Deciphering Deadly Minds in Their Native Language The Operational Codes and Formation Patterns of Militant Organizations in the Middle East and North Africa

Sercan Canbolat

Introduction

There is no firm scholarly consensus over the definition of terrorism within the literature on terrorist violence. Some scholars have argued that terrorism has a random and/or indiscriminate nature, in order to cause fear in individual members of a target audience (Kalyvas 2004). Others have noted that terrorism might be of a highly discriminate nature or a mixture of both discriminate and random violence (Crenshaw 1981). Another issue with a comprehensive definition of terrorism is the identity of the terrorist actor and/or perpetrator. Terrorism is aptly identified as a tactic that can be utilized by both states and a variety of nonstate actors (Hoffman 2006). The focus here is on the latter, violent nonstate actors (VNSAs) who employ terrorism for political purposes, such as the intimidation of a public audience (Sandler and Enders 2007) by the strategic targeting of civilians and/or governmental actors (Stanton 2013).

This definition comports with Crenshaw’s (1981) assertion that terrorism can have both a discriminate and indiscriminate outcome, with special emphasis on the conditions of intentional and strategic violence. However, the study of terrorist violence necessitates not only the study of specific acts of terrorism but also the study of the social context or milieu in which such acts are made. Research on terrorist violence should consequently be familiar with general psychological principles of terrorism to account for both its causes and effects, and also with the psychological backgrounds of influential terrorist leaders such as the ISIS ringleader Abu Bakr al-Baghdadi. The social context and leadership dynamics include the tendency for VNSAs usually to break down and/or split apart, with new groups emerging from the ranks of existing organizations.

Many terrorist groups afflicting the Middle East and North Africa (MENA) region from Syria to Iraq and from Afghanistan to Palestine have splintered and proliferated in the face of altering political and strategic conditions. These emergent “radical flanks” further complicate the landscape of the conflict-ridden MENA region (Haines 2013). For example, in their study on the Palestinian radical flanks’ response to the Israeli settlement movement, Krause and Eiran (2018) argue that some radical flanks succeed in turning the state-centric norms/constrains on territorial revisionism and war into advantages on the ground. These novel splinter groups and their leaders often have very different trajectories than their parent organizations such as Al-Qaeda (AQ). Some last for a long time such as AQ in Yemen; others quickly fall apart or sometimes morph into a more radicalized character, e.g., AQ in Iraq. In their counter-terrorism strategies, states generally pursue a divide and conquer strategy against VNSAs, aiming to splinter terrorist groups, hamper their organizational networks, and compel existing members to relinquish terrorist violence.

Strikingly, there has been a dearth of systematic analysis of how in-group fragmentation impacts the trajectories of terrorist organizations and how to deal with the subsequent splinter groups that devolve in different ways (Perkoski 2019). This chapter analyzes the variation among certain parent organizations versus radical flanks in the MENA, which reasserted themselves in the political realm in the aftermath of Arab uprisings. I focus on the following militant groups: (1) AQ central led by Zawahiri and the Muslim Brotherhood (MB) under Badie’s leadership in Egypt as parent groups; (2) ISIS in Syria and Iraq plus al-Nusra (also known as Hayat Tahrir al-Sham or simply HTS) in Syria as AQ’s splinter groups; (3) Hamas and Islamic Jihad in the Gaza Strip as the MB’s radical flanks. This study focuses on the role of individual leaders of certain VNSAs in the MENA region and poses the following research question: How do the operational codes of terrorist leaders of parent groups and radical flanks affect the lethality of terrorist-group violence?

This chapter is composed of five sections plus an Appendix. First, an overview of the literature on VNSA’s formation and propensities for violent behavior with special emphasis on leadership factors is presented. The second part expounds upon the theoretical and methodological framework of this research and introduces operational code analysis (OCA) as an at-a-distance leadership assessment tool. Third, this study’s proposed research design to study the belief systems of Islamist terrorism in the MENA is put forward. The fourth section presents the findings of an automated at-a-distance leadership analysis with a detailed interpretation and discussion. The fifth section ends the chapter by stressing the significance and overall contribution of this research project. The Appendix highlights the technical mechanics of constructing the Arabic coding scheme and identifies the data and language limitations of applying the novel Arabic scheme on a well-established research topic within the terrorism scholarship (Canbolat 2020a).

Literature Review

The collective identity of a group becomes an integral part of an individual militant’s perception through which the values are dictated by the leadership of a terrorist organization (Crenshaw 2013; Post 1998). There is a burgeoning debate within terrorism scholarship about the actual utility of studying the leader within these organizations (Crenshaw 2013). The caveats about studying leaders are as follows: (1) groups may tend to follow the paradigm of leaderless resistance; (2) an individual’s own goal or inclination is the most important motivator of becoming active in political violence (Rapoport 2001). Many groups do follow a leaderless or incentive-based organization structure, yet this does not mean all collectivities are organized this way. More recent scholarship has focused on how and why terrorist leaders are indispensable factors of group behavior and outcomes (Hermann and Sakiev 2011; Price 2012; Walker 2011).

It is safe to argue that the magnitude of terrorist violence committed by different terrorist organizations has a considerable degree of variance (Asal and Rethemeyer 2008). Earlier studies in the terrorist violence literature have focused mostly on organizational mechanisms that cause specific groups to be more violent than others (Asal and Rethemeyer 2008). The bulk of the scholarship has focused on structural and/or material mechanisms that influence terrorist violence levels, but there are few studies focusing on the political psychology of individual terrorist leaders as an alternative explanatory variable. These studies deal with certain questions such as what provokes individuals to pursue political violence (Crenshaw 1986), and whether psychology can predict when terrorists will use violence (Hermann and Sakiev 2011; Walker 2011). Other studies argue that terrorist leaders’ verbal outputs (or rhetoric) are radically different in their operational codes from average political leaders (Lazarevska et al. 2006).

The literature on terrorist violence shows that capabilities and ideology are empirically important predictors of lethality among terrorist organizations (Hoffman 2006). This study does not adjudicate previous hypotheses concerning the lethality of terrorist-group behavior. It instead places individual-level, actor-specific variables at the center of analysis when studying terrorist-group behavior. While a few proponents of leadership analysis looked at the political psychology of certain terrorist leaders (Picucci 2008; Walker 2011; Hermann and Sakiev 2011), they did not take issue with the research questions about the direction and magnitude or lethality of terrorist violence.

This study posits that political speeches, both scripted and spontaneous variants, of predominant terrorist leaders can be used to explain and forecast their propensities for the direction and magnitude of terrorist violence.

The analysis, therefore, is not only germane to theories of terrorism and leadership studies in the field of international relations. It is also politically relevant to the tools and strategies for countering terrorism in the world. This study employs OCA, a prominent actor-specific approach to leader psychology rooted in the field of foreign policy analysis (FPA), to assess the top leadership profiles of several militant groups in the MENA region.

Theory and Method

How should we study terrorist leaders presiding over VNSAs and their impact on political behavior? Although early leadership studies were psycho-biographical and anecdotal (see Dyson 2015 for a history), the state-of-the-art for some time now has involved quantitative content analysis (Hudson 2005). Some examples are the works of Axelrod (1976) and Bonham et al. (1978) on cognitive maps; Tetlock (1998) and Suedfeld et al (2005) on integrative complexity; Leites (1951), George (1969; 1979), and Walker et al. (1998) on OCA: Hermann (1980, 2005) on leadership trait analysis; Boulding (1956), Hermann (1976), and Cottam (1985, 1992) on image theory; Walker (1987), Thies (2010), Harnisch et al. (2011), and Thies and Breuning (2012) on role theory. OCA focuses on the beliefs of political leaders as causal mechanisms in explaining foreign policy decisions (Leites 1951; George 1969, 1979; Walker 1983, 1990; Walker and Schafer 2007).

