Greta Uehling. Ethnic and Racial Studies. Volume 27, Issue 1, 2004.
Political actors and government officials have a decided role to play in the production of collective identities by carving national, racial, and ethnic categories out of a diffuse spectrum of humanity. Statistics gathering is therefore one of the ways that the state enters the complex process of identity formation. Turning to the state’s role in categorizing a populace is not, however, to suggest that ‘the state’ is necessarily operating in a unified or coherent way. Ethnography can bring greater clarity to how various representatives of the state (and its citizens) are involved. This article considers the dynamics of the 2001 Ukrainian census in the Autonomous Republic of Crimea. I focus on census-takers’ interactions with respondents, conversations among census officials, and informal dialogues among census-takers to explore the complex nexus in which the practices and politics of census-taking were worked out Attention to the ‘micropolitics’ of statistics-gathering may affect the way we think about the numbers. Although statistics themselves can ‘enclose’ and ‘flatten’ (Appadurai 1993, p. 334) the process of gathering them is three-dimensional. As the census proceeded in Crimea, it was inflated with myth, lore, and humour.
This article is part of a larger project that examines post-Soviet censuses. In December 2001, three researchers visited Ukraine and in October 2002, twenty-one worked in the Russian Federation. In Crimea, my research was facilitated by the Vice Premier, who authorized me to work directly with the Department of Statistics. With the Deputy Head of the Department, I visited five uchastoks, or field offices, where I interviewed instructors and supervisors. I was also introduced to census-takers, three of whom I followed. One was polling a dormitory, another was in a so-called ‘private sector’ of apartments and a third was working in a big high-rise in the capital city of Simferopol, the administrative and geographic centre of the peninsula. Efforts to enlist censustakers who were representative of the peninsula’s diversity had not been undertaken. In order for the research to be as representative as possible, however, I followed one census-taker from each of the three major ethnic groups: Russian, Ukrainian, and Crimean Tatar. To flesh out this research, I interviewed the editors of three major newspapers, and spoke with the leaders of various ethnic organizations. While my work began in a top-down manner, it was complemented by interaction with consultants from 1995–1998 fieldwork. Living in a familiar neighbourhood gave me an opportunity to speak informally with people who had been polled.
This approach, of studying the census ethnographically, is new for the former Soviet Union. Soviet anthropologists and ethnographers have long made detailed use of the Soviet census as a platform for theorizing. Their works, as well as landmark studies by Western sociologists, demographers, and geographers, have resulted in a substantial literature that examines the implications of the Soviet census for nationality, ethnicity, language, and policy issues (Silver 1975, 1986; Wixman 1980, 1984; Anderson and Silver 1983; Hirsch 1997). While these valuable studies draw on field experience, the researchers affiliated with Brown University’s Watson Institute are the first to attempt a full-fledged ethnography of the census itself. This methodology brought enhanced awareness of the ways in which the newly formed states are being imagined by contemporary citizens. Silver observed in 1986 that, ‘the nationality question on the census forms has evoked little controversy or even discussion in recent years’ (1986, p. 94). By 2001, this situation no longer obtained. Not only did the Rusyns use the census as a platform for ethnic mobilization in western Ukraine, but in Crimea, the Tatars threatened to bring down local elites. While it has long been noted that language presented a challenge to the validity of the census and was a topic of debate among the groups concerned (Silver 1986), the ethnography of the 2001 census highlighted how the census can become a contentious ethnic and national issue as well.
The ethnographic approach to the census dovetails with aspects of historical institutionalism, concerned as it is with the way in which the institutional organization of a polity and its history shapes discourses and their effects (Hall and Taylor 1996; Hay and Wincott 1998; Dyson 1999). This has the potential to illuminate cultural identities and their manifestations not as ‘given’ but as the outcome of choices and preferences in particular contexts (Lecours 2000). Historical institutionalism critically examines the role of political institutions in shaping identities and outlooks. The implicit social ontology here is that institutions and ‘paths’ are not determining – there exists a dialectical relationship between structure and agency (Hay and Wincott 1998). This brings greater clarity to culturalist interpretations of ethno-nationalism by highlighting the role of the state without losing sight of individuals.
The literature on nationalism, colonialism, and modernity draws out the broader significance of the census. It is striking to recall that carefully enumerating cultural populations was not a part of premodern statecraft (Gellner 1983). Anderson saw statistics-gathering as part of the limited imaging of modern nation-states. He argued the census was an ‘institution of power’ and one way of imposing a ‘totalizing and classificatory’ grid, thereby claiming territories and constituencies (1991, p. 163). Appadurai describes how the colonial census helped to solidify communitarian and nationalist identities that would eventually undermine colonial rule (1993). These observations link up with thinking on modernity and statecraft in history and philosophy: Foucault argued that the modern state manages its population by extending surveillance over it. As Gupta and Ferguson have argued, we now need a clearer idea of how citizens imagine the state (2002, p. 981).
