Identity change in Bashkortostan:

Tatars into Bashkirs and back

Dmitry Gorenburg

Abstract
Students of ethnic identity have recently begun to recognize the role of the
state in causing identity shift. Constructivists, in particular, focus on the
importance of state institutions and policies in creating new identities and
transforming old ones. This article focuses on identity creation and change
in Bashkortostan, an ethnic region within the Russian Federation. It shows
how, at the beginning of the twentieth century, the Russian/Soviet state
created new ethnic identities from pre-existing regional, estate-based and
religious identities. It also shows how later changes in state institutions and
policies played a crucial role in determining the direction of identity change
among a mixed population, straddling the geographical and cultural bound-
aries between the Tatar and Bashkir ethnic groups. By tracing the impact of
the state on one ethnic group over an extended time-period, this article
shows that state actions can lead to both instrumental and culturally-ba sed
shifts in ethnic identity.

Keywords: Identity; Bashkortostan; Russian Federation; Bashkirs; Tatars;

ethnicity.

In recent years scholarship on the formation and maintenance of ethnic

identity has largely focused on the debate over whether students of eth-
nicity should focus their attention on the effects of cultural boundary
maintenance between neighbouring groups (Barth 1969; 1994) or on cul-
tural development among an ethnic group’s core members (Roosens
1994). In some ways this is an outgrowth of the earlier debate between sit-
uationalists, who assert that ethnicity is exible and is often used for econ-
omic and political gain, and primordialists, who argue that ethnicity is a
permanent and essential aspect of one’s identity.1 Barth and his followers
argue that ethnic identity is ‘a feature of social organization, rather than
a nebulous expression of culture. . . . This means focusing on the bound-
ary and the processes of recruitment, not on the cultural stuff the bound-
ary encloses’ (Barth 1994, p. 12). Roosens (1994, p. 84), by contrast, argues
that ethnic identity cannot exist without some cultural or genealogical
substance, which provides an internal source of identiŽcation for the

