This is an open access article distributed under the terms of the Creative Commons
Attribution License (Any account of the past is “a particular bundle of silences” (Trouillot, 1995, p. 27) that necessarily focuses on some events but excludes others. With respect to collective memory, representations of national history emphasize some aspects of the collective past and render others silent. Indeed, as Renan (1990, p. 11) famously observed, “Forgetting … is a crucial factor in the creation of a nation,” especially with respect to minority populations or previous occupants of the land whose existence would challenge the just-natural character of national identity. Forgetting or denial of minority experience not only affords a construction of national identity based on ethno-cultural criteria, but also may promote support for nation-glorifying action in service of national identity (Kurtiş, Adams, & Yellow Bird, 2010; Roccas, Klar, & Liviatan, 2006).
The purpose of this research is to examine the relationship between a particularly contested case of silence and construction of national identity at the level of both cultural representation and individual response. Official representations of Turkish history provide an appropriate case for the study of silence in collective memory and identity for several reasons. First, Turkey features prominently in international discourse as “an archetype of social amnesia” (Bakıner, 2013, p. 6) and a paradigmatic example for state-sponsored denial (Cohen, 2001) of past misdeeds. Even so, psychological research on implications of silence within Turkish history remains relatively scarce. Second, the majority of international attention about silence in Turkish history focuses on a particular event and era: specifically, conflicts with Armenians at the end of WWI. Yet, scholars in Turkey note how silence has been integral to the Turkish nation-building project since the founding of the state (e.g., Çetin, 2004; Keyder, 1997). Finally, history education in Turkey—a primary vehicle for the transmission of collective memory and national identity (e.g., Apple, 1989; Wertsch, 2002)—operates under the influence of central institutions of power (Uygun, 2008).
Besides expanding existing research on collective memory and national identity in the Turkish context, a novel feature of the present research is that it examines silence in collective memory at the level of both cultural representation and individual experience. At the level of cultural representation, we investigate manifestations of silence regarding minority populations in the history section of national university admissions examinations, which are based on the high school history curriculum. At the level of individual experience, we investigate associations between ethno-cultural constructions of national identity, tendencies to deny historical events about minority populations that official accounts of history omit, and support for policies that may threaten dominant conceptions of national identity (i.e., those related to extension of rights to minority populations, and freedom of expression that may constitute threats to dominant constructions of national identity).
A third important feature of the present research is that rather than adopting a unidimensional model of national identification, as in the case of most previous research, we focus on particular constructions of national identity. An ethno-cultural construction of national identity is a less inclusive form of identity that limits membership in the nation to blood, ancestry, or shared cultural characteristics such as language and religion (Pehrson, Vignoles, & Brown, 2009; Shulman, 2002; Wakefield et al., 2011). A civic construction of national identity offers a more inclusive base for membership in the nation that includes citizenship and adherence to the laws and civic institutions (Pehrson et al., 2009; Shulman, 2002; Wakefield et al., 2011). The present research specifically examines how endorsement of an ethno-cultural conception of national identity relates to denial of minority-related historical events (that are already omitted from official sources), and support for policies that may threaten this particular conception of national identity.
Collective Memory and National Identity
Social scientists have long emphasized how collective memory both shapes and serves interests of collective identity (e.g., Durkheim, 1912; Halbwachs, 1992). Similarly, social psychological research has suggested a bi-directional relationship between national identity and representations of history, which we define as repositories of collective memory embedded in cultural tools such as official memorials, museums, and commemorative holidays (Kurtiş et al., 2010; Liu & Hilton, 2005; Liu, Wilson, McClure, & Higgins, 1999).
In one direction, representations of history provide a basis for national identity. People rely on these representations to construct a national identity as a sense of imagined community with others distant in time and space (Anderson, 1983). Different representations of history afford particular constructions of national identity and forms of identity-relevant action. For instance, celebratory representations of history—such as observations of the U.S. Thanksgiving holiday that mention English colonists but omit mention of the atrocities that they committed against Indigenous Peoples—promote nation-glorifying forms of identification (Kurtiş et al., 2010). Conversely, critical representations of history challenge such forms of nation-glorifying identification and undermine associated forms of action (e.g., readiness to serve or die for nation, Reicher & Hopkins, 2001).
In the other direction, representations of history are not mere reflections of objectively recorded events. Instead, they are cultural tools that bear the identity interests of their producers. People do not engage with accounts of the past in neutral or objective fashion; instead, they do so in ways that resonate with their experience of identity. For example, strong collective identification is associated with the tendency for people to recall few instances of collective misdeeds (Sahdra & Ross, 2007), to prefer celebratory portrayals of the past over more critical portrayals that focus on past injustices (Salter & Adams, 2016), and to interpret events in other ways that defend against identity threats (e.g., collective guilt, Branscombe & Miron, 2004; Wohl, Branscombe, & Klar, 2006).
Cultural representations of history are often key sites of contestation as different identity groups reconstruct accounts of the past to advance present purposes (e.g., Kubal, 2008). Dominant representations of history emerge and evolve through differential exercise of power as people in institutional roles act on collective identity motivations to depict the past in ways that serve collective beliefs and goals. As a result, dominant representations typically share a common feature: namely, they tend to emphasize collective triumphs while silencing collective misdeeds (e.g., Loewen, 1995; Trouillot, 1995). Research suggests that collective silence and denial of past events are prevalent among groups who have committed atrocities against other groups (Baumeister & Hastings, 1997; Branscombe & Miron, 2004), resulting in a conspiracy of silence and denial (Cohen, 2001).