OCA was originally developed by Leites (1951) to analyze the decision-making style of the Soviet Politburo and was later developed and refined by George (1969; 1979), Holsti (1977), and Walker (1983, 1990). According to OCA, a leader’s cognitive schemata or belief system has two components. The first set is the five philosophical beliefs about the political universe in which the leader finds himself and the nature of the “other” he faces in this environment. Second, there are five instrumental beliefs that represent the image of “self” in this political universe and the best strategies and tactics one could employ to achieve one’s ends (George 1979; Walker 1990). Taken together operational code beliefs explain “what the individual knows, feels, and wants regarding the exercise of power in human affairs.” (Schafer and Walker 2006, 29).

The “automation turn” in political psychology has addressed many challenges associated with the study of political leaders from a distance such as the paucity and low quality of text materials as data (Walker et al. 1998). Automated at-a-distance analysis of verbal statements by political leaders to create leadership profiles has remained largely confined to English-language texts (Brummer et al. 2020). To overcome this limitation, a novel Arabic language coding scheme is employed, which is compatible with the Profiler Plus software and the OCA research program in FPA (Walker, Schafer, and Young 1998; Young 2001; Canbolat 2020a). For the main terrorist leaders in the MENA, three key variables from OCA are generated by utilizing this Arabic coding scheme1: P-1, Nature of the Political Universe; 1-1, Approach to Goals (direction of strategy); P-4, Control over Historical Development. According to Walker et al. (1998), there are three master beliefs shaping a leader’s operational code construct, namely, the P-1,1-1, and P-4 beliefs (see also George 1969).

P-1, Nature of the Political Universe, is a measure of the hostility or friendliness Self as an actor sees in the political environment. Lower scores indicate a more hostile worldview by a generalized Other. The score is calculated as the balance of verbs referring to other actors indicating hostile action versus verbs referring to other actors taking cooperative action. 1-1, Approach to Goals, is the counterpart for Self to P-1: how conflictual or cooperative is Self’s exercise of political power? The index is calculated as the balance of verbs referring to Self that indicate hostile action versus verbs referring to one’s Self taking cooperative action. The index for P-4, Control over Historical Development, evaluates the perceived degree of control by the actor over the political environment. The measure is a ratio of all verbs indicating action attributed to Self rather than Other as a proportion of the total verbs in a speech sample.

Research Design and Hypotheses

To execute the coding of the leadership psychology variables from OCA, this study employs an automated version of the Arabic content analysis scheme via the Verbs in Context System (VICS) and its software application Profiler Plus (Walker et al. 1998; Canbolat 2020a). Automated coding has significant advantages over human coding. The computer codes the same piece of text the same way over indefinite runs, guaranteeing replicability of results, and does not make mistakes in coding decisions due to human fatigue or bias. Locating and compiling enough speeches of MENA’s terrorist leaders in Arabic proved quite challenging. So far, a total of 45 public statements of six leaders have been compiled, which comport with the speech selection criteria set by the pioneers of the Profiler Plus school (Young 2001; Schafer and Walker 2006).

Table 4.1 lists the MENA’s high-profile, jihadist leaders included in this study’s data set. The databases utilized in this study to obtain verbal material for these leaders include LexisNexis, the Foreign Broadcast Information Service (FBIS), Al Jazeera Online Achieve, the official websites of the MB and Hamas, al-Mayadeen News, al-Masdar News, and other sundry websites and blogs. For coding the leaders’ public speeches and following the criteria for creating eligible data (see Schafer and Walker 2006), a minimum of five (for Badie) and a maximum of ten (for Meshal) public statements in Arabic are compiled in Arabic and their official English translations, the content analysis of which results are used as a robustness check. There were no statistically significant differences among the operational code scores in the Arabic and English speech data sets. Each public speech contains more than 1000 words and a minimum of 15 transitive verbs.

Table 4.1 The Main Islamist Terrorist Leaders and Organizations Included in This Study

Leader

Militant Organization(s)

Leadership Period

Ayman al-Zawahiri2

Al-Qaeda Central

2011-Present

Mohammad al-Jolani3

Al-Nusra/Fateh al-Sham

2012-Present

Abu Bakr al-Baghdadi4

ISIS (or ISIL, IS. DAESH)

2014-2019

Mohamed Badie5

Muslim Brotherhood (in Egypt)

2010-2013

Khaled Mashal6

Hamas (Gaza Strip)

2004-2017

Ramadan Shalah7

Palestinian Islamic Jihad (PIJ)

1995-2017

Number of Files

Min. Max. Tot. Words

Mean STDEV

45

1126 2768 79,200

1760 460

Three hypotheses are drawn from the terrorism literature on parent versus radical flank organizations within the subfield of international security. According to Kydd and Walter (2006), “outbidding” is one of the five main strategic logics of terrorism - along with attrition, intimidation, provocation, and spoiling. Outbidding refers to a situation where a terrorist group shows greater resolve to attack the enemy targets than rival terrorist groups to gain more public support and recruits for their cause (Kydd and Walter 2006, 51). Bloom (2004, 61) argues that while multiple terrorist organizations compete with other groups to increase their prestige, the extreme tactics of radical (or outlier) militant groups, e.g., suicide bombings, give such groups an upper hand in outbidding more mainstream groups and gaining constituents’ support.

There is a latent assumption in the literature on the outbidding theory about group decision-making regarding the use and lethality of terrorist violence, which can be observed and tested at the individual level as an actor-specific analysis. If a militant leader decides that outbidding behavior is the best way to pursue political goals, then the group’s lethal behavior is likely to escalate (Pape 2003; Kydd and Walter 2006). This assumption in the literature leads to the first two hypotheses given below, which relate to the beliefs of terrorist leaders about the strategic approach to goals (Hypothesis 1) and nature of the political universe (Hypothesis 2).

HI: When a fragmentation of a parent organization transpires, the beliefs of leaders of the radical flanks will be prone toward more violent strategies (1-1) than the beliefs of leaders of the parent organization to gain recognition and recruit more members from the latter (Outbidding theory).

H2: When a parent terrorist organization splinters into smaller groups, the new flanks will be presided over by leaders with more conflictual beliefs about the political universe (P-1) than the beliefs of the leaders of the parent organization (Outbidding theory).

Individual leaders with a higher sense of control over events accept the political environment around them and are generally more satisfied with the status quo. Individuals with a lower sense of this personal capacity are more reactive to and frustrated by the status quo within their surrounding environment. Some empirical research ties lower P-4 scores to the frustration-aggression hypothesis (Schafer et al. 2006). Thus, it is hypothesized that radical flank leaders should have a belief in lower historical control than parent group leaders, which will incite the former to use highly indiscriminate and lethal violent action against their enemy (Greenberg et al. 1997; Bar-Tal 2001; Schafer et al. 2006; Besaw 2014).

H3: When a parent terrorist organization splinters into smaller groups, the leaders of the new flanks will have a belief (P-4) in a lower level of control over historical development (Capacity Hypothesis).