Clearly, the census does more than passively reflect the ethnic, racial, linguistic, or religious constituency it seeks to count. In a pragmatic way, this is demonstrated in the US and Canada, where whether a group was listed as an example or not made a significant impact on the number of respondents to identify with it (Kertzer and Arel 2002, p. 17). Specific examples lead back to the more basic point that, as Pierre Bourdieu has argued, the act of naming helps to create social realities: naming structures the perceptions through which agents view the social world (1991, p. 105). This simple truth was not lost on colonial powers, and it appears to be a growing preoccupation among national elites. Activists quickly realize that entitlement entails a contest for achieving the ‘right’ numbers (Kertzer and Arel 2002, p. 30). While the census purports to count these subjective identities in an objective manner, it is far from a scientific exercise. This is what has led Kertzer and Arel to argue that the census is best compared to a national plebiscite or political campaign. When it comes to identity, censuses can be viewed as measuring preferences rather than objective ‘data’ (2002). This article attempts to push this line of thinking one step further to ask whose preferences are being recorded. In post-Soviet Crimea, the preferences of census-takers, census officials, and regional elites entered in.
Legibility and the census as a ritual of verification
The interactions I observed suggest that the first Ukrainian census was plagued by a number of difficulties. Further, the census failed as what Dunn would call a ‘ritual of verification’ or a tool that produces comfort for politicians, planners, and citizens (Dunn 1999, p. 43). She argued with regard to accounting, and a similar point could be made about the census, that it can mystify as much as it reveals. In Crimea, the census spawned a number of interesting myths and rumours in its wake. The census also served as a cipher for the relationshp between Ukraine and its citizens, and Ukraine and the international community. As Dunn put it:
accounting is mystifying as much as it’s revealing. It obscures power relations . . . in the production of an auditable face . . . [It] produce[s] the social institution of trust, a substitute for knowledge and a force which mystifies both the workings of the firm and the social relations which surround it. (Dunn 1999, p. 43)
In rendering social types more legible, the census provides a profile of society that then legitimizes decisions, whether the ‘auditable face’ is accurate or not. Whereas the accounting which Dunn studied produced trust as a substitute for knowledge, the census generated widespread doubts. Nevertheless, the census was inserted in narratives about democratic reform and progress. As one consultant put it, ‘what is needed [are] not the facts but the political context in which facts can be constructed.’ He went on to say that, ‘if the authorities understand they really need accurate information – it is going to be the highest achievement’. He felt Ukraine was on the verge of going in a democratic direction, but the outcome was still uncertain.
State officials saw the census as an exercise that would bolster (the admittedly faltering) image of Ukraine proceeding on a path joining democratic Europe. Marring the picture were botched communication about language and nationality, confusion surrounding the purpose of a ‘preliminary’ census, and the use of practice notebooks. The uncertainties these practices generated meant the exercise failed to reassure or affirm. So although the census could have produced an ‘auditable face’ and generated confidence (Dunn 1999, p. 43) it seemed to obscure and mystify. And so it should not be surprising that it turned into a source of tension and potential conflict. Crimean Tatar elites threatened statistics officials; Ukrainians had altercations with Russian-speaking census-takers; and Crimean Tatars expressed hostility towards the enumerators who claimed ignorance of their identity. As Zeman showed with regard to the Imperial Austrian censuses, the effort to record a cultural nationality for each individual contributed to the tension between nationalities (Zeman 1990). A significant amount of tension was displaced into humour that, while further undermining the credibility of the exercise, veiled enough hostility to give people a way to talk about the event.
One of the most important tasks facing Ukraine is consolidating Ukrainian national identity while taking into account the needs of multiple ethnic minorities. Another is ameliorating relations with Russians. The census provided practical ground on which to begin to realize these goals, but they are especially challenging in Crimea. Whereas in Ukraine as a whole the majority is Ukrainian (77 per cent) followed by Russians (17 per cent) the situation is the reverse in Crimea where the majority is Russian (58 per cent) followed by Ukrainians (24 per cent). The separatist inclinations of the Russian population are somewhat balanced by the presence of the Crimean Tatars (12 per cent) who have, since their voluntary repatriation in the late eighties and early nineties, traditionally allied themselves with Kyiv. Given the delicate balance, linguistic representation is an issue of particularly vital importance. Out of a total Crimean population of just under 2.5 million, an estimated 1.5 million are Russian speakers. Ukrainian respondents complained of linguistic Russification, but Russian elites counter that there is no historic basis for Ukrainian in Crimea. Tatars are promoting Crimean Tatar, but remain divided on the specifics of education policy. The results of the census will inform future debates about proportional representation in the government of Crimea, the status of the autonomy (whether it is national or territorial), the official languages, and the languages of education.
Myth or miscoding?