Silence in Official Representations of Turkish History and National Identity
The modern state is unique in its access to resources when it comes to shaping a usable past and restricting competing narratives (Wertsch, 2002). A central institution the state has at its service is formal education. Formal education serves a key function in the construction and legitimacy of the modern nation-state (Benavot, Cha, Kamens, Meyer, & Wong, 1991). History education assumes a particularly important role in socializing people to develop an experience of national identity that serves interests of political elites (e.g., Carretero & Kriger, 2011; Çetin, 2004).
Within international discourse, Turkey frequently appears as a prototypical case of state-sponsored denial and silence about past atrocities. Although most discussions focus on conflicts with ethnic minorities at the end of WWI, scholars of Turkish history situate silence in a broader context of events associated with Turkey’s transition from the multiethnic Ottoman state into a republican and secular nation-state. Central to this transformation was the re-writing of history and the re-imagining of a new national identity. In both processes, state elites assumed the role of primary architects, while “the masses remained silent partners” and “passive recipients” (Keyder, 1997, p. 43).
Constructions of Turkish History
A key agenda of the Turkish nation state in its early days was to sever all ties with the Ottoman past and promote secular, nationalist, and modernist tenets of Kemalism (e.g., basic ideas and principles of Mustafa Kemal Atatürk, the founder of the new republic) via history and social studies education (Swartz, 1997). The Ministry of Education, established in 1926, formulated and disseminated the Kemalist message via educational policies and curricula (Childress, 2001; Kubicek, 1999). In fact, formal education constituted the primary means to construct and disseminate a sense of Turkish national consciousness, identity, and patriotism (Tarman, 2006).
One urgent task for curricular revision involved reconfiguring the identity of the Turkish state in relation to religion. During the Ottoman era, Turks identified primarily with religion, in particular Islam (e.g., Turk = Muslim, see Waxman, 2000). Yet, the new Kemalist regime demanded that religion would play no role in the new identity of the Turkish state, and a key area of curricular change concerned a process of secularization to break ties to the “Islamic world” (see Eskicumalı, 1994). Several decades later, the transition to the multi-party regime in the 1940s coincided with a move towards a synthesis of Turkish nationalism and Islam as tenets of Turkish identity (Tarman, 2006). Particularly since the AKP (i.e., Justice and Development Party) government came to power in 2002, scholars observed a shift in official memory from the dominant Kemalist paradigm to a paradigm of neo-Ottomanism. Sakallıoğlu (1996) defines the new educational agenda as a case of “Turkish-Islamic synthesis” that reinterprets the past from the perspective of Muslim-Turkish nationalism. The recent rhetoric emphasizes ties with the Ottoman past, constructs the nation in ethno-cultural terms (e.g., as “Sunni and Turkish”), and promotes state-sanctioned “patriotic” versions of the past (e.g., Bakıner, 2013).
Conceptions of Turkish Identity
Popular discourse depicts Turkey as a “bridge between two worlds” (Kinzer, 2006) or a “country caught between two continents, between two traditions, two trends of history” (Financial Times, 1998, p. 4). Similarly, scholars within and outside of Turkey often frame the question of Turkish national identity in terms of an “identity crisis” (e.g., Ozay, 1990; Waxman, 1998). Some scholars attribute this crisis to the top-down manner in which state elites constructed Turkish national identity, like representations of Turkish history, in monolithic terms that disregarded popular sentiment. In particular, the 1924 constitution endorsed a civic definition of national identity based on shared citizenship and civic participation (Yeğen, 2004). Yet, scholars note how the predominant construction of national identity in everyday discourse is one that emphasizes an ethno-cultural definition based on “Turkishness”—or more aptly, Sunni Muslim-Turkishness—in spite of Turkey’s commitment to secularism and its past and present diversity (e.g., Özkırımlı, 2006). Moreover, discrimination based on religion, ethnicity, and language is a persistent problem in Turkey (Çağaptay, 2006; Kirişci, 1998; Oran, 2000; Yeğen, 2004).
The Present Research
This necessarily brief review reveals the extent to which silence—including selective emphasis and active omission—helps constitute constructions of national history in the particular case of Turkey (as one example among many). The present research examined how silencing of historical events is associated with constructions of national identity at the level of both cultural representation and individual experience. We first focused on how official accounts of history omit mention of diverse populations in Turkish history. We then investigated the relationship between civic or ethno-cultural conceptions of national identity, silence about historical events related to minorities at the individual level, and support for current policies that may threaten a particular understanding of national identity.
Study 1 considered silence at the level of cultural representation. To do so, we analyzed history sections of national university examinations, which reflect state-sponsored accounts of history. Building on research about the role of representations of history in constructing a particular (glorifying) understanding of national identity (Loewen, 1995; Trouillot, 1995), we examined the extent to which official accounts of Turkish history omit mention of minority populations.
In Study 2, we examined how silencing of minority-related historical information at the level of individual experience relates to national identity and support for identity-relevant policy. We extended previous research on the relationship between identification and reproduction of history (Sahdra & Ross, 2007; Salter & Adams, 2016) by examining whether individuals who endorse ethno-cultural conceptions of national identity—a less inclusive conception of belonging based on blood, ancestry, and shared cultural background (Cingöz-Ulu, 2008; Pehrson et al., 2009; Shulman, 2002; Wakefield et al., 2011)—showed an inclination to silence or deny aspects of national history that relate to past treatment of ethnic and religious minorities. Furthermore, we examined whether such tendencies predicted support for minority rights and freedom of expression.
