A census is an example of the social construction of knowledge and the politics of measurement. Measuring people assumes a political significance because it entails converting heterogeneous populations into numbers – stable pieces of knowledge that can be easily combined and manipulated. In constructing such numerical representations, census officials claim to be creating an objective portrait of the population. Censuses, however, also contribute to something less tangible by playing a key role in the creation of what Benedict Anderson has termed an “imagined community.”^[1] General censuses provide states with a unique opportunity to unify space and populations with a single instrument. Furthermore, in their quest to secure a statistical portrait of what their polity “is” census officials shape the resultant outcome on the basis of categories derived from their own conceptions of what their polity has been. Measurers’ agendas and biases become objectified in the construction of the census form, the creation of census categories, and the publication of census data. The census-taking component of the imagining process is itself divided between central census administrators (those who create census forms, rules for their completion, etc.) and census workers in the field – local census authorities and enumerators whose own conceptions of what is being counted intrude into the interpretation of rules and the enumeration of people. Ultimately, the numbers derived from the census process are used to reify or alter prior images of the polity and to evaluate, conceptualize and control.

At the same time, the census process contributes to the creation of imagined community/communities among the counted themselves. By drawing individuals into a general process a census, like mass media, closes spatial and social distance by creating an opportunity for persons to imagine themselves as part of a larger whole that can be placed in similar categories. Indeed, repeated censuses create what Anderson terms a “bound seriality” – a series of vignettes that together provide a population not only with a collective image of the present, but of the past and the future.^[2] Furthermore, the standard census practice of self-identification provides an opportunity for the counted to express their own sense of which official categorical communities they belong to within the whole. Censuses shape a sense of identity in the minds of the counted – either in terms of a population group’s sense of numerical (political) strength or in the sense of a cohesive group response to perceived mis-categorization by census-takers. Therefore, for all of its supposed objectivity, census taking is a political act that is inextricably linked to questions of power and identity (both individual and as part of an imagined community).^[3] A historical understanding of the process of census taking is thus crucial to understanding the categories through which we view ourselves as human beings – both in the past and in the present.

Historians, demographers and social scientists, in search of a statistical foundation for their own research, have looked to the first general census of the Russian Empire as an important source of information. As such, their discussion has focused on data quality and the methodological manipulations necessary to make them useful.^[4] However, “the question of the accuracy of these data is relevant only when the census is approached as a source of information; the problem becomes largely irrelevant if the census is studied not as a data source for research into other subjects, but as the subject of research itself.”^[5] Several studies have examined Russia’s tax censuses (revizii) and the first general census of 1897 as administrative tools and sites of social interaction, but none have presented a comprehensive and empirical study of this topic.^[6] Such a study is crucial to answering questions connected to identity in the late imperial period.

From eighteenth-century attempts to catalogue the “other” to the Slavophile-Westernizer debates and the shock of discovering Russians assimilating into Siberian cultures, identity issues were nagging questions within the Russian Empire.^[7] Many of the conflicts of the period centered on differences between the state’s shifting conception of the empire and the evolving identities of various groups in Russian society. Past historiography has generally portrayed this as a one-way street – as the autocratic state working (largely unhindered except for periodic outbursts) to bring order to an unruly and diverse territory and largely failing (under-government), or as an emerging “civil society” struggling to liberate itself from autocratic oppression. The history of the first imperial census suggests something more complex – that the dialogue between competing images of the empire pointed toward a vision that was neither that of the state nor its opponents. The census, as an official act, was a dialectic of social identity that gave both census-takers and subjects the opportunity to define and legitimate their interpretations of Russia’s socio-political organization.

For purposes of space, this paper focuses on the census process from the census-takers’ perspective – that of state officials in St. Petersburg, local servitors, and volunteers. Census categories provided the government with a means to construct society in “proper order,” and various groups in imperial society found these categories convenient (even if they had reservations about how they should fit into them). From this perspective, the census appears as a means to legitimate the dynastic empire by using a single instrument to define the latter as a single territorial entity with a single population (in Anderson’s words, stretching the “skin of nation” over the bulging body of empire). It provided the first opportunity for nearly all inhabitants to do this simultaneously. Given levels of literacy, and thus the diminished role that “print capitalism” could play among the population as a whole, this was perhaps the first way in which many could imagine themselves as people who shared the same larger space.^[8] Emphasis on the census as a complete enumeration of the entire population positions the census as an attempt to create a sense of horizontal integration and legitimation (a sense of what Jьrgen Habermas and Miroslav Hroch would call “the public sphere” or “civil society”) while at the same time maintaining the vertical order of the dynastic empire.^[9] Although Russia has been described (following Antonio Gramsci’s view that “in Russia the state was everything, civil society was primordial and gelatinous”) as a place where state and society were mutually exclusive entities, the census process suggests something more complicated. The state could in fact construct civil society for its own purposes and it did.^[10] From this perspective, the census can be placed within the context of the autocracy’s attempts to define “imperial citizenship” and within the context of its on-going attempts to rationally “order” the empire to facilitate pursuit of an imperial agenda.^[11] To investigate this we shall examine, first, the origins of the census project as a means of illustrating what the census meant to state officials. A discussion of the way in which the census project attempted to create a sense of imperial citizenship and the confines of that citizenship follows. The boundaries of citizenship (and the dilemma posed by geopolitical and cultural constraints) are illustrated with a number of examples from the census process itself.