Ethnic and Racial Studies Volume 22 Number 3 May 1999

© Routledge 1999 0141-987 0
 Identity change in Bashkortostan 555
groups’ members that is at least as important as the external source of
identiŽcation provided by group boundaries.
Scholars who study ethnic identity change have attempted to surmount
this debate by synthesizing the two approaches. Charles Keyes (1981), for
example, begins with the cultural core of ethnic identity, but argues that
a complete analysis of ethnic change must incorporate factors such as
social structure and inter-group interaction. Judith Nagata (1981) ex-
plicitly argues for combining primordialist and situationalist approaches.
Studies of ethnic change based on this synthetic perspective argue that the
way members of an ethnic group conceive of their identity may change as
the result of interactions between neighbouring ethnic groups, but that
newly modiŽed identities will remain anchored by a cultural core which
is based on the members’ self-perception. Keyes argues that primordial
identity serves as a gyroscope for individuals facing radical changes in
their political and economic environment. Individuals may change their
identities as a consequence of migration, changes in state boundaries, the
establishment of new government programmes, and revolutionary change
in the structure of society (1981, pp. 27–28). In an introduction to a col-
lection of articles on identity and change in Central Asia, Jo-Ann Gross
(1992) provides a similar list of factors that may lead to shifts in ethnic
identity, including changes in economic and political conditions in a
region, cultural differences between neighbouring groups, changes in the
interpretation of a group’s shared experiences, the emergence of national-
ist ideology, and changes in power relations between groups.
The sources of ethnic change described by these scholars focus on the
actions of group members, who change their identity to correspond with
various kinds of changes in their environment. Such a view plays down
the ability of outside forces to deliberately shape ethnic identity. In this
article I argue that such outsiders, particularly states that are not con-
trolled by any of the local ethnic groups, can create changes in ethnic
identity among the populations living in areas under their control. State
institutions such as the census and administrative boundaries, as well as
state policies such as preferential treatment for members of certain
groups, often lead to the formation of new ethnic groups and to changes
in the cultural content of existing ethnic identities. This point of view,
often labelled constructivism, has been articulated by Leroy Vail (1989)
and his collaborators in ascribing the creation of tribalism in Southern
Africa largely to European colonial rule. They argue that the Europeans
created ethnic divisions between cultural groups in order to make ruling
their colonies easier. I show how similar processes were at work in the
development of contemporary ethnic identities among the Tatars and
Bashkirs, two Turkic Muslim minorities inhabiting the Volga-Ural region
of Russia.
This change of focus is particularly important for students of ethnicity
in the territory that formerly comprised the Soviet Union. Until the 1990s
556 Dmitry Gorenburg
Soviet scholars were the strongest remaining proponents of a pri-
mordialist view of ethnicity. Ethnos theory, as championed by Yulian
Bromley and his colleagues at the Institute of Ethnography of the Soviet
Academy of Sciences, argued that all ethnic groups have a stable core,
called ethnos, that persists across time and space, despite all changes in
a group’s political and economic circumstances (Bromley 1981). Need-
less to say, such a point of view does not allow for the possibility that a
particular ethnic identity may be a construct created and spread by the
state. According to the ethnos theorists, Soviet nationalities policy simply
catalogued existing ethnic groups, allowing them to receive, usually for
the Žrst time, political and cultural institutions such as literary languages ,
native language school systems, and self-governing territorial units.
Although a few Russian anthropologists have recently begun to move
towards the constructivist argument (see Tishkov 1997), the primordial-
ist view of ethnicity continues to predominate among scholars in the
region. However, the evidence strongly disputes their claims that ethnic
identities in the former Soviet region have remained relatively
unchanged for many centuries. Tishkov (1997, pp. 15 –21), for example,
describes how deliberate ethnic engineering for political purposes on the
part of Soviet authorities created new ethnic identities by combining or
dividing previously existing ethnic groups. In this article, I go one step
further by using census data to show how particular government policies
led to changes in ethnic identiŽ cation at the district and village levels.
Before moving on to the case-study, it is necessary to distinguish
between two types of ethnic identiŽcation. Although the state is the most
important actor in the creation of new ethnic labels, it is left to the poten-
tial members of the newly established or transformed ethnic group to
accept or reject their new ‘identity’. In cases where rejection of the state-
sponsored labels may lead to negative consequences or hinder the acqui-
sition of beneŽts, affected individuals sometimes adopt the new ethnonym
as their public ethnic identity, while continuing to identify along tra-
ditional lines in the private sphere. Roosens (1994, p. 88) notes how
migrants to Belgium from Catalunya assume ‘Spanish’ as their public iden-
tity when interacting with Belgians or the state, while retaining ‘Catalan’
as their identity when interacting among themselves and with migrants
from other regions of Spain. Following Roosens, I distinguish between
public ethnic identity, which refers to the identity used by members of a
group in interactions with outsiders or with the government, and private
ethnic identity, which refers to the identity on which interactions within
the group and with neighbouring groups are based. As shown in this essay,
private and public ethnic identities are not always congruent. Changes of
ethnic identity may also be either public or private. Public identity change
usually occurs when members of a group accept newly created or changed
ofŽcial designations for use when interacting with outsiders without adopt-
ing them for use within the group. Private identity change occurs when
 Identity change in Bashkortostan 557
members of a group change their self-designation and adjust their social
interactions with others to correspond with the new identity.
The case described in this article shows how state policies and insti-
tutions create the conditions for changes in ethnic identity. The state is
shown to play a key role in both boundary shifts and changes in the cul-
tural cores of ethnic identity, creating conditions that lead to changes in
both public and private ethnic identity.

Ethnic re-identiŽcation in Bashkortostan: 1897–1989

Bashkortostan is a republic in the Volga-Ural region of the Russian
Federation. It is one of the largest and wealthiest ethnic republics of
Russia, with a territory of 143,000 sq km and a population of over four
million. It is unusual among Russia’s ethnic republics in that the titular
ethnic group, the Bashkirs, is only the third largest group, less numerous
than both Russians and Tatars. Furthermore, the relative percentages of
Bashkirs and Tatars have varied signiŽ cantly over the course of the last
century. Figure 1 shows two periods of signiŽcant change in the propor-
tions of Tatar and Bashkir ethnic groups in the total population. During
the Žrst period, which lasted from 1897 until 1939, the percentage of the
population which identiŽed itself as Bashkir fell by one-half, while the
absolute number of Bashkirs declined by a quarter. The Tatar population
more than tripled in number, increasing its proportion of the population
from 11 to 25 per cent. 2 A period of relative stability followed, but then,
during the 1979 –1989 intercensal period, the Bashkir population fell by
70,000 while the Tatar population rose by 180,000. Yet local ethnogra-
phers have shown that none of the usual sources of such uctuation, such

