Al Qaeda
international network, some of which occupy the new centrality roles. The removal of
other parts of the terrorist network does not inflict significant damage on them. There
are reasons to believe that the sparse interconnections between the Al-Qaeda
structure and local operatives, and the fact that they were largely based on reputation
(instead of direct interaction) made it more difficult for security forces to track down
their activities and destabilize the cluster through the removal of individuals filling key
                                             Graph 6
                                 Network after extracting FOG

Size of nodes: proportional to degree

        The analysis and visualization of networks provides many clues to devise
successful strategies to destabilize terrorist organizations (Carley, Lee and Krackart,
2001; Taipale, 2005). The elimination of key relationships and figures from the
network can curtail its ability to move knowledge and resources in an efficient way and
disable its capacity to undertake actions (Tsvetovat and Carley, 2005; Borgatti, 2004).
As G r a p h 6 A suggests, the removal of a few nodes and links may be sufficient to make
the network fall apart. The resultant fragmentation of the network can have large
destabilizing effects, given the importance of some pathways that could be disrupted.

        Incarceration of actor 63 is certainly the most important step. By removing him,
counterterrorist forces disable the only remaining link between international and local
operatives, trough actor 40. If the terrorist network is unable to find a replacement
rapidly, it is likely that this initiative will in itself deprive the local network from crucial
resources to execute its tasks and, therefore, can disable for some time its capacity to
effect harm. More than that, the removal of actor 63 also increases greatly the average
distance among actors within the local cluster by disrupting important linkages withing
the group, like those among actor #63 and actors 21, 22 and 34.
                                                 Graph 6A
                                           Action on Network

Lines: indicating actions upon relationships

           It may still be advisable to enfeeble the two resulting clusters by targeting key
individuals in each cluster. The two most relevant figures in each cluster are actor 24
and actor 7 respectively. The removal of actor 24 leaves many local operatives
completely isolated from the rest of the terrorist organization. Elimination of actor 7
destabilizes the international command and control structures by substantially
increasing the distance among the actors in the cluster.

           A bit larger intervention could additionally target actos 11, 18, 21 and 60. All
together, with those mentioned earlier and that have been already arrested, their
removal from the network would produce the largest fragmentation6, increase in the
average distance among actors 7, thereby making communication slower and more
costly, as well as reduce its cohesion.

    Based on results obtained applying the Fragmentation Criteria in Key Player.
    Based on results obtained applying the Distance-weighted Fragmentation Criteria in Key Player.
       The combination of these initiatives results in the disconnected network of
small substructures and isolated actors presented in G r a p h 6 B where its ability to
undertake action has been substantially curtailed. Centrality within the network makes
actors and relations more visible and therefore more vulnerable to interventions
facilitating the destruction of the network depending on them. Their power becomes
its weakness.

                                         Graph 6B
                              Resulting Network after Action

After removal of 7 actors (63, 24,7, 11, 18, 21 and 60)

       September 11th and March 11th terrorist attacks have shown the growing
importance of networks as organizational forms for collective action, especially
relevant in cases of secret societies and terrorism. They also point to a new network
model not based on intense relations, tight cohesion, hierarchical structures or cells
but rather on weak relationships. As a weak network it is less visible and more difficult
to be detected as well as easier to reconstruct as it does not require strong, and costly,
relationships. Therefore, in its weakness lies its strength.

       Our analysis of the March 11 attacks discovers that the Field Operative Group
(the network of direct perpetrators) is not a cohesive cell, as was expected, but rather a
disconnected structure. This attacking network becomes structured within a much
larger and diffuse network thanks to relations through non-operative people. The
larger network is the social space which produces the attacking network and facilitates
the communication between all its members. This larger network does not act, but it
makes action possible.

