Violence in Ideological and Non-Ideological Groups:
A Quantitative Analysis of Qualitative Data1
M D. M2
K E. B-A
S T. H
Penn State University
University of Oklahoma
D E
M S C
J E
Development Dimensions
International, Inc.
University of Oklahoma
Multiple models have been proposed to account for violence among ideological
groups. To identify critical variables contributing to violent behavior in these
groups, violent ideological groups were compared to relevant comparison groups. A
historically based content analysis was conducted to assess these groups with respect
to a number of variables examining leader, group, organizational, and environmental attributes held to influence violence. Discriminant analyses revealed that violent
ideological groups differed from comparison groups with respect to leader extremism, group righteousness, organizational indoctrination, and environmental conflict
and disruption. Regression analyses revealed that these discriminant functions predicted a number of notable violent and ideological criteria. The implications of these
findings for understanding the origins of violence in ideological groups are discussed.
The images we see on television seem to suggest that there is something
unique about ideology—extremist ideology—that gives rise to the propensity
for violence. Groups such as Al Qaeda (Jordan & Boix, 2004; Post, 2005),
White supremacists (Blazak, 2001; Henry, Sidanius, Levin, & Pratto, 2002),
the militia movement (Pitcavage, 2001), and the Christian Identity Movement (Sharpe, 2000) all espouse extremist ideologies that have been used to
justify violent acts ranging from the bombing of abortion clinics to the
destruction of the World Trade Center. These acts, and their impact on
people’s lives, have led scholars such as Moghaddam (2005) and Sternberg
(2003) to call for a new wave of research that will allow us to develop a
1
The authors thank Matthew Allen, Gregory Ruark, and Amanda Angie for their contributions to the present effort. We also thank an anonymous reviewer whose comments were
especially helpful in revising the manuscript. Parts of this effort were supported by a series of
grants from the U.S. Department of Defense, Michael D. Mumford, Principal Investigator.
2
Correspondence concerning this article should be addressed to Michael D. Mumford,
Department of Psychology, University of Oklahoma, Norman, OK 73019. E-mail: mmumford@
ou.edu
1521
Journal of Applied Social Psychology, 2008, 38, 6, pp. 1521–1561.
© 2008 Copyright the Authors
Journal compilation © 2008 Wiley Periodicals, Inc.
1522 MUMFORD ET AL.
systematic understanding of the origins and consequences of ideological
violence.
Although few would dispute the value of research along these lines, it
clearly represents a formidable undertaking. Despite their visibility, the emergence of violent ideological groups remains a rare event. As a result, it has
proven difficult to conduct the kind of systematic quantitative studies necessary to begin developing general models specifying the key variables giving
rise to extremist ideologies. One response, clearly the most common response
to this problem, has been for scholars to conduct in-depth qualitative analyses examining the psychosocial dynamic of groups in which espoused ideology has led to acts of violence (e.g., Hoffman, 1999; O’Boyle, 2002; Staub,
1996, 2003). An alternative approach has been to examine underlying psychological processes; for example, threat (e.g., Cameron, Duck, Terry, &
Lalonde, 2005; Navarrete, Kurzban, Fessler, & Kirkpatrick, 2004) and outgroup stereotyping (e.g., Haslam et al., 1998; Hewstone, Rubin, & Willis,
2002) that might give rise to violent behavior.
Both qualitative case analysis and experimental studies of group process
have proven useful in understanding the origins of ideological violence.
However, a third approach is available that might also prove of some value.
Simonton (1991) argued that rare events may also be studied using a historiometric approach. The historiometric approach entails the quantitative
analysis of multiple qualitative cases cumulated over time. Indeed, this
approach has proven useful in understanding other rare events, such as
eminent achievement in the arts and sciences (Simonton, 1999), as well as
outstanding, historically notable leadership (Mumford, 2006). Our intent in
the present study is to apply historiometric techniques in a quantitative
analysis of multiple case studies of ideological groups with the intent of
identifying leadership, group process, organizational, and environmental
variables that uniquely characterize violent ideological groups.
Ideological Groups
Groups form and create institutions for many reasons. For example,
groups may form for purely instrumental and pragmatic reasons. A case in
point may be found in industrial organizations where division of labor in the
group promotes economic gain (Katz & Kahn, 1978). Groups may also form
to facilitate association, as is the case for most clubs. Groups, however, may
also form to provide members with a sense of identity and a structure for
interpreting and acting on their world (Vigil, 2003).
Although groups may form for many reasons, ideology provides a noteworthy, and potentially important, basis for the formation of groups and
VIOLENCE AND IDEOLOGY
1523
institutions (Thompson & Bunderson, 2003). Ideology is reflected in the
formation, adoption, and articulation of a set of beliefs and values that are
held to represent a unique and universal good: an inherently right, correct,
and just worldview (Blau, 1964). Ideology, however, represents something
more than a set of beliefs about what is right. Rather, ideology entails an
integrated set of values and goals that assume unique status in which attainment of salient goals, or failure to attain these goals, is held to be determined
by a fixed, relatively rigid, set of beliefs about the causes influencing goal
attainment and failures in goal attainment (Murray & Cowden, 1999). In
other words, ideology involves a mental model specifying desirable goals and
the causes shaping goal attainment (Strange & Mumford, 2005). This mental
model is imposed as an interpretive structure for understanding and reacting
to events; a wide range of events in the case of powerful ideologies such as 20th
century communism or 21st century millennialism. Ideological groups arise
when a mental model defining goals and causes is adopted by a number of
individuals who apply this model as a vehicle for understanding and acting in
uncertain situations (Friedland, 2001).
Ideological groups may prove attractive to people for a number of
reasons. First, ideology provides people with a framework for understanding
their world and interpreting crises that occur in their lives (Mumford, 2006).
Second, ideology provides people with a sense of identity and meaning
(Shamir, House, & Arthur, 1993). Third, ideology, vis-à-vis valued goals,
provides a basis for acquisition of feelings of self-esteem (Aberson, Healy, &
Romero, 2000). Fourth, and finally, ideology allows people to reduce uncertainty and define viable courses of action through the normative prescriptions embedded in mental models (Hogg, 2003).
Understanding crisis events, reducing ambiguity, providing a sense of
identity, and enhancing self-esteem all appear to represent powerful forces that
attract people to ideological groups. However, a number of variables operating at multiple levels of analysis have been found to influence the emergence
and maintenance of ideological groups. For example, effective articulation of
an ideology by charismatic leaders who evidence an ascetic commitment and a
willingness to sacrifice for the ideology appears to be critical to the formation
of ideological groups (Yorges, Weiss, & Strickland, 1999). Leader articulation
and symbolic embodiment of an ideology, however, often arises from prolonged exposure to failed social models (Strange & Mumford, 2005), with
leaders evidencing a commitment to resolve injustice that has arisen from these
failed models (Moghaddam, 2005).
Although the creation and articulation of an ideology by leaders is a
necessary precondition for the emergence of ideological groups, the behavior
and maintenance of these groups is influenced by a number of processes
operating at the group level. For example, ideological groups often establish
1524 MUMFORD ET AL.
strong group boundaries whereby entry into the group requires some degree
of self-sacrifice and overt acceptance of certain symbolic commitments
(Byman, 1998; Gerstenfeld, Grant, & Chiang, 2003; Gordijn, Wigboldus, &
Yzerbyt, 2001): strategies intended to encourage adoption and internalization of the ideology. Once people have joined the group, social learning and
peer role modeling serve to reinforce and elaborate the ideology being advocated (Burdman, 2003; Crenshaw, 1988; Lee & Leets, 2002; Tindale &
Kameda, 2000). This social learning will often be associated with affective
framing of select events that articulate the “rightness” of the ideology (Wood,
2000) and the need for urgent action (Spoor & Kelly, 2004), with these
affective framing actions serving to create the commitments that lead the
group to become a salient aspect of members’ identity (Monroe, Hankin, &
Van Vechten, 2000).
Ideological groups not only establish processes that serve to inculcate
ideology, they also create institutional systems that serve to reinforce and
maintain the group. Thus, Blazak (2001) has found that to offset high
dropout rates, ideological groups will typically create strategies and structures to promote member recruitment. Not only do ideological organizations
stress recruitment, they adopt practices and policies (e.g., training, mentoring) intended to replace extant systems for understanding events (i.e., sense
breaking) with new interpretive frameworks (i.e., sense making) consistent
with the organizations’ ideology (Combs, 2003; Pratt, 2000; Schwartz &
Kaslow, 2001). These organizational practices are typically subject to tight
control and monitoring by a key leadership cadre (a form of clan control)
known to be highly committed to the ideology (Strange & Mumford, 2002;
Zawodny, 1983).
Ideological groups, and the organizational policies and practices that
maintain these groups, do not exist independent of their environment. One
common explanation for the emergence of these groups is a lack of direction,
structure, and opportunity induced by a chaotic environment in which traditional paths to success have been undermined by socioeconomic conditions
(Post, Ruby, & Shaw, 2002; Staub, 2003). Cultural dislocation and the loss of
traditional patterns of social interaction have also been linked to the emergence of ideological groups if the ideology provides a framework for understanding and responding to change (Combs, 2003; Hoffman, 1999).
Violence and Ideology
Although ideological groups display some distinctive characteristics—
characteristics that at times may predispose these groups to violence—it
should also be recognized that many ideological groups exist that do not
VIOLENCE AND IDEOLOGY
1525
exhibit violence. This observation, in turn, begs a question: What are the
unique characteristics of ideological groups that advocate and initiate violence? Given the influence of leadership on the formation and behavior of
ideological groups, it seems plausible to argue that leader characteristics may
represent a contributing factor to group violence. In fact, narcissism, a sense
of superiority, and feelings of pride have been linked to the propensity of
leaders to induce violence (O’Connor, Mumford, Clifton, Gessner, &
Connelly, 1995) and violence on the part of terrorist groups (Atran, 2003;
Cordes, 2001; Crayton, 1983). Associated with narcissistic pride, however, is
often a sense of fear or uncertainty (Atran, 2003; O’Connor et al., 1995),
which also appears to induce violence. Fear and narcissism, furthermore,
when combined with rigid application of an ideological mental model, may
lead to systemic distortion of information, which gives rise to misunderstanding of others’ intentions; misunderstandings that may promote violence.