Results and Discussion

An automated content analysis of the studied terrorist leaders’ speeches reveals consequential insights regarding the general patterns of militant leadership in MENA and the effects of group fragmentation on the lethality of radical flanks’ strategies. This section presents the results of disparate tests for each pair of leaders to answer four basic questions: are the three master belief scores significantly different (1) between each leader and a norming group? (2) between parent and radical flanks within each terrorist group? (3) between parent groups across different terrorist organizations and between radical flanks across different terrorist organizations?

Individual Leaders versus Average World Leader

Table 4.2 presents the scores of the militant leaders’ master beliefs compared with those of the world leadership norming sample via two-tailed difference of means tests. This table presents the three belief variables (on the left vertically), leaders and norming group (on the top horizontally), and the belief scores in a comparative format. The average world leadership group has the highest P-1 score while MB leader Badie’s score comes a close second and is followed by those of MB radical flank leaders Meshal and Shalah. The parent AQ leader Zawahiri’s P-1 is lower than all MB leaders but higher than radical AQ leaders Jolani and Baghdadi. In other words, the МВ-affiliated militant leaders view the political universe as more conflictual than the average world leader but as more peaceful than the AQ-atfiliated leaders.

One of the unexpected findings is that Badie’s 1-1 score is even higher than the norming sample score, which is followed by those of Zawahiri, Meshal, Shalah, Jolani, and Baghdadi. The most statistically significant differences between the MENA leaders and the norming group are between the 1-1 scores of AQ’s outlier leaders Baghdadi followed by those of Jolani compared with the other militant leaders. The MENA leader who is most like the norming group is Badie, who does not have statistically significant, different beliefs barring a more cooperative approach to strategy (1-1) than the average world leader. The belief in historical

76 Sercan Canbolat

Table 4.2 Master Operational Code Beliefs of Main Militant Leaders in MENA Compared with Average World Leadership Norming Group*

Leader

Zawahiri

Jolani

Baghdadi Badie

Meshal

Sltalah

Norm.

Grp.

Org. Tvpe

P-1

1-1

P-4a

Speech N

P-AQ

  • 0.099"
  • 0.267*
  • 0.133"
  • 8

R-AQ

  • -0.095*"
  • -0.101"
  • 0.207
  • 8

R-AQ

  • -0.146*“
  • -0.292"*
  • 0.218
  • 7

P-MB

  • 0.239
  • 0.569*
  • 0.229
  • 5

R-MB

  • 0.157"
  • 0.235*
  • 0.099"
  • 10

R-MB

  • 0.104"
  • 0.118"
  • 0.068"*
  • 7

Leader

Org. Tvpe

P-1

1-1

P-4a

Speech N

N/A

  • 0.25
  • 0.334
  • 0.212
  • 255

Zawahiri

P-AQ

  • 0.099"
  • 0.267*
  • 0.133"
  • 8

*Abbreviations: (1) Org: organization; (2) P: parent organization; (3) R: radical flank organization; (4) AQ: Al-Qaeda; (5) MB: Muslim Brotherhood; (6) Norm. Grp: norming group; (7) Asterisks indicate significant differences are at the following levels (two-tailed test): *p < 0.10, **p < 0.05, *"p < 0.01. The scores for the norming group sample (N = 255) of average world leaders are courtesy of Schafer and Walker (2006, 170, n.13).

control over events (P-4a) for Jolani, Baghdadi, and Badie is also not significantly different from the norming group’s average belief.

Regarding the P-4a distribution, Badie and Baghdadi’s belief about Self’s level of control is slightly higher than the average leader and Jolani comes a close third, which are all within one standard deviation from the norming group score. While Zawahiri and Meshal’s P-4a scores are lower than Baghdadi and Jolani and the differences are statistically significant from the norming group score, their scores exceed Shalah’s control score, the statistical difference of which stands out as the most significant from the norming group. In contrast to the P-1 and 1-1 distributions, there is no clear-cut pattern for the P-4 beliefs of the studied leaders whose individual idiosyncrasies might have made more difference than their organization types. For example, while the radical flank leaders of AQ have higher P-4 scores than the parent AQ leader, the other way around is true for the MB-atfiliated militant leaders. An emerging, yet under-tested, pattern could be that the AQ-atfiliated leaders perceive more control over political events than the MB leaders, who operate in drastically different political and socioeconomic contexts than the former.

Parent Groups versus Radical Flank Groups

Table 4.3 shows the results of an additional statistical analysis in the form of a difference of means t-test to further investigate similarities or differences between parent VNSA and radical flank leaders. The t-test results

Table 4.3 Mean Operational Code Beliefs of Generic Terrorist Groups in MENA Region*

Org. Type

Parent AQ (N = 8)

Parent MB (N=5)

Difference of Means

P-1

0.099

0.239

0.140'**

1-1

0.267

0.569

0.302'**

P-4a

0.133

0.229

0.096'*

Org. Type

Radical AQ (N = 15)

Radical MB (N=17)

Difference of Means

P-1

-0.121

0.131

0.252'**

1-1

-0.197

0.177

0.374'**

P-4a

0.213

0.084

0.129'**

Org. Type

Parent AQ (N = 8)

Radical AO (N=15)

Difference of Means

P-1

0.099

-0.121

0.220'**

1-1

0.267

-0.197

0.464'**

P-4a

0.133

0.213

0.080'*

Org. Type

Parent MB (N = 5)

Radical MB (N=17)

Difference of Means

P-1

0.239

0.131

0.108'**

1-1

0.569

0.177

0.392'**

P-4a

0.229

0.084

0.145“**

Org. Type

Parent Group (N=13)

Radical Flank ( N = 32)

Difference of Means

P-1

0.169

0.005

0.164'**

1-1

0.418

-0.01

0.428'**

P-4a

0.181

0.148

0.033'*

’Significant differences are at the following levels (two-tailed test): *p<0.10, ”p < 0.05, ”*p<0.01.

show that there are significant differences between the scores of parent group leaders and those of the radical flanks. Key leadership style distinctions between the parent VNSA and radical flank leadership styles can be seen in all three OCA variables: P-1, 1-1, P-4. In contrast to the parent group leaders, the radical flank leaders believe the political universe (P-1) is significantly less friendly, and they are significantly more conflict oriented in both their strategy (1-1). While both parent group and radical flank leaders’ mean self-control (P-4a) belief scores are lower than the norming group score (0.212), the radical flank leaders have a notably lower score for their beliefs in controlling historical events (0.148).

Within-group Differences

Tables 4.4 and 4.5 break down the aggregated operational code scores of parent VNSA and radical flank leaders and compare the master belief scores of the two groups: (1) parent AQ to radical AQ (Table 4.4) and

Table 4.4 Differences in Master Operational Code Beliefs of Parent Group and Radical Flanks Leaders of Al-Qaeda*

Org. Type

Parent AO (N = 8)

Parent MB (N = 5)

Difference of Means

P-1

0.099

0.239

0.140***

1-1

0.267

0.569

0.302***

P-4a

0.133

0.229

0.096**

Org. Type

Radical AQ (N=15)

Radical MB (N = 17)

Difference of Means

P-1

-0.121

0.131

0.252***

1-1

-0.197

0.177

0.374***

P-4a

0.213

0.084

0.129*“*

Org. Type

Parent AO (N = 8)

Radical AQ (N = 15)

Difference of Means

P-1

0.099

-0.121

0.220*"

1-1

0.267

-0.197

0.464***

P-4a

0.133

0.213

0.080**

Org. Type

Parent MB (N = 5)

Radical MB (N = 17)

Difference of Means

P-1

0.239

0.131

0.108**’

1-1

0.569

0.177

0.392***

P-4a

0.229

0.084

0.145*“

’Significant differences are at the following levels (two-tailed test): *p<0.10, ”p < 0.05, ”’p<0.01.