The December 2001 census officially included the Crimean Tatars. This represents a change from the post-World War II censuses, when anyone indicating he or she was Crimean Tatar was told the group did not exist and coded ‘Tatar.’ This was one of many mechanisms used to try to obliterate the cultural distinctiveness of this indigenous group. The word ‘Tatar’, has been used in a bewildering variety of ways to refer to many different traditionally nomadic peoples coming to Rus from the East. They do not, however, share a single culture and history. The Crimean Tatars feel they have about as much in common with the Kazan Tatars of Tatarstan who live in the Russian Federation (with whom they are often confused) as Serbs and Russians. Their struggle for cultural rights began after annexation by the Russian Empire in 1783, continued during the Soviet period and reached a crisis after World War II when they were collectively accused of treason and deported to the Urals and Soviet Central Asia. The Crimean Tatars began returning from exile in large numbers after the disintegration of the Soviet Union. In particular, their success in achieving equitable representation in official bodies and sufficient education in their native language depends on numerical presence. Long distrustful of Soviet statecraft, the Crimean Tatars have conducted their own unofficial censuses. They have also had to fight for official recognition. After the 1979 census, the Crimean Tatars objected to the practice of counting them as ‘Tatars’ so vociferously that they were then included in the 1989 census under the ‘Crimean Tatar’ ethnonym (Tishkov, personal communication).
In spite of the Crimean Tatars’ official status in the 2001 census, dynamics between the census-takers and the Crimean Tatar population refreshed memories of their historical erasure, raising new fears of their annihilation as group. From the first days of the census, the editorial offices of the major Crimean Tatar newspapers began receiving calls that Crimean Tatars were being told by census-takers that there was no such ethnic group, only ‘Tatars’. The behaviour of the census-takers led Crimean Tatars to believe that a hidden hand was operating behind the scenes, intent (not unlike the authorities of the Soviet Union) on their disappearance. As the acting head of the Crimean Tatar political body, the Mejlis put it: ‘This is a special, political genocide’. The Mejlis has been charged with advocating on Crimean Tatars’ behalf on issues of this nature.
On the ground, it was hard to ascertain how prevalent was the coding of ‘Crimean Tatars’ as ‘Tatars’. The Deputy Director of the Department of Statistics admitted that the census-takers ‘had their own shortcomings’. She elaborated on the limited training they received and did not exclude that such violations were possible. It seemed more likely, however, that the Crimean Tatars were especially sensitized to the possibility of this kind of treatment by their 1944 deportation and experiences of discrimination in the past. In the Soviet period, the experience of being denied their very existence (they were issued passports that read ‘Tatar’) was woven into the habitus of the Crimean Tatars, making it doubly difficult to assess the accuracy of these perceptions. When I followed a census-taker in a Crimean Tatar area, there were numerous instances in which it was the respondent who replied ‘Tatar’ to the question about nationality and it was the census-taker who sought clarification before writing Crimean Tatar. Similarly, when these respondents said they spoke ‘Tatar’ she sought clarification whether they meant Crimean Tatar. Here, it is partly the Crimean Tatar ethnonym that is responsible: the ‘Crimean’ part is dropped, particularly by youth, for the sake of brevity and ease of speech. However, it is also the case that the sharpness and immediacy of Crimean Tatar identity may be fading. National elites worry that there are Crimean Tatars who are not sufficiently conscious of their ethnic ‘roots’. This concern was evident when both the editors of the Crimean Tatar newspapers and the acting head of the Mejlis lamented that by failing to propagandize prior to the census, they had missed an opportunity to ‘work with the people’ and raise ethnic self-awareness.
The instances in which the Crimean Tatars were coded as ‘Tatars’ became such a concern over the course of the week that it coloured religious observances. The Muslim holiday of Yantar, which fell on 10 December just prior to the end of Ramadan, was supposed to have been marked off from politics. However the holiday celebration dissolved into a discussion of the threat the census posed. This led to a debate about the relative merits of changing their Crimean Tatar ethnonym. Census politics had so thoroughly penetrated the milieu that they were palpable at the ritual.
The preliminary results of the census show that the fears of the Crimean Tatars were only partially borne out. The 248,000 indicated in the preliminary census results is viewed as an undercount by the Crimean Tatars, whose unofficial censuses suggest that they number closer to 265,000 or even 270,000. However, the Crimean Tatars’ fears were not borne out with respect to miscoding. Except in Kherson oblast and Sevastopol where the number of Tatars rose while the number of Crimean Tatars fell, the relative proportion of the two groups remained at 1989 levels. The Kherson data are explained by the fact that Kherson was home to Crimean Tatars who migrated into the peninsula proper following the mass repatriation effort. Why Tatars rose in relation to Crimean Tatars in Sevastopol, however, remains unexplained. Whether or not Crimean Tatar concerns were well grounded, the more basic point remains that an exercise designed to make the division of resources more rational and equitable led to rumours and the hypertrophy of fear. Administrators and citizens alike were highly conscious of the long-term political outcomes that could evolve from these events. Much will depend on whether the Crimean Tatars, who are seeking representation at all levels of government, will win a say in the matters that affect them.