Not only can leader characteristics engender violence, violence may
become more likely in groups evidencing certain kinds of interactions.
More specifically, conflict and group rivalry can threaten group ideology.
This threat, in turn, can result in the acquisition of negative views of others
and the dehumanization of rival group members, thereby leading to violence (Paladino et al., 2002; Staub, 2004), especially when negative emotions are activated by events that threaten group members and the
legitimacy of their ideology (Spoor & Kelly, 2004; Stroessner, Mackie, &
Michalsen, 2005).
There are two other characteristics of ideological groups that may act to
increase the likelihood of out-group dehumanization and violence. First, by
virtue of the justification provided for dehumanization and violence, groups
that feel victimized and that evidence a strong sense of moral superiority may
be especially violence-prone (Moghaddam, 2005; Staub, 2004). Second, violence may be encouraged by the requirement that people surrender individual
identity in service to group identity (Post et al., 2002).
In addition to group process, certain characteristics of the structure and
practices of ideological organizations may serve to make some groups especially prone to violence. McCormick (2003), for example, argued that fractionalization, coupled with attempts to induce centralized control through
cliques may promote violence as a technique for acquiring political power
within the group. Although certain structural characteristics of organizations
may promote violence, it seems likely that organizational processes that act
to legitimize violence—for example, use of violence and punishment as
control mechanisms—will exert particularly powerful effects on the proclivity of certain ideological groups toward violence (Vigil, 2003).
Along with organizational structure and policy, it is possible that the
socialization practices applied by certain ideological organizations may
1526 MUMFORD ET AL.
influence their propensity for violence. Socialization into ideological organizations often involves sense breaking and sense making (Pratt, 2000).
However, if sense breaking is framed in terms of opposition to other existing
groups—groups that are viewed as exploitive—then the potential for dehumanization and violence will increase. As a result, violent ideological groups
and organizations often are characterized by the use of propaganda, extensive indoctrination, and the articulation of new cultural symbols (Burdman,
2003; Harré, 2004; Pitcavage, 2001).
As was the case with regard to the formation of ideological groups,
there is reason to suspect that certain environmental conditions may act to
promote violence among ideological groups. As might be expected based
on our foregoing observations, environmental conditions giving rise to
perceptions of victimization and injustice—such as corruption, unfair
distribution of resources, and limitations of opportunity (especially for
ambitious middle-class youths)—all seem linked to the emergence of violent
ideological groups (Ehrlich & Liu, 2002; Staub, 2004). In addition to socioeconomic conditions, social conditions giving rise to a climate of violence
(e.g., group rivalry, societal fragmentation, fear of status loss) appear to
encourage violence as a result of perceived threat (Cordes, 2001). In this
regard, however, it is important to bear in mind that for ideological groups,
where the existence and identity of a group is tied to a mental model
or worldview, the existence of competing ideologies in the society may
prove particularly threatening and a potentially powerful source of
violence.
Taken as a whole, it appears that multiple variables—that is, variables
linked to leadership, group process, organizational practices, and environmental conditions—may act to promote violence among ideological groups.
Accordingly, our intent in the present study is twofold. First, we hope to
demonstrate that violent ideological groups differ from (a) violent nonideological; (b) nonviolent ideological; and (c) nonviolent nonideological
groups with regard to leadership, group process, organizational practices,
and environmental conditions. To illustrate, we expect to find that violent,
ideological groups will be led by individuals evidencing an extreme, rigid
ideology, as well as a strong sense of superiority. These leader characteristics,
in turn, will likely perpetuate the internalization of superiority within the
group, resulting in higher levels of self-righteousness among members. We
also expect that groups willing to commit violent acts will have to sanction
and legitimize those acts through organizational policies and procedures.
Moreover, we anticipate that such violence will only occur in environments
characterized by conflict and turmoil. After examining these group differences, we hope to demonstrate also that these differences are, indeed, related
to relevant violent and ideological outcomes.
VIOLENCE AND IDEOLOGY
1527
Method
Sample
The sample applied in the present study consisted of 80 groups. The
groups included in the sample were selected such that half were ideologically
based, while the other half were nonideological. Half the ideological and half
the nonideological groups examined in the present study were known to have
been involved in violence, while the remaining half were not involved in
violence. Of the 20 groups examined in each cell of this 2 (Ideological: yes or
no) ¥ 2 (Violent vs. Nonviolent) design, half were to be drawn from Western
cultures and half from non-Western cultures to ensure some generalizability
in the conclusions flowing from the present study.
Group Definition
To identify the groups to be examined in the present study, the following
criteria were applied. A group was considered violent if at least 10 acts of
violence (e.g., robberies, murders) had resulted from the actions of group
members where (a) the act of violence could be directly attributed to group
members; (b) multiple group members were involved in the act; (c) the act
of violence was sanctioned by the group; and (d) multiple group members
were involved in planning the actions that led to violence. A group was
considered nonviolent if no acts of violence involving multiple group
members and group-based planning occurred over the course of the groups’
existence.
Ideological and nonideological groups were identified on the basis of
the following considerations. A group was considered ideological if (a) the
basis for formation of the group was a shared mental model or image about
the general good and the actions that must be taken to bring about this
good; (b) group members applied this mental model in interpreting a wide
range of events; (c) the group sought to recruit new members based on this
model; and (d) new group members were expressly educated and socialized
to ensure internalization of this model. A group was considered nonideological if the basis for group formation and member recruitment was
utilitarian and the group made no attempt to advance a global image of a
greater good.
In addition to these considerations, three other criteria were applied in
selecting the groups to be examined in the present study. First, a group was
considered for inclusion in the study only if it was based on voluntary
association. Second, these voluntary associations could not have been
1528 MUMFORD ET AL.
formed under the auspices of a broader institutional entity, a state, corporation, or religion, to ensure that the group represented a distinct entity. Third,
to ensure that adequate historically verifiable information was available, only
groups operating in the 20th century were considered as candidates for inclusion in the sample.
Group Selection
After defining these criteria, a review of bibliographic sources was used to
identify in-depth qualitative analyses examining the origins and practices of
various social groups. These in-depth qualitative analyses were to be presented in the form of a book or monograph seeking to provide a comprehensive description of the origins and practices of the group. To be
considered for use as a basis for describing a group, the book or monograph
was required to be at least 150 pages long and prepared as an objective
academic analysis. Thus, analyses conducted by or supported by the group
were not considered for inclusion in this study.
Application of these criteria led to the identification of some 250 qualitative analyses that might be used to describe relevant groups. The material
presented in these books was then reviewed by two psychologists. Based on
this material, the group under consideration was classified into the ideological versus nonideological, violent versus nonviolent, and Western versus
non-Western categories. In each cell that the analyses retained (only one per
group) were those that (a) presented multiple verifiable incidents of group
behavior; (b) provided adequate documentation for these observations; and
(c) received favorable academic reviews.
Figure 1 presents the groups identified for inclusion in the present study.
The reference list provides the citations for the source material used to
describe the groups. As can be seen, most violent ideological groups identified in this review were terrorist or political advocacy groups, while most of
the violent nonideological groups identified in this review were criminal
groups. Nonviolent ideological groups and nonviolent nonideological groups
included social service, issue advocacy, and political advocacy groups
involved in a range of issues. These characteristics of the group classification,
of course, speak to the construct validity of the selection procedures applied
with regard to common conceptions of groups (e.g., violent ideologies are
commonly held to be associated with terrorism). They also indicate that the
groups under consideration provided a plausible and appropriate basis for
making comparisons with respect to leadership, group process, organizational policies, and environmental conditions.
VIOLENCE AND IDEOLOGY
WESTERN
NONVIOLENT
VIOLENT
IDEOLOGICAL
1529
NONIDEOLOGICAL
1. The Revolutionary Organization 17 Novembers
2. Ku Klux Klan
3. ETA (Basque separatist movement)
4. Baghwan Shree Rajneesh (Osho)
5. Black Panthers
6. Earth Liberation Front
7. Aryan Republican Army
8. IRA
9. Molly Maguires
10. IRGUN
1. Hell’s Angels
2. Vice Lords
3. Latin Kings (New York)
4. Cardiff Soul Crew
5. The Tison Gang
6. The Purple Gang
7. The Chicago Outfit
8. La Cosa Nostra
9. American Mafia
10. Red Mafiya
1. Japanese Red Army
2. Al-Jihad
3. Al-Qaeda
4. Hizbullah
5. Aum Shinrikyo (aka Aum Supreme truth)
6. Tupac Amaru Revolutionary Movement
7. Palestinian HAMAS
8. Popular Front for the Liberation of Palestine
9. Shining Path
10. Communist Party of the Philippines
1. Hong Kong Triad
2. Zimbabwe Land Grabbers
3. Shower Posse
4. Chinese Triad
5. Chinatown Born to Kill Gang
6. Colombian Medellin cocaine cartel
7. Japanese Yakuza
8. Shanghai Green Gang
9. Jamaican Posse
10. Cali Cartel
1. Mothers Against Drunk Driving
2. Masons
3. YMCA
4. National Education Association
5. Salvation Army
6. National Rifle Association
7. Habitat for Humanity
8. Polish Solidarity Movement
9. Russian Liberation Movement
10. ACLU
1. Oxfam
2. American Red Cross
3. American Legion
4. Big Brothers/Big Sisters of America
5. NCAA
6. Lions Club
7. Farm Labor Organizing Committee
8. Doctors Without Borders
9. Friends of the Library
10. London Charity Organization Society
1. Revolutionary Association of Women in
Afghanistan (RAWA)
2. Organization of African Unity
3. Indonesia Solidarity
4. Ethiopian Revolutionary Party
5. Indian Women’s Movement
6. Dalit Movement
7. The Industrial and Commercial Workers’ Union
of Africa
8. May Fourth Movement
9. Iranian Islamic Women’s Movement
10. Society of the Muslim Brothers in Egypt
1. Mothers of the Plaza de Mexico
2. Olympic Organizing Committee
3. Red Crescent
4. Self-Employed Women’s Association
5. Jordanian Hashemite Fund for Human
Development
6.Working Women’s Forum
7. ECOMOG
8. East African Flying Doctor Service
9. Amazonian Indigenous Rights Movement
10. Green Belt Movement
Figure 1. Groups examined.