Table 4.5 Differences in Master Operational Code Beliefs of Parent Group and Radical Flanks Leaders of the Muslim Brotherhood*

Org. Type

Parent AQ (N = 8)

Parent MB (N=5)

Difference of Means

P-1

0.099

0.239

0.140***

1-1

0.267

0.569

0.302“*

P-4a

0.133

0.229

0.096“

Org. Type

Radical AQ (N = 15)

Radical MB (N=17)

Difference of Means

P-1

-0.121

0.131

0.252“*

1-1

-0.197

0.177

0.374*“*

P-4a

0.213

0.084

0.129*“*

Org. Type

Parent AQ (N = 8)

Radical AQ (N= 15)

Difference of Means

P-1

0.099

-0.121

0.220“*

1-1

0.267

-0.197

0.464***

P-4a

0.133

0.213

0.080“

Org. Type

Parent MB (N = 5)

Radical MB (N=17)

Difference of Means

P-1

0.239

0.131

0.108***

1-1

0.569

0.177

0.392*“

P-4a

0.229

0.084

0.145*“*

’Significant differences are at the following levels (two-tailed test): *p<0.10, **p<0.05, ”’p<0.01.

(2) parent MB to radical MB (Table 4.5). While the MB’s parent group leader Badie’s 1-1 score stands out across all the comparison groups, of both parent organization leaders’ instrumental belief scores are more positive and higher than the radical flank leaders’ scores. Likewise, there is a statistically significant difference between the P-1 scores of parent group leaders Zawahiri (parent AQ) and Badie (parent MB) on the one hand, and radical flank leaders Jolani, Baghdadi (radical AQ), and Meshal, Shalah (radical MB) on the other.

As for the P-4a scores, while both parent group leaders’ scores are significantly different from those of their respective radical flank leaders, the directions of P-4a scores of parent and radical flank leaders are different. Interestingly, radical AQ leaders’ mean P-4a score (0.213) is higher than both parent group leaders’ mean (0.133), and the norming group score (0.212), however slightly. In marked contrast to the AQ leaders, the mean P-4a score of radical MB leaders (0.084) is significantly lower than those of their parent group leaders (0.229) and those of average world leaders (0.212). In other words, the relationship between the P-4a scores of parent and radical MB leaders is compatible with the Capacity hypothesis, while the direction of two AQ-affiliated group leaders’ P-4a scores of contravenes the hypothesized relationship between parent groups and radical flanks. In sum, the results presented in Tables 4.4 and 4.5 substantiate that the three master belief scores significantly differ between parent and radical flanks for each terrorist group, but the direction of the P-4 score is qualified by differences between the main terrorist organizations.

Between Group Differences

Tables 4.6 and 4.7 present the results of difference of means tests comparing master belief scores of parent groups (Table 4.6) and radical flanks (Table 4.7) across different terrorist organizations. First, all the master belief mean scores of parent MB are higher and more positive than those of parent AQ. In other words, Badie perceives the political universe (Other) in more peaceful terms and believes in more cooperative tools in his strategic approach than the AQ chief Zawahiri. Likewise, Badie’s sense of political control (0.229) is stronger than both Zawahiri’s (0.133) and the average world leader (0.212). Second, the МВ-affiliated radical leaders have higher and more positive P-1, 1-1, and P-4a operational code scores than those of radical AQ leaders. While radical AQ leaders attribute more control to themselves (P-4a) than radical MB leaders, the former’s view of the Other (political universe) is more conflictual than the latter. The hostility perceived by radical AQ leaders is compounded with their salient propensity for more conflictual strategies compared with their MB counterparts. In brief, the operational codes of both parent organizations and radical flank leadership are markedly different between the two terrorist groups.

Table 4.6 Differences in Master Operational Code Beliefs of Parent Group Leaders Across Different Militant Organizations'

Org. Type

Parent AQ (N = 8)

Parent MB (N=5)

Difference of Means

P-1

0.099

0.239

0.140*"

1-1

0.267

0.569

0.302*“*

P-4a

0.133

0.229

0.096"

Org. Type

Radical AQ (N = 15)

Radical MB (N=17)

Difference of Means

P-1

-0.121

0.131

0.252*"

1-1

-0.197

0.177

0.374"*

P-4a

0.213

0.084

0.129***

Org. Type

Parent AQ (N = 8)

Radical AO (N = 15)

Difference of Means

P-1

0.099

-0.121

0.220*"

1-1

0.267

-0.197

0.464"*

P-4a

0.133

0.213

0.080"

Org. Type

Parent MB (N = 5)

Radical MB (N=17)

Difference of Means

P-1

0.239

0.131

0.108*"

1-1

0.569

0.177

0.392*“*

P-4a

0.229

0.084

0.145*"

'Significant differences are at the following levels (two-tailed test): *p<0.10, ”p < 0.05, ”*p<0.01.

Table 4.7 Differences in Master Operational Code Beliefs of Radical Flank Leaders across Different Militant Organizations*

Org. Type

Radical AQ (N = 15)

Radical MB (N=17)

Difference of Means

P-1

-0.121

0.131

0.252“"

1-1

-0.197

0.177

0.374*"

P-4a

0.213

0.084

0.129““*

Org. Type

Parent AQ (N = 8)

Radical AQ(N= 15)

Difference of Means

P-1

0.099

-0.121

0.220“"

1-1

0.267

-0.197

0.464*“*

P-4a

0.133

0.213

0.080"

Org. Type

Parent MB (N= 5)

Radical MB (N=17)

Difference of Means

P-1

0.239

0.131

0.108“"

1-1

0.569

0.177

0.392““*

P-4a

0.229

0.084

0.145“"

'Significant differences are at the following levels (two-tailed test): *p<0.10, *'p<0.05, ’”p<0.01.

Collectively, the results in Tables 4.2-4.7 corroborate the first two hypotheses concerning the militant leaders’ view of the political universe (P-1) and their strategic orientation (1-1), with robust empirical evidence. Tables 4.2,4.4, and 4.5 cast some doubt on the third (capacity) hypothesis as the control over historical development (P-4) scores of the AQ and the MB radical flank leaders show reverse directions compared with their respective parent groups leaders. While the radical flank leaders of the AQ (Jolani and Baghdadi) assign higher Self’s control to themselves compared with the parent group leader (Zawahiri), the MB-atfiliated radical flanks leaders’ (Meshal and Shalah) control scores are significantly lower than those of the parent organization leader (Badie). Table 4.3 supports the capacity hypothesis at the group level of analysis, showing statistically significant differences between the parent and radical flank groups’ the P-4a scores (p < 0.05); these results are qualified by differences in direction of the P-4 scores within each terrorist group.

Thus, it can be argued that when a parent organization splinters, the emerging leaders of the radical flanks are likely to have more violent beliefs (1-1) about strategy than the beliefs of the parent organization leaders. Likewise, once a parent terrorist organization fragments into smaller groups, the new flanks will be led by individuals with more conflictual beliefs about the other policy actors (P-1) within the political universe. This study hypothesized that the shifts in 1-1 and P-1 toward conflict are likely to be more lethal when accompanied by radical flank leaders’ lower capacity to control historical development (P-4a) compared with their parent group leaders. The exceptions are the AQ’s outliers: Baghdadi’s and Jolani’s propensities for using more lethal and cruel militant tactics to compete with other terrorist groups in the MENA are accompanied by their belief in historical control similar to the norming group (Friis 2015; Abrahms 2018).