Figure 1. Ethnic distribution in Bashkortostan

558 Dmitry Gorenburg
as migration or differences in rates of population growth, was present.
Instead, many Bashkirs re-identiŽed as Tatars during both periods. Why
was ethnic re-identiŽcation so frequent among Bashkirs in the Ž rst half
of the twentieth century and again in the 1980s? 3
This topic has political implications that go beyond the scholarly
debates outlined in the introduction to this article. Bashkir cultural and
political leaders have long blamed the relative decline of the Bashkir
population as compared to the Tatar population on assimilation through
Tatar cultural imperialism. Political tensions over this pattern of re-
identiŽcation existed throughout the twentieth century and, as I show
below, in uenced the cultural and census policies of the Bashkortostan
government from the 1920s to the 1980s. However, the political liberal-
ization of the Gorbachev and post-Soviet periods brought these tensions
into public discourse for the Žrst time, leading to conicts between
Bashkir and Tatar cultural élites. Debate on the causes of identity shifts
in northwest Bashkortostan has in recent years become closely tied to
the debate on the status of Tatars and the Tatar language within the
republic and therefore highly politicized. This environment has made
scholarly study of the topic by local scholars very difŽcult. In analysing
the factors leading to changes in ethnic identities in this region, I hope to
provide an outsider’s point of view that is not biased by the political
dimensions of this question.
In order to explain the factors behind Bashkir/Tatar identity shift, it is
important to consider the speciŽc area where most of this re-identiŽ-
cation took place. District-level ethnic breakdowns show that ethnic
identity was relatively stable in most of Bashkortostan throughout the
twentieth century (see Appendix 1). Almost all of the shifts in ethnic
identity were limited to northwestern Bashkortostan, indicating that
identity changes were even more pronounced in this area than aggregat e
census data for the entire republic would suggest. Since this area borders
Tatarstan, it is not surprising that it is home to most of the republic’s
Tatars (see Figure 2). It is also populated by many Bashkirs who consider
Tatar their native language. This group was Žrst noted in the 1926
Census, the Ž rst census to separate ethnic identiŽcation and native lan-
guage into distinct categories. The number and percentage of Tatar-
speaking Bashkirs among the total Bashkir population has been steadily
declining since that time (see Table 1). This sets up the second puzzle this
article will need to explain: considering the strong connection between
language and ethnic self-identiŽcation in communist ideology, how did a
Tatar-speaking Bashkir group form? How did members of the group
resolve the contradiction between these two ethnic identity markers,
retaining their identity as Tatar-speaking Bashkirs over several decades?
And what happened to them when they did change their identity? Did
they re-identify as Tatars or did they switch their linguistic identiŽcation
to correspond with their ethnic identiŽcation? Finally, how does this
 Identity change in Bashkortostan 559
Figure 2. Map of Bashkortostan
Note: Highlighted areas represent northwestern counties.

Verkhniye Tatyshly
Neftekamsk Askino

Kaltasy
Karaidel
Burayevo

Dyurtyuli Birsk
Maloyaz
Bakaly
Kushnarenkovo

Tuymazy Ufa

Buzdyak Uchaly

Belebey
Beloretsk

Kirgiz-Miyaki
Bizhbulyak
Ishimbay

Kumertau
Baymak

process interact with the general demographic trend towards Tatariza-

tion described above?