Predictors
A content analysis of the material presented in these qualitative analyses was used to assess the groups with respect to leadership, group processes, organizational policies, and environmental conditions. Initially, a
1530 MUMFORD ET AL.
psychologist familiar with the literature on group behavior was asked to
review the book or monograph summarizing the findings obtained in the
qualitative analysis. The psychologist was then asked to identify the two
chapters that provided the most complete, detailed description of the characteristic behavior of group members. A reliability check indicated greater
than 90% agreement in the chapters selected as having the greatest value in
describing the behavior of the group. These chapters were typically 20 to 30
pages in length, and contained 320 to 390 words per page.
The material presented in these chapters provided the basis for the
content analysis examining predictors of group violence and ideology. The
predictors applied in the present study were based on a review of the available
literature examining violent ideological groups (e.g., Moghaddam, 2005; Post
et al., 2002; Staub, 2004). The literature review was used to identify variables
linked to both ideology and violence with respect to leader characteristics
(e.g., superiority, narcissism, negative life themes, value crystallization, sense
of injustice), group processes (e.g., victimization, submergence of individual
identity, strong group boundaries, identity expression through group), organizational policies and practices (e.g., sense making, violence as control,
ritualized investment, systematic recruitment), and environmental conditions
(e.g., economic exploitation, presence of other ideological groups, loss of
institutions, strong traditional culture). Overall, 15 to 26 potential predictors
were identified at each level of analysis. Table 1 presents the list of predictors
to be examined at each level of analysis.
With regard to the predictors identified on the basis of the literature
review, four other points are worthy of note. First, predictor dimensions were
selected to cover a wide range of theoretical perspectives on both group
violence and ideological groups. Second, predictors were selected that had
received some attention in a number of different models of ideological violence. Third, an attempt was made to minimize redundancy among dimensions within a given level of analysis. However, because some variables
operate in different forms across levels (e.g., injustice, victimization), no
attempt was made to eliminate cross-level redundancies. Fourth, and finally,
predictors were defined in terms of the variables operating within the relevant
level of analysis.
After a list of potential predictor variables had been specified, a set of
rating scales was developed to be applied in content analysis of the relevant
qualitative material. Initially, behavioral or event-based operational definitions were developed for each predictor. After these operational definitions
had been developed, a set of potential behavioral or event markers were
developed for assessing expression of a predictor by the groups. These behavioral and event markers were generated using a variation on the procedures
recommended by Gessner, O’Connor, Clifton, Connelly, and Mumford
Table 1
Predictor Variables at Leader, Group, Organizational, and Environmental Levels
Individual
1. Negative life themes
2. Superiority/narcissism
3. Entitlement
4. Low openness
1. Victimization
2. Negative mental models
3. Group feelings of
superiority
4. Group rivalry
5. Group exclusivity
6. Strong group
boundaries
7. Selective interpretation 7. Strong group affect
of information
8. Information distrust
8. Submergence of
individual
9. Stereotypic normalcy
9. Strong group values
Organizational
Environmental
1. Violence as control
2. Punishment as control
3. Resource control
1. Social conflict
2. Loss of social patterns
3. Loss of institutions
4. Clan control
4. Middle class
marginalization
5. Loss of opportunity
6. Fear of social loss
5. Fractionalization
6. External conflict
7. Turbulence
8. Centralization
9. Formalization
7. Disappointed social
expectations
8. Disruptive
environmental events
9. Corruption
VIOLENCE AND IDEOLOGY
5. Reality distortion
6. Compartmentalization
Group
1531
Individual
Group
Organizational
10. Sensitivity to injustice
10. Extensive socialization
10. Isolation of members
11. Value crystallization
12. Personal self-sacrifice
11. Peer group influence
12. Group sacrifice
11.
12.
13. Disregard for time
13. Value-based
13.
recruitment
14. Symbolic commitments 14.
14. Commitment to
like-minded others
15. Redemptive life events 15. Identity expression
through group
16. Strong enculturation
17. Extreme anchoring
Environmental
10. Inequitable resource
distribution
Close supervision
11. Economic exploitation
Denigration of external 12. Economic displacement
institutions
Rituals
13. Strong cultural
traditions
Sense breaking
14. Threats to tradition
15. Sense making
15. Globalization
16. Sacrifice for
organization
17. Recruitment by
investment
16. Anomie
17. Rivalry
1532 MUMFORD ET AL.
Table 1 Continued
18. Channeled social
commitment
19. Exposure to
ideological leaders
20. Ascetics
21. Personal self-sacrifice
22. Interest in ideological
material
18. Emphasis on
organizational
commitment
19. Training in ideology
19. Multiple ideological
groups
VIOLENCE AND IDEOLOGY
20. Mentoring
21. Value-based policies
22. Rewards for value
congruence
23. Extended life influence
of organization
24. Value-based
communication
25. Idolization of
leadership
26. Protection of
ideological core
18. Fragmentation
1533
1534 MUMFORD ET AL.
(1993), and Mumford, Costanza, Connelly, and Johnson (1996). In this
procedure, a panel of four psychologists—all doctoral students familiar with
the groups literature—were asked to generate behaviors or events that would
indicate the operation of a given dimension using the relevant operational
definition. The pool of potential markers was then reviewed by two senior
psychologists for construct relevance, construct coverage, likely appearance
of these markers in qualitative analyses, and clarity of interpretation.
Markers that failed to meet these criteria after editing were eliminated.
Typically, 20 to 30 markers were developed for each dimension, of which 10
to 20 survived the review process.
This pool of candidate markers was then presented to a panel of four
psychologists. Panel members were asked to rate each marker on a 5-point
scale with respect to the substantive significance of the relevant behavior or
event for assessing the dimension of interest. These substantive significance ratings produced interrater agreement coefficients above .90. The five
markers that revealed the highest average substantive relevance ratings, and
had the lowest standard deviations, were to be applied in assessing the
qualitative data describing group behavior. Table 2 presents examples of
these markers.
In applying these markers to assess the group’s status on the relevant
predictor dimensions, four judges—again, all doctoral students in psychology familiar with the group’s literature—were asked to read through each
chapter. After reading through a given chapter, they were asked to rate on a
5-point scale the extent to which behavior or events similar to the critical
markers identified for a predictor dimension appeared in the chapter. The
average of these ratings across the two chapters drawn from the qualitative
analysis obtained for a group provided the group’s score on each predictor
dimension.
Prior to evaluating the qualitative material using these ratings scales, the
four judges received 40 hours of training. In the training, the judges were
familiarized with the operational definition of each dimension and all of the
relevant behavioral and event markers. They were then asked to practice
applying the rating scales to a set of sample chapters. After making their
ratings of the sample chapters, judges met to discuss their ratings and to
resolve any discrepancies. It is of note that in making the ratings, judges were
rotated across chapters to minimize fatigue and order effects. Application of
these procedures resulted in adequate interrater agreement coefficients.
Using the procedures suggested by Fleiss and Shrout (1979), the mean interrater agreement coefficient obtained for the leader, group, organizational,
and environmental dimensions were .82, .85, .75, and .78, respectively.
Some evidence for the construct validity of these ratings was obtained by
examining the correlations among the predictor dimensions. For example,
Table 2
Examples of Markers for Leader, Group, Organizational, and Environmental Predictors
Individual
Organizational
Victimization
Violence as Control
Members who disobey
Ridiculed by public
organizational rules are
Members killed/beaten by
other groups
beaten or killed
Forced to meet in hiding
Members who attempt to
Petitions government about
leave the organization
unfair treatment
are threatened with
Talks about unfair
violence
treatment at rallies
Culture includes stories of
past violence
Leaders carry weapons
Violent acts are committed
against families of
disobedient members
Environmental
Social Conflict
Competing religious groups
in region
Rival ethnic groups in
region
Groups make claims for
same objects
Discontentment over large
differences in
socioeconomic status
Reign of unpopular leader
VIOLENCE AND IDEOLOGY
Sensitivity to Injustice
Contacts government
officials regarding unjust
events
Boycotts organizations for
unjust acts
Speaks of what is “right”
and “wrong”
Gives resources to those
who have been treated
unfairly
Avoids business dealings
with people who are
viewed as unjust or
improper
Group
1535
Individual
Low Openness
Group
Group Feelings of
Superiority
Sudden break in
Advocates extreme policies
communication with
Rallies reference a need for
others who have different
radical change
views
Propaganda uses terms like
Owns very specific and
“destiny” or “ordained”
narrow collection of
Identifies specific targets
books or literature
necessary for change
Avoids debate about beliefs
tooccur
and values
Propaganda states that
Members of individual’s
group will change the
social circle are similar in
world
nationality or culture
Performs actions for group
without asking for
reasons
Organizational
Environmental
Sense Breaking
Economic Exploitation
Propaganda uses explicit
examples of
“inequalities”
Members are told they are
not respected by society
Members are required to
think of, or meditate on,
unfulfilled aspects of life
Repeatedly blames the
same organization for
problems
Members are forbidden to
associate with others
outside of the
organization
Infrequent changes in
dominant group
leadership
Dominant group withholds
resources
Few laws protecting
workers’ rights
Few educational
opportunities for lower
class
Extremely low wages for
lower class
1536 MUMFORD ET AL.
Table 2 Continued
Strong Group Culture
Performs rituals/
ceremonies that
emphasize beliefs and
values
Distributes materials that
articulate beliefs and
values
Attire reflects belief system
Recites oaths at meetings
Refers to values/principles
when making group
policy decisions
Rituals
Rituals and ceremonies
are considered “sacred”
or “holy”
Rituals kept secret from
outsiders
Use of symbolic items at
organizational meetings
Ceremonies or rituals for
successful completion of
missions
High degree of affect
during rituals
Strong Traditional Culture
Region has multiple
monuments to historical
leaders
Majority of region
participates in same
rituals for number of
years
Common hierarchy is
evident in non-work
venues (e.g., churches)
Members of population
engage in customs of
past
Current problems in legal
system are interpreted
using past framework
VIOLENCE AND IDEOLOGY
Sense of Asceticism
Practices self-disciplined
religious activities (e.g.,
fasting)
Mentions importance of
purity of lifestyle
Home and office are simple
and nonextravagant
Only purchases simple and
necessary foods
Leisure activities require
self-discipline (e.g.,
martial arts)
1537
1538
MUMFORD ET AL.
value crystallization among leaders was positively related to sensitivity to
injustice (r = .53) and exposure to other ideological leaders (r = .32), but not
to information distrust (r = .15). Identity expression through the group was
positively related to submergence of the individual (r = .47) and group sacrifice (r = .34), but not value-based recruitment (r = .08). At the organizational level, sense breaking was positively related to isolation of members
(r = .38), but not factionalization (r = .16) or the use of violence as a control
(r = .18). Similarly, at the environmental level, social conflict was positively
related to fear of social loss (r = .53), but not strong cultural traditions
(r = .23). Additional evidence pointing to the meaningfulness, or construct
validity, of these ratings was obtained by examining the cross-level correlations. Here, it was found that social conflict was positively related to group
rivalry (r = .40), but not peer group influence (r = .16); while isolation of
members was positively related to strong group boundaries (r = .56), but not
value-based recruitment (r = .11).