Figure 4.1 classifies all the militant decision makers and their respective groups in terms of leadership styles and roles. These are classified as operational code types А, В, C, or DEF (Holsti 1977; Walker 1990), which correspond to the four quadrants in this figure. The VICS indices for the master beliefs, P-1 (nature of the political universe), 1-1 (strategic approach to goals), and P-4 (ability to control historical development), are mapped on intersecting vertical (P-l/I-1) and horizontal (P-4) axes to locate the leader’s image of the Self and Other in one of the four quadrants. The coordinates for Self (1-1, P-4a) and Other (P-1, P-4b) lead to predictions regarding strategic preferences over the goals of settle, submit, dominate, and deadlock (Walker and Schafer 2006).

Leaders who fall into the same zones within each quadrant share the similar roles and relations of friend and alignment, partner and cooperation, rival and conflict, or enemy and domination attributed to Self and Other (Malici and Walker 2017). These locations with corresponding roles and relations are inferred from comparing the scores for the MENA terrorist leaders to the mean P-1, 1-1, and P-4 indices of the norming

82 Sercan Canbolat

Operational Code Types and Roles of Terrorist Leaders and Groups in the MENA Region

Figure 4.1 Operational Code Types and Roles of Terrorist Leaders and Groups in the MENA Region

Sources: Holsti (1977); Walker (1990); Malici and Walker (2017). Abbreviations: SD = Standard deviation; Bag = Baghdadi; В = Badie; J = Jolani; M = Meshal; S = Shalah; Z = Zawahiri. S = Self; О = Other; Triangle Symbol = Parent Group; Square Symbol = Radical Flank Group group of world leaders from a variety of geographical leaders and historical eras (Schafer and Walker 2006; Malici and Walker 2017). To render the comparisons more robust, the norming group scores are used as the reference points (origin) of the coordinate system in Figure 4.1. A Z-score measurement is used to plot all the militant leaders on Figure 4.1. Table 4.8

Table 4.8 Standard Deviation Z-Scores for Deviations in Beliefs from Norming Group for Militant MENA Leaders0

Leader

Z

J

Bag

В

M

Norm. Grp.

Org. Type:

P

R

R

P

R

R

Ave. (SD)

P-1

-0.47

-1.09

-1.25

-0.06

-0.28

-0.47

0.25 (0.32)

1-1

-0.13

-0.91

-1.32

+0.51

-0.21

-0.45

0.33 (0.47)

P-4a

-0.67

0.00

+0.08

+0.17

-0.91

-1.17

0.21 (0.12)

P-4b

+0.67

0.00

-0.08

-0.17

+0.91

+1.17

0.79 (0.12)

aZ score = (observed score minus sample mean) divided by standard deviation of sample. Org. = Organization; Z = Zawahiti; J = Jolani; Bag = Baghdadi; В = Badie; M = Meshal; S = Shalah; P = Parent Group; R = Radical Flank Group; Norm. Grp. = Norming Group (N = 255).

Source: Schafer and Walker (2006, 170, n. 13).

presents the Z-scores for the standard deviations, respectively, from the norming group means for 1-1, P-1, and P-4 (N = 255).

Strikingly, only the MB leader Badie’s image of Self falls under the Type C operational code system, who attributes himself a partner role of cooperation in the foreign policy realm with somewhat strong control over events. Nevertheless, Badie views Other as enacting a rival role of conflict with a Type DEF leadership style. The other parent leader Zawahiri sees both Self and Other as rivals in a conflictual political universe, yet he perceives Other as the stronger rival (Type B), while he wields less power (Type DEF) in any strategic interaction between Self and Other.

The difference between the MB leader Mohamed Badie and the AQ’s Zawahiri speaks volumes about the distinct characters of parent VNSAs. While the AQ views itself as an armed group fighting the West and secular Muslim regimes in MENA and elicits a similar image in the Western countries, the MB has a more convoluted modus operandi in which the organization claims to wear two hats: (1) a democratic player in electoral regimes (such as in Egypt and in the Gaza Strip) which operates within the parameters of democratic politics; (2) maintaining a military wing and resorting to terrorist violence via the radicalized cells when necessary. Thus, wearing two hats at the same time, the MB leader Badie has both a democratic and a militant ringleader image, which manifest themselves in his public speeches. These distinctions in the organizational structures of the MB and AQ account for the dissimilarity between al-Zawahiri and Badie’s operational code types and different role attributions to Self and Other.

Figure 4.1 indicates that the studied radical flank leaders have a high propensity to waver between the leadership types В and DEF regarding both their conceptualizations of Self and Other, which influence their preference rankings for different political outcomes. All the radical flank leaders are situated on the lower half of Figure 4.1, which means that both their Self and Other images are more conflictual (1-1) and hostile (P-1); they see Self and/or Other in rival or enemy roles. Nonetheless, the МВ-affiliated, radical flank leaders project less conflictual and hostile leadership styles than the AQ’s radical wing leaders. Shalah assigns rival foreign policy roles to both Self and Other, yet views himself as the weaker rival (Type DEF) with a tendency to pursue more conflictual strategies than Other (Type B), who has stronger control over events. While Meshal has less conflictual and hostile beliefs about the political universe and a strategic approach to foreign policy than those of Shalah, his leadership style and role attributions are the same as those of Shalah.

The AQ-affiliated radical flank leaders’ operational code types and roles dwell in the lowest zones of Figure 4.1, which correspond to their extremely negative P-1 and 1-1 belief scores compared with the rest of the militant leaders. Jolani attributes a very hostile rival role of conflict to Self and an enemy role of domination to Other; his perception of hostility is stronger for Other (within the -2 standard deviation zone) compared with those of his self-assigned, “almost enemy” role (just within the -1 standard deviation zone). Because of the Z-scores distribution of his P-4 beliefs, Jolani’s leadership style dwells on the vertical axis, which could shift between the Type DEF and Type В quadrants and render his operational code type less predictable compared with other militant leaders. Baghdadi overtops all the studied militant leaders in having the most negative 1-1 and P-1 scores; his role assignments to both Self and Other correspond to enemy roles (both within the -2 standard deviation zone) in a struggle for domination.

Intriguingly, Baghdadi construes Other as a (however marginally) weaker enemy with Type DEF leadership particularities, while he attributes a stronger “enemy” role and a Type В leadership style to Self in dealing with others in the political realm. Put differently, the ISIS chief stands out as the most conflictual individual in the entire group, and he has the second highest self-control (P-4a) score after Badie. While this finding is not surprising because ISIS is viewed as “one of the most lethal terrorist organizations in today’s world,” it is still interesting to confirm that argument by a systematic analysis of speeches (Friis 2015: 745). This research also bears out a scholarly argument that very lethal terrorist organizations, e.g., ISIS, sow the seeds of their own destruction by using extremely vitriolic rhetoric and resorting to cruel terrorist violence (Abrahms 2013, 2018).

Conclusion

Analyzing the splinter groups that break away from AQ and the MB, i.e., ISIS, al-Nusra (currently known as HTS), Hamas, and Islamic Jihad, is integral to explaining and forecasting their terrorist behavior. The results of this study concur with the argument that the formation of militant splinter groups influences their organizational evolution and future behavior (Perkoski 2019). The results lend a nuanced support to the hypotheses in this chapter that there are consequential differences even among leaders who hail from the same Islamic creed and ideological training. In addition to their idiosyncratic beliefs, the type of their militant organizations also influences an individual leader’s inclinations for making strategic choices and their lethality in dealing with others in the political universe.