Study 1
Previous authors have documented the extent to which Turkish history textbooks promote nation-glorifying views of the past and leave out instances of wrongdoing (e.g., see Tarman, 2006). In our first study, we extended the focus beyond textbooks to consider examples of silencing in another cultural product: national university admissions examinations. Wertsch (2002) suggests that national exams constitute some of the key tools states have for carrying out the task of creating a loyal citizenry. In the Turkish education system, high school graduates take national university admissions examinations to enter a higher education institution. The exam is based on the high school curriculum and administered once a year nationally. We examined the content of history items on the exam, as these items reflect topics that state representatives consider as the most important pieces of information about Turkish history that high school graduates should know. The content and wording of history questions on national examinations can provide an insight into state-sponsored accounts of history that construct a particular understanding of national identity.
We examined not only what themes or topics the history section of national examinations gave greater emphasis, but also what themes or topics received less emphasis. In light of previous research suggesting the rise of neo-Ottomanism in Turkey’s contemporary educational era (e.g., Sakallıoğlu, 1996), a tentative hypothesis for this study was that official representations of history embedded in national exams would focus to a greater extent on the history of the Ottoman State and the modern Turkish Republic than on the history of diverse civilizations that lived in present-day Turkey (e.g., Hittites, the Byzantine Empire) or the history of pre-Islamic (e.g., Huns, Göktürks) and pre-Ottoman (e.g., the Selçuk Empire) Turkish societies. A related hypothesis was that, given conceptions of national identity based on the ethnically Turkish and Sunni Muslim population of Turkey, official accounts of national history would likely de-emphasize the role of ethnic and religious minority populations. Finally, past research (Tarman, 2006) suggests the hypothesis that the majority of items on the history examination would highlight instances of collective glories (such as militaristic achievements and expansion of power) and de-emphasize or omit instances of collective wrongdoing, particularly regarding state treatment of minority populations.
Method
To examine these hypotheses, we conducted content analyses of 112 history items from national university entrance exams between the years 2007-2011i. Each item included a question stem and five response options. We (the first and second authors, both native speakers of Turkish) first coded all items (both the stem and response options) into different eras of Turkish history. Next, we developed an inductive coding scheme (Thomas, 2006) by reading each item multiple times and independently identifying common themes or categories to condense the data. Given our hypotheses, we paid particular attention to glorifying themes (e.g., military victories, political successes), critical themes (e.g., military failure, political and social hardships), and references to ethnic and religious identities or populations. We then discussed emerging themes and created labels to represent each theme.
After finalizing the coding scheme, we used a scale from 0 (not at all emphasized) to 4 (strongly emphasized) to independently rate the extent to which each of the 112 test items—again, both the stem and five responses—emphasized each of the themes. Intra-class correlation coefficients for each coding dimension (range from .71 and .94), suggested acceptable inter-rater agreement, so we created total emphasis scores for each theme. To do so, we computed the mean of the two coders’ ratings on each theme for each test item. We then computed a sum of the mean ratings of all test items that constituted a theme. Finally, we converted this sum into a percentage relative to the highest emphasis score possible. For instance, if one coder gave a rating of 2 for all n test items that constituted a theme, and the other coder gave a mean maximum rating of 4 for all of these items, then the mean rating for each test item would be 3, the sum of mean ratings would be 3n, and the percentage emphasis score would be 3n out of a possible 4n, or 75%.
Results
Analysis of Eras
We first examined the distribution of items across different historical eras. Out of the 112 items, 5 did not pertain to any particular era (e.g., items asking for definition of terms or interpretation of ideas that did not make reference to any historical event or era) and 7 referred to global history without direct relevance to Turkish history. For this particular analysis, we categorized the remaining 100 items into 6 main eras of Turkish history (Table 1). The largest percentage of items focused on Ottoman history (41%), suggesting the extent to which current depictions of the Turkish nation state emphasize ties with its Ottoman past. The next most frequent set included items focusing on the Republican era (21%) and the War of Independence (14%), marking the transition to the current Turkish nation state. Few items referred to the history of pre-Ottoman (6%) and pre-Islamic (10%) states. Only the items referring to pre-Islamic Turkish societies acknowledged the presence of a Turkish identity in the absence of Islam. For instance, one item in the 2008 exam focused on the pre-Islamic “era of the Huns within the Turkish history”, and asked students to identify which developments—establishment of an army, establishment of a state body, or discovery of print—took place during that period. A set of questions focused on the history of Islam (8%) independent of Turkey, reflecting the centrality of Sunni Islamic identity for the construction of Turkishness. Overall, the majority of items (90%) focused on Turkish societies after conversion to Islam.
Table 1
Distribution of Items per Era
| Historical Era | Number of Items |
|---|---|
| Pre-Islamic Period | 10 |
| History of Islam | 8 |
| Pre-Ottoman Turkish States | 6 |
| Ottoman State | 41 |
| War of Independence | 14 |
| Republic Era | 21 |
| Total | 100 |
To summarize, patterns were in line with our tentative hypothesis. For the most part, exam items represented “Turkish” history as continuous with Ottoman history. Moreover, Islam was a defining feature of Turkish history and identity. There was very little mention of “Turks” before Islam. Official representations of history embedded in national exams thus seemed to define Turkish identity in relation to Islam and narrate Turkish history from the standpoint of majority Muslim-Turks.