Criteria and Controls
The criterion information was drawn from either the prologue or epilogue
chapters of the qualitative analysis of group activities. These prologue or
epilogue chapters provide a summary of the social consequences of the
group’s activities. These summaries of the group’s social impacts were
reviewed by four judges—again, all doctoral students in industrial and organizational psychology—to assess the kind of violent behaviors manifested by
a group and effective dissemination of the group’s ideology. With regard to
violence, judges were asked to appraise (a) the amount of violence engaged in
by the group; (b) the severity of the group violence; (c) whether the targets of
the violence were other groups; (d) whether the targets of the violence were
institutions; and (e) whether the targets of the violence were cultural. With
regard to ideology, judges were asked to appraise (a) initiation of mass
movements; (b) effectiveness of member recruitment; (c) amount of change
involved in institutions; (d) amount of change induced in society; and (e)
number of positive contributions.
Appraisals of the violence and ideology criteria were based on 5-point
ratings made by judges following their review of the material presented in the
prologue or epilogue. It is of note that different judges were used to evaluate
this material than those who reviewed the predictors. Additionally, it should
be noted that judges were unaware of the study predictions, group selection
procedures, or group cell breakdown (i.e., violent and ideological; nonviolent
and ideological, and so forth), thereby limiting some of the potential biasing
effects arising from the knowledge of study intention. Moreover, the judges
VIOLENCE AND IDEOLOGY
1539
were exposed to an 8-hour training program prior to making the ratings
where (a) the nature of the rating scales was described; (b) judges practiced
making ratings using sample material; and (c) judges discussed discrepancies
to clarify the nature of the ratings. The average interrater agreement coefficient obtained for the violence ratings was .83, while the average interrater
agreement coefficient obtained for the ideology ratings was .81.
At the time judges were appraising group outcomes with regard to violence and ideology, requisite control variables were also obtained. These
controls examined either characteristics of the qualitative analysis or characteristics of the group. The use of these control variables allowed for the
statistical removal of effects that may have influenced ratings. For example,
by controlling for length of the materials rated, potential fatigue effects can
be removed from the observed results. Selection of control variables was
based on previous studies employing similar methodology (e.g., Mumford,
2006; Simonton, 1991).
The qualitative analysis controls included the following: (a) amount of
detail presented in the book; (b) strength of documentation provided by the
author; (c) evidence of firsthand reports (e.g., use of quoted speeches, communications); (d) author’s or authors’ origin in the same country as the
group; (e) whether the author’s or authors’ first language was or was not
English; (f) overall length of the book; (g) length of chapters selected for
predictor rating; and (h) length of chapters selected for criteria rating. The
group controls included time of group establishment, whether the group was
currently in existence, size of group at the time of formation, group’s origin
in an industrialized or non-industrialized country, and group’s social basis
(e.g., political, religious, social, military, business). These control variable
ratings produced interrater agreement coefficients above .90.
Analyses
To examine whether leadership, group process, organizational practices,
and environmental variables could be used to identify violent ideological
groups and account for violence and ideological outcomes, a two-stage analytical strategy was applied. Initially, a set of discriminant analyses was
conducted to identify the variables that accounted for the unique characteristics of violent and ideological groups. These discriminant analyses were
conducted within a given level of analysis (e.g., leader, group, organizational,
environmental) where all analyses were based on a comparison of four
groups: (a) violent ideological groups; (b) nonviolent ideological groups; (c)
violent nonideological groups; and (d) nonviolent nonideological groups.
The resulting discriminant functions provided dimensions summarizing the
1540
MUMFORD ET AL.
differences between the groups. Scores on these functions provided the basis
for the second set of analyses.
In the second set of analyses, scores on the violence and ideological
outcome measures were regressed on discriminant function scores. In these
regression analyses, scores on the qualitative analysis and group characteristics controls were entered as the first block of predictors to ensure that
potential artifacts were taken into account before examining the relationship
of discriminant function scores to violence and social outcomes. It is of note
that in these blocked regression analyses, a control variable was retained only
if it was significant beyond the .05 level.
Results
Discriminant Analyses
Table 3 presents the results obtained in the discriminant analysis examining the ability of leadership predictors to account for the four groups under
consideration. Of the three functions obtained in this analysis, two produced
significant discrimination of the four groups. The variables yielding sizable
loadings on the first function include sensitivity to injustice (r = .37), disregard for time (r = -.37), low openness (r = -.35), and sense of superiority
(r = -.32). Urgent feelings about injustice and the awareness of others implied
by these loadings led this function to be labeled social concern. As might be
expected, violent ideological groups (M = -.65) and particularly violent nonideological groups (M = -3.05) evidenced lower scores on this dimension
than did nonviolent ideological groups (M = 1.76) or nonviolent nonideological groups (M = 1.94).
Although violent ideological groups differed from violent nonideological
groups in that they evidenced somewhat greater social concern, violent ideological groups differed from other groups with respect to leader attributes in
terms of ideological extremism. This ideological extremism function was
defined by commitment to like-minded others (r = .54), reality distortion
(r = .43), value crystallization (r = .42), sensitivity to injustice (r = .40), selective interpretation of information (r = .36), and sense of superiority (r = .34).
As noted previously, violent ideological groups (M = 2.47) obtained far
higher scores on this dimension than did nonviolent groups—both ideological (M = -0.42) and nonideological (M = -0.78)—and violent nonideological groups (M = -1.27). Apparently, ideological violence is uniquely
linked to reality distortion and rationalizations associated with adoption of a
rigid belief and value structure. Cell means, and their respective comparisons
to the other group means, can be seen in Table 4.
VIOLENCE AND IDEOLOGY
1541
Table 3
Summary of Discriminant Analysis Results for Leadership, Group, Organization and Environment Predictors
Social Concern
(R = .90),
c2 = 210.65, p < .001
Disregard for time
(r = -.37)
Sensitivity to injustice
(r = .37)
Low openness
(r = -.35)
Sense of superiority
(r = -.32)
Sense of asceticism
(r = .26)
Compartmentalization
(r = -.22)
Oppositional Bonding
(R = .89),
c2 = 168.94, p < .001
Group rivalry (r = .59)
Negative mental
models (r = .55)
Strong group
boundaries (r = .40)
Group exclusivity
(r = .35)
Submergence of
individual (r = .26)
Leadership
Ideological Extremism
(R = .83), c2 = 98.10,
p < .001
Commitment to
like-minded others
(r = .54)
Reality distortion
(r = .43)
Value crystallization
(r = .42)
Sensitivity to injustice
(r = .40)
Selective interpretation
of information
(r = .36)
Sense of superiority
(r = .34)
Group
Ideological
Righteousness
(R = .73), c2 = 61.60,
p < .001
Group feelings of
superiority (r = .68)
Strong group values
(r = .56)
Victimization (r = .53)
Strong group affect
(r = .40)
Group sacrifice
(r = .32)
Ideological Exposure
(R = .50), c2 = 98.10,
p > .10
Sense of entitlement
(r = -.32)
Exposure to
ideological leaders
(r = .28)
Information distrust
(r = .22)
Symbolic Expression
(R = .34), c2 = 9.31, p
> .10
Group rivalry (r = .34)
Value-based
recruitment (r = .21)
Symbolic commitments
(r = .16)
1542
MUMFORD ET AL.
Table 3 Continued
Extensive socialization
(r = .25)
Violence as a Control
(R = .91),
c2 = 213.31, p < .001
Denigration of external
institutions (r = .50)
Violence as control
(r = .48)
Recruitment by
investment (r = .45)
Clan control (r = .45)
Punishment as control
(r = .39)
Isolation of members
(r = .36)
External conflict
(r = .30)
Social Conflict
(R = .70),
c2 = 100.17, p < .001
Multiple ideological
groups (r = .62)
Fragmentation
(r = .54)
Rivalry (r = .53)
Symbolic commitments
(r = .29)
Organization
Ideological
Indoctrination
(R = .81),
c2 = 100.65, p < .001
Sense breaking
(r = .42)
Sacrifice for
organization
(r = .33)
Sense making (r = .30)
Value-based
communication
(r = .26)
Training ideology
(r = .24)
Environment
Social Disruption
(R = .53), c2 = 54.98,
p < .05
Inequitable resource
distribution
(r = -.38)
Fear of social loss
(r = .33)
Multiple ideological
groups (r = .31)
Organizational
Framing (R = .64),
c2 = 33.30, p < .10
Training in ideology
(r = .41)
Isolation of members
(r = .29)
Value-based
communication
(r = -.26)
Denigration of external
institutions (r = -.22)
Institutional
Replacement
(R = .62), c2 = 22.67,
p >.10
Loss of institutions
(r = .30)
Corruption (r = -.29)
Inequitable resource
distribution (r = .25)
VIOLENCE AND IDEOLOGY
1543
Table 3 Continued
Social conflict (r = .47)
Social conflict (r = .30)
Fear of social loss
(r = .43)
Threats to tradition
(r = .41)
Economic
displacement
(r = .40)
Loss of social pattern
(r = .28)
Fragmentation
(r = .23)
Middle-class
marginalization
(r = .22)
Fragmentation
(r = -.16)
Note. R = canonical correlation; c2 = chi square for the discriminant function;
r = loading of variables on discriminant function.