       At the center of the larger complete network there are two large cohesive and
connected sub-structures (or clusters): the one that prepared and carried out the
attacks (center right) and a more cohesive cluster linked to Al-Qaeda (center left). The
first sub-structure takes shape thanks to interaction based trust and especially thanks
to the repeated encounters at the Telephone Calling Shop. All FOG members get linked
to each other and are surrounded by a “second circle” of actors who support the
preparation of the attacks. Reliability ties (based on linkages to Al Qaeda and its
operations) produce the other cohesive center. This structure links the central FOG
figures to the Al Qaeda network thanks to the key connecting role played by I.E.B (the
head of the Spanish Al Qaeda’s network).

       The goal of protecting the network as a whole and ensuring the viability of the
attack is achieved by minimizing inter-cluster direct relations through a limited
number of actors, fundamentally through the three central FOG figures. The rest of the
FOG members remain loosely connected to the core group to reduce the danger of
being tracked down.

       Trust (interaction based) is key in creating a rather large (connects 80% of the
actors) and open network that embeddes the attacking group and provides support for
its actions. Reliability (based on links to Al Qaeda and its activities) produces a smaller,
closed, and more cohesive cluster key in linking the attackers to the larger Al Qaeda

       While most of the operatives (FOG) and the surrounding people are Moroccan,
the majority of the cohesive cluster linked to Al Qaeda has other Arabian origins
(Syrian, Pakistanis, Egyptian, etc). The March 11 is, therefore, an attack organized and
carried out by mostly Moroccan terrorist guided by Al Qaeda people, who are mostly
non Moroccan.

       In this network there are two types of central actors playing key roles. One type
is represented by the three terrorists (1, 3, 7?) who are central both in the FOG network
and the Complete larger network and that at the same time are the linking border
between the interaction-based-trust and the reliability networks. They are the main
link between the attack and the international Al Qaeda network. Actors that facilitate
communication and hold together the network are also very important. Here we should
point to actor #19, who plays a key role linking isolated FOG members, and to actor
#7, who facilitates the linkage between the attackers and the Al Qaeda network.

       Our analysis also uncovers another interesting element: the power of loose and
weak ties. Even though strong relations form the nerve center of the network and
especially the link to the Al Qaeda’s network, they are very uncommon and only
account for 12% of the relations only linking one forth of the people in the entire
structure. The network as a whole rises thanks to non-intense ties, which account for
88% of the relations and facilitate both the action of the FOG group and its support

       Indirect ties add and immense relational power to the standard direct ties
network and raises the relational power of certain actors. The addition of weak ties
triples the relational capability of the network as a whole. The network of “hidden” ties
reveals the “hidden” centrality of previously peripheral actors, becoming potential key
players. A. Berrak, A. Lamari and S.B. Abdelmajid are the people who most increase
their networking capacity and share potential centrality with the central actors of the
standard network. Weak ties (indirect relationships) form a back-up network (a
potential network) with new central key figures.

       In the last part of the paper we have taken a step further devising possible
actions on the network, deploying network analysis’ indicators and tools (centrality,
key players), to weaken and de-structure it. The removal from the network of few key
central and intermediary figures and relations, that create cohesion and maintain the
network together, will break it apart and leave it unfunctional (at least temporally).
Powerful and central actors and relations holding the network functional are at the
same time its weakest points since they are more visible and therefore more prone to
be objects of action. The removal will weaken, or even destroy, the entire network as
such. In their power lies its weakness.