Analysis of Themes
We categorized the total set of 112 exam items into 16 themes, based on the emphasis scores of coders (Table 2).
Table 2
Themes in Order of Emphasis
| Coding category | Examples of constituent events | Emphasis score |
|---|---|---|
| Triumph | acquisition of power or influence | 22.6 |
| Threat | internal or external threats to national security | 18.9 |
| Military | war, armed conflict | 18.4 |
| Failure | loss of power or influence | 16.0 |
| Nation | mentions of nation, national identity | 13.4 |
| Islam | Islamic figures, institutions, practices | 11.3 |
| Land | possession, conquest or loss of land | 10.7 |
| Ethno-cultural | ethnicity, ancestry, common religion or language | 10.4 |
| Democracy | democratic reforms and institutions | 7.8 |
| Civic | citizenship, institution, laws | 6.4 |
| Oppression | oppressive forces or institutions | 6.0 |
| Minority Negative | separatist acts by minorities | 4.1 |
| Diversity | linguistic, religious, ethnic, cultural diversity | 3.0 |
| State Benevolence | state protection of or tolerance towards minorities | 2.6 |
| State Harmdoing | forced displacement, restricted rights | 0.8 |
| Minority Positive | positive acts by or contributions of minorities | 0.3 |
More emphasized themes
As anticipated, our analysis revealed that the most strongly emphasized theme in history items was triumph. This theme included successes, victories, and glorifying representations of historical events. A sample item with a high emphasis on the triumph theme stated that “Turks have made important contributions to the development of Islamic civilization,” and then asked which of several scientists “provided the biggest contribution of Turks to Islamic civilization in the domain of medicine” (2007 exam). This item was rated as high in triumph due to its emphasis on Turks’ “important contributions” to (world) civilization. Other items that received high scores on this theme referred to the expansion of political or economic affluence of the Ottoman State, contributions of Turkish-Muslim scholars to European civilization, or similar celebratory themes.
The next most emphasized themes were threat (e.g., items regarding internal or external challenges and dangers to national security), military (e.g., items regarding wars, armed conflict or military personnel), and failure (e.g., items regarding the Ottoman State’s loss of economic independence or political influence). A sample item with a high emphasis score on the threat theme mentioned “the partition of Ottoman lands through a secret contract between the UK, Russia, France, and Italy, made after World War I” (2009 exam). A sample item with a high emphasis score on the military theme included a quote from Atatürk: “In response to an idea such as destroying us, it is natural to protect and defend our existence through military means. No action can be more justified and natural than this” (2008 exam). A sample item with a high emphasis score on the “failure” theme asked students to identify “the reasons why the economy of Ottoman State became dependent on foreign nations” (2009 exam).
Other frequent themes included nation (e.g., mentions of the Turkish nation), Islam (e.g., mentions of Islam, such as Turks’ conversion to Islam and contributions to Islamic civilization, or mentions of Islamic concepts), and land (e.g., mentions of control over specific territories such as the Bosphorus Strait, or Ottoman territories in general). A particularly noteworthy theme referred to ethno-cultural characteristics of Turkishness (e.g., mentions of Turkish language, culture, and ancestral bonds). A sample item with a high emphasis score on this theme stated that, “Those who speak the language of another nation start thinking like that nation,” and one of the response options mentioned that “common language is the most important element in developing a sense of national consciousness” (2009 exam).
The following themes were also emphasized, but to a relatively lesser extent: democracy (e.g., mentions of development of democracy and democratic institutions), civic activities (e.g., mentions of elections and the parliament, laws and regulations), and oppression (e.g., items referring to colonialism, or mentioning foreign pressures on the Ottoman Empire or Turkish Republic). A sample item that emphasized the former two themes focused on the implications of an early law, passed after the foundation of the republic, regarding the eligibility of male citizens above the age of 18 to the parliament (2008 exam).
Overall, analyses suggest that items emphasized both glorifying themes such as triumph and negative or critical themes such as threat. Triumph and military themes prompt a sense of pride or positive affect that affords nation-glorifying identification and action (Roccas et al., 2006). Threat and failure themes may also prompt responses that motivate nation-unifying forms of identification and action (Li & Brewer, 2004).
Less emphasized themes
As anticipated, less emphasized themes included diversity (e.g., items referring to existence of different cultures and civilizations within the Ottoman/Turkish territories), minorities (e.g., positive and negative references to minority populations or individuals), and state treatment of minorities (e.g., items referring to acts of benevolence and negative treatment toward minorities). Out of the 112 items, only a total of 16 (14%) mentioned minorities. Among these 16 items, none mentioned minorities in the Republican era. Minorities were completely left out of official history of the current republic (1920s onwards)ii.
Besides low frequency of mention of minorities in general, we noted that only one item in this subset of 16 items included a positive mention of minorities. This item (2009 exam) highlighted the military contributions of one ethnic minority person to the War of Independence, stating that this person “was effective in preventing the military advancement of Greeks in Western Anatolia, in the establishment of the parliament, and the suppression of uprisings.” However, the item continued to state that “despite these contributions”, the person “later engaged in an uprising against the parliament.” The net effect was to cancel out the initial positivity toward an ethnic minority person by emphasizing betrayal against the state.