Why a Census? Origins, Context and Meaning

The 1897 census was not, of course, the Russian state’s first population enumeration. When seventeenth-century household taxes and military recruitment procedures proved inadequate for waging war in the early eighteenth century, the state instituted periodic censuses of the population liable for the new poll tax and military recruitment. The first of these censuses of the taxable population, or revizii, took place in 1719-21 and the last (the tenth) was finished in 1857/8.^[12] By the time of the last revision, however, developments were already rendering this type of partial population enumeration obsolete. As Academician and statistician P. I. Keppen, noted in his 1857 monograph, The Ninth Revision, such partial censuses failed to provide “either material for scientific conclusions, or other bases for the discussion of many legal and administrative questions which require information on the distribution of the entire population by age, social estate, religion, family position, tribe, literacy, occupation, and ability to work.”^[13]

The idea of conducting one-day censuses of the entire population on a standard form reflected the “rise of statistical thinking”, a belief – fostered by the Belgian astronomer Adolphe Quetelet and others – that the compilation of mass data would ultimately yield general social laws or a “social physics.” This mania for numbers became institutionalized in 1853 with the convocation of the first International Statistical Congress in Brussels. The Brussels congress began compiling a standard census form, and subsequent congresses built on this work. Beginning in 1857, Russian representatives from the Imperial Russian Geographic Society regularly participated in these congresses. This connection between the Statistical Congress and the reform-minded enlightened bureaucrats of the Geographic Society solidified these reform-minded officials’ faith in data collection as a necessary tool for reforming and properly directing imperial society. In 1860, the Geographic Society placed the idea of an empire-wide census on the public agenda by calling on all interested parties to submit census plans for a juried contest. Several enlightened bureaucrats remained closely tied to proposals to carry out a census throughout the second half of the nineteenth century. P. P. Semenov-Tian-Shanskii, who first attended a congress in 1863, ultimately became de facto chair of the Main Census Commission in 1895. The 1897 census was thus in many respects a belated act of the Great Reform era. Some officials even suggested that the census be announced in the form of an imperial manifesto similar to that issued to proclaim the end of serfdom.^[14]

Like the Great Reforms themselves, the census was necessary for the “correct direction of the state.” The Interior Ministry’s preface to the census project it submitted for State Council approval in 1895 not only presented the census as a continuation of the state’s practice of counting its subjects, but specifically linked the revisions to Peter the Great’s reforms and accomplishments. The Russian Empire could not hope to maintain its position without the ability to keep track of its subjects in terms of numbers and location.^[15] As Semenov argued in a brief supporting the census project:

All states of the enlightened world, both Old and New, have already long ago come to the full realization that for the correct direction of the state it is impossible to proceed without accurate information – not only on the general size of its population and its territorial distribution, but even more so on its composition by age, family make-up, religion, nationality, estate and occupation, literacy, etc. …[A] proper population count can be founded only on the periodic production of one-day, state-wide censuses of the population by name...^[16]

The state could not fulfill its paternalistic function without relevant data, data that revisions of only those liable for the poll tax and conscription could not provide. Knowledge-based power would facilitate the practical demands of administration and, in turn, enhance the population’s well being. Indeed, supporters of the census project argued that the possession of such data might have enhanced state performance during the 1891-92 famine years. As one official pointed out, “the absence of any amount of satisfactory data on the size of the population and especially its age distribution presented not a little difficulty for the correct establishment of government relief…”^[17] Having such data, the autocracy would surely serve its subjects better in the future. Finally, as the Interior Ministry noted in a memorandum supporting passage of a census statute, “even parts of Asia where state administration is somewhat well-established” had managed to conduct censuses; the Russian Empire (by implication) would remain in the ranks of great powers only if it had vital information at its fingertips. After all, a general census “had even been carried out in India.”^[18]

More specific concerns also strengthened this association between statistics, control and autocratic paternalism. The process behind the 1861 emancipation of the Empire’s servile population revealed to officials a dearth of important data on the lives of the Empire’s majority peasant population. As Interior Minister Tolstoi noted in his September 1882 petition for the Emperor’s support of a census project, the census was a “dire necessity” as such figures would provide a basis for the peasantry’s redistribution of allotment lands between households and for the correct assessment of state and local taxes.^[19] Indeed, because peasant practice connected past revisions to the repartition of communal land, Tolstoi argued, the absence of a population count since the Tenth Revision was a key factor sustaining the peasantry’s “vain hopes for an additional [land] allotment.” At the same time, a soslovie-based population count would also cause rural disorder because of its association with tax levies.^[20] Maintenance of rural social and economic stability thus required a census of the entire population.

The military reform of 1874 provided an even more compelling reason for conducting a complete census of the entire population. The reform required that the population be considered as a whole, without distinction between social estate, by making all twenty-one year old males subject to conscription. The revisions thus only accounted for part of the draft-age population. This, combined with a new reserve system, required that the government have some idea of how many males would be eligible for service each year. In addition, omissions in the revisions and their infrequent nature meant that they were less than an ideal basis on which to calculate the defense of the empire. As Semenov later noted, “Unfortunately, revisions, as population counts, do not withstand even the most lenient criticism.” The revisions

did not provide answers for the most important questions connected to the composition of the population; they did not transfer to anywhere any basic material for statistical analysis, even important information such as the age distribution of the population... For this reason, when the question of how many persons in Russia would reach the age of twenty-one each year arose in connection with the law project on military obligations no one could provide an answer.^[21]

The temporary rules for discovering the number of draft-age individuals also proved unsatisfactory. Between 1878 and 1882 the government attempted to register and count those subject to military service by compiling family lists (semeinye spiski). For the taxable estates, this was done on the basis of extrapolations from lists compiled as a part of the Tenth Revision; other estates were asked to submit this information voluntarily. Semenov explained the situation in greater detail. The government calculated the size of the taxable population for 1874 as a base year by multiplying figures from the Tenth Revision by the rate of population growth. This put the entire process on an imprecise footing, especially as each provincial statistical committee carried out the revision, calculated the rate of population growth and extrapolated differently. Initial inaccuracies were then multiplied with each passing year.^[22] Besides this problem, Semenov noted, the family lists compiled in the period 1873-1889 were also highly suspect because they were regularly corrected only for families having draft-age sons. Thus, “one family on the list was registered as it was in 1873, another as it was in 1878, a third as in 1882, a fourth as in 1889, etc.” Semenov concluded by noting that, “For an experienced statistician this data provides only the surety that in Russia there are no fewer than eighty, and no more than 120 million persons.”^[23] As the Interior Minister noted, the existing system had “deprived” the compilers of this data of “any control,” a view with which Defense Minister Dmitrii Miliutin, himself an enthusiast for statistics, agreed.^[24]