Using OCA to construct psychological profiles of key terrorist leaders from a content analysis of their written statements reveals a set of master political beliefs, which are hypothesized to motivate their terrorist behavior. These political beliefs are instrumental in understanding the VNSAs’ strategies because many terrorist leaders in the MENA region encounter very few institutional/structural constraints within their organizations. These MENA terrorist groups perceive the political world through religious and ideological lenses, and therefore, their political behavior rests on the specific formation (or fragmentation) patterns of these movements.

Put differently, top decision makers of these jihadist radical flanks - e.g., Baghdadi of ISIS, Jolani of al-Nusra, and Shalah of Islamic Jihad - interact with the political world solely based on their judgment of the capabilities of the jihadist movement rather than more concrete structural and material conditions. What terrorist leaders say about others in their respective environment does matter. When terrorist leaders say they view violent strategies as the only way toward gaining political objectives, their organization’s violent behavior, by and large, reflects that strategic belief. Moreover, militant leaders who show more hostility in their utterances will be more likely to promote increased lethality in their group’s violence.

Evidence to support this argument is in the comparison of relevant political beliefs about the political universe and political strategies across parent and splinter terrorist organizations and with the operational codes of reference groups, e.g., average world leaders. Preliminary results suggest that OCA provides a set of scientific tools for linking differences in terrorist behavior to differences in belief systems. Finally, by comparing observed operational code variables to expected findings informed by commonly held assumptions within the field of terrorism studies, OCA provides a rigorous analytical means of testing the validity of such assumptions.

While automated at-a-distance leadership assessment research programs have made steady progress since the late 1990s (Walker et al. 1998), there is still a gap in the leadership research program concerning the study of non-Western decision makers using languages beyond English (Brummer and Hudson 2015). Hinnebusch (2015) argues that while the leaders operating in Western liberal democracies have been studied extensively and in a comparative fashion, non-Western leadership cases including terrorist leaders have received less attention in the FPA field (for exceptions, see Malici 2008; Malici and Buckner 2008; Duelfer and Dyson 2011; Ozdamar and Canbolat 2018; Canbolat 2020a, 2020b; Ozdamar et al. 2020). This chapter has aimed to address this gap by expanding the North America- bounded FPA approach toward the Global South by studying in their native language a MENA “brand” of terrorism from the vantage point of at-a-distance leadership assessment tools.

In that sense, future avenues of research in this area include the following: (1) tracking political beliefs of key decision makers over time within the same terrorist organization; (2) assessing the internal belief system differences within the same parent and/or splinter groups (Perkoski 2019); (3) developing an actor-specific framework to anticipate the timing and lethality of terrorist attacks from the operational codes of individual terrorist leaders (Walker 2011); (4) analyzing the “audience effect” on the militant leaders’ strategic rhetoric and operational codes, which might differ depending on the type of audience including domestic versus regional versus international (Canbolat 2020a, 2020b).

In a nutshell, this research aims at helping both the scholars of FPA with more rigorous theoretical advancement and MENA specialists with more far-sighted policy recommendations. Providing them with an inroad into the mindsets of key terrorist groups and drawing inferences on the future trajectory of terrorist groups’ behavior and lethality in the MENA region, this study avoids making sweeping or normative statements about possible solutions to the vicious circle of terrorist violence in MENA. This chapter also casts doubt on key assumptions that are central to counterinsurgency policies of regional and international players in MENA such as the United States, Russia, and Israel.

Accordingly, the foremost policy recommendation in this chapter is that regional and world leaders should give heed to the leaders of terrorist organizations. By and large, most of what terrorist leaders say has been censored or overlooked. Eliminating terrorist violence and hampering terrorist movements are by nature convoluted issues, the solutions of w'hich are dependent on a variety of factors. Nevertheless, putting White’s (1991) notion of realistic empathy, not a naively construed sympathy or tolerance, into practice to understand certain militant actors - who mostly make group decisions singlehandedly in the MENA region - should be part and parcel of regional and global counterterrorism efforts.

Appendix: Challenges of Coding in Native Tongues:

Data and Language Limitations

Studying non-mainstream leaders, who mostly speak in their native languages and who run rather clandestine organizations, from a distance has proved onerous. Major methodological hurdles include, (1) the availability of text, (2) the veracity of authorship problem, (3) the lost-in-trans- lation issue, and (4) the danger of the individualistic fallacy.

First, the availability of text - transcripts of words spoken by the subject of study - determines the shape of the research. The ideal is an online archive of carefully curated text, sortable by date, topic, and audience, maintained on behalf of a leader who spoke regularly. Occasionally, research will be driven by the appearance of an enticing text corpus: the availability of new tapes from the Nixon presidency, for example, or the discovery that Saddam Hussein kept his own taping system to track his meetings and phone calls (Dyson and Raleigh 2014). In this research, the most pressing problem was finding sufficient speech data to establish foreign policy profiles for the terrorist leaders. Particularly, ISIS leader Al-Baghdadi’s and Nusra leader Al-Jolani’s public speeches on politics and foreign policy have been elusive and so locating and retrieving them was the most time-consuming part of the research. This is because such militant leaders’ public appearances are rather scant and so their speeches often lack confirmation and reliability and such leaders’ speeches are subject to broad censorship in both their countries and in the world media.

Second, the common question about using a leader’s public speeches is whether a leader pens their own speeches or just reads a speechwrit- er’s script from the prompter. This study’s rejoinder to this criticism is twofold: (1) all leaders, even the terrorist leaders under investigation, are accountable for what they say in front of the public, and this constrains their policy behaviors later on; and (2) no leader will publicly utter a political statement without approving it.

Third, there is the “automation problem”: missing the particular contextual meanings of speech due to the context-free nature of automated coding in all the languages including Arabic. This is an undeniable challenge to content analysis methods, but I still opt for an “at-a-distance” text analysis since it is more feasible than doing hand coding in Arabic or traveling to war-torn MENA countries to conduct field research and interviews with the militant groups.

Fourth, any agent-centered approach runs the risk of reductionism - explaining the behavior of a complex phenomenon like terrorist behavior and militants’ motives for resorting to violence through a simple recourse to the idiosyncratic will of its leaders. At some level, this possibility is methodologically impossible to eliminate because of lacking access to an alternate reality in which terrorist organizations have other leaders. One cannot definitively test the impact of the selected group of militant leaders on policy outcomes. In FPA research programs, counterfactual analyses based upon assuming leaders with different measured characteristics are usually the best available remedy.

At this early stage, the author merely signals an awareness of the need to perform this in-depth analysis in future iterations of this project. That said, using “at-a-distance” leadership analysis tools powered through the Arabic scheme to unveil terrorist leaders’ psychological profiles mitigates the financial and logistical costs associated with studying leaders, i.e., a lack of direct access to the leader, the potential security risks of visiting worn-torn countries such as Syria and Iraq, and the linguistic and cultural problems associated with conducting field research.

Developing an Arabic Coding Scheme for OCA

Creating automated content analysis coding schemes for texts in English and Arabic requires particular attention to each language. While the basic strategic task is similar for each language, the linguistic details prompt the adoption of different tactics.