The results obtained in the discriminant analysis conducted for the group
predictors are also presented in Table 3. The functions obtained in this
analysis produced significant discrimination of the four groups for two of the
three discriminant functions. The first function defined by group rivalry
(r = .59), negative mental models (r = .55), strong group boundaries (r = .40),
and group exclusivity (r = .35) was labeled oppositional bonding. Violent
groups, both ideological (M = 1.03) and nonideological (M = 2.58), received
higher scores on this dimension than did nonviolent groups, both ideological
(M = -1.79) and nonideological (M = -1.82). Apparently, “strong” groups
that disparage other groups give rise to violence.
It appears, however, that certain unique characteristics of ideological
groups may make them especially prone to violence. The second function
obtained in this analysis—defined by group feelings of superiority (r = .68),
strong group values (r = .56), victimization (r = .53), strong group affect
(r = .40), group sacrifice (r = .32), and symbolic commitments (r = .29)—was
labeled ideological righteousness. This sense of ideological righteousness
characterizes violent ideological groups (M = 1.60), but not violent nonideological groups (M = -1.04), nonviolent ideological groups (M = 0.21),
or nonviolent non-ideological groups (M = -0.78). Group means as well as
their respective comparisons to the other group means can also be seen in
Table 4.
Table 3 presents the results obtained in the discriminant analysis examining the organizational predictors. The three resulting functions were found
to produce significant discrimination of the four groups for all three of the
discriminant functions. The first function obtained in this analysis, labeled
violence as control, is defined by denigration of external institutions (r = .50),
use of violence as a control mechanism (r = .48), recruitment by investment
1544 MUMFORD ET AL.
Table 4
Tukey’s HSD Cell Mean Comparisons for Leadership, Group, Organization,
and Environment Level
Violent
Leadership level
Social concern
Ideological
extremism
Group level
Oppositional
bonding
Ideological
righteousness
Organization level
Violence as control
Ideological
indoctrination
Organizational
framing
Environment level
Social conflict
Social disruption
Nonviolent
Ideological
Nonideological
Ideological
Nonideological
-0.65b
2.47c
-3.05a
-1.27a
1.76c
-0.42b
1.94c
-0.78ab
1.03b
2.58a
-1.79c
-1.82c
1.60c
-1.04a
0.21b
-0.78a
2.31a
1.78c
1.86a
-1.97a
-1.43b
0.02b
-2.73c
0.16b
0.15b
0.20b
-1.29a
0.94c
1.17a
0.75b
-0.06b
-0.48a
0.35b
-0.72a
-1.46c
0.45b
Note. HSD = honestly significant difference. Means that do not share a subscript
differ significantly at p < .05.
(r = .45), clan control (r = .45), punishment as control (r = .39), isolation of
members (r = .36), and external conflict (r = .30). As might be expected,
violent groups, both ideological (M = 2.31) and nonideological (M = 1.86),
produced higher scores on this dimension than did nonviolent groups, both
ideological (M = -1.43) and nonideological (M = -2.73).
With respect to organizational-level variables, violent ideological groups
differed from other groups with respect to systematic attempts to inculcate a
new extremist mental model. Thus, the ideological indoctrination dimension
VIOLENCE AND IDEOLOGY
1545
defined by sense breaking (r = .42), sacrifice for the organization (r = .33),
sense making (r = .30), value-based communication (r = .26), and training
in ideology (r = .24) produced high scores for violent ideological groups
(M = 1.78), but not violent nonideological groups (M = -1.97), nonviolent ideological groups (M = 0.02), or nonviolent nonideological groups
(M = 0.16).
The third and final function obtained in this discriminant analysis was
rather complex. Training in ideology (r = .41) and isolation of members
(r = .29) produced positive loadings, while value-based communication
(r = -.26) and denigration of external institutions (r = -.22) produced negative loadings. Because inspection of the pattern matrix indicates that limited
clan control and limited idolization of leadership were also linked to scores
on this dimension, it seems plausible to argue that this dimension reflects an
organizational or institutional framing of issues, as opposed to a leaderbased framing. Nonviolent, nonideological groups (M = 0.94) obtained
higher scores on the organizational framing dimension than did nonviolent
ideological groups (M = -1.29), violent ideological groups (M = 0.15), and
violent non-ideological groups (M = 0.20). Cell means and relative comparisons to other cells are presented in Table 4.
The results that were obtained when the environmental-level variables
were used to discriminate these groups can be seen in Table 3. The significant
discrimination among all three functions in this analysis indicates that environmental conditions did shape the emergence of these groups. More specifically, in accordance with the observations of Moghaddam (2005) and Staub
(2004), social conflict appears to give rise to ideological groups, especially
violent ideological groups. Accordingly, the social conflict dimension, defined
by the presence of multiple ideological groups (r = .62), fragmentation in
society (r = .54), rivalry (r = .53), social conflict (r = .47), fear of social loss
(r = .43), threats to tradition (r = .41) and economic displacement (r = .40)
was associated with high scores for violent ideological groups (M = 1.17), but
not for violent nonideological groups (M = -0.06), nonviolent ideological
groups (M = 0.35), and nonviolent nonideological groups (M = -1.46).
In addition to social conflict, however, social disruption also appears to be
linked to the emergence of violent ideological groups. Social disruption, as
reflected in fear of social loss (r = .32), social conflict (r = .30), loss of social
pattern (r = .28), fragmentation (r = .23), and middle-class marginalization
(r = .22), as opposed to economic disruption or inequitable distribution of
resources (r = -.38), was associated with the emergence of violent ideological
groups (M = 0.75), but not nonviolent ideological groups (M = -0.72) or
violent nonideological groups (M = -0.48). The tendency of nonviolent, nonideological groups to emerge under these conditions (M = 0.45) may be
related to the role of ideology in reducing uncertainty.
1546 MUMFORD ET AL.
Regression Analyses
Table 5 presents the results that were obtained when the various violence
criteria were regressed on the discriminant function scores. Across criteria,
the origins of the group in social interaction (i.e., a control variable) was
found to be negatively related to the various violence criteria (β = −.36).
When discriminant function scores were added to social groups, along with
the other controls that were significantly related to a particular violence
criterion, it was found that the function scores resulted in significant
( p ⱕ .01) gains in production for all of the violence criteria under
consideration.
With regard to the amount and severity of violence, it was found that
social concern on the part of leaders tended to inhibit violence (b = -.37),
while institutional sanctioning of violence, as reflected in organizational use
of violence as a control, tended to contribute to violence (b = .34). It was also
found that the social conflict in the environment was positively related to the
amount of violence (b = .21) observed in these groups.
Social conflict and social disruption, conditions giving rise to ideological
groups, were also found to be related to group violence (b = .21) and cultural
violence (b = .31). Moreover, ideological extremism (b = .42), along with
oppositional bonding (b = .44) and organizational framing (b = -.29), were
found to be related to institutional violence.
With regard to the ideological criteria, a somewhat different pattern of
findings emerged. In considering the various controls across criteria, it
was found that the nature of the group—political (β = .31) or social
(β = .34 )—was positively related to criteria such as initiation of mass movements, changes in society, changes in institutions, and positive contributions
made. Apparently, social and political groups are especially likely to effect
social outcomes. These results can be seen in Table 5.
Even when these controls were taken into account, however, the discriminant functions were found to be effective predictors of three of the five
ideological criteria. With regard to recruitment, it was found that using
violence as a control (b = .43) was positively related to recruitment effectiveness, while ideological righteousness (b = -.48) apparently inhibited effective
recruitment. The ability of ideological groups to induce changes in institutions, however, was negatively related to the use of violence as a control (b = -.57), as well as social conflict (b = -.27), and social disruption
(b = -.24). Ideological extremism, on the other hand, was positively related
(b = .37) to institutional changes. Finally, the number of positive contributions made by the group was positively related to leader social concern
(b = .29), but negatively related to group oppositional bonding (b = -.38) and
ideological indoctrination (b = -.26). Apparently, using violence as a control
Table 5
Results of Regression Analyses Examining Violence Criteria
p
.34
.002
Severity of
violence
.35
.002
Group violence
—
—
Violence criteria
Amount of
violence
R
p
Rc2
pc
Social group
(b = .-.34)
.91
.001
.72
.001
Social group
(b = -.35)
.84
.001
.59
.001
—
.75
.001
.56
.001
Significant
discriminant functions
( p ⱕ .10)
1547
Social concern
(b = -.39)
Violence as control
(b = .33)
Social conflict
(b = .21)
Organizational
framing (b = .13)
Social concern
(b = -.35)
Violence as control
(b = .35)
Social concern
(b = -.48)
Social conflict
(b = .27)
VIOLENCE AND IDEOLOGY
R
Significant controls
( p ⱕ .10)
R
p
Institutional
violence
.38
.005
Cultural violence
.33
.009
Ideological criteria
Recruitment
effectiveness
—
—
.70
.001
Initiation of mass
movement
Significant controls
( p ⱕ .10)
R
p
Rc2
pc
Social group
(b = -.33)
Business group
(b = -.32)
.72
.001
.37
.001
Biographer from
country (b = -.21)
Military group
(b = .23)
.71
.001
.40
.001
—
.52
.029
Political group
(b = .43)
Industrialized country
(b = .40)
Social group (b = .35)
.73
.001
.05
.659
Significant
discriminant functions
( p ⱕ .10)
Ideological extremism
(b = .42)
Oppositional bonding
(b = .44)
Organizational
framing (b = -.29)
Social conflict
(b = .22)
Ideological
righteousness
(b = -.48)
Violence as a control
(b = .43)
—
1548 MUMFORD ET AL.
Table 5 Continued
Changes in
institutions
.35
.02
Social group (b = .31)
.001
.32
.001
.59
.008
.13
.271
.84
.001
.58
.001
Political group
(b = .18)
Changes in
society
.47
.001
Number of
positive
contributions
.36
.001
Political group
(b = .29)
Business group
(b = -.12)
Biographer from
country (b = -.19)
Amount of criterion
information
(b = .26)
Social group (b = .35)
Ideological extremism
(b = .37)
Violence as control
(b = -.57)
Social conflict
(b = -.27)
Social disruption
(b = -.24)
—
Note. R = multiple correlation; p = significance level; Rc2 = change in variance accounted for; pc = significance of change.