Nationality
In the first independent Ukrainian census, it was the questions on nationality and language that generated the most controversy. This is predictable given the history of Soviet nationality politics. By the middle of the twentieth century, Soviet ethnographers, Marxist-Leninist social scientists, and Soviet officials had crafted a system of classifying national and territorial units that not only worked as neatly as a set of nesting dolls (Slezkine 1994) but was perhaps more clearly articulated than in any other country in the world (Hirsch 1997). The irony is that in 2001, many individuals were simply not asked about their nationality. Census-takers often failed to ask the question that respondents had been waiting for, given over half a century of conditioning by nationality politics. I watched one census-taker automatically identified several respondents as Russian, based on the fact that they said Russian was their native language. This involved skipping the nationality question. Later, when asked, she said they had been instructed not to ‘press’ the nationality issue. Furthermore, it had been ‘apparent’ to her that they were Russian. Similar observations were made in newspaper editorials (Kravchenko 2001).
The eliding of national identity shocked and incensed the Crimean Tatar respondents I spoke with. One woman related that she had confronted the census-taker saying, ‘First you ask, then you fill out the form’. The census-taker reportedly replied, ‘Why ask when I know what you are?!’ In this incident, the census-taker based her action on the individual’s surname, which was a Crimean Tatar one. This kind of assumption, extrapolating nationality from surname, is fairly typical of the former Soviet Union. In the context of the census, however, it shows that while the social ritual of a census presents an opportunity to generate support and feelings of patriotism, Ukraine missed her chance. The census failed as a ritual of reassurance or ‘verification’ because the manner in which it was carried out prompted suspicion and resentment.
The leaders of the Russian Society and other Russian ethnic organizations in Crimea were also concerned. They felt that many respondents, particularly those with some combination of Russian and Ukrainian heritage, would be inclined to answer ‘Ukrainian’ so as to align themselves with the Ukrainian state. This would make them eligible for the status and benefits accorded the titular nationality. The actual door-todoor recording of the census suggested their concerns were wellfounded. Many respondents seemed to have trouble answering the question about nationality. Often, the question was followed by a long and awkward silence before some reply was provided. At other times, the matter was decided in consultation with the census-taker. For example, when a woman did not know what to say about her son’s nationality, the census-taker suggested that, ‘Since nationality is determined by father in Russia, and by mother in Ukraine, we will write down “Russian” ’. The logic was presumably that even though we were located in Ukraine, the boy’s father was from Russia so the ‘Russian rule’ would apply. The choice between Russian and Ukrainian ethnicity was an especially forced one for respondents who had lived their entire lives on the peninsula and developed a sense of regional identity. With Ukrainian independence only a decade old, the process of realignment and rearrangement is far from complete. But at a broader level, the difficulty of choosing an ethnicity underscores the extent to which the categories are subjective, moulded by processes that are inherently political. Thus the technical exercise of counting was profoundly influenced by the social dynamics between census-taker and respondent. Based on these kinds of interactions, the results will be as much a reflection of the census-takers’ preferences as the respondents’.
While the Soviet conception of nationality has been deeply internalized, it is not immune to reflection. As one Crimean Tatar explicitly stated: ‘it’s like a national referendum’. The propensity to be self-reflexive was also apparent in two toasts made at the Karaim Society. I met with representatives of the Karaim Society because they were about to undertake an independent census with funding from the Soros Foundation. They brought out a traditional dish of kobete, and proposed we drink to being more like Americans: Americans say they are all ‘American’ even though they know they are Japanese-Americans, German-Americans, and all sorts of other ‘hyphenated’ Americans. My presence undoubtedly generated some reactivity here, but it appears the Karaims were willing to entertain the possibility that the nationality question could some day be removed from the Ukrainian census (cf: Arel 2002, p. 9). As they put it in the second toast, ‘Let us have enough pride to feel equal to all the rest and not have to declare our nationality’. They suggested that it is out of a feeling of inequality that they feel compelled to focus on difference. These two toasts encapsulate a tension inherent in the first Ukrainian census. On the one hand, there was a certain recognition that they might be better off ‘hyphenated’, that is, without the rigidly defined national identities that set them apart from one another. On the other hand, it was still necessary to have these identities to lay claim to rights, power, and resources, for a classificatory logic so thoroughly pervades the milieu.
Language
The Crimea is a predominantly Russian-speaking region, and the imposition of Ukrainian as the state language presented the population – and the government – with a new challenge. One official complained that every state office now needs translators and interpreters to communicate with Kyiv. ‘Imagine the cost!’ he exclaimed. The question of language instruction in the schools was also a primary concern. The Russian society asserted that based on ‘scientific polls’ there was a broad base of support for more Russian language in the schools, although Ukrainian respondents complained of widespread linguistic Russification. As the head of the Russian society put it, ‘What kind of rebirth of Ukrainian language can there be if there was never a basis of Ukrainian culture here in Crimea? It’s contrived’. This issue was played out in a dramatic way during the census.