       Given the difficulties to obtain complete and extensive information on terrorists
and terrorist network (especially before their actions), our analysis the utility of
network analysis providing a more complete view of a network unveiling an invisible
structure. In addition, indirect and loose relations provide a view of a hidden but
potentially very powerful network.
       It seems now clear that terrorism has taken a new organizational form as its
bases for action. Counter-terrorist activity must therefore take this into account and
look for relational information (whether encounters and contacts or reliability ties) to
be able to view the networks supporting terrorist activities and to be able to act upon
them. Additionally, the unveiling of the potential network (indirect ties’ network) might
provide counter-terrorism with some advantage in a fight against a normally invisible
         ANEX 1
                                                                        I nteraction                            “Weak”
                                                   Complete                                    based on                             Network
                                  FOG Network                              Based                             (indirect ties)
                                                    NetworK                                International                          without FOG
                                                                           Trust                               Network
                                                                                           Network Links
             Size                      13                70                   70                   70             70                    57
            Density                   29,5            10,06                 6,79                  4,6            27,2                 7,96
  Average Geodesic Distance           1,6                2.7                 3.2                  1,9             1,7                 3,43
                                     3 (and                                                      4 (and
 Maximum Geodesic Distance                                6                    7                                   3                     7
                                  disconnected)                                            disconnected)
     Non reachable actors              2                  6                    9                   45              6                     8
Network Centralization (degree)       63,6             32.9                20,38                 22,12          28,64                 23,2
                                                  1 , 3 , 7 , 1 1 , 4 1 ,1 , 2 4 , 1 9 , 3 1            4 0 , 3 0 , 3 1 ,41, 7 , 1 1 , 1 5 , 1 6 , 1 8 ,
  Actors with highest Degree         1,3,41                                                 3,7,11,1,18
                                                   24,18,19 ,30,33, 3                                         28,50                   63

                                                  1 , 3 , 4 1 , 7 , 3 1 ,1 , 3 0 , 3 1 , 3 ,
 Actors with highest Closeness         1                                                        3,7,1,11     40,41,28,31,6 40,7,63,11,18
                                                         40                24,19

                                                  6 6 , 1 , 7 , 4 0 , 3 ,7 , 1 , 6 3 , 3 1 ,
Actors with highest Betweenness       1,6                                                        3,25,7      40,31,3,63           63,40,7,22,56
                                                         31                24,61

   Actors with highest Flow                       66,7,64,31,1
                                      1,6                    7,63,1,61                          25,3,8,7     3,1,40,31,23           63,40,7
         Betweenness                                  9
Actors with highest Eigenvector   1,3,31,30,41,                         1,24,19,31                                     9
                                                                                              5 0 , 6 , 3 0 , 4 0 , 8 ,7 , 1 1 , 1 8 , 1 5 ,
                                                    1,7,3,11                      7,3,1,11,18
           Centrality                  28                                  ,30                         ,2                     16,
     Actors in cliques of 5            6                 43                   22                   21             55                    30
  Actors in largest number of     1,3,31,30,41,                         21,64,65,
                                                  1,3,11,7,41                              1,18,11,3,7 50,2,6,57,61                  4,8,15
         cliques of 5                  28                                  66,67
       Maximum Degree                  10                29                   18                   18             38                    17

                                                                        1                  ,
                                                  1 , 1 1 , 3 , 1 3 , 1 4 , 3 , 4 , 5 , 1 91 , 18,3,12,13, A l l w i t h t h e
        Actors in Core               1,3,41                                                 6
                                                  , 5 8 , 7 , 1 2 , 1 8 ,2 4 , 2 8 , 3 0 , 3 1 , 7 , 1 6 , 1 1 , exception of 6
                                                      16,15              1,33,41                 14,15       disconnetced
F O G N E T W O R K : Material perpetrators + Dead at Leganes’s explosion
C O M P L E T E N E T W O R K : ADDITION OF DICHOTOMIZED RELATIONS: Kinship, friendship and contact relationships, Telephone Calling Shop,
relationships with the international terrorist network (with Al Qaida, O. Bin Laden, participation in actions in Pakistan, Afghanistan and Bosnia
as well as in previous terrorist attacks).
I N T E R A C T I O N B A S E D T R U S T : Aggregation of Binding Ties (kinship, friendship and contact relationships) and Repeated Encounters (at
theTelephone Calling Shop)
R E L I A B I L I T Y B A S E D O N I N T E R N A T I O N A L N E T W O R K L I N K S : Based on membership in Al Qaida networks, Al Qaida relations, training in Afganistan
and Pakistan, participation in the War in Chechenya, involvement in terrorist actions prior to March 11th
“ W E A K ” ( I N D I R E C T T I E S ) N E T W O R K : Network of indirect ties only (only two steps)
N E T W O R K W I T O U T F O G : Complete network without FOG actors.