First period: 1897–19394

During the nineteenth century, the territory that is now northwestern
Bashkortostan was the northwestern region of the Ufa province
[guberniia] of the Russian empire. This region of the Ufa province
560 Dmitry Gorenburg
Table 1. Native language among Bashkirs in Bashkortostan

Language 1926 1939 1959 1979 1989

% % % % %

Bashkir 53.8 54.3 57.6 64.4 74.7

Tatar 46 45.4 41.4 32.9 20.7

Sources: Murzabulatov 1995, p. 20 (1939). Calculated from Yuldashbaev 1995, Tables 6&7
(1926). Yuldashbaev 1995, Table 3 (1959–1989). 1939 Tatar percentage approximate, based
on subtracting total number of Tatars from total number of Tatar-speakers, then dividing
by total number of Bashkirs. The actual percentage may be slightly lower.

consisted of the Belebey, Birsk and Menzelinsk districts [uezdy]. Ufa

province also included several districts to the south and east where ethnic
identity remained relatively stable. After the revolution, the administra-
tive division of this territory was radically altered by the creation of the
autonomous ethnic republics of Tatarstan and Bashkortostan. Ufa
province was divided between the two republics in 1922. Most of the
province, including Belebey and Birsk districts, became a part of Bash-
kortostan, while Menzelinsk district was annexed by Tatarstan.
Until the middle of the nineteenth century, northwestern Ufa province
attracted large numbers of Tatar-speaking migrants. During this period,
the Tatar ethnonym had not yet been universally adopted by Tatar-
speaking Turks in the area. All Muslims living in the Volga-Ural region
still shared a common identity based on their religion and a myth of
descent from the Bulgar state of the tenth to thirteenth century (Frank
1998). Only gradually did the people who identiŽed with this common
identity come to refer to themselves as Tatars. Furthermore, while
sharing this common regional identity, the migrants were divided into
two groups based on their land of origin. The two groups were called
Kazanly (people from Kazan) and Mishars.5 There were two other
groups already established in northwestern Ufa province: the Bashkirs
and the Teptiars. These were considered both ethnic groups and estates
[soslovie], stemming from a sixteenth-c entury royal decree giving
Bashkirs special land-owning privileges on their territory.6 Teptiars were
also declared a land-owning estate, although they were entitled to a
smaller portion of land and therefore had less prestige than the Bashkir
estate. While the Bashkir estate as a whole was dominated by members
of the Bashkir ethnic group, the Teptiar estate included members of most
ethnic groups in the Volga region, including non-Turkic groups such as
the Mari and Udmurts (Yakupov 1992, p. 168). Ethnic minorities who
were not members of the Bashkir or Teptiar estates could not own land
and were required to rent it from members of one of the estates. During
the late nineteenth century, the land-ownership system was reformed,
ofŽcially eliminating the estate system and placing members of all ethnic
 Identity change in Bashkortostan 561
groups on an equal footing. On the ground, however, the last vestiges of
this system did not disappear until the 1917 revolution .7
Ethnic identity in northwestern Ufa province was highly unstable
throughout this period. Korostelev (1994) has compiled a database which
includes information on the dominant ethnicity in all villages in north-
western Ufa province/Bashkortostan at various times from 1870 to 1989.
This database shows that the dominant ethnicity in 47 per cent of all vil-
lages changed between 1870 and 1913. These changes included not only
switches from Mishar, Teptiar and Tatar to Bashkir, but also from
Bashkir to the other groups (see Table 2).
After 1917 ethnic identity changed again. The census administration
gradually eliminated the Teptiar and Mishar categories, asking members
of these categories to choose among ethnic labels recognized by the
Soviet government. For Muslim Turkic-speakers, this allowed a choice
between Tatar and Bashkir. Korostelev shows that approximately two-
thirds of Mishars and Teptiars identiŽed themselves as Tatars, while one-
third identiŽed as Bashkirs. In addition, during the 1920s many Bashkirs
re-identiŽ ed as Tatars (1994, pp. 77 and 91).
The distinction between ethnic identiŽcation and native language, Žrst
made in the 1926 Census, allows us to see two distinct patterns in Bashkir
identity change. In those parts of Ufa province which were joined to
Tatarstan after the revolution, the number of Bashkirs declined from
123,000 in 1897 to none in 1926 (Table 3). Bashkirs who found themselves

Table 2. Changes in predominant village identity in NW Bashkortostan,

1870–1913 (number of villages)

Identity in 1913
———————————————————————————
Identity in 1870 Bashkir Mishar Teptiar Tatar

Bashkir 264 21 71 14
Mishar 109 30 10 1
Teptiar 47 2 31 9
Tatar 18 3 2 20

Source: Korostelev 1994, p. 90.