When the abolition of the poll tax in 1886 removed the Finance Ministry’s continued need for a soslovie-based population count a series of commissions began work on a census project. The State Council finally passed a census law on June 5, 1895.^[25] Just as the Great Reforms began to dismantle the particularism of imperial society, the census intended to provide a measure of the population based on a standard form. Like the Great Reforms, policy makers argued that a modern census was necessary to enhance and maintain the dynastic state and the Empire’s position in the world. A complete census would allow the state to maintain order, manage public welfare and project its power abroad. In short, it provided a means of harnessing the Empire’s demographic resources for its own ends. Ultimately this meant that census officials would have to maintain a precarious balance between their definition of citizenship and that of the population.

Constructing an Imperial Identity

“Othering” was a process essential to maintaining the dynastic state; difference legitimated the emperor as an imperial adhesive. However, as much as categorizing the population facilitated the “othering” process it also pointed toward a horizontality that challenged the vertical structure of state and society. Enumerating individuals by name according to their own indications (those of the head of household), not according to official documents (enumerators were not allowed to ask for them) undercut the official conception of imperial society as a hierarchy of social estates or soslovia (sing. soslovie). The census counted the “developed” segments of the population equally (i.e., using the same questions) and reduced the entire population (including the imperial family) to numbers devoid of rank and, ultimately, holes on a punch card manipulated by one of Herman Hollerith’s tabulators.^[26] Nicholas II was the only person who could record his primary occupation as “Khoziain zemli Russkoi,” but, when aggregated, he became one of many Orthodox male heads of household who was a native speaker of Russian. Although the punch-card tabulator promised to infuse autocratic power with the electrical current of the modern age, Nicholas himself became nothing more than a manila card capable of inanimate manipulation and control.

In addition, the census defined the empire not as the dynasty, the state, or its territory but as the population (naselenie). This invention of the social was by far the most significant and long-lasting achievement of nineteenth-century statistics.^[27] The official title for the census, “First General Census of the Population of the Russian (Rossiiskaia) Empire,” had inclusive, dynastic implications; census categories amounted to official recognition of the empire’s diversity, and it was thus perfectly normal that Nicholas II recorded German as Empress Alexandra’s native language on the imperial family’s census form. At the same time, however, it elevated the population at large to center stage and highlighted the empire’s demographic strength – as opposed to its government or particularistic social structure – as its most important resource. This was the focus of the report on preliminary census results delivered by the chair of the Main Census Commission (MCC), Petr Petrovich Semenov (later Semenov-Tian-Shanskii). The government’s newspaper, Pravitel’stvennyi vestnik, and local newspapers perpetuated this first statistical representation of the imagined imperial community in its entirety, focusing on key demographic data related to their province.^[28]

The same horizontality that challenged the vertically ordered dynastic state legitimated a state-sponsored conception of an imperial civil society (“citizenship” or horizontality) within it. Semenov delivered his statistical portrait of the empire to a general session of the Geographic Society. This act can be interpreted as an attempt by the state to legitimate its imperial project before civil society – an attempt perpetuated by local newspapers and their focus on how their locale fit into and compared with the imperial whole. Dal’nyi vostok, quoting from Pravitel’stvennyi vestnik, offered its readers a statistical comparison of the Far East and other parts of the empire, paying particular attention to the region’s disparity of female inhabitants (and thus mirroring official demographic concerns in the region).^[29] In addition, even though officials became increasingly uncomfortable with self-identification (and thwarted it in certain circumstances – see below), they generally trusted the population to self-identify and, in doing so, made them more “citizens” than subjects of the empire.^[30]

The census also implied citizenship in that, for the first time, it attempted to mobilize the entire population for a single project. The MCC initially planned to conduct the census in imperial terms in the sense that it planned for all census work to be completed by local officials and representatives of leading sosloviia who were either required by the census law or expected by tradition to participate. As the Statistical Council discussed the census project in March of 1894, its members (including the chair, Semenov) included in the list of potential enumerators noble land owners (pomeshchiki), clerics of various types, rural teachers, tax inspectors, members of Peasant Affairs bureaus, and perhaps (if necessary) literate peasants. Ideally, approximately 4,200 enumerators would count 25,000 people each.^[31]

These projections, however, proved to be grossly out of touch with reality. The expected number of volunteers from the ranks of local officials and notables fell far short of the original estimate, and the MCC found itself deluged by frantic notes from provincial census commissions asking for money to hire additional enumerators. For some reason it did not occur to the Statistical Council and the MCC that traveling through the Russian countryside in January (plus participating in the sub-totaling of results) might not be an attractive proposition. The MCC responded by first making it clear that all persons deemed sufficiently literate (including women and sometimes Jews) could serve as enumerators. The MCC also pursued additional enumerators – especially volunteers – along a more traditional dynastic path. At the end of October 1896 Interior Minister Goremykin wrote to Count I. I. Vorontsov-Dashkov, the Minister of the Imperial Household, requesting that a medal for volunteers be struck in the Tsar’s name. The MCC planned to use the medal as a lure to recruit volunteer enumerators and as a reward to those volunteers who met a certain quota for a given area.^[32] Assessing the conduct of the census in St. Petersburg, the head of the city census commission noted that the medal and the Emperor’s accompanying decree were instrumental in convincing a significant number of persons of the need to perform their “civic duty (grazhdanskii dolg).”^[33] The fact that copies of these medals circulated on the open market in at least one locale is perhaps testimony to the census’ creation of a sense of horizontal integration among at least some segments of the population not traditionally included in the traditional dichotomy of state and society.^[34] Nearly 150,000 enumerators (many of them peasants) and other census officials eventually participated in the census, and the census project itself took on heroic proportions in numerous enumerator accounts.^[35]

