Saturday 25 March 2017, 03:55

Ukrainian Institute
of Research of Extremism

Agent-Cultural Model of radicalization of immigrants milieus in the West


O. Kutsenko, Professor of Sociology, Head of Social Structure and Social Relations Department, Taras Shevchenko National University of Kyiv (Ukraine)

L. Berlyand, Professor of Mathematics, Department of Mathematics, Penn State University, University Park, PA (USA)

A. Gorbachyk, Doctor of Mathematics, Dean of Sociology Faculty, Taras Shevchenko National University of Kyiv (Ukraine)

The modeling of the terrorists activities emerging in the West, which is based on network analysis and psychological approach, has recently become popular as a starting point for greater theoretical and methodological recognition of the dynamic structural properties of terrorist groups [Horgan 2004; Borum 2004; Dalgaard-Nielsen 2008, 2010; Sageman 2004, 2007; Hoffman2010; Mullins andDolnik 2010 etc.]. However, applications of this approach to empirical studies and monitoring are complicated essentially due to latent state of the radical networks and groups, which become apparent only through the terrorist actions (or other political violence), and their close functioning. The fact that accessibility of the group’s (network’s) representatives in empirical studies is very limited sets a limitation over research and practical use of this theoretical and methodological approach. By contrast with this approach, we are going to attack the problem from other angle. We propose to explore the utility of the Agent-Cultural sociological model of radicalization (ACSMR) of social milieus of Muslim migrants in the West through statistical (multi-level and log-linear regression, path analysis) and mathematical (using differential equations) modeling and empirical validation of the model. The ACSMR model is developed by O. Kutsenko, V. Sherstoboev in collaboration with L. Berlyand in several research projects during 2004-20111 to monitor and predict the development over time of social milieus of highly religious immigrants towards support of terrorism and other political violence [Berlyand, Kutsenko 2005; Kutsenko 2010; Sherstoboyev 2012]. The model is applicable not only to monitoring of general situation with intolerance in a country but also to defining of social milieus, which potentially can support the politically justified violence.

The model is grounded on conceptual interpretation of radicalization developed within the research finding by M. Sageman [Sageman 2004]. This interpretation distinguishes the four phases (stages) in critical changing in individual beliefs, feelings and behaviors: from the moral outrage as an initial trigger spreading in a social milieu (the phenomenon of the ‘morality play’) up to fusing between the global and the local, and finally an individual joining a terrorist cell, which becomes a ‘surrogate family’ [Sageman 2004, 2008; see also: Horgan 2009, 2008; McCauley 2008; Silber and Bhatt 2008; Slootman and Tillie 2008; Kühle, Lene and Lindekilde 2010;Kydd 2011].  As the basic indicators (dependent variables) of radicalization of social milieus our model uses the propagation in the milieu of beliefs and attitudes at supporting and approving violent ways of upsetting the existing political and social order in a society (at local, national, regional or global levels) based on the cultural justification (political, ideological, religious, moral) of the violent ways. We conceptualize the radicalization as a non-linear networked social process of self-organization in large-scale social milieus on the bases of critical transition in beliefs and attitudes, which lead to a formationof radically-oriented social networks or even radical group organizations. Two factors, namely: a social milieu (‘nutrient medium’ in physical sense) and information\cultural transmissions (from a moral authority, via mass-media and Internet) forming a learned behaviour, – play the key role as a possible catalyst of the radicalization process.

It is well-known that the rich spectrum of networked internal processes and their interactions result in multiple complexities. For a reduction of this complexity we propose to study the compactly supported factors and the critical conditions for transition of the social milieu to the radical state.  On the background of the theoretical achievements both the so-called ‘cultural turn’ in sociological methodology [Alexander 1999; Harrison & Hantington 2001, etc.] and the agency and structural theoretical synthesis in explanatory sociological models (P.Burdieu; M.Archer; G.Therburn, etc.) we develop a sociological explanatory model of radicalization of social milieu of Islamic migrants, which can be verified on cross-cultural empirical data. One of the basic concepts in this explanatory model is ‘cultural kinetics’ [Moody and White 2003; White 2011]. The concept describes a transmission of cultural patterns and relationships as learned behaviour (e.g. of migrants) via interpersonal interactions within a large-scale community (or social network) that happen even under the strong pressure from the side of the control social institutes (e.g. of ‘accepting society’). By using this concept so it is able to study networked social processes, emergent phenomena in large networked communities, and formation and transmission of cultural patterns and social structure as learned behavior. The cultural patterns emerge out of the lives of ordinary persons, which interact and are networked within a certain social milieu, and produce these patterns in the course of their everyday life. The cultural patterns are continually updated with respect to the field of interdependencies between the environment, interpersonal relationships, bundles of behaviours/affects/cognitions, and access to resources and information through groups. These interdependencies form structures (‘shared’ culture, parts of which may be distributed differentially), which may change relatively slowly, but may also change abruptly as they reach critical transition thresholds. The study of complex systems demonstrates that it is a mistake to regard structures as fixed, immobile or recurrent in exactly repetitive form.  Rather, these structures tend to be ‘vibrant’ within ‘oscillatory’ fields of motion. Studies of large-network phenomenon show important properties of self organization, including those of redundancy and resilience, in networks of low density that are central to understanding the emergence and functioning of processes of social and political influence and social organization.