Of the items that mentioned minorities, the majority of items (12 out of 16) involved negative reference. Of these, 10 items referred to negative acts committed by minorities, including separatist attempts, potential partnership with foreign enemies, and destructive military actions towards the Ottoman State. For instance, one item (2011 exam) asked students to identify which minority groups—Arabs, Armenians, or Greeks—engaged in separatist activities in the Eastern provinces of the Ottoman Empire. The remaining two items portrayed minority rights as a threat to national independence and majority interests.
In contrast to the negative portrayal of minorities, 6 of the 16 minority-related items focused on Ottoman benevolence toward minorities (e.g., protection and privileges of minority populations living in the Ottoman State). A representative example stated that, “Populations that differed in terms of language, religion, and traditions lived in the lands under the sovereignty of The Ottoman Empire. The Empire did not adopt policies to try to influence the characteristics of these nations” (2010 exam). The item mentioned minorities with the purpose of highlighting state benevolence toward them, emphasizing policies of multiculturalism and tolerance.
Only three items mentioned negative treatment of minorities (e.g. failure of reforms to promote civil rights), but all did so alongside mention of state benevolence. For instance, one item focused on the reform package passed in the later stages of the Ottoman Empire (“Tanzimat Fermani”) that aimed to prevent the empire from dissolving (2011 exam). The question stem explained that the reforms evoked reaction from the public, and asked which reasons—“Muslims reacted against the privileges that the state provided to non-Muslim populations”, “The privileges provided to non-Muslims were not satisfactory”, or “Non-Muslims continued to engage in separatist efforts”—were the source of this reactance.
Discussion
Results of Study 1 suggest that official representations of Turkish history in national exams emphasized nation-glorifying (e.g., triumph and military) and nation-uniting (e.g., threat and loss) themes. These representations made no mention of the diversity of peoples who historically inhabited the lands of present-day Turkey. Moreover, they made no mention of minorities in contemporary Turkish society. Exams tended to silence positive contributions of minorities or negative treatment of minorities. In rare cases where they did mention minorities, they depicted minority groups in negative terms (e.g., as harming the state). In other words, official representations of Turkish history embedded in national exams not only omitted instances of past wrongdoing, but also implied a construction of national identity in ways that homogenized contemporary Turkish society and portrayed minority groups as a threat to national unity (see Çetin, 2004).
Study 2
If we interpret absence from national exams as an indication of silence about events and themes in the state educational system more broadly, then how—if at all—do people access alternative (e.g., silence-breaking) accounts of history? All citizens of Turkey with a high school degree encounter official representations of history through formal schooling. However, people may access counterhegemonic accounts of history via university education and international media. Whether and how people engage with alternative accounts might be a function of different identity needs (Sahdra & Ross, 2007). To the extent that silence regarding (mis)treatment of minorities serve identity-enhancing purposes, people who endorse particular conceptions of the national identity might be less likely to acknowledge, or more likely to resist, alternative accounts.
Although most research on engagement with identity considers variation in strength or degree of identification with the category, people also vary in the construction of national identity they endorse (e.g., Brubaker, 1992; Kosterman & Feshbach, 1989; Shulman, 2002). One important dimension of variation in constructions of national identity concerns the distinction between civic (i.e., based mainly on citizenship) and ethno-cultural (i.e., based mainly on ancestry and cultural background) understandings (Brubaker, 1992; Ignatieff, 1993; Pakulski & Tranter, 2000; Smith, 1991). Although the Turkish constitution endorses a civic construction of national identity, results of Study 1 suggest that official exams define national identity based on the dominant ethnic-religious identities. This distinction has important implications for intergroup outcomes. Civic constructions are associated with more inclusive attitudes towards immigration (e.g., Rothi, Lyons, & Chryssochoou, 2005; Wright, Citrin, & Wand, 2012), but ethno-cultural constructions are associated with intolerance towards immigrants (e.g., Jones, 1997; Wright, 2011).
The present study examined the link between national identity and tendencies to reproduce silence about (treatment of) minorities through active denial of relevant historical facts available in sources outside the national curriculum. We hypothesized that endorsement of ethno-cultural, but not civic, conceptions of Turkish identity would predict denial about silenced, minority-relevant history and, in turn, predict support for two identity-relevant policies: extension of minority rights and freedom of expression. We focused on these two policies as they relate to issues that have the potential to disrupt ethno-cultural understandings of Turkish national identity. The primary hypothesis suggests a mediation model such that people who show a greater tendency to endorse an ethno-cultural conception of Turkish national identity (a) tend to deny “silenced history” accounts that reference negative acts against minorities, and, in turn, (b) deny the need (or show less support) for identity-relevant policies.
Method
Participants
Participants included 60 Turkish adults (40 women, 17 men, 3 unspecified, mean age = 27.14, SD = 5.52) whom we recruited online through social media (e.g., postings in psychology forums). Of the 43 participants who reported ethnicity, 35 identified as Turkish, 2 identified as Albanian, 1 as Kurdish, 1 as Turkish-Kurdish, 1 as Circassian, 1 as Laz, 1 as Turkmen, and 1 as Tatar iii. Two participants indicated regional identification (e.g., Black Sea region) instead of ethnicity. Two participants indicated religious sect instead of ethnicity (e.g., Sunni Muslim). The remaining 13 participants did not indicate ethnicity. Regarding education, 56 participants (93.3%) reported college or post-graduate level, 1 (1.7%) reported having a high school degree, and 3 (5%) did not report education.
Procedure
Participants completed an online survey in Turkish. The survey included a history quiz, a measure assessing constructions of national identity, support for minority rights, and support for freedom of expression.