In the practical execution of the census, the status accorded Russian was reproduced. In fact, it appeared that certain state actors, perhaps the census-takers’ instructors, were interested in elevating the proportion of Russian speakers. Census-takers probed respondents who answered ‘Ukrainian’ was his or her native language, while taking all other answers at face value. In other words, if a person responded ‘Russian’ or ‘Crimean Tatar’ (speaking in Russian) it was simply written as stated. If, on the other hand, the answer was Ukrainian language, the census-takers were instructed to ask for clarification. This created a certain pressure for respondents in Crimea to scale back their level of Ukrainian.
What is important for the validity of the question is that each of the census-takers asked for clarification in his or her own idiosyncratic manner. For example, one of the census-takers asked ‘fluently, fluently?’ A second census-taker intimidated her respondents by asking, ‘Are you able to translate?’ A third tack was to choose a Russian word and test the respondents’ ability to identify the correct Ukrainian equivalent. Given that only a small portion of the population would know the Ukrainian for words like ‘drill’, most would simply use Russian.
Thus in the practical execution of the census, the high status accorded Russian was reproduced. This general tack undermined trust. Altercations arose between the census-takers and respondents over perceived slights on the language issue. According to my consultant at the Department of Statistics, there were some instances in which people refused to answer questions until they were approached in Russian, while others insisted on Ukrainian. While the Ukrainian census-taker I followed shifted between Ukrainian and Russian depending on the respondent, the Russian census-taker was less flexible, speaking only in Russian. Some initially refused to answer her questions. She nevertheless prevailed by telling her respondents that, ‘in Crimea, the forms must be filled out in Russian so that the machines can read them’. This should have been taken as preposterous, given the machines in question do not read any handwriting and the forms themselves are printed in Ukrainian.
The belief that ‘the machines’ could only read Russian handwriting is one of the ways in which the process of statistics-gathering produced its own mythology. This mythology leads straight to a major inconsistency in the approach of the Ukrainian state: in western Ukraine, where Ukrainian language usage predominates, census-takers received the instructions that the machines only read Ukrainian handwriting. Even though the state’s approach was far from monolithic, the practical politics of the census mirrored large-scale demographics and tensions inherent in Ukrainian statehood. The explanation about ‘the machine’ worked for the census-taker I observed. However, the instructors at the various field offices had less favourable outcomes to report. Respondents had seized documents and complained to Kyiv. Respondents had also taken issue with the wording of the question. For example, one of my census-takers related a woman who, when she was asked ‘native language’ (rodnoi yazyk) objected that the census-taker was asking the parenthetical version of the question. The respondent ripped the forms from the census-taker’s hands and crossed them out.
Here it should be explained that the stem of the language and nationality questions had been reworded, leaving the traditional language in parentheses. In practice, census-takers simply ignored the rephrasing. They asked the most abbreviated form they could, using ‘native language’ (rodnoi yazyk) and ‘nationality’ (natsional’nost) rather than longer forms, ‘the language in which you speak’ or ‘your ethnic background’. The use of words from the parentheses was a source of tension with respondents who wanted the long form of the question read to them. When I asked officials, national elites, and census-takers, however, they felt that respondents would not understand the question unless abbreviated. Thus, the nuances of language and nationality debated at the planning level fell into the gap between census-takers and respondents.
Given these micropolitics, it should not be surprising that the preliminary results suggest an isomorphism between language and ethnicity that does not obtain on the ground. For example, the preliminary census indicates that among Crimean Tatars 92 per cent speak language of nationality, 6 speak Russian, 1.8 speak other language, and 0.1 per cent speak Ukrainian. These results obscure the marked linguistic Russification of the Crimean Tatar people and lack validity. The question ‘native language’ evokes for respondents a political category of belonging, not the language in which they are most comfortable or speak most fluently.
There was a certain amount of resistance to the language question from the Crimean Tatar contingent. As Kertzer and Arel have pointed out, ‘Ever since censuses began, state efforts to pigeon-hole each individual into a single category of identity … have faced resistance’ (2002, p. 27). One day while I was sitting in the waiting room of the Department of Statistics, I listened to the statistics officers commiserating. They lamented some of the difficulties they were having. For example, one woman related an incident to the others in which the census-taker was polling a Crimean Tatar.
Census-taker: Native language?
Respondent: Crimean Tatar.
Census-taker: Any others?
Respondent: No, no other languages.
The difficulty? The entire conversation took place in Russian. This ‘punch line’ was received by the other statisticians (who had their own stories to recount) with rolling eyes and raised brows.
This consideration of the nationality and language questions demonstrates how plagued by difficulties was the first Ukrainian census. What makes these difficulties particularly significant for Ukrainian statebuilding is not so much the question of validity, as the larger question of trust. As Refat Chubarov indicated when I interviewed him at his office at the Verkhovna Rada, the Ukrainian Upper Parliament in Kyiv, ‘The biggest problem is the distrust of the people toward the government’.