Table 3. Bashkirs in neighbouring regions of Ufa province (in thousands)

1897 1897% 1913 1913% 1926 1926%

Menzelinsk 123 32% 154 34% 0 0%

Belebey 233 54% 213 37% 118 21%
Birsk 262 53% 233 39% 178 33%

Sources: Yuldashbaev 1995, Table 8 (1926), Table 49 (1897), Table 56 (1913). Vsesoiuznaia
Perepis Naseleniia 1926 goda, vol. 3, pt. 1, Table X, page 281.
562 Dmitry Gorenburg
living in Tatarstan re-identiŽed as Tatars. In areas which became part of
Bashkortostan, some Bashkirs re-identiŽed as Tatar, but many retained
their Bashkir identiŽcation in the 1926 Census. Bashkirs who re-identi-
Žed as Tatar primarily consisted of individuals who had adopted the
Bashkir self-designation recently, previously calling themselves Mishar
or Teptiar. Bashkirs who maintained their identity were concentrated in
those villages which were considered Bashkir prior to the 1865 land
reform (Rodnov 1995, p. 76). However, in the northwest, almost all of
these Bashkirs declared Tatar to be their native language. Whereas 89
per cent of the Turkic population in this area declared Bashkir to be their
native language in 1897, 94 per cent declared Tatar to be their native lan-
guage in 1926 (Korostelev 1994, p. 80). Among Bashkirs, Tatar-speakers
made up 46 per cent of the total population in 1926, including 89 per cent
of northwestern Bashkirs. 8
The period between 1926 and 1939 witnessed a continuation of the
previous pattern with a gradual tendency towards stabilization. The
Tatar percentage of the population continued to increase at the expense
of the Bashkirs, but at a slower rate than before (see Table 4). Most
Bashkirs in the northwest maintained their ethnic identity while at the
same time retaining Tatar as their native language .

Causes of ethnic re-identiŽcation: 1897–1939

As shown above, ethnic identity was highly unstable during this period.
This instability was caused primarily by changes in state policies towards
non-Russian minorities and by the establishment of new ethnic insti-
tutions in the aftermath of the 1917 revolution. The tsarist system of popu-
lation classiŽcation based on estates encouraged most local non-Russians
to identify as Bashkirs in the 1897 Census. The gradual elimination of
estate-based privileges, combined with the establishment of ethnic admin-
istrative units by the new communist government, encouraged many local
inhabitants to switch to Tatar identiŽcation during the 1910s and 1920s.
As a result of government decisions in creating a Bashkir literary lan-
guage, a sizeable number of those inhabitants who retained their Bashkir
identity declared Tatar to be their native language .
Estate-based privileges. At the end of the nineteenth century, ethnic
identity for local inhabitants was closely connected to their estate, which
determined their place in the local socio-economic hierarchy. In this hier-
archy, Bashkirs occupied the top rung, followed in order by Teptiars and
Mishars, with everyone else at the lowest level. Bashkirs were the only
group allowed to own land until 1865, when land reform gave some land-
ownership rights to Teptiars and Mishars. The 1865 reform also made it
signiŽ cantly easier for individuals to change their estate status. After
1865, Teptiars and Mishars who changed status became New Bashkirs
[NovoBashkiry ], while those who continued to rent land from Bashkirs
 Identity change in Bashkortostan 563
retained their old label. Increasingly in the 1890s, members of the lowest
level of the estate hierarchy, consisting mostly of former crown serfs,
became known as Tatars (Rodnov 1995, pp. 64 –6 and 77).
It seems likely, therefore, that the number of Bashkirs according to the
1897 Census has little to do with the number of people who considered
themselves culturally Bashkir or who spoke the Bashkir language ,9 which
were found largely in areas to the south and east of Belebey and Birsk.
Many of the people who identiŽed as Bashkirs in the 1897 Census con-
sidered themselves members of the Bashkir estate while sharing the
culture and dialect common to all groups in the area. The predominance
of Bashkirs was due to the economic desirability and prestige of the
Bashkir label and the 1865 institutional reforms which made re-identiŽ-
cation possible for many Mishars and Teptiars. This explains the rapid
growth in the number of Bashkirs in the west as compared to the east,
where a distinct Bashkir culture and language did exist and which had
never experienced an inux of migrants10 (see Table 5).
The 1917 revolution changed the economic and political environment
for the inhabitants of northwest Bashkortostan. All estate-based privi-
leges, which had been in decline since the 1890s, were revoked by the
communist government, giving all inhabitants the same rights to own
land. (Collectivization would not take place until the 1930s.) Further-
more, the political administration of the region underwent a decisive
transformation with the creation of autonomous ethnic republics for the
Tatar and Bashkir ethnic groups. In 1922 the Ufa province was divided