The similarities are as follows:

  • 1 Content words in both languages are set olf by spaces in texts. This is nontrivial, since some languages do not (e.g., Chinese).
  • 2 The basic task is to identify the distinguishable content forms in the most efficient way. This involves a robust grammatical analysis for both languages, so that purely (or partially) grammatical indicators may be separated from content forms.
  • 3 The scheme design is similar for both languages. Grammatical analysis is done first, isolating the content forms but preserving the grammatical context. Content forms are assigned values, which are turned into data. Each analyzed text (called a “token”) has a text form, a lemma form, and various values associated with it (part-of- speech, truth-value, etc.). One can use these values in the analysis, change them, create new ones, etc.
  • 4 Particular content values are determined by native reader judgments, for both languages.

The differences are,

  • 1 English tends to have consistent stem content forms (which are called “lemmas”), with endings attached for grammatical differences. There are some exceptions, such as “man/men.” Arabic makes much more use of internal consonant and vowel pattern distinctions than English.
  • 2 English grammatical function words (prepositions, pronouns, particles, auxiliary verbs, etc.) are separate words. In Arabic, some of these are attached either at the beginnings or ends of words (such as bi-. Н-, wa-, and -kum).
  • 3 English has a predominantly subject-verb-object (SVO) declarative word order. The normal Arabic Order is verb-subject-object (VSO).
  • 4 The English lexical grammatical analysis (called “tagging”) primarily involves recognition of word forms, grammatical endings, and labeling of parts of speech. It proceeds from word matches in dictionaries. The content form (stem form without endings) is cited as the “lemma” in the scheme. These forms, labels, or other values are then available for later parts of the scheme to access.
  • 5 The Arabic scheme is different because there are so many different morphemes, clitics, and word-internal distinctions.8 The procedure here is to treat unanalyzed Arabic words as strings of individual characters set off by spaces.

6 Clitic strings (mostly prepositions and pronouns) can be recognized, separated, and collapsed, treating them as separate words, but in the same order. Then the content word can be recognized, grammatically and lexically analyzed, and a lemma form can be created. The result is collapsed into a single analyzed word (“token”). So, for example, (“to the United Nations”) in Arabic is separated

into 5 separate forms: J ,Ji ,ji .боли by this process.

Notes

  • 1. Courtesy of Michael D. Young and Doug Fuller of Social Science Automation, who graciously helped me develop the Arabic coding scheme and generate the three OCA variables via the scheme. The Arabic coding scheme’s mechanics are detailed in the Appendix (see also Canbolat 2020a).
  • 2. Most of the Al-Zawahiri speeches are retrieved from the AQ’s primary media platform (al-Sahab): https://archive.org (accessed 12/15/2019).
  • 3. There is a dearth of speeches associated with al-Jolani. The author had to rely on a few long interviews with the Al-Nusra leader, many of which are broadcasted by Al-Jazeera, available at https://www.aljazeera.com/ (accessed 05/15/2020).
  • 4. Like al-Jolani, the ISIS leader al-Baghdadi’s speeches are retrieved from various clandestine illicit jihadi blogs and/or the ISIS’s media platform, called Dabiq, on Twitter (accessed 04/15/2020).
  • 5. Badie’s speeches can be accessed at the MB’s official website: https://www. ikhwanonline.com/ (accessed 12/15/2019).
  • 6. Mashal’s speeches and interviews are available at A1 Jazeera: https://www. aljazeera.com/ and at A1 Hadath (a local newspaper in Gaza Strip): https:// www.alhadath.ps/ (accessed 12/15/2019).
  • 7. Shalah’s speeches are retrieved from the website of Saraya al-Quds (the military wing of Islamic jihad movement in Palestinian territories): https:// saraya.ps/ and from the A1 Mayadeen (a well-known Lebanese media platform): http://almayadeen.net (accessed 12/15/2019).
  • 8. In linguistics, the difference between morpheme and clitic is as follows: morpheme is the smallest linguistic unit within a word that can carry a meaning, such as “in-,” “escape,” and “-able” in the word “inescapable” while clitic is a morpheme that is always attached to following or preceding words instead of an independent word.

References

Abrahms, M. 2013. The credibility paradox: Violence as a double-edged sword in international politics. International Studies Quarterly 57(4): 660-671.

Abrahms, M. 2018. Rules for rebels: The science of victory in militant history. New York, NY: Oxford University Press.

Asal, V, R. Rethemeyer. 2008. The nature of the beast: Organizational structures and the lethality of terrorist attacks. The Journal of Politics 70(2): 437 449.

Axelrod, R. 1976. Structure of decision: The cognitive maps of political elites. Princeton, NJ: Princeton University Press.

Besaw, C. 2014. Deadly premonition: Does terrorist-leader psychology influence violence lethality? (Unpublished M.A. Thesis). Retrieved from http://stars.library. ucf.edu/etd/4494

Bar-Tal, D. 2001. Why does fear override hope in societies engulfed by intractable conflict, as it does in the Israeli society? Political Psychology 22(3): 601-627.

Bloom, M. 2004. Palestinian suicide bombing: Public support, market share, and outbidding. Political Science Quarterly 119(1): 61-88.

Bonham, G., D. Heradstveit, O. Narvesen, M. Shapiro. 1978. A cognitive model of decision-making: Application to Norwegian oil policy. Cooperation and Conflict 13(2): 93-108.

Boulding, K. 1956. The image: Knowledge in life and society. Ann Arbor, MI: University of Michigan Press.

Brummer, K., and V. Hudson. Eds. 2015. Foreign policy analysis beyond North America. Boulder, CO: Lynne Rienner Publishers, Incorporated.

Brummer, К., M. Young, O. Ozdamar, S. Canbolat, C. Thiers, C. Rabini, A. Mehvar. 2020. Coding in tongues: Developing non-English coding schemes for leadership profiling. International Studies Review. IT. 1051-1055. Online first: https://doi.org/10.1093/isr/viaa001.

Canbolat, S. 2020a. Profiling leaders in Arabic. In Forum Coding in Tongues: Developing Non-English Coding Schemes for Leadership Profiling. International Studies Review. Online first: https://doi.org/10.1093/isr/viaa001.

Canbolat, S. 2020b. Understanding political Islamists’ foreign policy rhetoric in their native language: A Turkish operational code analysis approach. APSA- MENA Politics Newsletter 3(1): 13-16.

Cottam, M. 1985. The impact of psychological images on international bargaining: The case of Mexican natural gas. Political Psychology 6: 413-440.

Cottam, M. 1992. The Carter Administration’s policy toward Nicaragua: Images, goals, and tactics. Political Science Quarterly 107(1): 123-146.

Crenshaw, M. 1981. The causes of terrorism. Comparative Politics 13(4): 379-399.

Crenshaw, M. 1986. The psychology of political terrorism. Political Psychology 21(2): 379-413.

Crenshaw, M. 2013. Decisions to use terrorism: Psychological constraints on instrumental reasoning. In Terrorism studies, 272-281. New York, NY: Routledge.

Dueller, C. and S. Dyson. 2011. Chronic misperception and international conflict: The US-Iraq experience. Internationa! Security, 36(1): 73-100.

Dyson, S. 2015. Leaders in conflict: Bush and Rumsfeld in Iraq. Manchester: Manchester University Press.

Dyson, S., A. Raleigh. 2014. Public and private beliefs of political leaders: Saddam Hussein in front of a crowd and behind closed doors. Research & Politics 1(1): 1-7.