1549
Social concern
(b = .29)
Oppositional bonding
(b = -.38)
Ideological
indoctrination
(b = -.26)
VIOLENCE AND IDEOLOGY
.67
1550 MUMFORD ET AL.
is useful for attracting new members, but limits the ability of the group to
achieve certain outcomes. Moreover, it appears that conflict and disruption
in the environment may inhibit change, while a strong commitment to an
extreme ideology helps to bring about change, at least in some forms.
Discussion
Before turning to the broader conclusions flowing from the present study,
certain limitations should be noted. To begin with, the present study was
intended to contrast violent ideological groups with other relevant comparison groups. Although this kind of design is useful in determining the unique
characteristics of these groups, it is not a design that makes it possible to
draw conclusions about the natural rate at which groups falling into these
categories emerge in a society (Strange & Mumford, 2005). Similarly, the
design and sampling of the study produced cell sizes that precluded withingroup analyses. Thus, it should be noted that future studies focusing on
specific groups (e.g., violent, nonideological groups) may be necessary to gain
a full understanding of the mechanisms operating within those unique
groups.
Along related lines, our intention in the present study was to identify the
dimensions, or attributes, that distinguish violent ideological groups from
other relevant comparison groups. Accordingly, we examined a number of
potential predictors; predictors operating at different levels of analysis and
reflecting different theoretical models concerning the sources of ideological
violence (e.g., Crayton, 1983; Moghaddam, 2005; Staub, 2003, 2004). Nonetheless, it seems likely that some potential markers of ideological violence
exist that were not examined in the present study.
Finally, these predictors were derived from a content analysis of groups
that had been subject to prior qualitative analysis. The number of groups
examined was similar to that applied in prior studies involving a quantitative
analysis of qualitative data (e.g., O’Connor et al., 1995). However, it is also
true that the groups for which requisite qualitative data were available represent historically noteworthy, prototypical exemplars of a given type of
group. Thus, some caution is called for in extending our findings to less
extreme groups. It should also be noted that the use of these extreme groups
may have resulted in some form of range restriction among a few of the
criteria examined. Such range restriction, however, is likely to have resulted
in more conservative, rather than inflated estimates (Greer, Dunlap, Hunter,
& Berman, 2006). Nonetheless, this is a notable study limitation.
Even bearing these limitations in mind, we believe that the results
obtained in the present study do lead to some noteworthy conclusions about
VIOLENCE AND IDEOLOGY
1551
the origins of violent ideological groups and the consequences of violent
ideology. Most current theories of ideological violence stress the multilevel
nature of the phenomenon (e.g., Moghaddam, 2005; Post et al., 2002). In
fact, the results obtained in the discriminant analysis indicate that variables
operating at the leadership, group, organizational, and environmental level
all contribute to ideological violence on the part of groups.
Most theories examining the environmental conditions giving rise to ideological violence stress the importance of social disruption and social conflict.
Social disruption, of course, obviates extant mental models while promoting
the development of new, potentially violent ideologies based on the apparent
failure of past models (Strange & Mumford, 2005). Social conflict involves
the threat held to give rise to violence (Cordes, 2001).
The results obtained in the present study indicate that not only are violent
ideological groups likely to emerge under conditions of social conflict and
social disruption, but the amounts of social conflict and social disruption
evident in the group’s environment are relatively powerful predictors of
violence, even when relevant controls are taken into account. Moreover, the
regression results indicate that social conflict may serve to inhibit the inducement of change: change that may be necessary to alleviate some of the tension
building in these types of environments. This, in turn, may fuel growing
resentment and push individuals into the extreme actions associated with
violent ideology.
Although these findings with regard to the environmental conditions are
of some importance, the question remains as to whether certain unique
characteristics of groups, their mode of organization, and their leadership act
to promote the acquisition and dissemination of violent ideologies. Consistent with the argument that ideology involves the creation and dissemination
of shared mental models, the organizational-level analysis produced an ideological indoctrination dimension defined by sense breaking and sense making
(Pratt, 2000). Violent ideological groups were apparently more likely to
engage in this indoctrination than were other groups. However, the results
obtained in the regression analysis indicate that the sense breaking and sense
making associated with indoctrination did not necessarily lead to violence
unless institutional sanctioning of violence occurred as reflected in the use of
violence as a control mechanism.
These findings seem to be consistent with the results obtained for the
group and leadership predictors. Violent ideological groups, in comparison
with other groups, were characterized by an expression of ideological righteousness at the group level, and expression of ideological extremism on the
part of the leaders. Ideological righteousness, of course, through victimization and a sense of superiority, acts as a potential justification mechanism
for violence (Moghaddam, 2005). Ideological righteousness may, moreover,
1552 MUMFORD ET AL.
provide a logic justifying adoption of a new, potentially violent mental model
being advocated by extremist leaders. Ideological righteousness, coupled
with indoctrination or sense breaking, causes people to be willing to comply
with institutional policies sanctioning acts of violence: policies that may arise
as part and parcel of the extremist positions commonly advocated by the
leaders of violent ideological groups.
Taken as a whole, it appears that violent ideological groups are characterized by a unique syndrome of characteristics. More specifically, these
groups arise under conditions of social conflict and social disruption, with
violence emerging from institutional policies; potentially policies linked to
extremism. People’s willingness to execute these policies, however, depends
on both indoctrination and creation of a sense of righteousness. This syndrome of characteristics giving rise to ideological violence, moreover, is
distinct from the syndrome of characteristics that were evidenced by violent
non-ideological groups. Violent nonideological groups were characterized by
a lack of social concern and oppositional bonding, rather than ideological
extremism, ideological righteousness, and ideological indoctrination. In fact,
the only attribute that these violent ideological and violent nonideological
groups shared was institutional sanctioning of violence through the use of
violence as a control mechanism. Indeed, social conflict and social disruption
did not seem to be especially likely to give rise to violent nonideological
groups. Thus, ideological violence may represent a distinct phenomenon
from group violence in general.
Along similar lines is the finding that violent ideological groups apparently
did not share a great deal in common with nonviolent ideological groups,
again illustrating the unique nature of violent ideological groups. It seems that
nonviolent ideological groups arise from attempts to replace failed institutions
stressing the social value of the institution being built through symbolism and
organizational framing. These results are consistent with Moghaddam’s
(2005) argument that ideology does not necessarily lead to violence.
It should be noted, however, that although violent ideological and nonviolent ideological groups were differentiated by a unique set of characteristics,
the present study illustrated an important lack of distinction among groups;
namely, between ideological groups and nonideological groups broadly.
Based on the pattern of group means, ideological and nonideological groups
did not appear to distinguish themselves substantially from one another. This
point highlights the importance of examining ideology and violence in combination when attempting to research and understand extreme violence.
These findings and the other findings discussed previously are of some
importance for practical as well as theoretical reasons. At a practical level,
they suggest that it may not be desirable to apply models developed to
account for criminal activity in attempts to understand violent ideological
VIOLENCE AND IDEOLOGY
1553
groups, such as those involved in terrorism. In point of fact, applying such
inappropriate models may result in misguided efforts to limit the violence and
actions of criminal groups.
The results also bring to the fore important theoretical questions regarding current conceptualizations of ideological violence. Specifically, Moghaddam’s (2005) theoretical metaphor of an ever-narrowing staircase contends
that as individuals proceed up the pathway, their options become fewer until
a select few ultimately choose some form of ideological violence. Thus, in
applying this metaphor to the current study, a natural conclusion may be that
ideological violence resides at the top of the staircase, while nonideological
violence resides along lower portions of the staircase.
The results of the present effort, however, may be taken to indicate that
violent ideological groups should not be viewed as the final stage of a continuum, but rather as a unique and distinct entity. This is an entity distinct
even from those who share strong ideologies. Although the design and nature
of the present effort makes such a contention speculative, nonetheless it
appears to be noteworthy. Clearly, such theoretical investigation stands as an
important and necessary avenue for future research.
The findings obtained in the present study also bring to the fore theoretical issues related to how righteousness (i.e., sense of group superiority),
indoctrination, and sense breaking, vis-à-vis leader extremism, make possible
the conditions for violence on the part of ideological groups. Put more
directly, we need studies examining mediators and moderators of ideological
violence, especially at the level of the leader, the group, and the organization.
In sum, the results of the present research suggest that violent ideological
groups are characterized by a unique syndrome of characteristics. Such characteristics are derived from and exist at multiple levels of analysis, including
leadership influences, group processes, organizational sanctions, and environmental conditions. Such characteristics, moreover, appear to be useful in
understanding both violent and ideological outcomes. The results also underscore the importance of considering both violent and ideological influences—
as well as their combination—when attempting to understand such relevant
topics as global terrorism and organized crime. It is our hope that the present
study has contributed to this understanding and that it stimulates future
efforts along these lines.
References
(*Used in study coding and analysis.)
Aberson, C. L., Healy, M., & Romero, V. (2000). In-group bias and selfesteem: A meta-analysis. Personality and Social Psychology Review, 4,
157–173.
1554 MUMFORD ET AL.
*Andreyev, C. (1987). Viasov and the Russian Liberation Movement. New
York: Press Syndicate of the University of Cambridge.
*Arunachalam, J., & Landwehr, B. (Eds.). (2003). Structuring a movement
and spreading it on: History of growth of the Working Women’s Forum
(India), 1978–2003. London: IKO-Verlag für Interkulturelle Kommunikation.
Atran, S. (2003). Genesis of suicide terrorism. Science, 229, 1534–1539.
*Barger, W. K., & Reza, E. M. (1994). The Farm Labor Movement in the
Midwest: Social change and adaptation among migrant farmworkers.
Austin, TX: University of Texas Press.