Subjects and objects
As early as 1926, ethnographers and state planners began to discuss the right of an individual to choose his or her own national identity (Hirsch 1997, p. 260). Hirsch points out that by 1939, the principle of national identity had become so firmly linked to national self-determination (and sovereignty) that there was a certain amount of back-peddling on the part of authorities who wanted to avoid the issues raised by multiple and competing claims. This led ethnographers and government representatives to question whether more ‘empirical’ accounting might be more expedient. They considered having nationality determined on the basis of ‘objective’ criteria identified by the census-taker, but the course of Nazi Germany dissuaded them from adopting this approach. The idea eventually sanctioned was that choosing one’s national identity was part of being Soviet (Hirsch 1997, p. 274).
The state’s willingness to take nationality and other information at a person’s word produced a divided reaction during the 2001 census. Intellectuals and elites readily accepted nationality as an eminently subjective dimension of identity that was liable to shift within a person’s lifetime. However, the institutionalization of this category during previous censuses, on passports, and on other official documents still concretized this aspect of identity for many. Once nationality was surrounded by such a factual aura, the census-takers’ lack of interest in documents became a source of great concern. In many of the homes I visited, respondents rushed to get their passport, only to be told it would not be necessary. This was unnerving: as one consultant put it, ‘It’s not serious! If they had done it according to documents, then it would have been serious. But they are doing it by the words of person!’ Many people were so troubled by the state’s apparent willingness to accept such important data ‘from the words of a person’ that it undermined their confidence in the census. As a political scientist at Tavrida National University, a respected educational institution in the capital of Simferopol put it, ‘It’s going to be so inaccurate! I doubt we can take the results seriously’. Scott’s (1998) observations about the simplifications of the state are apt here because state-created institutions (such as the passport and the propiska) have come to orient thinking and structure experience. Hence my consultant at the Department of Statistics was inclined to see it as something of a contradiction that the passport was not required for the census, even though this approach is fully in line with European standards (Arel 2002, p. 10).
Could the preference for documents over ‘the words of a person’ represent the opposite of what Handler (1988) identified as a ‘possessive individualism?’ In possessive individualism, people feel a sense of ownership about their identity, which is something they themselves produce (Kertzer and Arel 2002, p. 34). Does the desire to fall back on passports, work records, and other documents indicate a reluctance to actively claim an identity? Not if we consider the discomfort with the use of notebooks. In the preliminary census beginning 23 November (the official census began on 5 December), census-takers were sent out with notebooks to be used as training devices. Realizing it was difficult to get through a single interview without revisions, many census-takers kept using the notebooks for the census itself in order to preserve the cleanliness of the documents. They complained that people changed their answers in the course of the interview, and they had to sign for their blank census forms, which were numbered. To avoid wasting official documents, they recorded the information in a notebook, and then recopied it at their field office. However, it did not take much imagination for the individuals polled in such a manner to suspect gross violations. Their concern signals not only suspicion of state representatives, but a desire to have the power to designate their own identities. The use of the notebooks was at the head of the list of items for discussion when my consultant from the Department of Statistics organized a meeting for the heads of the field offices. However, it was only two days before the end of the census that she planned to preclude their use.
Another topic of concern was that an individual did not need to provide data on him or herself. Anyone co-habitating was acceptable. This was a perfectly reasonable approach in single-family homes, but it became untenable in the dormitories where respondents may or may not have known their absent room-mates. In the dormitory, a few gave the information they could, while others declined. If it is in fact preferences that are registered by censuses, then again we must ask whose preferences: in the 2001 census we can discern traces of census-takers, censustaker’s instructors, and room-mates of respondents.
Informal conversations among the census-takers revealed other emergent questions. When the census-taker I was following happened upon people who were also employed by the census, their discussion suggested that the seemingly repetitive process of taking down census information was anything but methodical. They had been issued an instruction booklet, but found that it was riddled by a ‘play of words’. The instructions coming from the heads of the field offices were only a marginal improvement. For example, just what constituted an ‘independent household’ was never made clear. Census-takers were especially uncomfortable with the question about income because they could not guarantee confidentiality. It was feared the census would be used to trace and track the Ukrainian citizens who rely on the informal economy to survive. Others were troubled by this question because they suspected a high level of inaccuracy. A consultant chuckled that, ‘In the process of answering the census, people rethink what they should say. This shows that first, the questions are not always clear. And second, they are lying somewhere: you have to think to make everything correspond. Lying is complicated! You have to have brains to lie!’ His comment reminds us that far from exhibiting pure primordialism, there is a level of self-reflexivity on the part of respondents.
Most vexing for the census-takers was the problem of whether or not to turn in their forms for individuals who claimed to have been polled in the preliminary census at another location. Since some census-takers attempted to beat the clock by using the preliminary census as an actual census, this was important. Turning in their forms based on the place where people were residing on 5 December, the official start of the census, could lead to double-counting.