Friis, S. 2015. ‘Be}'ond anything we have ever seen’: Beheading videos and the visibility of violence in the war against ISIS. International Affairs 91(4): 725-746.

George, A. 1969. The "operational code": A neglected approach to the study of political leaders and decision-making. International Studies Quarterly 13(2): 190-222.

George, A. 1979. The causal nexus between beliefs and behavior and the operational code belief system. In Psychological models in international politics, 95- 124. Boulder: Westview Press.

Greenstein, F. 1967. The impact of personality on politics: An attempt to clear away underbrush. The American Political Science Review, 61(3): 629-641.

Greenberg, J., S. Solomon, T. Pyszczynski. 1997. Terror management theory of self-esteem and cultural worldviews: Empirical assessments and conceptual refinements. In Advances in experimental social psychology, 61-139. Cambridge, MA: Elsevier Academic Press.

Harnisch, S., C. Frank, and H. Maull. Eds. 2011. Role theory in international relations. New York, NY: Routledge.

Haines, H. 2013. Radical flank effects. In The Wiley-Blackwell encyclopedia of social and political movements. Chichester: John Wiley & Sons.

Hermann, M. 1976. Circumstances under which leader personality will affect foreign policy: Some propositions. In Search of globed patterns, 326-332. New York: Free Press.

Hermann, M. 1980. Explaining foreign policy behavior using the personal characteristics of political leaders. International Studies Quarterly 24(1): 7 46.

Hermann, M. 2005. Assessing leadership style: A trait analysis. In The psychological assessment of political leaders, 178-212. Ann Arbor, MI: University of Michigan Press.

Hermann, M., A. Sakiev. 2011. Leadership, terrorism, and the use of violence. Dynamics of Asymmetric Conflict 4(2): 126-134.

Hinnebusch, R. 2015. Foreign policy analysis and the Arab world. In Foreign policy analysis beyond North America, eds. K. Brummer, V. Hudson, 77-100. Boulder, CO: Lynne Rienner.

Hoffman, B. 2006. Inside terrorism. New York, NY: Columbia University Press.

Holsti, O. 1977. The "operational code" as an approach to the analysis of belief systems. Durham, NC: Duke University.

Hudson, V. 2005. Foreign policy analysis: Actor-specific theory and the ground of international relations. Foreign Policy Analysis 1: 1-30.

Kalyvas, S. 2004. The paradox of terrorism in civil war. The Journal of Ethics 8(1): 97-138.

Krause, P, E. Eiran. 2018. How human boundaries become state borders: Radical flanks and territorial control in the modern era. Comparative Politics 50(4): 479-499.

Kydd, A., B. Walter.,. 2006. The strategies of terrorism. International Security 31(1): 49-80.

Lazarevska, E., J. J. Sholl, M. Young. 2006. Links among beliefs and personality traits: The distinctive language of terrorists. In Beliefs and leadership in world politics, 171-184. New York, NY: Palgrave Macmillan.

Leites, N. 1951. The operational code of the politburo. New York, NY: McGraw-Hill.

Malici, A. 2008. When leaders learn and when they don’t: Mikhail Gorbachev and Kim II Sung at the end of the cold war. Albany, NY: SUNY Press.

Malici, A., A. Buckner. 2008. Empathizing with rogue leaders: Mahmoud Ahmadinejad and Bashar Al-Asad. Journal of Peace Research 45(6): 783-800.

Malici, A., S. Walker. 2017. Role theory and role conflict in US-Iran relations: Enemies of our own making. New York, NY: Routledge.

Ozdamar, O., S. Canbolat. 2018. Understanding new Middle Eastern leadership: An operational code approach. Political Research Quarterly 71(1): 19-31.

Ozdamar, O., S. Canbolat, and M. Young. 2020. Profiling leaders in Turkish. In Forum Coding in Tongues: Developing Non-English Coding Schemes for Leadership Profiling. International Studies Review, online first: https://doi. org/10.1093/isr/viaa001.

Pape, R. 2003. The strategic logic of suicide terrorism. American Political Science Review 97(3): 343-361.

Perkoski, E. 2019. Internal politics and the fragmentation of armed groups. International Studies Quarterly 63(4): 876-889.

Picucci, P. 2008. Terrorism's operational code: An examination of the belief systems of al-Qaeda and Hamas (Doctoral dissertation, University of Kansas).

Post, J. 1998. Terrorist psycho-logic: Terrorist behavior as a product of psychological forces. In Origins of terrorism: Psychologies, ideologies, theologies, states of mind, ed. W. Reich, 25—42. Washington, DC: Woodrow Wilson Center Press.

Price, B. 2012. Targeting top terrorists: How leadership decapitation contributes to counterterrorism. International Security 36(4): 9 46.

Rapoport, D. Ed. 2001. Inside terrorist organizations. New York, NY: Routledge.

Sandler, T, W. W. Enders. 2007. Applying analytical methods to study terrorism. International Studies Perspectives 8(3): 287-302.

Schafer, M., and S. Walker. Eds. 2006. Beliefs and leadership in world politics: Methods and applications of operational code analysis. New York, NY: Palgrave Macmillan.

Schafer, M., S. Robison, B. Aldrich. 2006. Operational codes and the 1916 Easter Rising in Ireland: A test of the frustration-aggression hypothesis. Foreign Policy Analysis 2(1): 63-82.

Stanton, J. 2013. Terrorism in the context of civil war. The Journal of Politics 75(4): 1009-1022.

Suedfeld, P, K. Guttieri, P. Tetlock. 2005. Assessing integrative complexity at a distance: Archival analyses of thinking and decision making. In The psychological assessment of political leader, 246-270. Ann Arbor, MI: University of Michigan Press.

Thies, C. 2010. Role theory and foreign policy. In Oxford research encyclopedia of international studies. DOI: 10.1093/acrefore/9780190846626.013.291.

Thies, С, M. Breuning. 2012. Integrating foreign policy analysis and international relations through role theory. Foreign Policy Analysis 8(1): 1 4.

Tetlock, P. 1998. Social psychology and world politics. In Handbook of social psychology 868-912. New York, NY: McGraw-Hill.

Walker, S. 1983. The motivational foundations of political belief systems: A re-analysis of the operational code construct. International Studies Quarterly, 27(2): 179-202.

Walker, S. Ed. 1987. Role theory and foreign policy analysis. Durham, NC: Duke University Press.

Walker, S. 1990. The evolution of operational code analysis. Political Psychology 11: 403-418.

Walker, S., M. Schafer, M. Young. 1998. Systematic procedures for operational code analysis: Measuring and modeling Jimmy Carter’s operational code. International Studies Quarterly 42(1): 175-189.

Walker, S., M. Schafer. 2007. Theodore Roosevelt and Woodrow Wilson as cultural icons of US foreign policy. Political Psychology 28(6): 747-776.

Walker, S. and M. Schafer. 2007. Theodore Roosevelt and Woodrow Wilson as cultural icons of US foreign policy. Political Psychology, 28(6): 747-776.

Walker, S. 2011. Anticipating attacks from the operational codes of terrorist groups. Dynamics of Asymmetric Conflict‘ 4(2): 135-143.

White, R. 1991. Empathizing with Saddam Hussein. Political Psychology 12(2): 291-308.

Young, M. 2001. Building worldview(s) with Profiler-к In Progress in communication sciences: Applications of computer content analysis, ed. M. West, 17-32. Westport, CT: Ablex Publishing.

 
Source
< Prev   CONTENTS   Source   Next >