*Bayless, P. (2002). The YMCA at 150: A history of the YMCA of Greater
New York, 1852–2002. New York: YMCA of Greater New York.
*Beiswinger, G. (1985). One to one: The story of the Big Brothers/Big
Sisters movement in America. Philadelphia: Big Brothers/Big Sisters of
America.
*Black, M. (1992). The cause for our times: Oxfam the first 50 years. Oxford,
UK: Oxford University Press.
Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.
*Blake, D. (2003). Shower Posse: The most notorious criminal organization.
New York: Diamond.
*Blazak, R. (2001). White boys to terrorist men: Target recruitment of Nazi
skinheads. American Behavioral Scientist, 44, 982–1000.
*Bortolotti, D. (2004). Hope in hell: Inside the world of Doctors Without
Borders. New York: Firefly Books.
*Bouvard, M. G. (1994). Revolutionizing motherhood: The mothers of the
Plaza de Mayo. Wilmington, DE: Scholarly Resources.
*Bowden, M. (2001). Killing Pablo: The hunt for the world’s greatest outlaw.
New York: Penguin.
*Brodsky, A. (2003). With all our strength: The Revolutionary Association of
the Women of Afghanistan. New York: Routledge.
Burdman, D. (2003). Education, indoctrination, and incitement: Palestinian
children on their ways to martyrdom. Terrorism and Political Violence, 15,
96–123.
Byman, D. (1998). The logic of ethnic terrorism. Studies in Conflict and
Terrorism, 21, 149–169.
Cameron, J. E., Duck, J. M., Terry, D. J., & Lalonde, R. N. (2005). Perceptions of self and group in the context of a threatened national identity: A
field study. Group Processes and Intergroup Relations, 8, 73–88.
*Chalmers, D. (2000). Hooded Americanism: The history of the Ku Klux Klan.
Durham, NC: Duke University Press.
*Charles, J. (1993). Service clubs in American society: Rotary, Kiwanis, and
Lions. Chicago: University of Illinois Press.
VIOLENCE AND IDEOLOGY
1555
*Chen, J. (1971). The May Fourth Movement in Shanghai. Leiden, The Netherlands: E. J. Brill.
*Chespesiuk, R. (2003). The bullet and the bribe. Westport, CT: Praeger.
*Chu, Y. (2000). The Triads as business. London: Routledge.
*Cirtauras, A. (1997). The Polish Solidarity Movement. London: Routledge.
Combs, C. C. (2003). Terrorism in the 21st century (3rd ed.). Upper Saddle
River, NJ: Pearson Education.
Cordes, B. (2001). When terrorists do the talking: Reflections on terrorist
literature. In D. C. Rapoport (Ed.), Inside terrorist organizations (pp.
150–171). Portland, OR: Frank Cass Publishers.
Crayton, J. W. (1983). Terrorism and the psychology of the self. In L. Z.
Freedman & Y. Alexander (Eds.), Perspectives on terrorism (pp. 33–44).
Wilmington, DE: Scholarly Resources.
Crenshaw, M. (1988). The subjective reality of the terrorist: Ideological and
psychological factors in terrorism. In R. Slater & M. Stohl (Eds.), Current
perspectives on international terrorism (pp. 12–46). New York: St. Martin’s.
*Cubert, H. M. (1997). The PFLP’s changing role in the Middle East.
London: Frank Cass & Co.
*DeLan, F. (1984). Trial of blood: The Tison Gang. Victoria, TX: CrossRoads.
*Donham, D. (1999). Marxist modern: An ethnographic history of the Ethiopian revolution. Berkeley, CA: University of California Press.
Ehrlich, P. R., & Liu, J. (2002). Some roots of terrorism. Population and
Environment: A Journal of Interdisciplinary Studies, 24, 183–192.
*English, T. J. (1995). Born to kill. New York: Avon.
*Esfandiari, H. (1997). Reconstructed lives: Women and Iran’s Islamic Revolution. Washington DC: Woodrow Wilson Center Press.
*Farrell, W. (1990). Blood and rage: The story of the Japanese Red Army.
Lexington, MA: Lexington Books.
Fleiss, J. L., & Shrout, P. E. (1979). Intraclass correlations: Uses in assessing
rater reliability. Psychological Bulleting, 86, 420–428.
*Foner, P. S. (1970). The Black Panthers speak. Cambridge, MA: Da Capo.
Friedland, R. (2001). Religious nationalism and the problem of collective
representation. Annual Review of Sociology, 27, 125–152.
*Friedman, R. (2000). Red Mafia: How the Russian mob has invaded America.
New York: Burkley Publishing Group.
Gerstenfeld, P. B., Grant, D. R., & Chiang, C. P. (2003). Hate online: A
content analysis of extremist Internet sites. Analyses of Social Issues and
Public Policy, 3, 29–44.
Gessner, T. L., O’Connor, J., Clifton, T. C., Connelly, M. S., & Mumford,
M. D. (1993). Situational influences on destructive acts. Current Psychology: Developmental, Learning, Personality, Social, 13, 303–325.
1556
MUMFORD ET AL.
*Giri, A. (2002). Building in the margins of shacks: The vision and projects of
Habitat for Humanity. New Delhi, India: Orient Longman.
Gordijn, E. H., Wigboldus, D., & Yzerbyt, V. (2001). Emotional consequences
of categorizing victims of negative out-group behavior as in-group or
out-group. Group Processes and Intergroup Relations, 4, 317–326.
*Gorri, G. (1999). The shining path: A history of the millenarian war in Peru.
Chapel Hill, NC: University of North Carolina Press.
Greer, T., Dunlap, W., Hunter, S. T., & Berman, M. (2006). The effects of
skew on internal consistency. Journal of Applied Psychology, 91, 1351–
1358.
*Guaratna, R. (2003). Inside Al Qaeda: Global network of terror. New York:
Berkley Publishing Group.
*Gunst, L. (1995). Born fi’ dead: A journey through the Jamaican Posse
underworld. New York: Henry Hold & Co.
*Gutman, B. (1993). The history of NCAA basketball. Avenel, NJ: Crescent.
*Hamilton, R. (1998). Hellbent for enlightenment: Unmasking sex, power, and
death with a notorious master. Ashland, OR: White Cloud Press.
*Hamm, M. (2002). In bad company. Boston: Northeastern University Press.
Harré, R. (2004). The social construction of terrorism. In F. M. Moghaddam
& A. J. Marsella (Eds.), Understanding terrorism: Psychological
roots, consequences, and interventions (pp. 101–121). Washington, DC:
American Psychological Association.
Haslam, S. A., Turner, J. C., Oakes, P. J., Eggins, R. A., Nolan, M., &
Tweedie, J. (1998). When do stereotypes become really consensual? Investigating the group-based dynamics of the consensualization process.
European Journal of Social Psychology, 28, 755–776.
*Haug, H. (1993). Humanity for all: The International Red Cross and Red
Crescent Movement. Bern, Germany: Paul Haupt Publishers.
Henry, P. J., Sidanius, J., Levin, S., & Pratto, F. (2002). Social dominance
orientation, authoritarianism, religious identification, and support for terrorism in the Middle East (Working Paper #186). New York: Russell Sage
Foundation.
Hewstone, M., Rubin, M., & Willis, H. (2002). Intergroup bias. Annual
Review of Psychology, 53, 575–604.
Hoffman, B. (1999). The mind of the terrorist: Perspectives from social
psychology. Psychiatric Annuals, 29, 337–340.
Hogg, M. A. (2003). Social identity. In M. R. Leary & J. P. Tangney (Eds.),
Handbook of self and identity (pp. 462–479). New York: Guilford.
*Humphreys, R. (2001). Poor relief and charity, 1869-1945: The London
Charity Organization Society. New York: Palgrave.
*Jones, D., & Rivers, T. (2002). Soul Crew: The inside story of Britain’s most
notorious hooligan gang. Bury, UK: Milo Books.
VIOLENCE AND IDEOLOGY
1557
Jordan, J., & Boix, L. (2004). Al-Qaeda and western Islam. Terrorism and
Political Violence, 16, 1–17.
*Kaplan, D., & Dubro, A. (2003). Yakuza: Japan’s criminal underworld.
Berkeley, CA: University of California Press.
*Kassimeris, G. (2001). Europe’s last Red terrorists. New York: New York
University Press.
Katz, D., & Kahn, R. (1978). The social psychology of organization (2nd ed.).
New York: John Wiley & Sons.
*Kavieff, P. (2000). The Purple Gang: Organized crime in Detroit, 1910–1945.
Fort Lee, NJ: Barricade Books.
*Keiser, R.L. (2002). The Vice Lords: Warriors in the streets. Belmont, CA:
Wadsworth/Thomson Learning.
*Kenny, K. (1998). Making sense of the Molly Maguires. New York: Oxford
University Press.
*La Botz, D. (2001). Made in Indonesia: Indonesian workers since Suharto.
Cambridge, MA: South End Press.
*Lagermann, E. C. (1983). Private power for the public good: A history of the
Carnegie Foundation for the Advancement of Teaching. New York: College
Entrance Examination Board.
Lee, E., & Leets, L. (2002). Persuasive storytelling by hate groups online:
Examining its effects on adolescents. American Behavioral Scientist, 45,
927–957.
*Levine, D. (1991). David Raziel, the man and the legend: The birth of the
Irgun Zvai Leumi, a Jewish liberation movement. Jerusalem: Gefen Publishing House.
*Lia, B. (1998). The Society of the Muslim Brothers in Egypt: The rise of an
Islamic mass movement, 1928–1942. Reading, UK: Garnet.
*Maathai, W. (2004). The Green Belt Movement: Sharing the approach and
the experience. New York: Lantern Books.
*Magyar, K., & Conteh-Morgan, E. (Eds.). (1998). Peacekeeping in Africa:
ECOMOG in Liberia. Basingstoke, UK: Macmillan.
*Martin, B. (1996). The Shanghai Green Gang: Politics and organized crime,
1919–1937. Berkeley, CA: University of California Press.
*McCormick, G. (1993). Sharp dressed men: Peru’s Tupac Amaru Revolutionary Movement. Santa Monica, CA: Rand.