The process of creating a statistical portrait of the empire – like Repin’s “Religious Procession in Kazan’” – became empire-wide in a social, as well as territorial, sense. The conduct of the census in Odessa illustrates this most lavishly. On January 22 city governor and census commission chair, N. A. Zelenoi, launched the census with a solemn mass in the great hall of the stock exchange. According to the correspondent for Odesskii listok, all twenty-nine census parcel leaders and all 640 enumerators (several of whom sang with the choir) attended the mass. Many wore their credentials (a pin), and all were treated to a patriotic blessing. Besides this sacral send-off, census officials in Odessa took the unprecedented step of publishing the names and addresses of all census parcel leaders in local newspapers to ensure that inhabitants knew where to turn with their inquiries. The census in Odessa thus became not only a sacred, patriotic duty, but also a public act of society as a whole.^[36]

The government itself encouraged the image of the census as a social act. As a MCC pamphlet prepared to explain census goals to the general public noted, the “undoubted increase” of the population and

the gradual development of living and economic conditions and other phenomena of public life – both favorable and unfavorable – are leading people of science and persons standing at the head of state administrations to a recognition of the necessity of direct observation over the life of the people, such that by means of reasonable and expedient measures extended by life itself, they might facilitate the development of public welfare.^[37]

The state thus asked the population to participate in a process that legitimated the dynastic empire by representing its order in a “scientific” manner – by incorporating the population into an “objective” system of representation that could be used to ensure the efficacy of state actions vis-а-vis its subjects. By extension, subjects in a sense became citizens responsible for the improvement of their own welfare. This was a new imperial vision in that it charged the dynastic state and subjects alike (as opposed to enlightenment/reform from above) not only with the task of maintaining order but of creating progress.

The legitimating power of the census explains why census-takers became so concerned with the accuracy and incontestability of the census and insisted on measures to ensure that the census process remained free of fraud. Oberprocurator of the Holy Synod K. P. Pobedonostsev was especially concerned that some persons would misrepresent their rank or social position. He found this possibility disturbing because the census, as an official act, legitimated any recorded data. In his words:

[I]n many cases persons who illegally misappropriate for themselves...the rank of priest or (according to the Austrian hierarchy) bishop, and, in addition, presbyters and preceptors in unrecognized sects, can use the revision [census] as an official act to receive indirect recognition of a rank which is not recognized by law. It might easily happen that persons [registrars]...(who are not rarely from the ranks of the young and inexperienced – students, middle-school pupils, teachers, etc.), owing to their inexperience…will make notations about these individuals according to these persons’ indications, whereas the rank which ought solely to be recorded is their soslovie (meshchanin, peasant, etc.)….^[38]

Thus, the Oberprocurator feared that the power of the census to define social identity would be used to further the interests of Old Belief and “sects” to the detriment of the official church. Manufactured consent – in this case, statistical representation of the demographic predominance of Orthodoxy – required diligence. The MCC took his fears seriously, and passed a resolution imposing stiff fines for malicious misrepresentation.^[39] Based on a fear of inaccuracies (intended or not) it also denied requests by local administrative bodies (state, city and zemstvo) to process or make use of any census data prior to its official publication.^[40] In addition, officials deemed the publication of census figures based on questionable raw data as not only “undesirable” but also “dangerous.”^[41]

Concern with the legitimating power of the census persisted after the completion of data collection and well into the process of aggregation and publication. As one member of a commission reviewing census processing noted in his post-census critique, any inaccurate figures would “injure the authority of official figures and would provide reason to doubt them.” Because it would be difficult to “re-establish trust” in government figures once it had been abused, commission member N. Osipov argued that, in view of the poor quality of some of the data, the Central Statistical Committee (CSC) should refrain from presenting a detailed analysis of several aspects of the first census.^[42] Eliminating certain planned tables (e.g., tables pertaining to the distribution of the peasant population in terms of their pre-1861 status, tables on children less than one year of age, tables on persons having completed military service, and tables listing the place of permanent residence for persons absent on the day of the census) would ensure “more trust in the census,” and hence, in the government.^[43] Finally, the CSC’s inability to complete census processing and publication in a timely, detailed and useable fashion (the last volumes appeared in 1906 and the individual publications for each province became increasingly simplified) became a source of embarrassment for the state because it undermined the very legitimacy that officials intended the census to create.^[44] The policy of banning zemstvo and city statistical work prior to the census, combined with the tardy publication of census results, erased much of these organizations’ original enthusiasm for the census project. Within five years of administering the census the government had to concede requests from both the St. Petersburg and Moscow city dumas to conduct censuses (in 1900 and 1902 respectively). At the same time, zemstvos and other cities deluged the CSC with petitions for access to census data.^[45]

The census, then, provided an opportunity for the imperial state to create a civil society compatible with its imperial project of ordering its domains (“seeing like a state”^[46]) and maintaining its great power status. At the same time, however, the all-inclusive nature of the census process – which subjected even the imperial family to enumeration and categorization – itself undermined the empire’s vertical structure. The state’s inability to fulfill the civic promise of the census – access to data vital to improving general welfare – further damaged the credibility of the Empire’s vertically ordered society. Furthermore the categorization of the population, although it provided an opportunity for officials to construct the empire according to their own vision, also created an arena for imperial citizens to contest their official place in the imperial order.