Lies and jokes
Having observed the kinds of difficulties I have been discussing, one consultant exclaimed: ‘It’s not going to be objective! Have you heard the old saying about statistics? “There are lies, there are big lies, and there are statistics”’.
Defining statistics as an official ‘lie’ is one view of the 2001 census. But there was also a lighter version evident in the jokes and general hilarity that surrounded the process. Just as the census was met with indignation and incredulity in some quarters, it was made light of in others. For example, in one dormitory room where the census-taker requested information on an absent room-mate, the girls and their boyfriends broke into laughter. The girls quipped that, ‘we can make something up, and if it’s not right, we’ll say “she was drunk”.’ This was followed by laughter that even the census-taker had trouble resisting.
The attitude towards the census in the press alternated between official pronouncements admonishing the public to open their doors and fulfil their obligation to be counted, and more buoyant reporting. For example, one article described how a man thought the census-taker was actually a prostitute he had ordered, just imaginatively posing as a census-taker. A Kievskie Vedomosti took a darker approach in noting the final results are still a long way away. To fill the gap, Kievskie Vedomosti supplied some preliminary ‘results’, as follows:
- 18% of the People’s Deputies were unable to provide the source of their income.
- 15% informed the census-takers that they had already been recorded that week, after which the VCR disappeared. Incidentally, at the time of the census, 475,000 wristwatches disappeared and 500,000 pens were stolen.
- 75% of the women in Ukraine are under 30.
- 47% of the population stated that they speak Ukrainian, but at the same time told the census-takers ‘where to go’ in Russian.
- 5% of those questioned demanded money for accurate answers.
- At the end of the census 5% of the census-takers were missing, 7% had the flu, and 3% had married respondents.
- 12% of the Ukrainian population remained uncounted because they could not tear themselves away from the television serials.
- 20% of Ukrainians were recorded by sight – through the peephole.
- 16,000 apartments are mistakenly registered as one room due to the fact that the census-takers were not allowed any further.
The humour in these findings derives from the anxieties they draw upon. For example, widespread reluctance to respond to the question about income combined with distrust of officials appears to have yielded the ‘statistic’ on People’s Deputies who could not respond to questions about their source of income.
Similarly, popular images of the census were mixed with reports of misdemeanours. During the census, newspapers reported that criminals were posing as census-takers in order to execute robberies. Hence we are offered ‘statistics’ on the number of VCRs, wristwatches, and pens that were taken. There is also the statistic that 20 per cent of Ukrainians responded through the peephole. The statistic on the number of apartments that were mistaken for one room further confirms the fear of being robbed or harmed in some way, which was shared by respondents and census-takers alike. The fear experienced by the census-takers, most of whom were women, should not be underestimated. The census-takers I observed felt that it was a frightening task, and expressed gratitude that uniformed soldiers had been provided. Unfortunately, the soldiers’ day ended at 7:00 p.m., a peak time for census-takers who wanted to find people at home. Incidentally, the soldiers themselves were not above making light of the census. For example, when one woman called out from behind her door ‘Who’s there?’ The soldiers answered ‘Santa Claus’ before the census-taker could interject ‘census’.
The Kievskie Vedomosti article refracts the tension surrounding the nationality question through the tensions concerning the language question: the census-takers were told ‘where to go’ in Russian. The ‘statistic’ that 75 per cent of women are under thirty draws on the anxieties of women and the equation between youth and beauty. The humour surrounding the first independent Ukrainian census could be further unpacked, but I think it is clear that even as state officials struggled to project an image of competence and technical expertise, the Ukrainian population, when they did not resist or become angry, expressed their opposition obliquely. Census humour is part of a much wider tradition of humorous cynicism, much of it directed towards the state, that has pervaded public life especially since Glasnost. While this humour does not produce trust, it does constitute a kind of social ritual that identifies and affirms collective suffering.
Who calls the tune?
Enumerating a population is a technology of power that provides some interesting openings for resistance. Observing how state-defined categories shape individual self conceptions, Scott argued that ‘State officials can often make their categories stick and impose their simplification’ (Scott 1998, p. 82). It is true that in Crimea, the spectrum of identities had been subsumed under state-sanctioned categories. To census respondents, the categories had a truth value. However, we should not go as far as Scott does to say that these categories constitute ‘an authoritative tune to which most of the population must dance’ (1998, p. 83). It was plain that in post-Soviet Crimea, who called the tune was often unclear. Two brief examples dramatize this point. On the third day of the census, there was an uproar over a high-rise where approximately three hundred people had refused to answer the census. The residents of the apartment building had been without heat for some time, but with the worst winter weather to date, the situation had become unbearable. They wanted the heat in their building restored, and the flooding in the basement resolved. My consultant and her colleagues immediately sent a brigade that restored heat and began to address the flooding problem. At this point, all but a few of the building’s residents relented and agreed to be polled.