McCormick, G. H. (2003). Terror decision making. Annual Review of Political Science, 6, 473–507.
*Meredith, M. (2002). Our votes, our guns: Robert Mugabe and the tragedy of
Zimbabwe. New York: Public Affairs.
Moghaddam, F. M. (2005). The staircase to terrorism: A psychological
exploration. American Psychologist, 60, 161–169.
1558 MUMFORD ET AL.
*Moloney, E. (2002). A secret history of the IRA. New York: W. W. Norton
& Co.
Monroe, K. R., Hankin, J., & Van Vechten, R. B. (2000). The psychological
foundations of identity politics. Annual Review of Political Science, 3,
419–447.
*Morell, L. (2003). The Mafia of Havana: The Cuban Cosa Nostra. New
York: Public Affairs.
Mumford, M. D. (2006). Pathways to outstanding leadership: A comparative
analysis of charismatic, ideological, and pragmatic leaders. Mahwah, NJ:
Lawrence Erlbaum.
Mumford, M. D., Costanza, D. P., Connelly, M. S., & Johnson, J. F. (1996).
Item-generation procedures and background data scales: Implications for
construct and criterion-related validity. Personnel Psychology, 49, 361–
398.
Murray, S. K., & Cowden, J. A. (1999). The role of “enemy images” and
ideology in elite belief systems. International Studies Quarterly, 43, 455–
481.
Navarrete, C. D., Kurzban, R., Fessler, D. M. T., & Kirkpatrick, L. A.
(2004). Anxiety and intergroup bias: Terror management or coalitional
psychology? Group Processes and Intergroup Relations, 7, 370–397.
*Noorani, A. (2002). Islam and Jihad. London: Zed Books.
ɺɺ sse, A. (1998). Muslim Palestine: The ideology of Hamas. London:
*Nυ
Routledge Curzon.
O’Boyle, G. (2002). Theories of justification and political violence: Examples
from four groups. Terrorism and Political Violence, 14, 23–46.
O’Connor, J. A., Mumford, M. D., Clifton, T. C., Gessner, T. E., & Connelly, M. S. (1995). Charismatic leaders and destructiveness: An historiometric study. Leadership Quarterly, 6, 529–555.
Paladino, M. P., Leyens, J. P., Rodriguez, R., Rodriguez, A., Gaunt, R., &
Demoulin, S. (2002). Differential association of uniquely and nonuniquely human emotions with the in-group and the out-group. Group
Processes and Intergroup Relations, 5, 105–117.
Pitcavage, M. (2001). Camouflage and conspiracy: The militia movement
from Ruby Ridge to Y2K. American Behavioral Scientist, 44, 957–981.
Post, J. M. (2005). When hatred is bred in the bone: Psycho-cultural foundations of contemporary terrorism. Political Psychology, 26, 615–636.
Post, J. M., Ruby, K. G., & Shaw, E. D. (2002). The radical group in context:
I. An integrated framework for the analysis of group risk for terrorism.
Studies in Conflict and Terrorism, 25, 73–100.
Pratt, M. G. (2000). The good, the bad, and the ambivalent: Managing
identification among Amway distributors. Administrative Science Quarterly, 45, 456–493.
VIOLENCE AND IDEOLOGY
1559
*Reader, I. (2000). Religious violence in contemporary Japan: The case of Aum
Shinrikyō. Honolulu, HI: University of Hawaii Press.
*Reppetto, T. (2004). American Mafia: A history of its rise to power. New
York: Henry Holt and Company.
*Ridley, J. (1999). The Freemasons: A history of the world’s most powerful
secret society. New York: Arcade Publishing.
*Robb, P. (Ed.). (1993). Dalit Movements and the meanings of labour in India.
Delhi, India: Oxford University Press.
*Robertson, F. (1977). Triangle of death: The inside story of the Triads—the
Chinese Mafia. London: Routledge.
*Robinson, C. (1999). Tradition and liberation: The Hindu tradition in the
Indian Women’s Movement. New York: St. Martin’s Press.
*Rose, K. (1992). Where women are leaders: The SEWA Movement in India.
London: Zed Books.
*Rosenbraugh, C. (2004). Burning rage of a dying planet: Speaking for the
Earth Liberation Front. New York: Lantern Books.
*Rumer, T. (1990). The American Legion: An official history, 1919–1989. New
York: M. Evans & Co.
*Russo, G. (2001). The outfit: The role of Chicago’s underworld in the shaping
of modern America. New York: Bloomsbury.
*Saad-Ghorayeb, A. (2002). Hizbu’llah: Politics religion. London: Pluto.
*Sadoff, M. (1990). Americans get MADD! Irving, TX: Mothers Against
Drunk Driving.
*Sanchez, R. (2000). My bloody life: The making of a Latin King. Chicago:
Chicago Review Press.
Schwartz, L. L., & Kaslow, F. W. (2001). The cult phenomenon: A turn-ofthe-century update. American Journal of Family Therapy, 29, 13–22.
*Senn, A. (1999). Power, politics, and the Olympic Games: A history of
the power brokers, events, and controversies that shaped the Games.
Champaign, IL: Human Kinetics.
Shamir, B., House, R. J., & Arthur, M. B. (1993). The motivational effects of
charismatic leadership: A self-concept-based theory. Organizational
Science, 4, 577–594.
Sharpe, T. T. (2000). The Christian identity movement: Ideology of domestic
terrorism. Journal of Black Studies, 30, 604–623.
Simonton, D. K. (1991). Psychology, science, and history: An introduction to
historiometry. New Haven, Connecticut: Yale University Press.
Simonton, D. K. (1999). Talent and its development: An emergenic and
epigenetic model. Psychological Review, 106, 435–457.
Spoor, J. R., & Kelly, J. R. (2004). The evolutionary significance of affect in
groups: Communication and group bonding. Group Processes and Intergroup Relations, 7, 398–412.
1560 MUMFORD ET AL.
Staub, E. (1996). Cultural–societal roots of violence: The examples of genocidal violence and of contemporary youth violence in the United States.
American Psychologist, 51, 117–132.
Staub, E. (2003). The psychology of good and evil: Why children, adults, and
groups help and harm others. New York: Cambridge University Press.
Staub, E. (2003). Notes on cultures of violence, cultures of caring and
peace, and the fulfillment of basic human needs. Political Psychology, 24,
1–21.
Staub, E. (2004). Understanding and responding to group violence: Genocide, mass killing, and terrorism. In F. M. Moghaddam & A. J. Marsella
(Eds.), Understanding terrorism: Psychological roots, consequences, and
interventions (pp. 151–168). Washington, DC: American Psychological
Association.
*Staving, W. (1999). The world of Tupac Amaru: Conflict, community, and
identity in colonial Peru. Lincoln, NE: University of Nebraska Press.
Sternberg, R. J. (2003). A duplex theory of hate: Development and application to terrorism, massacres, and genocide. Review of General Psychology,
7, 299–328.
Strange, J. M., & Mumford, M. D. (2002). The origins of vision: Charismatic
versus ideological leadership. Leadership Quarterly, 13, 343–377.
Strange, J. M., & Mumford, M. D. (2005). The origins of vision: Effects of
reflection, models, and analysis. Leadership Quarterly, 16, 121–148.
Stroessner, S. J., Mackie, D. M., & Michalsen, V. (2005). Positive mood and
perceptions of variability within and between groups. Group Processes
and Intergroup Relations, 8, 5–25.
*Sugarman, J. (1992). NRA National Rifle Association: Money, firepower, and
fear. Washington, D.C.: National Press Books.
*Talal, B. (1999). Rethinking an NGO: Development, donors, and civil society
in Jordan. London: I. B. Tauris.
*Thompson, H. (1995). Hell’s Angels. New York: Ballantine.
Thompson, J. A., & Bunderson, J. S. (2003). Violations of principle: Ideological currency in the psychological contract. Academy of Management
Review, 28, 571–586.
Tindale, R. S., & Kameda, T. (2000). “Social sharedness” as a unifying theme
for information processing in groups. Group Processes and Intergroup
Relations, 3, 123–140.
*Underwood, J. (2003). More than a pink Cadillac: Mary Kay Inc.’s nine
leadership keys to success. New York: McGraw-Hill.
*Urban, W. (2000). Gender, race, and the National Education Association:
Professionalism and its limits. New York: Routledge Falmer.
Vigil, J. D. (2003). Urban violence and street gangs. Annual Review of Anthropology, 32, 225–242.
VIOLENCE AND IDEOLOGY
1561
*Walker, S. (1999). In defense of American liberties: A history of the ACLU.
Carbondale, IL: Southern Illinois University Press.
*Wallace, S. L. (1962). Friends of the library. Chicago: American Library
Association.
*Weekley, K. (2001). The Communist party of the Philippines, 1968–1993: A
story of its theory and practice. Quezon City, Philippines: University of the
Philippines Press.
*Wickens, P. L. (1978). The Industrial and Commercial Workers’ Union of
Africa. Cape Town, South Africa: Oxford University Press.
*Wilkins, D., & Wilmer, F. (Eds.). (2003). The globalization of contentious
politics: The Amazonian Indigenous Rights Movement. New York: Routledge.
*Winston, D. (1999). Red-hot and righteous: The urban religion of the Salvation Army. Cambridge, MA: Harvard University Press.
*Wolfers, M. (1976). Politics in the organization of African unity. Norwich,
UK: Methuen & Co.
Wood, W. (2000). Attitude change: Persuasion and social influence. Annual
Review of Psychology, 51, 539–570.
*Woodworth, P. (2001). Dirty war, clean hands. New Haven, CT: Yale University Press.
*Wynne, B. (1968). Angels on Runway Zero 7: The story of the East African
Flying Doctor Service. London: Souvenir Press.
Yorges, S. L., Weiss, H. M., & Strickland, O. J. (1999). The effect of leader
outcomes on influence, attributions, and perceptions of charisma. Journal
of Applied Psychology, 84, 428–436.
Zawodny, J. K. (1983). Infrastructure of terrorist organizations. In L. Z.
Freedman & Y. Alexander (Eds.), Perspectives on terrorism (pp. 61–70).
Wilmington, DE: Scholarly Resources.