Categories and Conceptions of Imperial Order

Sources – mainly the journals of the MCC – reveal little of the actual debate behind the selection of census categories. Given the dedication of Semenov and other officials to the International Statistical Congress’ project (Russia hosted the event in 1872), it is possible that the Congress’ categories elicited little discussion.^[47] The information to be collected, as stipulated in Article 2 of the census statute, demonstrated a direct relation between the census project and the Statistical Congress’ general principles. Registrars would ask each person for their name, relationship to head of household, age, religion, native language, place of birth, occupation, literacy, and physical disabilities. The census would also follow the Congress’ recommendations on registering persons in terms of their permanent or temporary residence and collecting data by household. The census project authors also deviated from international proposals by ignoring the Congress’ recommendation to collect information on dwelling size and attached gardens, and adding a question on gender. In addition, the MCC divided the question on literacy into the questions “Can you read?” and “Where did you study?” It also added a question on military service (under pressure from the War Ministry) and divided the question on occupation into a request that respondents state their primary and secondary occupations. The final project aimed at dividing the Empire’s population according to eighteen different characteristics.^[48]

The nature of other categories added by the MCC – questions or “rubrics” on the census forms – reveals much about the state’s conception of the empire and how the census process would legitimate this image. For all of the horizontality implied in the census process itself, census categories aimed to concretize a hierarchical social structure in numbers. Questions on soslovie, sostoianie, or zvanie; place of registration (pripiska – for those required to be registered); and place of permanent/temporary residence represented a population that was predominantly rural where everyone could account for his or her social and geographic place in the vertical structure of the empire. More than any other type of information, the collection of such data distinguished the Russian from other European censuses. As St. Petersburg University Statistics Professor I. I. Kaufman noted in his summary of census difficulties, even though social estate was no longer a “topic of statistics” in Western Europe it was still an important category in Russia. “Social estate still has a great significance in Russia,” he noted, “because, on the one hand, the nobility has maintained and desires to maintain its state importance; it continues to live and desires to live by its full estate life. On the other hand, the peasantry, too, continues to live as a social estate, the maintenance of which…it considers the basic condition for the protection of its vitality...”^[49] The collection of these data thus represented an attempt to manufacture consent for the decaying (as the census process itself revealed) soslovie system and solidify this conception of the Empire’s evolving social order via the census.^[50] Indeed, the collection and tabulation of data on soslovie marked a pointed attempt to vertically order the Empire.

The most graphic explication of this appears in the projected tables for the first two publications of census material. These publications consisted of the sixteen tables listed in Table 1.^[51] Note the census categories being compared and numerically fixed in this tabulation process: age, social estate, native language, and religion. The breakdown and comparison of the population by social estate was a somewhat frivolous endeavor when peasants comprised such an overwhelming majority of the population. This choice of comparisons seemed no less puzzling to contemporaries such as A. Kotel’nikov, whose critique of the census asked, “Why…has there not yet been given a compilation of the most important descriptors of the population, for example by occupation, family position, or a combination of occupation with place of origin?” Why did the CSC “with great pains” wish to publish such information as the “distribution of children less than one year old” by social estate, native language and religion?^[52]

Table 1. Composition of Tables in First Two Census Publications
1. General tabulation of population by locale
2. Distribution of population by household
3. Dist. of pop. by sex & age (10-yr. age gps.)
4. Distribution of population by sex, age & literacy
5. Distribution of children < 1 year old by age in mos.
6. Distribution of population by soslovie
7. Dist. of pop. by soslovie & age (10-yr age gps.)
8. Dist. of children < 1 yr. old by soslovie & age in mos.
9. Distribution of population by native language
10. Distribution of pop. by native lang. & soslovie
11. Dist. of pop. by native lang. & age (10-yr. age gps.)
12. Dist. of children < 1 yr. by parental native lang. & age in mos.
13. Distribution of population by religion.
14. Dist. of pop. by religion & age (10-yr. age gps.)
15. Dist. of children < 1 yr. by religion & age in mos.
16. Dist. of pop. by place of birth and soslovie

The answer was that, for a government whose political authority rested on Orthodoxy and the cooperation of the noble estate – both those born to it and those who earned it through government service – the comparison of these data by social estate can be interpreted as an attempt to discern its fate in the census. The combination of social estate and age provided clues to the demographic vitality of the ruling class. Similar comparisons for religion and native language indicated the demographic vitality of Orthodoxy and the Great Russian people. The distribution of children less than one year old by all three of these categories provided an indication of where the reproduction of rulers, Russians and Orthodox stood in comparison to the ruled, non-Russians, and the non-Orthodox. Even after census processing fell behind schedule and over budget most of these original combinations remained or were expanded (for example, literacy received a more prominent place). In fact, the CSC only rejected the tables on children less than one-year old because they relied on questionable data. From this perspective the collection and tabulation of census data on soslovie emerges as an attempt to concretize an existing (though recent and developing) category for purposes of control.^[53]

The question, “place of registration” was also unique to Russia’s census and also represents an attempt to use the census to order the Empire. In theory, the question simply required that respondents indicate the village, community, township, district and province in which they paid taxes. The purpose of the question was, in part, to be able to trace the migration of the empire’s labor supply. Although acknowledging the new reality of a mobile labor force, however, the question’s very nature aimed more at a Russia where, in spite of labor mobility, every person “knew their place” – even if they did not happen to be living there on the day of the census. Four different census forms (A, B, V, and A/B) reinforced this idea. The census counted persons on one of these forms based on the distinction between rural and urban areas, land tenure, and local circumstances.^[54] The compilers of the census did not anticipate that subjects would not know and willingly divulge their place of legal registration. For persons not residing in their place of registration, this information should have been readily accessible from internal passports (which they could consult, but enumerators could not). Census-takers, however, discovered that although a majority of the population (probably those still residing in or near their place of birth) could state the province, district and volost’ to which they were registered, many had problems providing answers as one moved down the scale of territorial units.
Attempts to pin down the rural population via the census (the statute required that a third copy of the results be used to update volost’ records) were among the least successful – most likely because some peasants attempted to use the census to escape the bounds of their communes (although the official explanation attributed problems to the fact that many villages had changed membership between neighboring communes on more than one occasion) or legitimate illegal family divisions.^[55] Indeed, one suspects that the census was seen and used by many as an opportunity to liberate themselves from their official place in society – to be citizens of the imagined space created by the census (part of a broader civil society) rather than subjects. This was especially true for the Empire’s official outcasts, exiles and their progeny, who not only sought to change their status in the census process, but also served as enumerators.^[56]