Another incident illustrating the extent to which the state lacked a monopoly on the power to ‘call the tune’ concerns the disgruntled Crimean Tatar population. As the official census was drawing to a close, the Crimean Tatar political body or Mejlis found out that the preliminary totals of Crimean Tatars did not match the Mejlis’s statistics. In a personal interview, Remzi Ilyasov (acting head while Mustafa Dzhemilev was in Kyiv) threatened that ‘heads would roll’ if the Department of Statistics did not go out and find the Tatars necessary to attain the numbers they believed were there. As of 13 December, only 184,700 had been counted. According to the Mejlis, this was only 70 per cent of a total they believed to be closer to 265,000. This threat on the part of the Mejlis was surprisingly effective: when I went to the Department, I found the officials there mobilized by their own conversations with Ilyasov. My consultant at the Department was scurrying to find a car so she could go and hopefully placate Ilyasov before the tension escalated any further. Then, when she took me out on her ‘rounds’ of the various field offices, she intentionally went to a Crimean Tatar neighbourhood and admonished the head of the field office to make sure to locate all the Crimean Tatars in her census tract. She drilled the head of this field office about their practices, and expressed concern that the Crimean Tatars who had land there but were living elsewhere might not be counted. Knowing the propensity of Crimean Tatars to take to the streets and demonstrate, or carry political weight in international circles, attending to the Crimean Tatar contingent seemed particularly urgent.
In these two instances, it was state actors who were ‘dancing’, to a tune they neither called, nor expected on the play list. Of course, in the first instance, the protest was not about ethnic categories or the census per se. The census simply provided a convenient opportunity for the apartment residents to gain leverage they otherwise lacked. In the second instance, the issue was not so much the outlines of the categories that Scott was referring to but their contents, as manifest in the dispute about undercounting. The Mejlis advocated a number of strategies for reaching their totals: coding anyone stating they had been born in Central Asia as Crimean Tatar, doing a recount, and calling the Crimean Tatar people to go to the field offices themselves and verify that they had been counted. In the end, a recount was unnecessary. In the last days of the census, additional Tatars were counted and a total of 248,000 was reached.
This discussion of the micropolitics surrounding the census suggests that there may be better metaphors than Scott’s metaphor of dancing. To be fair, Scott emphasizes that state simplifications of the population are always a work-in-progress. The ‘project of legibility’ is one that is never fully realized (1998, p. 80). State simplifications are always marked by inaccuracies and omissions. And while complete legibility is an aim, it is often undermined by rivalries within the state, as well as the resistance of its subjects.
Conclusion
The ethnography of the first independent Ukrainian census illustrates that no matter how ‘flattened’ or ‘enclosed’ (Appadurai 1993, p. 334) a population may be by the statistics that represent them, the process of statistics-gathering is imbued with a full spectrum of emotions. Administrators and respondents alike sensed the process of enumeration was embedded in a larger process of making meaning that linked social identities to political goods and material rewards. The legibility of the Ukrainian population was a goal, and officials aimed for a rational accounting of a diverse, multiethnic, and multilingual population, but no one seemed to believe legibility was achieved.
One consultant speculated that however inaccurate, the census was a learning exercise. He was among those who inserted the census in a narrative about the transition, suggesting that while there is still a danger of going in the Soviet direction, it appears that Ukraine is moving in the ‘right’ direction: ‘We are on the edge of going in the direction of real democratic reforms’. Having subscribed to neoliberal ideologies of transparency that suggest individuals, and by extension societies, can be made legible, it appeared just a matter of time and practice. But based on the tension that arose between the Crimean Tatar leadership and the Department of Statistics, I suspect this is too optimistic an assessment. Correspondence with consultants in Crimea reveals festering resentment. Rather than constituting a ritual of reassurance, the census provided new terrain on which future disagreements could arise.
National statistics-gathering is perhaps the paradigmatic instance of knowledge as power. As Urla has argued, the professionalization of statistics-gathering entails ‘a uniquely privileged way of knowing the social body and a central technology in diagnosing its ills and managing its welfare’ (1993, p. 819). The information generated gives those in power not just a picture of the population or rational criteria for designing interventions: it offers concrete justification for the language, housing, and news media policies put in place.
If a challenge facing the Ukrainian state is consolidating its national identity while taking into account the needs of multiple constituencies, then the census presented an opportunity. The preliminary results show an increase in Ukrainians by 0.3 per cent. Russians declined by 26 per cent. But if we take an ethnographic approach to more closely examine the micropolitics involved, a more complicated picture emerges. In Crimea, the census revealed that citizens have taken defensive, sarcastic, and aggressive stances towards a state they neither trust nor entirely respect. The distress of individuals who were neither approached in the state language, nor asked about their national identity seems to suggest that in this case, Ukraine missed an opportunity to reassure its ethnic minorities and consolidate Ukrainian identity. While the offences and insults that came with the census have faded, and Ukrainians have moved on to more pressing issues, ethnography reveals the extent to which the census is an inherently political process. Neither dry nor predictable, it comes with difficulties and surprises.