Measures
History quiz
The history quiz consisted of 22 official history items that we adapted from the pool of national examination items in Study 1. These included 13 items that official representations of Turkish history consider to be “true” (e.g., “Up until the Kanuni Sultan Süleyman era, the Ottomans had a continuous friendship with the French) and nine items in which we inserted “false” information to statements from actual exams (e.g., “The first text including the work ‘Turk’ was written in the Uyghur alphabet”). We supplemented these official history items with four silenced history items that we constructed using scholarly sources outside of the official high school history curriculum. All silenced history items referred to past events regarding negative treatment of minorities that scholars of Turkish history agree to be to “true” (e.g., Hirschon, 2003; Toledano, 2014). These items were “Greeks living in Turkey were forced to migrate due to a policy protecting the economy from the influence of non-Turkish communities”, “The Lozan Pact declared that there were no linguistic or ethnic minorities with the exception of Greeks, Armenians, and Jews”, “During the Tanzimat era, the Ottoman State used force and military power to control the Alevi Turk and Sunni Kurdish tribes as part of its centralization policy”, and “In the 18th and 19th centuries, most slaves in the Ottoman Empire were Ethiopian, while there were also Circassian and Georgian slaves”. Participants indicated whether the information presented in each item was true (or false) and reported their level of confidence about this judgment using a 7-point scale (0=not at all confident to 6=extremely confident).
For the four silenced history items, we calculated denial scores. Denial of silenced history refers to the tendency to claim silenced history items to be false with high certainty. Rather than mere lack of knowledge, we conceptualized denial as an identity-defensive reactance to information that is actually accurate. To create denial scores, we first coded participants’ true answers as -1 and false answers as 1, and we then multiplied this coefficient with the level of confidence associated with each item. For instance, if a participant indicated that a silenced item was false (coded as 1), but also indicated low confidence in her answer (e.g., confidence level of 1), she would end up with a low denial score (i.e., 1). This suggests that her (false) answer likely stems from lack of knowledge. However, if the participant indicated high confidence in her wrong answer (e.g., confidence level of 7), she would end up with a high denial score (i.e., 7). This score suggests that her answer likely stems from active denial of silenced information rather than mere lack of knowledge. We averaged denial scores for each item to create a composite denial score (α = .61). High scores on this measure indicated the tendency to claim that consensually true “silenced history” items are false with high certainty. Rather than a reflection of ignorance or simple lack of knowledge, we propose that claiming “silenced history” items to be false with a high degree of confidence constitutes a case of identity-relevant denial. In other words, denial scores constitute a form of individual level silencing of history. For the “official history” items, we calculated raw scores by granting participants one point for each correct response and averaging all responses, without taking confidence levels into accountiv.
Constructions of national identity
As a measure of identity construction, participants indicated which factors are most important to be considered “truly Turkish” based on items that Cingöz-Ulu (2008) created to assess ethno-cultural and civic constructions of identity. Twelve items (α = .85) referred to ethno-cultural markers (e.g., “Being Muslim”, “Speaking Turkish as one’s mother tongue”, “Having a Turkish name”), and five items (α = .74) referred to civic markers of national identity (e.g., “Being a Turkish citizen”, “Having a Turkish ID”, “Owning and respecting Turkish laws and obligations”). Participants indicated their responses using a 6 point-scale (0=not at all important to 5=extremely important).
Support for minority rights
Participants completed a ten-item measure (α = .92) assessing support for minority rights (Verkuyten & Yıldız, 2006) using a 7-point response scale (1=strongly disagree to 7=strongly agree) v. A sample item was “Minorities should be allowed to establish their own schools”.
Support for freedom of expression
Participants completed a five-item measure (α = .81) assessing support for freedom of expression using a 7-point response scale (1=strongly disagree to 7=strongly agree). A sample item was “It is okay to limit the expression of opinions that could disrupt public order (reverse-scored)”.
Results
Correlations between measures appear in Table 3. There were no significant correlations between knowledge of official history and any of the identification measures or outcome variables. In contrast, denial of silenced history was positively correlated with national identification measures, but negatively correlated with support for minority rights and support for freedom of expression. These findings suggest that it is acknowledgment (versus denial) of silenced history, rather than knowledge of official history, that is relevant to identity concerns. Accordingly, we focused subsequent analyses on denial of silenced history.
Table 3
Correlations, Means (Standard Deviations), and Range of Key Variables
| Variable | 1 | 2 | 3 | 4 | 5 | M (SD) | Range |
|---|---|---|---|---|---|---|---|
| 1. Ethno-cultural ID | 2.26 (.86) | 1.00 - 4.50 | |||||
| 2. Civic ID | .542** | 3.05 (1.06) | 1.00 - 5.00 | ||||
| 3. Denial of historical events | .430** | .317* | -1.49 (3.02) | -7.00 - 4.75 | |||
| 4. Official knowledge of history | .004 | -.087 | -.152 | .67 (.10) | .45 - .86 | ||
| 5. Minority rights | -.326* | -.228 | -.531** | .148 | 5.50 (1.40) | 1.50 - 7.00 | |
| 6. Freedom of expression | -.424** | -.166 | -.382** | .174 | .631** | 6.05 (1.30) | 1.50 - 7.00 |
*p < .05. **p < .01.