Table 4. Ethnic distribution in Bashkortostan

1897 1920 1926 1939 1959 1970 1979 1989

% % % % % % % %

Bashkir 40.9 30.2 23.5 21.2 22.1 23.4 24.3 21.9

Tatar 11.2 21.0 23.3 24.6 23.0 24.7 24.5 28.4
Slavic 38.2 38.5 43.6 44.2 45.5 43.0 42.7 41.6

Sources: Yuldashbaev 1995, Table 5 (1920), Table 8 (1926), Table 9 (for 1939), Table 10
(1959 –89). Also Khismatullin 1992, p. 1 (1897). Tatar population includes Teptiar and
Mishar for 1897 –1926. These categories were eliminated from the census after 1926. Slavic
population includes Russians, Ukrainians and Belorussians.

Table 5. Total Bashkir population: 18th–19th centuries

1726 1849 1897

Eastern 168,000 280,000 478,000

Western 89,000 216,000 832,000
Source: Ganeeva 1992, pp. 60 –1.
564 Dmitry Gorenburg
between Tatarstan and Bashkortostan, with Menzelinsk canton joining
the former while the rest of the province became part of the latter. These
changes together had a decisive impact on ethnic identiŽcation.
Census data can help to separate the political and the economic factors
causing ethnic re-identiŽcation. The 1913 rural Census shows the effect
that the decline of estate privileges was having on ethnic identiŽcation.
The proportion of Bashkirs in Ufa province fell from 41 per cent in 1897
to 32 per cent in 1913. The 1920 Census, conducted after the revocation
of estate privileges but before the consolidation of the ethnic republics,
shows a slight further decline, to 30 per cent11 (Yuldashbaev 1995, Tables
5 and 56). Altogether , the end of estate privileges led to an 11 per cent
decline in the relative proportion of Bashkirs in Ufa province. This
decline was primarily caused by the return of New Bashkirs to their
previous Mishar and Teptiar identities. Mishars increased from one per
cent to 6 per cent of the total population between 1897 and 1920, while
Teptiars increased from 1.8 per cent to 5.3 per cent. Tatar identiŽcation
also rose, but to a much lesser extent, from 8.4 per cent to 9.7 per cent.
Altogether, the three groups’ proportion of the total population
increased by 10 per cent. The remaining one per cent decline can be
attributed to Russian migration into the region (Khismatullin 1992;
Yuldashbaev 1995, Table 5).
Institutional effects. After 1920 political factors played the most impor-
tant role in causing ethnic re-identiŽcation. The creation of Tatar and
Bashkir ethnic republics reinforced these ethnic identities. At the same
time, Soviet nationalities policy declared Mishar and Teptiar as un-
acceptable ethnic labels, asking all Teptiars to choose another desig-
nation beginning with the 1926 Census and identifying all Mishars as
Tatars beginning with the 1939 Census. 12 Starting with the 1920s, Soviet
nationalities policy called for privileges for members of the titular ethnic
group in ethnic republics. These privileges included control of republic
administration, preferences in hiring, promotion and acceptance by uni-
versities, and priority in cultural development funding. These privileges
made it advantageous for individuals to identify as members of the titular
ethnic group. For members of most ethnic groups, changing one’s iden-
tity to the titular ethnicity was impossible because linguistic and cultural
differences between groups made it relatively easy to detect ‘inauthen-
tic’ titulars. However, for groups as closely related as Tatars and
Bashkirs, detection was quite difŽ cult, making such a choice not only
possible, but even quite easy to accomplish.
The division of Ufa province between Tatarstan and Bashkortostan in
1922 allows us to study the effect of belonging to an ethnic republic on
identity choice. In this division, Menzelinsk canton became a part of
Tatarstan, while the rest of Ufa province, including neighbourin g Birsk
and Belebey cantons, joined Bashkortostan. Table 6 shows the sharp fall
in the Bashkir population in the region as a whole between 1897 and
 Identity change in Bashkortostan 565
1926. 13 Part of this fall can be attributed to the decline of estates dis-
cussed above. But while a sizeable Bashkir population remained in the
two regions that joined Bashkortostan, by 1926 not a single Bashkir
remained in Menzelinsk.14 Menzelinsk Bashkirs shared the language and
culture of Bashkirs in Belebey and Birsk. The difference in their rate of
assimilation can most plausibly be attributed to the difference in their
political circumstances after 1922. Finding themselves in Tatarstan,
Menzelinsk Bashkirs recognized the advantages of belonging to the
titular nationality and quickly re-identiŽed as Tatar. They were able to
do so because of the extensive cultural similarities between them and
Kazan Tatars. In addition, they were undoubtedl y encouraged to re-
identify by the Tatarstan government’s cultural and linguistic policies,
which fostered Tatar education and culture at the expense of Bashkir lan-
guage and culture. Similarly, Bashkirs in Belebey and Birsk were
restrained from re-identifyin g by their recognition that the titular ethnic
group would be privileged in Bashkortostan.
If being a member of the titular ethnic group provided cultural and
political beneŽts, why did a sizeable number of Bashkirs in Belebey and
Birsk still re-identify as Tatars between the 1913 and 1926 Censuses?
While no certain answer to this question can be given, Rodnov (1995,