The primary dimensions of ‘cultural kinetics’ are the horizontal plane of multiple agent/environmental interactions and the vertical plane of single agent/interior process interactions. The lag-time for results of internal processes to re-emerge into an overt behaviour far exceeds the time scale at which agents affect one another or agent/environment reactions take effect. The ‘cultural kinetics’ can be resulted in producing of threshold phenomena in the networked histories of people’s lives, that produce  phase transitions or structural transformations of milieu, which may happen very rapidly.

The social milieus, characterized by the high level of support and approving of political violence, having well-developed interpersonal trust network, can potentially become the source for recruiting of violent acts performers.

Our previous study (2004-2011) reveals that the push factors (such as: economic and civic frustration, social and cultural distance and status deprivation, pinched personal dignity, conflict of personal identities, social isolationism) can form the sensitiveness of milieu to radicalization and can lead to the lost of durability and tolerance by social milieu. Thus, the pull factors perturb the milieu and catalyze the emergence of coherent large scale structures of radical attitudes and beliefs. The following principal pull factorswere identified in our study [Sherstoboyev 2012]:

-         the ideological values (cultural, political, religious, moral);

-         justification of the political violence;

-         extensive social ties within a milieu;

-         intolerant and aggressive attitudes towards outer  social groups,

-         authoritarian leadership (“agents of radicalization”),

-         cultural / information influence.

However, the theoretical model tested on Ukrainian empirical data should be verified more thoroughly at least taking into account its the cultural sensitiveness. The empirical study of social milieus has to be oriented to the certain limited parameters of social milieus of migrants considered as a prototype of a milieu with high probability of radicalization.

According to  the M. Sageman,  Khosrokhavar, Hoffman and other researchers [Sageman 2004, 2007; Khosrokhavar 2006; Dalgaard-Nielsen 2008, 2010; Hoffman2010,  etc.], there have been essential changes in radicalized groups in the West in ‘post-Iraq’ period after the transition phase of 9/11/01 – 3/03, i.e., the increase of number of poorly educated persons and homegrown dominance,  as well as the  decrease of the average age in such groups (from 25,6  to 20 in the current period). Taking into account these recent research findings about radicalization in the West and the peculiar features of its subjects in so-called post-Iraq generation, a further field study has to be focused on young migrants from Islamic states, in age of 16-30 attending Islamic religious or cultural centers in their city of residence. Personal devotion to Islam should be a crucial feature of the milieu; however, it can be combined with such religious as secular practices in everyday life.

Paying attention to the conceptual focus, the collecting of empirical data has to be conducted in cities of the Northern American and European continents with Muslim migrants’ communities. To minimize the influence of the cultural and institutional national specifics in the model the study needs comparison of national cases distinguishing in three-dimensions, namely: the national economic well-being, the immigration flows and the terrorism index. Each of the selected national cases has to have a moderate rate on the Global Terrorism Index (GTI), and strong representation of Islamic migrants in the immigrant flows, but they have to be differing on indicators of national wellbeing, political (un-)stability, the migration net flows as well as their historical preconditions.

The preliminary list of selected national cases matching the dimensions mentioned above is the following: the U.S., Austria (or German), Ireland, Portugal (or Spain), and Ukraine. Their profiles are:

  1. The U.S. – the GTI is 41 from 158 (in 2011); immigration from Asia to the U.S. has increased considerably since the 1965 US Immigration and Nationality Act. In 1960, the Asian born immigrants accounted for just five percent of the foreign-born population in the US, while in 2000 they made up over a quarter of those born outside the United States. Today, the Asian born are the country’s second largest foreign-born population by world region of birth; e.g. half of the Pakistani-origin foreign born lived in the New York, Houston, Chicago, Washington, DC, and Los Angeles greater metropolitan areas (according to 2006 data – [Migration Information 2013].
  2. Austria – located in the central Europe; high political stability and national wellbeing; the GTI is 69 from 158 (2011 data); the net migration in the country in absolute numbers is one of the largest in the EU (+61,000 in 2005) after Spain, Italy, France, Germany and Portugal; has one of the largest  in the EU foreign-born population (15.1%); the number of immigrants to the country increased considerably during the 1990s and early 2000s [The Global Terrorism Index 2012; Muenz 2006].
  3. Ireland – located in the northern Europe; facing deep economic and political crises since 2008 (just after the period of economic boom) and the high level of unemployment (especially among the youth; the male unemployment rate in 2012 was 16.9%);  the GTI is 68 from 158 (2011 data); the net migration in the country in absolute numbers is one of the largest in the EU (+47,000 in 2005); also has one of the largest in the EU foreign-born population (14.1%); the number of immigrants to the country also increased considerably during the 1990s and early 2000s.
  4. Portugal – located in the southern Europe; facing deep economic and political crises since 2008 and the high level of unemployment (especially among the youth; the unemployment rate in total civilian labor force in 2012 was 12%); the GTI is 77 from 158 (in 2011), much higher than in Spain; the number of foreign-born population is significant; the number of immigrants to the country also increased during the 1990s and early 2000s.
  5. Ukraine - located in the eastern (post-Soviet) Europe; its geographical position has made it one of the main transit countries along the path of the migration flows from the Asia and Eurasia to the EU, moreover the flow of illegal migrants from the East countries grows due to the porous borders with the Russian Federation and Belarus;  it faces permanent deep economic and political crises since the middle of the 1990s, and the high level of unemployment (especially among the youth and men, according to the Mundi index the youth unemployment rate in 2012 was 14.9%) [Index Mundi]; the GTI is 56 (2011 data); according to a report of UN Secretary General on monitoring of the world population, Ukraine is the fourth in the world in terms of number of international migrants; the portion of immigrants among the total population was 14.7% in 2008; there has been a visible upward trend in the net migration during the 2000s [Migration in Ukraine 2008] with growing number of immigrants of Chechen, Uzbeks and Azeri origins.

To clarify the structural dimensions of the model (AGM) and estimate quantitatively its parameters, we propose to explore the following set of the data mining methods in all national cases listed above:

  • expert surveys of public leaders and civil servants dealing with immigrant, religious, public defense issues,; a purpose of this expert surveys is to determine the range of possible values for model quantitative parameters, to define the intervals of normative values for model basic parameters, to specify the parameters of population and sample for survey on the next stage;
  • sampling survey in the social milieus, potentially (theoretically) inclined to radicalization;
  • content-analysis and qualitative text analysis (for the period of scrutiny) of speeches, publications by public, religious, political, nongovernmental and other organizations leaders posted on internet (web-sites) with focus at the organizations potentially capable to network the milieus inclined to radicalization (“points of attraction”).

The main result of the proposed work will be a development of a formal model of radicalization of a migrant’s milieu and quantitative and qualitative validation of the model. This model will include a set of empirically based criteria for monitoring of radicalization of social milieus and will allow one to further justify policy-making suggestions. Combining the results of empirical study and novel theoretical modeling we expect to explain mechanisms of radicalization, or critical transition in social milieu with the clearly identified parameters that will allows early diagnostic and prevention of radicalization of social milieu.



Berlyand L., Kutsenko O., Sherstoboev V. (2005) Development of Theoretical Model Explaining the Support of Terrorism. In: Meythodolkogy, Theory , and Practics of Sociological Analyses of Contemporary Society.

Chaudhry, Saleha Zaffir (2012) Ten years later: Exploring the lived experiences of college-aged Muslim American women after September 11 \ A dissertation. Northerneastern University, Boston, Massachussets

Dalgaard-Nielsen, Anja (2010) Violent Radicalization in Europe: What We Know and What We Do Not Know. In:Studies in Conflict & Terrorism, Volume 33, Issue 9. Pp. 797-814

Dalgaard-Nielsen, Anja (2008) Studying violent radicalization in Europe II. The potential contribution of socio-psychological and psychological approaches, Copenhagen: DIIS Working Paper 2008.

Epstein, Joshua M. (2002) Modeling civil violence: An agent-based computational approach / Proceesings of the National Academy of Sciences of the United States of America 99 (Suppl 3): 7243-50

Harison, Lawrence E. and Samuel P. Huntington (eds.) (2001) Culture matters: How values shape human progress. Boulder CO: Basic Books.

Hammond, Ross A., and Robert Axelrod (2006) the evolution of ethnocentrism. In: Journal of Conflict Resolution 50 (6): 926-36

Horgan, J. (2005). Psychology of Terrorism. London: Routledge.

Horgan, J. (2008). ‘From Profiles to Pathways and Roots to Routes: Perspectives from Psychology on Radicalization into Terrorism,’ ANNALS, American Association of Political and Social Sciences, 618, July.