Dichotomized relations
                                                      ANEX 2

                                                                                                  Trainning   Links       Links to
                                                                            Links to Previous
#                  NAME         Nationality   Born       Profession                               Camps /     to Al       Terrorist
                                                                            Terrorists Activity
                                                                                                      Wars    Qaeda       Groups
1    Jamal Zougam              Morocco        1973   Small business owner   9/11, Cassablanca     X           X       B, E, G, K, M, T
2    Mohamed Bekkali           Morocco        1972   Mechanic
3    Mohamed Chaoui            Morocco        1969   Construction worker    9/11, Cassablanca                 X       B
4    Vinay Kholy               India          1977   Small business owner
5    Suresh Kumar              India          1972   Small business owner
6    Mohamed Chedadi           Morocco        1966   Small business owner                                             B, M
     Imad Eddin Barakat (Abu
7                              Syria          1965                          9/11                              X       A, B, C, J, L, S
     Dahdah)                                                                                      X
8    Abdelaziz Benyaich        Morocco                                      Casablanca            X                   B, L, S
9    Abu Abderrahame           Morocco                                      Casablanca
10   Omar Dhegayes             Morocco                                                            X
11   Amer Azizi                Morocco        1968   Construction worker    9/11, Casablanca      X           X       A, B, D, L
12   Abu Musad Alsakaoui       Jordan         1967                          9/11                                      A, H, I
13   Mohamed Atta              Egypt          1971                          9/11                  X                   C
14   Ramzi Binalshibh          Yemen          1973                          9/11                  X           X       A, C
15   Mohamed Belfatmi          Argelia        1978                          9/11                  X                   B, C
16   Said Bahaji               Morocco               Salesman (cars)        9/11                  X                   C
17   Alí Amrous                Argelia
18   Mohamed Galeb Kalaje                            Construction worker    9/11                              X       A, B
19   Abderrahim Zbakh          Morocco        1971   Chemist                                                          M
20   Farid Oulad Ali           Morocco        1970   Construction worker                                              M
21   José Emilio Suárez        Spain               1977   Exminer
22   Khalid Ouled Akcha        Morocco             1973   Construction worker
23   Rafa Zuher                Morocco             1980
24   Naima Oulad Akcha         Morocco             1963
25   Abdelkarim el Mejjati     Morocco             1968                           Casablanca       K, M
26   Abdelhalak Bentasser      Morocco
27   Anwar Adnan Ahmad         Palestinian State                                                   B, S
28   Basel Ghayoun             Syria               1979   Construction worker
29   Faisal Alluch             Morocco             1970                                            N
30   S B Abdelmajid Fakhet     Tunisia             1968   Salesman (real state)                    B, K, S
31   Jamal Ahmidan             Morocco             1970   Small business owner                     B
32   Said Ahmidan              Morocco
33   Hamid Ahmidan             Morocco
34   Mustafa Ahmidan Hichman   Morocco             1968   Waiter
35   Antonio Toro              Spain               1977   Miner
36   Mohamed Oulad Akcha       Morocco             1975   Cleaner                 AVE              B
37   Rachid Oulad Akcha        Morocco             1970   Construction worker     AVE
38   Mamoun Darkazanli         Syria                                                               C
39   Fouad El Morabit Anghar   Morocco             1976   Construction worker
40   Abdeluahid Berrak         Morocco             1971   Small business owner                 X   N
41   Said Berrak               Morocco             1972   Messenger                            X   A, B, M
42   Waanid Altaraki Almasri   Syria
43   Abddenabi Koujma          Morocco             1975                           AVE              B
44   Otman El Gnaut            Morocco             1974   Construction worker                  X
45   Abdelilah el Fouad        Morocco             1969   Small business owner
46   Mohamad Bard Ddin Akkab   Syria
47   Abu Zubaidah                Palestinian State   1974                                       A
                                 Bosnia and
48   Sanel Sjekirika                                 1981   Student
49   Parlindumgan Siregar        Indoneyesa                                                     B, L
50   Allekeme Lamari (El Emir)   Argelia             1964                            AVE        F, J
51   Anuar Asri Rifaat           Morocco                                             AVE
52   Rachid Adli                 Morocco             1979   Butcher
53   Ghasoub Al Albrash                                     Small business owner                A, B
54   Said Chedadi                Morocco                    Small business owner                B, L
55   Mohamed Bahaiah             Syria                                                          A, B
56   Taysir Alouny               Syria                      Journalist (Al Yazira)              A, B, T
57   OM. Othman “Abu Qutada”,    Palestinian State   1961                                   X   A, B, F, J, L, Q
58   Shakur                      Morocco             1968                            9/11       B
59   Driss Chebli                Morocco             1973                                       A, B
60   Abdul Fatal
     Rabel Osman El Sayed
61                               Egypt               1971   Exmilitary                      X   D, K, M
     (Mohamed El Egipcio)
62   Nasredine Boushoa           Argelia             1965
63   Semaan Gaby Eid             Lebanon
64   Emilio Llamo                Spain                      Watchman
65   Ivan Granados               Spain               1982   Miner
66   Raul Gonzales Perez         Spain               1979   Miner
67   El Gitanillo                Spain               1988
68   Moutaz Almallah             Syria
                                                            Washing machine
69   Mohamed Almallah            Syria
70   Youself Hichman                   Syria