Questions on age, place of birth, occupation and military service not only represented the human resources at the state’s disposal and their movement over time, but also underlined this vertical structure (even as the census placed the population itself at center stage). In particular, the process of collecting data on primary and secondary occupations indicated a conception of the peasantry little changed from previous centuries. The question of primary and secondary occupations created much confusion, in part because it was an alien concept to peasant respondents. Part of the problem resulted directly from the time of the census; in January, many peasants working in factories probably answered that they were workers, even though their primary occupation was on the farm. Another difficulty arose from the fact that many peasants did not specialize, which meant that there was a great possibility that the secondary occupation identified in January 1897 was not the same a year (or less) later. Indeed, just as some peasants went to the factories in the winter, rural off-farm labor also varied with the seasons.^[57]

There are indications that, when enumerators explained “primary occupation” as “main means of subsistence,” peasants generally understood and provided accurate information. However, when these actual responses differed from those expected by census officials (i.e., when peasants listed their primary occupation as otkhod, carting, etc.), officials changed entries to “farming” based on the soslovie of the respondent. In such cases the “proper” primary economic place for peasants lay in agriculture. At the same time, when urban respondents classified their primary occupation as farming and secondary occupation as factory or some other urban work, census processors sometimes switched the two. When inhabitants in the towns of the Akmolinsk region attempted to more accurately describe their status as “urban peasants (gorodskie krest’iane)” enumerators resorted to physical appearance to determine their soslovie. Such incidents often resulted in what Kotel’nikov termed “the proletarianization of the rural population.” In addition to this, “unemployed” was not included among the acceptable responses. The rules thus obliged registrars to paint a picture of full employment by recording some sort of occupation for each person.^[58] In all of these cases, “ordering” the empire via the census provided a means to create the “well-ordered” state to which officials long aspired.

At the same time that questions related to soslovie and occupation elicited a portrait of the empire’s social order, questions on religion, native language, knowledge of Russian, literacy, and source of education aimed at a statistical portrait of the imperial order in a more traditional colonial sense. Data collected under these headings promised officials a measure of the population’s level of development within the empire, indicating both the civilizing empire’s achievements and areas in need of attention. From another perspective, such data would allow the state to gauge the demographic strength of the metropolitan people (i.e., Russians) – to assess the vitality of the state religion, the state language, and the main carriers of these civilizing influences. That there was no specific question on nationality should come as no surprise and cannot necessarily be attributed to imperial motives. Given the MCC’s reliance on international standards (much influenced by representatives from ethnically homogeneous states), the absence of nationality as a specific category could be innocuous. Unfortunately, other than post-census explanations of the intent to divine nationality by cross-referencing data on religion and native language, archival records again reveal little of the decision-making process.^[59]

Yet, excluding nationality as a census category ultimately served state interests, as doing so placed more power over who belonged to what nationality in census officials’ hands. On the one hand, the dynastic state made gallant efforts to ensure that headings of census forms were translated into the local languages of as many subjects as possible, fostering inclusion on particularistic terms within the imperial polity – a sense of belonging to a single entity. One of the MCC’s first acts consisted of a circular to provincial governors asking that they submit detailed lists of the languages that census forms should appear in and the quantity of each translation required.^[60] On the other hand, census forms reinforced a more typical sort of particularism. The census statute excluded several regions of the North, Central Asia, the Caucasus and Siberia from the general count, deeming them too “uncivilized” to fall under the general rules.^[61] Such areas were so “underdeveloped” that they could not be “ordered” in a civilized fashion. They were to be counted on form A/B; in especially “backward” areas the MCC instructed local census commissions to eliminate several questions on the form (e.g., religion for various steppe nomads).

Form A/B, however, also came to objectify all peripheral areas’ subordination to the center. It became the norm not only in undeveloped areas, but in many developed parts of the empire as well – particularly the Polish provinces. Here, the existence of the all-estate gmina – the lack of soslovie-based particularism – rendered the distinction between forms A and B unnecessary. In this way, the census united the most “civilized” western and “least civilized” eastern parts of the empire, contributing to what one recent study has described as a broadening conception of the term inorodtsy.^[62] Poles and Germans found themselves being counted on the same form (albeit with a full compliment of categories) as Kazakhs (“Kirgiz”), and Turkmen. Ironically, the assistance of “undeveloped” enumerators from the indigenous population ultimately determined the success of the census in these areas.

The absence of a specific census category for nationality became especially important for the process of statistically ordering the Empire in the Western Provinces, where the category nationality and its association with demographic power mattered most. The sensitive nature of the Polish provinces entered MCC discussions early as a request from the Warsaw Governor General that enumerators in the region be predominantly Russian and that commanders of the region’s fortresses be allowed to conduct the census in such a way that vital defense information not leave fortress walls in the guise of enumerators with suspect loyalties.^[63] The region (especially parts of Sedlets and Liublin provinces – the future Kholm province) was also a key demographic battleground.^[64] Russian claims on much of the area rested on a demographic foundation, namely the existence of “Little Russians” – Ukrainian peasants whose demographic weight comprised an important component not only of Russia’s regional presence but of the proportion of Russians in their own empire. Without Ukrainians, Russians became a ruling minority rather than a majority.^[65]

It was also a region where the relationship between religion, native language and nationality was problematic. Given the policy of self-identification, the Uniate population of Sedlets and Liublin provinces threatened the ability of the census to represent Russian demographic power in the area. These persons became subjects of the Empire as a result of the partitions of Poland. Tsarist officials spent much of the nineteenth century attempting to reincorporate these “lost” Russians spiritually, as well as territorially. By 1875 the Uniate Church was officially non-existent and its members, now technically reunited with Orthodoxy, fully “Russian” again. Yet, evidence indicated to local officials that this was not the case. At the beginning of July 1896 the Warsaw Governor-General submitted a note to the MCC pertaining to the enumeration of the former Uniate population. The note suggested that religious data in several areas be registered not according to their own responses but according to information held by the local administration. The Warsaw Statistical Committee agreed, noting that, as many former Uniates did not recognize themselves as Orthodox, they would “undoubtedly identify themselves to enumerators as [Roman] Catholics.” The MCC expressed little sympathy for his plan, but invited the Governor General to St. Petersburg to discuss the matter.^[66] By August Pobedonostsev was familiar with the request from Warsaw; he wrote to inform the MCC that he would be sending a representative to meet with the MCC and the Governor General in order to ensure that “Catholic priests and Polish agitators” were not allowed to convince the local population to register as Roman Catholics.^[67]

The existence of a large Uniate population in Austrian Galicia made defining the population in this region even more crucial.^[68] Rumors that the census was a first step toward religious freedom – a benevolent gift agreed upon by Emperors Franz Joseph II and Nicholas II – permeated the region and increased tensions between census officials and the population.^[69] Pressured by the Warsaw Governor General and Pobedonostsev, the MCC instructed enumerators to record respondents according to their own indications, but to have the entries on religion corrected by local officials as needed.^[70] However, as enumerators entered the villages of these “former Uniates” (a term that the MCC used itself), they soon encountered difficulties – in many cases because enumerators failed to heed these instructions. The archives include numerous petitions from individual Uniate families and – in most cases – from entire villages. Appealing to the Interior Minister, the Minister of the Imperial Household and the Tsar’ – the “Father of the People” – petitioners explained how enumerators and local officials ignored their wishes to register as Roman Catholics (in direct violation of the rules printed on the census form). In some cases, petitioners noted, enumerators used their self-identification with an officially non-existent religion as a justification for recording marriages and births as illegitimate.^[71] To emphasize that their complaint was not based on any fickleness of faith connected to the census, many villages emphasized the fact that they had been practicing Roman Catholics for nearly thirty years (i.e., since the final official abolition of the Uniate Church). Indeed, there is evidence that local gmina officials sometimes “corrected” some families’ official registration from Orthodox to Catholic.^[72] Ultimately nearly 300 pages of petitions, and several arrests could not overcome the state’s desire to win the demographic battle between Orthodox Russians and Catholic Poles in the region (made even worse by the fact that many of these “Russian” petitions were in Polish).^[73] Although the heading “Uniate” appeared in the initial drafts of proposed tables on the population’s religious composition, it soon disappeared.^[74] In this case, the “ordering” of the dynastic empire inherent in the census process did not allow the “Father of the People” to preserve his role as the empire’s glue. This sensitive geo-strategic area required an adhesive more Russian (national) in composition.

A similar “ordering” on religious grounds that had implications for demographic power also occurred in relation to Old Believers and members of various religious sects. In this instance, the “russifying” tendencies of those on the ground were themselves often at odds with the dynastic conception of the empire – undermining the unified particularisms of the dynastic state with a russifying agenda. Religion was an important category in Russia. Although the state deemed it crucial as an ethnic marker in the Western Provinces, official policy generally leaned in the direction of pragmatic toleration. Both census-takers and census-subjects alike saw the census as an opportunity to define the religious state of the Empire. The official Orthodox Church, Old Believers and sect members sought to enhance or maintain their own numerical significance. For Old Believers, the stakes were high: census figures could portray them either as a numerically significant and vibrant group or as simply a collection of fringe elements. Official accounts of the census emphasized the fringe element image by highlighting irrational sectarian reactions to the census. The account of the official assigned as the census coordinator for the upper-Volga provinces devoted twelve pages to this subject, most of which portray the Old Believers and sect members as absurdly superstitious. Thus we learn that many had the effrontery to attempt to register as “Orthodox” (which, in their own minds, they were). Others viewed the census as apocalyptic, the third and “final” census. Others simply refused to be counted by the anti-Christ.^[75] In effect, the account emphasized the established connection between religious and civic deviancy.

In addition, some registrars (often Orthodox priests) openly under-counted Old Believer and non-Orthodox villages or refused to recognize the legitimacy of their marriages and children. Three months after the census, the MCC received a letter of complaint on behalf of Old Believers in four census districts claiming that they were counted improperly. Because of the local Orthodox priest’s influence on the process, the letter stated, “husbands and wives were registered as unmarried and their children were registered as illegitimate.” In this manner, the out-riders of Orthodoxy used the census to legitimate their own non-recognition of the validity of schismatic rites.^[76] For these reasons, those outside the official church, especially Old Believers, questioned their numerical representation, especially after publication of the two census volumes on religion in 1901: The Distribution of the Population of the Empire by Main Religions and The Distribution of Old Believers and Members of Sects by Persuasion and Sect. The numbers in these volumes gave rise to a heated debate in the press. According to these census figures, Old Believers, members of various sects and “deviations” comprised 2,173,738 persons out of the empire’s population of 127 million – less than two percent of the population. However, according to other data, these religious non-conformists numbered fourteen to fifteen million, a much more significant eleven-twelve percent of the population.^[77]

One commentator argued that the census had significantly undercounted Old Believers and sect members. He noted that the Tenth Revision in 1858 had defined the number of Schismatics as 805,000, and that several studies from this period indicated that more than seven million additional members were not counted, refused to admit that they were not Orthodox, or practiced the Old Belief secretly. The author thus questioned the census count of just over two million, when the Schismatics had numbered eight million forty years earlier.^[78] We have no evidence that these late 1850s figures were any more accurate than the census. However, we also have no evidence that the government sought to make a more accurate count of these persons in its plans for the second census and must also conclude that, at least on the local level, there was a concerted effort to undercount Old Believers and sect members.^[79] In this way, the census diminished the importance of the dissenters, and enhanced that of the official church (or implied some sort of “reunion” with Orthodoxy). At the same time it undermined state attempts to manage the horizontality created by the census process in a way that would maintain existing power relations.

