Issues of Orientalism in Terrorism Research: A Critique of Barros and Proenca’s Research Terminology

Terrorism has become an important topic, especially after the 9/11 terrorist attacks on several locations within the United States. Since then, many articles in journals have attempted to discuss the new terrorism wave without fully understanding it. The ‘orientalist’ approach to the issue seems to be one of the major obstacles that prevent people from comprehending this phenomenon and accomplishing rigorous research on this matter.

Barros and Proenca’s (2005) article entitled Mixed Logit Estimation of Radical Islamic Terrorism in Europe and North America seems to reflect this epistemological problem. As do many other writers, the Portuguese economists Barros and Proenca continuously use the terms ‘Islamic terrorist,’ and ‘Islamic terrorism’ in their article, albeit often indiscriminately prefaced by the word radical or extremist. Furthermore, they define the terrorist actions using the labels ‘Islamic event’ and ‘Islamic attack’ (2005, p. 299, 304). They were, however, unable to distinguish religiously motivated terrorists from secular ones although they proposed to examine characteristic features of religiously inspired terrorists’ attacks. They do acknowledge that “this is a statistical, rather than a behavioral model. . . . We therefore use no motivational variables, rather relying on a statistical model that fits the probability of Islamic terrorist actions, given their particular attributes” (p.306).

Thus, it seems that they treat secular terrorist organizations as being radical Islamic if they emerge in Muslim-populated countries, or if their members’ ethnicities belong to those countries. Even if the authors do not have the intention of doing so, these kinds of misconceptions lead to stereotyping and present a negative connotation to a specific part of the world, that which was once called ‘the Orient’ and more recently refers to societies in the Middle East, North Africa, India, and Pakistan.

Despite the evolution, and as such, confusion, over time, of the Orient referring to the geographical far East (e.g., Japan or China) or the near East, this ‘middle Eastern’ geographical region is still referred to as such in orientalism literature. This essay will focus on some aspects of Barros and Proenca’s (2005) study that might be indicators of such orientalism. This type of epistemological problem can be one of the challenges encountered by any researcher in studying terrorism, but not a difficult one. The problem, as I see it, comes to the surface with the language used by the authors.

Naming or classifying terrorist attacks as “Islamic events,” is problematic, as is using labels such as “Marxist-Leninist events” or “liberalist events”, when referring to terrorist attacks. By choosing such a general term for a specific action, various groups who follow the religion or adhere to the ideology but who are not terrorists still seem to be indicted under the label.

Indeed, it may be important to avoid such ideological labeling for the simple purpose of not leading them to radical actions or not pushing them into extremism. Religion, however, is more complex. Religion is a more inclusive term and different than ideology. There are religions that have millions of followers. For instance, followers of the two largest religions of the world, Islam and Christianity, account for over 50% of the world’s population--Islam with an estimated 1.5 billion adherents and Christianity with 2.1 billion adherents (Wikipedia, 2005). However, within these large organizations of faith, there exist many variations and sects, so that the umbrella term of Christianity, for example, may be referring only nominally to groups who have very different beliefs and creeds (i.e., Pentecostals vs. Unitarians). Therefore, the identification of a religiously inspired but specific group’s terrorist attacks as “Christian events/attacks” and “Islamic events/ attacks” does not seem to be useful, as it is too broad, and does not clearly identify the specific set of beliefs the group has, nor the justifications used for the violence. Indeed, religion itself is generally not a cause of terrorism; rather, in order to take advantage of its popularity, religion is used as a tool or hook by many terrorist groups. These groups emerge as if they are acting on behalf of particular religions, such as Islam, Christianity, Judaism, or Sikhism (e.g., Ku Klux Klan, Al- Qaeda, Jewish Defense League, and Babbar Khalsa). Within each, religion is used as one of the group dynamics (ideology) that keeps members united for the achievement of the goal of the group, and it is also used as a tool for recruiting new members as well as gaining supplies and support from a larger group or society. Therefore, some terrorist groups that have a clearly religious creed use religious terminology in their manifestos and identify themselves with names that encode a religious word, title, or even the name of a religion, such as the Lord’s Resistance Army (formerly known as Uganda People’s Democratic Christian Army); the Islamic Great East Riders/Front; and the Jewish Defense League. Even terrorist organizations with a secular ideology use religious language to their advantage in specific regions. For instance, the PKK (Kurdistan Workers Party) which adheres to a Marxist-Leninist ideology established a subgroup called the Kurdistan Islamic Movement (KIH) under its command in the early 1990s in the areas of southeast Turkey in order to obtain new recruits, support, and materials (Demir, 2007; see also Belgenet.com; Sucbilimi.org).

Although the Muslim religion is not alone in being victimized by terrorists, it is apparent, especially after 9/11, that Islam and its abusers have become the focus of public interest. The new wave of terrorism (ostensibly in the name of Islam) has hit many cities worldwide, such as Bali (Indonesia), Mombasa (Kenya), Riyadh (Saudi Arabia), Casablanca (Morocco), Istanbul (Turkey), Madrid (Spain), Amman (Jordan), Cairo (Egypt), and London (UK). The casualty in human life and destruction has been enormous compared to previous terrorist incidents.

Much of this is attributed to Al-Qaeda and Osama Bin Laden. In the beginning Bin Laden ‘declared war’ against the United States (US) and its western partners; later, however, he also threatened Muslim countries, telling them not to cooperate with the US and the ‘West’. Thus, his network organization and connected cells in the Muslim countries hit their targets and killed hundreds of innocent people in the Mideast indiscriminately, heedless of the victims’ religions or ethnicities. There has been no differentiation between countries in these terrorist attacks.

Within this context, Barros and Proenca’s 2005 article focused on terrorism and terrorist incidents that abused the religion of Islam and occurred between 1979 and 2002. The econometric-model study tries to accomplish several things: (1) to estimate the characteristics of those kinds of radical Islamic terrorist attacks; (2) to try to find the probability of a terrorist attack being perpetrated by an extremist Islamic organization based on the attacks’ characteristics; and (3) compare this kind of terrorism in some European countries, in the US, and in Canada (p. 299). The authors note that “the choice of countries is based on the observed emigration from Islamic countries to the industrialized western nations” (2005, p. 305).

Overall, Barros and Proenca’s language in the article reminds readers of the separateness that comes from the orientalist perspective that some academic studies suffered from in decades past, and still surfaces today. The Orient is a Western description of “the other” world, and, as stated previously, includes areas of the Middle East, North Africa, India and Pakistan which were colonized by European countries. Orientalism has been defined as “a Western style of dominating, restructuring and having authority over the orient through a system of knowledge which develops a positional superiority” (Said, 1979, p.3). Said argues that “European culture gained strength and identity by setting itself off against the orient as a sort of surrogate and even underground itself” (1979, p.3). Turner (1989) considers this process as “othering” and says that it is an attempt to create stereotypes, such as east versus west, us versus non-Europeans, rational versus irrational, or progressive versus unprogressive. The reality is then misinterpreted through these stereotypes. In other words, it is a process of depreciation, of seeing the other as a negative and ‘othering’ those people in the Orient. Expressions such as “Islamic attacks” and “Islamic events” serve for this process, often thought of in simple terms as The West (or Christians) versus The Orient (or Muslims).

This reduction to ‘us (non-Muslims, the West) being attacked by them (Muslims, the East)’ is clearly too simplistic, and yet, once the terminology is in place, is challenging to correct. The erroneous use of the terms ‘Islamic attack’ and ‘Islamic event’ create a misconception, especially to the Western mind, that Islam equals terrorism, promotes violence, and encourages hatred of anything non-Muslim. Thus, these expressions serve as a means of negating Muslims even though the writers may not have that intention.

However, to put an end to such erroneous thinking, researchers must know that inherently, terrorist attacks can be neither Christian nor Islamic. And, as this paper specifically addresses, terrorists and acts of terrorism are not possible, nor should they be addressed as such, as actions within the Islamic faith, since the Islamic faith does not support either terrorists or terrorism (for Islamic faith and terrorism see Capan, 2004; see also Bloom, 2005). “Islamic events” according to my Muslim faith, are celebrations, prayers, and tasks which are performed as requirements of the religion.

The other error that Barros and Proenca make is that they do not distinguish religiously inspired terrorists from secular ones, although the dependent variable in their study is a dummy variable of whether the terrorist organizations are, according to them, “Islamic or non-Islamic”. They attempted to predict the probability of the occurrence of this type of terrorism and to predict the characteristic of this type of incident with this dependent variable. However, many Marxist-Leninist, and other secular organizations, some of whom have been identified as the most active terrorist organizations, are lumped in with and treated as radical Islamic organizations. Therefore, the findings in this study based on this dependent variable are biased.

The secondary data that are subject to the analysis of the article do not distinguish perpetrator organizations or individual perpetrators relative to their ideologies; the writers did this in order to form the dependent variable. Barros and Proenca (2005, p.306) mention that they excluded the PKK (identified as the Kurdish People’s Liberation Front; note however that the accepted, correct translation of the name for this terrorist organization is Kurdistan Workers Party) from the radical Islamic groups because it is a well known secular organization. However, they wrongly included many other well-known secular terrorist organizations as “radical Islamic terrorist organizations”, such as Yaser Arafat’s Fatah and PLO. Although a complete, comprehensive exploration of the groups was beyond the scope of my critique, through a random internet search, I found that the following organizations should be excluded from their list because none of them has a religion-based ideology. Instead some are nationalist, some Marxist and Leninist and so on, but the primary point for consideration is that all are secular.

• Popular Front for the Liberation of Palestine (PFLP)
• Al Fatah
• BSO Black September
• PFLP-GC PFLP-General Command
• Palestine Liberation Organization (PLO)
• Fatah Force 17
• Fatah-Revolutionary Council Abu Nidal Gr [sic]
• Al Asifah headed by Abu Nidal
• Black June headed by Abu Nidal
• Movement for Rebuilding Fatah
• Revolutionary Organization of Socialist [sic]
• Indeterminate Abu Nidal guerrillas
• Palestine Liberation Front (PLF)
• Martyrs of Palestine Abu Nidal group
• Jammu and Kashmir Liberation Front

It appears that the authors selected organizations which emerged in Muslim populated countries, or chose organizations whose members came from Muslim societies.

Similarly, the writers do not distinguish radical religious groups from non-radicals. Regarding their dependent variable, they infer that if the group is on the ITERATE list, is fits the variable’s requirement of an attack being perpetrated by radical Islamic terrorists:

We have chosen the Islamic groups. Some of these Islamic groups may not be radical in their nature, but since we are considering terrorism in Europe, the USA and Canada, we assume that almost all Islamic groups perpetrating terrorist attacks in the western countries are radical (p. 306).

This is quite a serious subject for such a broad assumption. Indicating that some of them may not be radical implies that nonradicals also may commit terrorist acts. I wish the writers had more specifically and clearly defined radicalism in order to understand what this expression really means as used within their context. As I understand it, all terrorist organizations that use religion as an ideology are considered to be radical and extremist. If we do not distinguish the difference between the radical and non-radical and treat them the same, the identical concepts again comes to mind, i.e., that all Muslims are terrorists, since both non-radical and mainstream Muslims might commit terrorism accordingly.

In addition, the authors restrict their examination to incidents which occurred only in the US, Canada, and some European countries, but do not explain this narrow focus as well as they might. Barros and Proenca provide reasons why certain countries are chosen: (1) the concept of Western/capitalist societies being prime targets; (2) an analysis of “terrorist events on both side of Atlantic, which are tied by the same market system as well as the same threat” (2005, p.299); and (3) “the choice of countries based on the observed emigration from Islamic countries to the industrialized Western nations” (2005, p. 305.)

However, there is no exact definition of ‘Western society’ in the article. Whether all or just certain societies who have a capitalist system are considered as Western is not clear. Although the authors focus on the societies which have the same market system, the study also includes non-capitalist countries (at the time of events) too, such as the former German Democratic Republic, the totalitarian socialist state of East Germany; and a theocratic state, The Vatican. Therefore, it seems that Barros and Proenca talk about a larger identity which forms Western society. The other countries in this group include Austria, Belgium, Canada, Denmark, Finland, France, (including Corsica), Germany, Greece, Iceland, Ireland, Italy, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom (including Northern Ireland), and the United States.

Similarly, the authors do not provide information about how those countries are chosen considering Muslim emigration. Do they have data of immigration numbers? Which countries have what ratio of emigration? Neither answer is mentioned. It seems that they intend to reach some conclusions regarding countries which receive ‘the others’, Muslim immigrants. Thus, it sounds as if it is part of the othering process- the others (Muslims) attack the West. As we add up some of the incidents with mass casualties, we discover that religiously motivated terrorism hit not only certain European countries and the US, but also countries in the Middle East, in Africa, and in the Far East. Indeed, it seems that countries having a majority Muslim population did not suffer fewer attacks than the other societies during this period. Examining it in a wider range all over the world could be more helpful for the prediction purpose.

Another point to consider is that Barros and Proenca seem to approach the issue from an historical perspective, but provide a very cursory review of the historical background. For example, the article mentions that the first terrorist groups in the Middle East were affected by confrontations during the cold war. However, it does not address other influences of historical episodes that occurred in the Middle East. These influences, such as occupations of territories, colonization, and oppressive governments, must also be analyzed to provide a more complete, accurate view. Nevertheless, they provide only one way of looking at the historical factor. This narrow, economic approach might suggest that the authors may not want to consider the effect of the west upon terrorism since this could be viewed as the negation of ‘West’. The negated subject must be the Orient, not the West, according to the orientalist approach.

Other problems in Barros and Proenca’s (2005) study can be stated as follows: Examining terrorist events based on type of attacks, such as armed attacks, arson, bombings, kidnappings, letter bombs, and skyjackings may not provide an accurate understanding of the characteristics of this kind of terrorism. Terrorist organizations usually imitate other organizations’ actions (Oots, 1986; Pape, 2003). For example, Pape (2003) asserts that after the 1983-Beirut suicide attacks against the US Marine barracks, many terrorist organizations, from radical religious groups to secular and ethnic groups, copied and employed suicide attacks in order to compel certain countries to withdraw from certain territories. Thus, relying on the examination of these incident’s characteristics only does not seem to be a comprehensive method to fully analyze the characteristics of a certain type of terrorism. Terrorism is used as a means to achieve certain political goals, and its specific tools are not exploited by only certain groups. It is true that Barros and Proenca’s (2005) goal is not to understand the characteristics of this kind of terrorism. Rather, they include this variable (type of attack) to their regression model in order to obtain predicted probabilities regarding expected terrorists attacks of a specific type of terrorism, and note, as mentioned earlier, that motivation is beyond the scope of their research (p.305). However, these kinds of predictions might also require more information, including an understanding of these types of terrorist organizations (e.g., their “modus operandi,” ideology, capability, leadership style and personality of leader) and knowledge regarding sociopolitical conditions of a certain region that is under examination.

Also, Barros and Proenca seem to have a generalization problem. Based on the unspecified dependent variable, the authors compare terrorist attacks in these countries between 1979 and 2002, and state that “radical Islamic terrorist attacks constitute 13.8 % of all terrorist attacks” (p. 300). Of course this number can not reflect the exact rate because of the measurement error that is mentioned above. Furthermore, even if the study did not have such an error, this kind of generalization would be worrisome since the study covers only transnational terrorist events, not ‘all terrorist incidents’. Mickolus et al.’s (2004) data set (the ITERATE project) that is used for the analysis of the study does not include other kinds of terrorist incidents such as state terrorism or domestic terrorism. We need to consider that domestic terrorist incidents usually comprise the majority of terrorist activities in a country, and the data include transnational terrorist incidents which are conducted by individuals or groups only. Therefore, making this kind of generalization under these limitations does not seem to be appropriate.

Conclusion

First, we should avoid using such terms as Islamic attacks, Islamic events, and Islamic terrorists because this promotes the conception of an othering process against Muslims and the religion of Islam. In general, scholars must be careful in their usage of language that sounds stereotypical of a nation, an ethnicity, or adherents of a religion to avoid conveying a sense of orientalism or othering. Second, concepts that are used in a research project must be well defined before being put into operation. The critique above shows that Barros and Proenca (2005) were not specific enough in their definitions regarding the types of terrorism that they address as ‘radical Islamic terrorism’. Because of this flawed initial premise, they were not able to properly identify the unit of analysis of the study. This type of measurement error will most certainly lead to a variety of biases.

Finally, the type of terrorist attack is an important piece of any research that may be trying to obtain predictions regarding terrorists’ future attacks. However, imitation among terrorist organizations makes it unreliable for predictions of those attacks based on this variable alone. For the purpose of prediction, we need more comprehensive models that include many more variables such as characteristics of particular terrorist organizations (e.g., leadership style, ideology, capability, modus operandi) and the sociopolitical conditions of the target country.

*4th Degree Turkish Police Chief and Ph.D. student at Kent State University, Department of Political Science.

References

Barros, C. P., & Proenca, I. (2005). Mixed logit estimation of radical Islamic terrorism in Europe and North America: A comparative study. Journal of Conflict Resolution, 49(2), 298-314.

Bloom, M. (2005). Dying to kill: The allure of suicide terror. New York: Columbia University Press.

Capan, E. (Ed.). (2004). An Islamic perspective: Terror and suicide attacks. New Jersey: The Light.

Demir, H. (2007). PKK dini nasil propaganda haline getirdi. Retrieved on September 16, 2007, from http://www.turquie-news.fr/spip. php?article293

Mickolus, E. F., Sandler, T., Murdock, J. M., & Flemming, P. A. (2004). International terrorism: Attributes of terrorist events, 1968-2003. Data code book for the data compiled for ITERATE project.

Oots, K.L. (1986). A political organization approach to transnational terrorism. Connecticut: Greenwood Press.

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

Said, E. (1979). Orientalism. New York: Vintage.

Turner, B. S. (1989). Research note: From orientalism to global sociology. Sociology 23(4), 629-638.