Terrorist Groups:

A : AQ,                                                             B : AbuDahdah’s Cell (Spain)
C : Hamburg’s Cell (11S)                                            D : Al Qaeda’s structure in Europe
E : Excombatientes Afganos                                          F : GIA Grupo Islámico Armado (Argelia)
G : Al Oussoud Al Khadidine "Leones Eternos" (marroquis afganos)    H : Ansar Al Islam (AlQaeda)
J : GSPC Grupo Salafista para la Predicación y el Combate (Túnez)   K : Salafiya Jihadija (Casablanca), L : Harakat ul-Mujahedin (Pakistán)
M : GICM o HASM Grupo Islámico Combatiente Marroquí (Marruecos)     N : Mártires para Marruecos (origen:cárcel Topas, Salamanca)
Q : Combatiente Tunecino                                            S : Soldados de Ala
T : Hermanos Musulmanes

Mariam Abou Zahab and Oliver Ray, Islamic Networks: The Pakistan-Afghan Connection
     (Eastbourne, England:Gardners Books, 2004).
Haizam Amirah Fernández, “¿Tiene Al-Qaeda una estrategia global?”, Real Instituto El Cano,
     ARI nº74, 2004.
John Arquilla, David Ronfeldt (editors), Networks and Netwars: The Future of Terror, Crime,
     and Militancy (Santa Monica, Calif.: Rand Corporation, 2001).
Steve Borgatti, KeyPlayer 1.41 (
S. Borgatti, "The Key Player Problem." In Proceedings of CASOS 2002 Conference, Pittsburgh,
     PA (2002).
S. Borgatti, “Flow Processes and the Strcutural Importance of NODES” (Boston College, Power
      Point Presentation, 2004).
Steve Borgatti, Everett, and Freeman, UCINET 6, Social Network Analysis Software
Pedro Canales and Enrique Montanchez, En el nombre de Alá: La red secreta del
     terrorismo islámico en España (Barcelona: Planeta, 2002).
Kathleen .M. Carley, Sung Lee and D. Krackart, "Destabilizing Networks", Connections,
     volume 24, issue 3 (2001).
James Coleman, “Social Capital in the Creation of Human Capital”, American Journal of
     Sociology 94, Supplement 95-120 (1988).
Bonnie Erickson, "Secret Societies and Social Structure" Social Forces, volume 60, number 1
P. Vos Fellman and R. Wright, "Modeling Terrorist Networks"
     ( (2204).
P. Vos Fellman, D. Sawyer and R. Wright, "Modeling Terrorist Networks- Complex Systems
     and First Principles of Counter-Intelligence” presentado en NATO and Central Asia:
     Enlargment, Civil-Military Relations and Security, 2003).
Casimiro García-Abadillo, 11-M, La Venganza (Madrid: La Esfera de los Libros, 2004).
Mark S. Granovetter, "The Strength of Weak Ties", American Journal of Sociology 78 (1973).
R. Goolsby, "Combating Terrorist Networks: An Evolutionary Approach." In Proceedings of
     the 8th International Command and Control Research and Technology Symposium,