The fact that officials also intended to extrapolate nationality from data on native language increased the possibility for honest error and, as with the case of religion, manipulation by local enumerators.^[80] In official explanations, a separate question on nationality was not only impossible (because of the large number of nationalities within the empire), but also unnecessary. Answering post-census criticism of the lack of a specific question on nationality, an official of the CSC responded that, “In locales… where the population has a mixed character (and such places in the empire are not few) [the population’s] answers to the given question by far would not always correspond with reality.” Furthermore, he averred, language, as opposed to nationality, was an indicator that was “completely objective and real.” Perhaps more significantly, in a multiethnic empire knowing the population of individual “tribes” was “necessary neither for economic nor state needs.” Information on native language was ultimately more useful. Combined with data on knowledge of Russian, it also provided a method by which officials could define “the degree of russification [obruseniia] of inorodtsy.”^[81]

The MCC acknowledged that in certain instances native language could be problematic as a determinant of nationality. One problem consisted of the fact that respondents were likely to give the language they used most often as their native tongue, and that this might have no bearing on their nationality. For example, Estonians might respond that their native language was German, Lithuanians might claim Polish as their native tongue, and Jews might claim Russian. In the case of the later, a cross reference with religion could settle the issue, but in the case of the Estonians and Lithuanians, “Lutheran” and “Catholic” could not serve to distinguish them from Germans or Poles. A similar problem existed for some steppe peoples who identified their native tongue as Tatar. In these cases, the registrar was to attempt to record the subject’s true native tongue based on other indicators, such as the subject’s surname, the name of the village, or his response to questions on local dialects (for which the MCC compiled a helpful chart).^[82] Thus, in spite of the supposed objectivity of native language as an indicator, the very manner of determining nationality allowed much room for error and abuse.
Some abuse took place. This was particularly true in the Empire’s western provinces, where some non-Russians were beginning to resent increasing demands to conform to Great Russian standards in language and culture. The Governor-General of Kiev, Podolsk and Volhynia provinces attempted to use the census as a means to increase this pressure by refusing to acknowledge the need to translate the census form into German, Czech, Polish, and other local languages, claiming that the population knew Russian “well enough.”^[83] Encumbering the population with the task of answering questions in Russian opened the door for abuse by registrars. A census-day report to the MCC noted that the chair of one Polish census district had instructed the registrars in his jurisdiction to mark Russian as a person’s native language if they answered the question “Do you speak Russian?” in the affirmative.^[84] At the very least, this was a conscious attempt to show the increased use of Russian by non-Russians. In a more serious sense, the district chair was attempting to show a decrease in the area’s Polish population. Similar cases probably went unreported, a fact census officials acknowledged themselves.^[85] As with religion, even when officials in St. Petersburg took steps to ensure an imperial “big tent” policy, local officials with a more zealous russifying agenda (as well as the tense and competitive nature of the western borderlands) often thwarted their efforts.

Conclusion

This brief look at one facet of the first imperial census illustrates a fundamental dilemma of empire in the late nineteenth century. On one hand, the census provided an opportunity for the state to legitimate traditional social structures such as soslovie. Census categories reaffirmed the vertical ordering of the dynastic empire; the forms and questions were designed to count a population in which everyone knew their places – family, social, geographic, economic and political – a society ordered in a vertical fashion with Nicholas II at the top. At the same time, the census process itself created a sense of horizontality – a sense of imperial citizenship. The “universal” nature of the census – its “empire-wide” character – created the impression of a horizontally integrated polity by making it possible to imagine the empire (and its population) as a single unit rather than the sum of its parts. Guidelines on self-identification, the mobilization of society as enumerators, census coverage in the press, the translation of census forms, the rhetoric of “scientific” administration and national rejuvenation contributed to the process of creating a horizontal legitimacy for the dynastic state.

This attempt to “stretch the skin of nation over empire” was, however, an ambiguous (and politically sensitive) process. Like a balloon, “the skin of nation” became weaker as officials expanded it to fit the empire. The same aspects of the census that made it possible to imagine the empire as community – a state-sponsored and categorized civil society – also undermined the legitimacy of the imperial project. Translating census forms into the empire’s languages in the name of legitimating its dynastic nature affirmed (or perhaps created) the aspirations of language groups themselves. Enumerating inhabitants by name undercut the idea of imperial society as a collection of hierarchically organized sosloviia. The universal nature of the census – the enumeration of the bulk of the population with the same instrument – also contradicted a legal structure based on privileged estates. At the same time, the census provided the state with a means to solidify the developing soslovie system and assess the demographic strength of the metropolitan people and its elite.

The process of ordering the empire became more overtly political in the process of enumerating the population by native language and religion. A conflict emerged between self-identification and where the state (either in the guise of the MCC or local census officials) thought people belonged. Indeed, mobilizing society meant in many cases mobilizing persons with agendas more (Russian) national than imperial. The cases of the Uniates and Old Believers show not only the increasing importance of Orthodoxy as an indicator of “Russianness,” but also how other concerns and the agendas of servitors and subjects outside the metropole undercut the ability of the census to create a sense of horizontal integration within the dynastic state. A continued examination of the census experience from the perspective of the counted will ultimately reveal both the extent to which this was the case and the meaning they attached to the census (i.e., the extent to which imperial categories and the imperial project worked or failed).