p. 73) proposes an interesting hypothesis. He argues that a signiŽcant
part of the instability in ethnic self-identiŽcation in this region can be
explained by changes in the meaning of the ethnonyms, rather than
changes in individual self-identiŽ cation. People changed their identiŽ-
cation depending on changes in the popular perception of what an eth-
nonym stood for, while keeping their private identity constant. The
creation of ethnic republics solidiŽed the popular understanding of the
Tatar and Bashkir ethnonyms, leading many individuals to re-identify as
Tatar. Since local customs and dialects were more closely related to those
of neighbourin g Tatarstan than to the core Bashkir area in the south-
eastern part of Bashkortostan, many local inhabitants changed their
identity to correspond with their perception of the meaning of these
ethnic labels. The change in label occurred because individuals recog-
nized that their culture and language Žtted better with their perception
of Tatar identity than Bashkir identity. While this hypothesis cannot be
proved conclusively, it does make sense given the confusion surrounding
ethnonyms in the area.
Language policy. The language policies of the Bashkortostan govern-
ment may also help to explain why Bashkirs re-identiŽed as Tatars. These
policies are also instrumental in explaining the development of a group
of Bashkirs who considered Tatar to be their native language.
As mentioned above, before 1926, censuses determined ethnic identiŽ-
cation according to a question about native language. The 1926 Census
thus provides the earliest opportunity to study differences between native
language and ethnic identity. Language and ethnic identity coincided for
566 Dmitry Gorenburg
over 97 per cent of respondents from most ethnic groups and regions
covered by this census. Bashkirs in Belebey and Birsk cantons were the
major exception to this rule. Only 11 per cent of the Bashkirs in these
areas declared Bashkir to be their native language, whereas 92 per cent
of Bashkirs in the rest of the republic did so. Altogether, there were
264,000 Tatar-speaking Bashkirs in Belebey and Birsk, with another
24,000 in the rest of the republic.
This disparity was created by the language policy of the Bashkir
government. In the early twentieth century, the Bashkir language was
divided into three major dialects: the southern, eastern, and northwest-
ern. Of these, the northwestern dialect was most similar to Tatar. After
the formation of the Bashkir republic in 1919, local ofŽcials and scholars
began to develop a Bashkir literary language to replace the Turko-Tatar
literary language that was previously used by all Muslim Turks in the
Russian empire. Their purpose was to assist the consolidation of Bashkir
identity by establishing a Bashkir high culture that was distinct from
Tatar or Russian high culture (Bikbulatov 1992, p. 45). These scholars
were divided into two groups. The Žrst sought to recognize the common
features of the Bashkir and Tatar language by basing the Bashkir liter-
ary language on a combination of all three major Bashkir dialects. The
second wanted to emphasize the uniqueness of Bashkir by distinguish ing
it as much as possible from literary Tatar. The latter group won, creating
a literary language which incorporated those traits shared by the eastern
and southern dialects that distinguished them from the northwestern
dialect (Yuldashev 1968, p. 70; Ishberdin 1989, p. 140; Kuzeev 1994,
p. 121). The new literary language was introduced in 1923, quickly
becoming the language of ofŽcial business and cultural activity through-
out the republic (Bikbulatov 1992, p. 45).
This context sheds considerable light on the twin puzzles of Bashkir
ethnic re-identiŽcation and language choice. Faced with a literary lan-
guage which differed signiŽcantly from the language that they spoke,
Bashkirs in the northwest selected the literary language that most closely
resembled their dialect, Tatar. Many of them chose Tatar ethnic identity
as well, probably out of a sense that language and ethnicity should corre-
spond. Others, however, chose to adopt the Tatar language but not the
Tatar identity. These people may have felt a deeper sense of Bashkir
ethnic identity, which allowed them to dismiss the contradiction between
ethnic identiŽcation and native language. They may also have more clearly
recognized the beneŽts of being a member of the titular ethnic group.
Combined, these factors allowed a signiŽcant percentage of the local
inhabitants to identify themselves as Tatar-speaking Bashkirs. This expla-
nation is consistent with the observed differences between Menzelinsk
and the northwestern cantons of Bashkortostan. For Bashkirs in Men-
zelinsk, political, economic, and linguistic incentives all encouraged
assimilation into the Tatar group. Bashkirs in Belebey and Birsk, on the
 Identity change in Bashkortostan 567
other hand, were torn. On the one hand, their culture and language
pointed towards Tatar identity. On the other hand, there were political and
economic incentives to identifying as Bashkir. In this situation, almost half
of the Turkic population of the region resolved this dilemma by choosing
Bashkir as their ethnic identity and Tatar as their native language .