Identity Relevance of Denial
To examine how national identity dynamics relate to denial of silenced history, we conducted a regression analysis with ethno-cultural and civic constructions of national identity as predictors of denial scores. When we simultaneously entered both predictors in the model, ethno-cultural construction of national identity emerged as the only significant predictor of denial, β = .33, t(57) = 2.32, p = .02. This suggests that denial of historical information regarding state treatment of minorities is driven by an ethno-cultural construction of identity based on ancestry and cultural background, as opposed to a civic conception of identity based on citizenship. We used ethno-cultural identification as our independent variable in the analyses that follow.
Implications for Policy Support
Analyses revealed a negative correlation between ethno-cultural constructions of Turkish identity with support for minority rights and freedom of expression (Table 3). To determine whether denial of silenced history mediated this negative relationship, we used the PROCESS macro for SPSS (Preacher & Hayes, 2004) and conducted two separate mediation analyses, one for each dependent variable, with 5000 bootstrapped samples.
Support for minority rights
Ethno-cultural identification significantly predicted support for minority rights, β = -.46, t(57) = -2.55, p = .01. Including denial of silenced history as a mediator in the model, we found evidence for full mediation (Figure 1). Ethno-cultural identification significantly predicted denial of silenced history, β = .43, t(57) = 3.60, p <.001, and denial of silenced history significantly predicted support for minority rights. β = -.68, t(56) = -3.876, p <.001. The direct effect of ethno-cultural identification on support for minority rights was no longer significant, β = -.16, t(56) = -.90, p = .37. Participants who endorsed an ethno-cultural construction of national identity tended to deny the truth of silenced events, which predicted lower support for minority rights, 95% CI for indirect effect [-.5507, -.1256].
Figure 1
Indirect effect of endorsement of ethno-cultural conceptions of identity on support for minority rights.
Note. Figure includes standardized coefficients.
**p < .01. ***p < .001.
Support for freedom of expression
Ethno-cultural identification significantly predicted support for freedom of expression, β = -.52, t(58) = -3.30, p = .002. With denial of silenced history as a mediator, we found evidence for partial mediation (Figure 2). Ethno-cultural identification positively predicted denial of silenced history, β = .43, t(57) = 3.60, p < .001, and denial of silenced history negatively predicted support for freedom of expression, β = -.33, t(56) = -1.961, p = .05. The direct effect of ethno-cultural identification on support for freedom of expression remained significant, β = -.38, t(56) = -2.20, p = .03. Participants who endorsed an ethno-cultural construction of national identity tended to deny the truth of silenced information, which predicted lower support for freedom of expression. The direct link between ethno-cultural identification and support for freedom of expression remained significant, 95% CI for direct effect [-.717, -.035], 95% CI for indirect effect [-.359, -.0003].
Figure 2
Indirect effect of endorsement of ethno-cultural conceptions of identity on support for freedom of expression.
Note. Figure includes standardized coefficients.
*p < .05. **p < .01.
Testing Alternative Models
Although results support the idea that denial of silenced history mediates the effect of identity (specifically, endorsement of ethno-cultural constructions) on support for policy, the correlational character of evidence means that other accounts of relationships between variables are possible. Previous research shows that people may engage in denial of facts relevant to a social issue in order to legitimize their current position regarding that social issue (Campbell & Kay, 2014; Prasad et al., 2009). Applied to our study, people may have relatively well-formed positions on policies regarding minority rights and freedom of expression (informed by endorsement of ethno-cultural forms of national identification), which may guide their responses to historical events (e.g., motivate denial) that potentially have implications for present policies. In other words, it is possible that people may deny silenced historical information regarding minorities to legitimize their stance on the said policies. To examine this alternative account, we conducted mediation analyses where we switched the mediator and dependent variable in our models. Results of these alternative models confirmed the plausibility of this account. For instance, ethno-cultural identification predicted support for minority rights, β = -.46, t (57) = -2.55, p = .01; support for minority rights predicted denial of silenced history, β = -.31, t (56) = -3.87, p < .001; and the indirect relationship between ethno-cultural identification and denial of silenced history through policy support was statistically significant, 95% CI for direct effect [.063, .521], 95% CI for indirect effect [.034, .340]. We observed a similar pattern for support for freedom of expression.vi In this case, the direct link between ethno-cultural identification and denial of silenced history remained significant, suggesting that policy support only partially mediated the direct relationship between identification and denial. Although evidence for full mediation in the initial set of analyses (for support for minority rights in particular) might suggest preference for the original “denial as cause” model over the “denial as consequence” model, we emphasize that results of the present study do not provide a strong basis for preferring one account over the other. A more definitive comparison of the different possibilities awaits future research.
Discussion
Our analyses yielded no relationship between knowledge of official history and constructions of national identity or support for policies concerning minority rights. Instead, consistent with our hypotheses, denial of silenced history (i.e., regarding state treatment of minorities) was significantly related to, and mediated the relationship between, ethno-cultural constructions of identity and opposition to identity-relevant policies (minority rights and freedom of expression). Ethno-cultural constructions of Turkish identity were associated with denial of the role of minority populations in Turkish history and identity. In turn, this denial was associated with opposition to minority rights and limitations to freedom of expression.
General Discussion
We conducted this research to examine silence in representations of history at the cultural level (i.e., official accounts of Turkish history) and individual level (i.e., implications for national identity and identity-relevant action). Study 1 addressed silence at the level of cultural representations via an analysis of cultural products: official accounts of history embedded in national university entrance exams. Consistent with past research within Turkish and other national contexts (e.g., Tarman, 2006; Trouillot, 1995) results revealed that the official narrative embedded in these exams emphasized nation-glorifying (e.g., triumph and military) and nation-uniting (e.g., threat and loss) themes but silenced experiences of minorities in past and present Turkish society.