     conference held at National Defense War College, Washington, DC (Vienna, VA:
     Evidence Based Research, 2003). Available: [March 17, 2005].
Mark S. Granovetter, "The Strength of Weak Ties: A Network Theory Revisited" in Peter V.
     Marsden and Nan Lin (eds.) Social Structure and Network Analysis (Beverly Hills, Sage,
R. Gunaratna, Inside Al Qaeda: Global Network of Terror (New York: Columbia University
      Press, 2002).
José M. Irujo, La red islamista en España (Libros Electrónicos El País, 2004).
Javier Jordan and Nicola Harsburgh, “Mapping Jihadist Terrorism in Spain”, Studies in Conflict
     and Terrorism 28:169-191 (2005).
Evan F. Kohlman, Al-Qaida’s Jihad in Europe: The Affhan-Bosnian Network (Oxford: Berg
      Publishers, 2004).
Vladis E. Krebs, "Mapping Networks of Terrorist Cells", Connections, volume 24, issue 3
Vladis E. Krebs, "Uncloaking Terrorist Networks", First Monday, volume 7, number 4 (april,
      2002) (
Peter Klerks, "The network paradigm applied to criminal organizations", Connections,
     volume 24, issue 3 (2001).
Fernando Reinares and Antonio Elorza, El nuevo terrorismo islamista: del 11-S al 11-M
     (Madrid: Temas de Hoy, 2004).
D. Ronfeldt and J. Arquilla, "Networks, Netwars and the Fight for the Future," First Monday 6,
     10. Available: [March 17,
David Ronfeldt, “Al Qaeda and its affiliates: A global tribe waging segmental warfare?”, First
     Monday, March 2005.
Marc Sageman, Understanding Terror Networks (Philadelphia: University of Pennsilvania,
Malcolm K. Sparrow, "The application of network analysis to criminal intelligence: An
     assessment of the prospects", Social Networks, volume 13 (1991).
Cynthia Sthol and Michael Stohl, “Networks, Terrorism, and Terrorist Networks”. Paper
     preseted at NCA Organizational Communication Preconference, New Orleans, 2002.
Kim Taipale, “Destabilizing Terrorist Networks: Disrupting and Manipulating Information and
     Information Flows” presented at the Global Flow of Information, Yale Information
     Society Project, Yale Law School, April 2005.
Maksim Tsvetovat and Kathleen M. Carley, “ Structural Knowledge and Success of Anti-
     Terrorist Activity: The Downside of Structural Equivalence”, Journal of Social Structure.
     Volume 6, 2005.
Austin T. Turk, “Sociology of Terrorism”, Annual Review of Sociology, 2005.
Stanley Wasserman and K. Faust, Social Network Analysis: Methods and Applications (New
    York: Cambridge University Press, 1994).
Duncan J. Watts, "Netwoks, Dynamics, and the Small World Phenomenon", American Journal
     of Sociology, volume 13, number 2 (1999).
Duncan J. Watts, Six Degrees: The Science of a Connected Age (New York: W.W. Norton,

Shared By: