upa-admin 10 Aralık 2016 1.443 Okunma 0

Political Scientists try to understand and explain the voting behavior of citizens for a time being. Obviously, there are several dimensions on electoral voting and political affiliation but maybe the most important one is ideology. As religiosity can be seen as an ideology, one’s religious tendency presents a good variable in social research of understanding and explaining voting behaviors. Çarkoğlu, in his article “Political Preferences of the Turkish Electorate: Reflections of an Alevi–Sunni Cleavage”, tries to look religiosity in sectarian glasses while comparing to major sects of Islam in Turkey, Alevism and Sunnism, and their possible political tendencies as a sect. As it is hard to identify sects in Islam, he comes up with a distinctive idea by using criterion-related validity and construct validity in order to diagnose being an Alevite or a Sunni. While it has got also its own obstacles and insufficiencies, Çarkoğlu’s way can be seen as a step forward for political psychology. Yet, in this paper, while summarizing Çarkoğlu’s study, I also will criticize it mostly from a political/social practice level as well as giving credit to the study in a fair sense.

Early researchers have found that Alevites are more inclined to leftist-secular parties while devoted Sunnis are most likely to vote for rightist and Islamist parties. This polarization -if you will- is also deepened by the rise of Political Islam in the second half of the 1990s. Because of the incidents took place in Maraş, Sivas, and some other places, they had a huge impact on the polarization between Alevites and Sunnis. Apart from Alevites, seculars as well, are very much suspicious about Political Islam and its leadership cadres. Secular citizens view against Recep Tayyip Erdoğan, and his governing party also something to do with the incidents occurred during the 1990s.

Ali Çarkoğlu, with the help of the early studies, forms a “multi-dimensional religiosity index” by assigning three dimensions of religiosity. He is also on the same page with the early researchers, defining religiosity is a tricky business and working has to be very detailed (Çarkoğlu, 2005: 274). These three dimensions are; faith, religious practice and religious attitudes. Every dimension has its own variations as questions asked during a survey in order to define a sample (a participant) whether he or she is a religious person or not. In faith dimension we have; “Do you believe in God?” while in religious practice the most common question is; “How often do you go to Mosque?” and in religious attitude dimension there are some questions, from “Do you support girls covering their heads in college” to “Do you support sharia law?”, in which samples mark their answers by ratio measure examples (Çarkoğlu, 2005: 277-279).

Çarkoğlu uses few questions while understanding whether a sample has a faith or not. However, in those questions, one important one is lacking. I think, “Do you believe in Muhammad is the prophet and messenger of Allah?” as well as “Do you believe that Quran is Allah’s holy book?” should be part of the evaluation in order to understand sample’s religiosity.

Although, statement regards to religious practice seem to be sufficient one can think of some other solid questions in order to evaluate religiosity of a sample. As the nominal measurement is more reliable than interval or ordinal measurement, the researcher should have asked “Do you read Quran periodically?” or one might also have asked to woman samples “Do you think head-covering is essential in Islam?” since headscarf seems to be a part of religious expressionism. Maybe headscarf question is manipulative or delicate in a political sense but still one should also have asked: “Do you fast?” (Addition of during Ramadan or Muharrem may be needed while dealing with different religious sects), “Do you practice necessities of Kurban Bayramı?”.

Some of the questions asked by Çarkoğlu to evaluate the religious attitude of samples are also not valid in the sense of religiosity. For example, the question “Do you support girls should wear a headscarf in the university if they want to?” lacks to measure samples’ religiosity (Çarkoğlu, 2005: 277). A liberal minded person may also support a girl’s freedom to cover her hair in university premises. On the other hand, answers to the question about whether a pious Muslim is more trustworthy compared to a non-religious person seems shocking to me. Only 22.6 percent of the participants totally agree on the assumption while 27.1 percent of them totally disagree that a religious person is more trustworthy in commerce than a non-religious person. Since “being a devoted Muslim” can open several doors in the country, as well as facilitate to become an MP, the answer that Çarkoğlu had got from this particular question simply baffled me. Besides the specific dimensions and their potential statements, as Çarkoğlu also suggested, the measurement is highly subjective to a Sunni tendency (Çarkoğlu, 2005: 277). Because of the secular tendencies of Alevites particularly resourced from the incidents in the 1990s and the polarization of the two sects, it is highly expected that an Alevite sample would  do less in religiosity measurement. With help of the multi-dimensional measurement, it is unlikely that a pious Sunni would score little while an Alevite score high. However, there is also a challenge to identify an Alevite or make a sample to reveal his or her sectarian choice without any hesitation (especially when he or she is from a minority sect).

Çarkoğlu’s next move is to portray being an Alevite. Since direct question whether a sample is being an Alevite or not has its own complexities, sample hiding his or her genuine identity or even not knowing well about his or her own identity (Çarkoğlu, 2005: 281), Çarkoğlu came up with an idea which enables him to understand the sectarian affiliation by asking indirect questions. First, Çarkoğlu forms two groups of religious figures which belong to Sunnism and Alevism. While in the first group, we have names like Caliph Omer, Caliph Ebubekir, Caliph Osman and Said Nursi; in another group, one can find names like Caliph Ali, Hacı Bektaşı Veli, and Ahmet Yesevi. The idea is that a sample from an Alevite background would choose Caliph Ali, Hacı Bektaşi or Ahmet Yesevi as an important figure in Islam while a sample from a Sunni/Hanafi orientation would choose figures from Group 1 (Çarkoğlu, 2005: 282). After this exercise, it’s been seen that even if the direct questioning of the identity assumes 3.3 percent of the participants are belongs to Alevite sect, the result of the second exercise gives us an outcome of being attached to Alevite tradition is nearly 15 percent (Çarkoğlu, 2005: 282).

Since participants are allowed to choose two characters and the possibility of a Sunni might have chosen Caliph Ali also as an important figure in its religion, even the outcome of this measurement can be perceived as reasonable, other nominal measurements are in need. According to Babbie, yes or no questions (or nominal measure dimensions) are more reliable in order to understand the true opinion of a sample (Babbie, 2007: 140). For example, questions like, “Do you favor Ali or Muaviye?”, “According to you, Hacı Bektaşi is an important figure in Islam or not?”. I think the questions with multiple selections, like Çarkoğlu, asks in his survey, might be complicated for a participant to answer according to his/her sectarian affiliation but still the results that Çarkoğlu received from this exercise seemed very much common-sensical.

According to Çarkoğlu, another possible way to differ an Alevi from Sunni is to ask whether participants have pictures of Al-Aqsa, Kaaba or some characteristic religious sites hanged upon their wall or not. It is most likely to expect that an Alevi does not have a picture of these place while a pious Sunni probably have. Nevertheless, it is very much likely a non-partisan Alevite to have a picture of a mosque on his or her wall while it is also likely to have no pictures of any religious site by non-partisan Sunnis. From Çarkoğlu’s point of view, another possible question might be “Do you have religious art is your house?” (since Islam does not allow art other that architect or calligraphy a researcher could ask “Do you have religious calligraphy in your home?”) in order to differentiate questions and break down the defense mechanism of the sample, we can ask some other questions like, “Do you wear an amulet of a charm?”, “Do you have a car decoration shaped as a verse of the Quran?” or even “Do you have an Ottoman tuğra bumper stick in your car?” because it seems like if a sample answers yes to these questions one can be highly suspicious of him or her being a Sunni. Maybe bumper stick question is little on the edge. However, bumper sticks are very much hip nowadays and a researcher could easily assume a driver with an Ottoman tuğra sticker can belong to Sunni sect (even if it has got its own complexities).

Çarkoğlu then asks samples if they have a picture of a religious leader in their homes in order to single out the Alevite population. Since it is frequent for Alevite community to have picture of Caliph Ali’s or 12 Imams’ in their households, while Sunnis normally do not have any religious figures pictures hanging on their walls,  it can be said that the idea is clever enough to identify Alevite sectarians without making them uncomfortable (Çarkoğlu, 2005: 283). However, as Çarkoğlu also points out a Sunni might also decorate his house with a picture of Mevlana or today’s religious Sunni leaders, the idea has loopholes. Besides, I am very much suspicious about some people in Turkey could consider a particular prominent political figure’s picture as the picture of a religious one, since he is very much in the spotlight with his skills of traditional Sunni Islam.

After collecting data, Çarkoğlu has created a ratio measurement for samples in order to their Alevite tendencies. Out of three questions, one can either have “0” or “3” as his or her Alevite score, according to their positive or negative answers regarding statements which are asked by the researcher. After analyzing the collected data, Çarkoğlu finds out that 73 percent of the participants have “0”, 23 percent of them scores “1”, 4.7 percent are associated with “2” and 1.4 percent of the samples scored “3” (Çarkoğlu, 2005: 283). Despite 3.3 percent of the samples present themselves as Alevites in direct questioning only 1.4 percent of them turned out to be associated with all three directions. The idea basically is that samples that get “3” are most likely to be Alevites while one who get “0” are definitely not Alevites (Çarkoğlu, 2005: 283). As a matter of fact, there is no way to know who is Alevite and not from this survey. Questions may be good enough so far, neither measurement is necessary causes for being an Alevite or even Sunni.

According to Çarkoğlu, his work is very much successful in construct validity and he further wants to examine his results with three religiosity dimensions and being an Alevite. From his point of view, it can be said that when a sample’s Alevite direction augments his or her likelihood to go to mosques or generally conservatism, decreases (Çarkoğlu, 2005: 285). Although Çarkoğlu also mentions in his article, dimensions that promotes the cleavage between two sects are more likely to draw distinctions between secularists and pious Muslims (Çarkoğlu, 2005: 285-286), still percentages that are contracted from the survey tend to be appropriate. As already mentioned, these dimensions are heavily linked with Sunni practices and it is quite problematic for me to understand an Alevite’s religiosity while using those dimensions.

Lastly, Çarkoğlu studies on “criterion-related validity check” for Alevites and their potential tendencies to the political parties which entered to 2002 general elections. The study was more of a post-election survey and “which party did you vote for 2002 general elections?” are asked to all 1174 samples (Çarkoğlu, 2005: 287). Çarkoğlu unifies his early survey with this new one and he comes up with a resolution that when Alevite directions are augmenting sample’s lean towards AKP decreases and CHP increases in replacing. Alevites’ share in AKP voting turns out to be 12.5 percent while it is 66.7 percent in CHP’s voting phenomenon (naturally, Alevite votes are deducted according to 0-3 ratio Alevite directions in the earlier study) (Çarkoğlu, 2005: 287).  Some other results from the study are; AKP holds a more significant non-Alevite electorate from any other party and while one-third of the CHP voters turn out to be from Alevite tradition, only one-fourth of the AKP voters tend to be Alevites (Çarkoğlu, 2005: 287).

To sum up, Çarkoğlu mentions the difficulty separating a sample’s sects and their possible party affiliations (Çarkoğlu, 2005: 288), and also as a number of questions increases he is confident that the finding will be more appropriate and valid (Çarkoğlu, 2005: 288). I also mention in the paper that if we multiply the questions we can stumble upon to more reliable and valid answers and analyzing party affiliations and religious sects of samples also becomes easy. On the other hand, Çarkoğlu also suggests that in the next studies, three dimensions (faith, religious practices, and attitudes) need to be reshaped according to Alevite context (Çarkoğlu, 2005: 288). In this study, three dimensions that supposed to measure religiosity of a sample are highly biased and they are formed to understand from a Sunni religious perspective. A researcher has to appreciate that there is a different religious practice, faith, and attitudes between a Sunni pious and a traditional Alevite, and it has to be clarified in next studies.

Although Çarkoğlu, claims that Alevites constitute a significant amount of voters in two largest parties and they can easily push for changes or implement their own policy agendas (Çarkoğlu, 2005: 288), day by day the cleavage between sects either in social or political life opens. Nevertheless, relatively strong Alevite civil societies have necessary networks to make policy changes in political parties (mostly in CHP) and yet, the possibility for a resolution through this channel stays ambiguous. Then again, Çarkoğlu’s opinion on AKP’s “awareness of sectarian sensitiveness” (Çarkoğlu, 2005: 289) might be the case in the early 2000s, but since EU had stopped being an anchor for Turkey’s democracy and human rights, minority rights and level of democracy tend to be put in deep in dark. As AKP governments develop more authoritarian characteristic, polarization between Alevites and Sunnis, between Kurds and Turks, Conservatism and Secularism, also strengthens. Like all freedoms, religious freedom also crumbles and Turkey faces a new modeling with Sunni Turkish-Islam connotations.

Scientific sides and methodology of Çarkoğlu’s “Political Preferences of the Turkish Electorate: Reflections of an Alevi–Sunni Cleavage” are in order while one can see some insufficiencies from a practical perspective. Some questions which are asked and measurements that are created tend to be not addressing to a distinction between an Alevite and a Sunni/Hanafi good enough. Although the percentage of belonging to Alevite tradition and their leaning towards to CHP sounds accurate the index of next surveys has to be reshaped to be more appropriate in the sense to understand the religiosity of an Alevite and a Sunni origin. Çarkoğlu also seems to be naïve while contextualizing AKP’s opinion towards religious freedom and democracy, his out of luck is mostly resourced by AKP’s different outlook in the early 2000s while EU was a pressure point to Turkey’s democratization route, and new studies have to consider this new political atmosphere we are living in Turkey.

Basri Alp AKINCI


  • Babbie, Earl, “The Practice of Social Research”, Thompson Wadsworth, 2007.
  • Çarkoğlu, Ali, “Political Preferences of the Turkish Electorate: Reflections of an Alevi–Sunni Cleavage”, Turkish Studies, Vol. 6, No. 2.

Leave A Response »

Time limit is exhausted. Please reload the CAPTCHA.