Horgan, J. (2009). Walking Away from Terrorism: Accounts of Disengagement from Radical and Extremist Movements. London and New York: Routledge.

Hudson, Rex A., and Marilyn L. Majeska (1999)  The sociology and psychology of terrorism: who becomes a terrorist and why? Washington D.C.: Library of Congress: Research Division

Khosrokhavar, Farhad (2006) Terrorism in Europe. In Terrorism and international relations, Edited by: Hamilton, D. 23–38. Washington, DC: Center for Transatlantic Relations.

Kydd, Andrew H. (2011) Terrorism and Profiling. In: Terrorism and Political Violence, 23:3, 458-473.

Kutsenko O. (2010) Turkish Minorities in Europe: Towards Societal Integration or the Rise of “Parallel Societies”?  / Olga Kutsenko. In: Neinz Fassmann, Max Haller, David Lane.  Migration and Mobility in Europe. Edward Elgar, Cheltenham, UK. 2010. Pp.191-208.

Kühle, Lene and Lasse Lindekilde (2010) Radicalization among Young Muslims in Aarhus \  Aarhus: Centre for Studies in Islamism and Radicalisation, January

Index Mundi

Leiken, Robert S., and Steven Brooke (2006) The quantitative analysis of terrorism and immigration: An initial exploration. In: Terrorism and Political Violence 18 (4): 503-21.

Macy, Michael W., and Robert Willer (2002) From factors to actors: Computational sociology and agent-based modeling. In: Annual Review of Sociology 28 (1): 143-66

McCauley, Clark & Sophia Moskalenko (2008) Mechanisms of Political Radicalization: Pathways Toward Terrorism. In:Terrorism and Political Violence, Volume 20, Issue 3. Pp. 415-433

Migration Information / DataHub / US Focus, 2013.

Migration in Ukraine: A Country Profile \ Migration: A World on the Move \ UNFPA. ZOM, Geneva, 2008

Moody, James and Douglas R. White (2003) Structural Cohesion and Embeddedness: A Hierarchical Concept of Social Groups. In: American Sociological Review, Vol. 68, February: Pp.103–127

Muenz, Rainer (2006) Europe: Population and migration in 2005

Mullins, Sam  andAdam Dolnik (2010) An exploratory, dynamic application of Social Network Analysis for modelling the development of Islamist terror‐cells in the West. In: Behavioral Sciences of Terrorism and Political Aggression,  Volume 2, Issue 1.

Sageman M. (2004) Understanding Terror Networks (Philadelphia: University of Pennsylvania Press.

Sageman M. (2008) Leaderless Jihad: Terror Networks in the Twenty-First Century (Philadelphia: University of Pennsylvania Press.

 Sedgwick, Mark (2010) The Concept of Radicalization as a Source of Confusion. In:Terrorism and Political Violence.Volume 22, Issue 4, 2010. Pp. 479-494

Silber, Mitchell D., and Arvin Bhatt (2007) radicalization in the west: The homegrown threat. New York,: NY: New York City Police Department Intelligence Division

Slootman, Marieke and Tillie, Jean (2006) Processes of radicalisation. Why some Amsterdam Muslims become radicals, Amsterdam: Institute for Migrations and Ethnic Studies, University of Amsterdam.

START (2010) Global terrorism database, national consortium for the study of terrorism and responses to terror (START). –

Sherstoboev, VladV. The Specific of Social Milieu Forming Attitudes towards Political Extremism. PhD Theses. Kharkiv National Unicversity, Kharkiv, 2012.

The Global Terrorism Index \ Institute for Economics and peace ‘START’, 2012

The Radical Dawa in Transition: The Rise of Islamic Neoradicalism in the Netherlands \ AIVD (Netherlands General Intelligence and Security Service. The Hague: AIVD, 2007. –

Toscani, G. (2006) Kinetic models of opinion formation. In: Comm. Math. Sci., 4, pp. 481-496

TRC (2010) American terrorism study, terrorism research center at the University of Arkansas. –

White, Douglas R. (2011) Social Networks, Cognition and Culture. Chapter 18.  Blackwell Companion to Handbook of Cognitive Anthropology.  Pp. 331-354

Vidino, Lorenzo (2007) The Hofstad Group: the new face of terrorist networks in Europe. In: Studies in Conflict and Terrorism 30 (7): 579-92


The paper is prepared within the NATO\CLG projects ‘Radicalization of Social Communities towards Support of Terrorism: sociological investigation and  mathematical modeling’ and ‘Critical transition in social networks supporting terrorism: social empirical study and modelling by random graphs’ (2005-2013, Ref. Nо.983046)