Study 2 investigated silence at the level of individual experience in the form of denial about the truth of incidents implying harm toward minorities. Participants who endorsed an ethno-cultural construction of Turkish identity tended to deny historical knowledge about negative treatment of ethnic minorities, and this denial was associated with opposition to minority rights and freedom of expression. The relationship between greater endorsement of ethno-cultural identity and reduced support for freedom expression is consistent with attempts of the Turkish state to constrain freedom of expression to protect Turkish identity (e.g., Article 301 which makes it illegal to offend the Turkish nation). However, constraining freedom of expression due to identity concerns is not limited to silencing of minority experience. It further involves silencing of critical perspectives that might pose a threat to the nation-state. This might help explain why the direct effect of ethno-cultural identification on support for freedom of expression remained significant in mediation analyses.
To summarize, the present studies suggest that identity-defensive silence and denial about historical treatment of minorities was evident both in cultural products and in individual action. Thus, one important contribution of the present research is illustrating the role of identity-motivated denial of critical or potentially threatening historical information at the individual level for support of progressive policies.
Limitations and Future Directions
As an initial study examining psychological associations of silence in official representations of Turkish history, the present research has limitations that suggest directions for future research. One limitation of Study 1 is the relatively narrow time-span of exams. An analysis that considered a longer time-span might reveal different themes or illuminate changes in emphasis over time. A related limitation is the narrow focus on one set of artifacts, national exams. A focus on other artifacts, such as textbooks (Leone & Mastrovito, 2010) or educational displays (Salter & Adams, 2016) might reveal a more diverse set of themes.
Limitations of Study 2 include the small sample size and the small number of silenced history items. A particularly noteworthy limitation is the narrow focus on participants who identified as ethnically Turkish. The observed relationship between denial of silenced history and support for policy may be attenuated among people who do not identify as ethnically Turkish, whether because they endorse minority rights policy regardless of historical knowledge or acknowledge critical historical accounts regardless of policy preferences (e.g., due to greater familiarity with marginalized knowledge). Finally, it is important to emphasize that our sample consisted of mostly college-educated individuals. If the tendency to deny troublesome facts is a consequence of engagement with official history education materials (e.g., in preparation for college entrance exams), then it suggests the somewhat ironic possibility that ignorance about past treatment of minorities may be greater among highly educated than less educated students. The investigation of this possibility awaits future research.
Broader Implications
Despite these limitations, the work offers interesting implications for the study of collective memory and identity. With respect to the latter, the research moves beyond the typical focus on strength or degree of identification to consider variation in constructions of identity (i.e., ethno-cultural versus civic). This has important implications for measurement. Once consideration turns from identification with an abstract category to the content and meaning of categories, it becomes relevant to consider how the content and meaning of categories may vary across contexts. In this regard, it is noteworthy that we used a measure of identity construction that Cingöz-Ulu (2008) designed in Turkish language for use in a Turkish context. Although some items on the scale might translate well to different settings, other items may express things that are more peculiar to Turkish constructions of ethno-cultural identity and may translate less well in other settings.
With respect to collective memory, the research extends consideration beyond patterns of mind in brain to consider patterns of mind in context: identity-relevant beliefs and desires inscribed in cultural representations of history associated with such official products as national university entrance exams. From this perspective, memory is a collective process not merely because it involves collective identity, but more profoundly because it extends beyond individual minds to include cultural manifestations inscribed in practices and artifacts. This sense of collective becomes clear when one considers the role of history education in particular in shaping intergroup relations. Similar to other nation-states, history education has been central to the formation of Turkish identity and citizenry since the founding of the Turkish nation-state. Atatürk emphasized the extent to which history education can serve purposes of social transformation and the formation of a new national consciousness (see Tarman, 2006). Yet, findings of the present research suggest the extent to which national history curricula might constitute “epistemologies of ignorance” (Mills, 1997)—that is, forms of knowing that afford lack of consciousness (and therefore inaction) about disturbing facts (e.g., past injustice and oppression in contemporary society) that might otherwise be obvious—that serve to maintain the status quo.
Implicit in the idea of mind in context (but not something that we tested directly) is the idea that patterns of mind evident in cultural products exist in a dynamic relationship of mutual constitution with patterns of mind evident in individual psychological tendencies. The point is not only that silence about minorities in official curricula promotes ignorance and denial about minorities among students who engage the curricula, but also that silence in official curricula comes about through acts of selection by institutional actors who decide to include information that they think is relevant and to omit information that they think is not. From this perspective, research must consider not only what people select for representation and thereby contribute to collective memory, but also what they fail to select and thereby contribute to collective forgetting.
Finally, the emphasis on silence highlights implications for intergroup relations and social justice. One possibility for coming to terms with the past involves the recovery of historical memory (Martín-Baró, 1994). Historical accounts that acknowledge past wrongdoing not only counteract tendencies of denial and forgetting, but also enable new imaginations of national identity (Anderson, 1983) that affirm the common humanity of diverse peoples and mobilize action toward democratic innovation. As such, recovery of historical memory can help reduce the risk of perpetuating past mistakes (Barkan, 2000; Pennebaker, Páez, & Rimé, 1997) and serve as a means of promoting healing, reconciliation, and restorative justice (Cole, 2004; Liem, 2007).