* * *

We can now summarize by answering the questions presented at the

beginning of this section. The predominance of Bashkir identiŽcation in
1897 resulted from the economic beneŽts of belonging to the Bashkir
estate and from the land reforms of 1865, which made changing estate
easier. The revocation of land-ownership privileges for Bashkirs led to a
gradual decline in Bashkir identiŽcation, as many non-Bashk irs returned
to their previous designation . Government policies contributed to this
trend in two other ways. In those parts of the area that became part of
Tatarstan, beneŽts for members of the titular ethnic group made re-
identiŽ cation as Tatar highly advantageou s. In addition, all speakers of
the local dialect were affected by the creation of a Bashkir literary lan-
guage on the basis of other dialects. This language policy, combined with
the beneŽ ts for belonging to the titular ethnic group, led many people in
northwestern Bashkortostan to choose the best of both worlds by declar-
ing Bashkir as their ethnic identity and Tatar as their language .

Second period: 1959–1989

During this period, censuses recorded two periods during which the
number of Tatars declined while the number of Bashkirs grew rapidly.
This was followed by one period of rapid Bashkir decline and Tatar
growth which more than offset the earlier Bashkir gains. (See Tables 4
and 6.) District-level data show that the oscillations in the republic-level
data are entirely the result of events in the northwest. Areas outside of
the northwest experienced a gradual decline in the Russian percentage

Table 6. Population of Bashkortostan (in thousands)

1897 1920 1926 1939 1959 1970 1979 1989

Bashkir 900 833 626 671 738 892 936 864

Tatar 246 580 621 777 769 945 940 1121
Russian 840 950 1065 1281 1418 1546 1548 1548

Sources: Yuldashbaev 1995, Table 5 (1920), Table 8 (1926), Table 9 (for 1939), Table 10
(1959 –89). Also Khismatullin 1992, p. 1 (1897). Tatar population includes Teptiar and
Mishar for 1897 –1926. These categories were eliminated from the census after 1926. 1897
Russian population includes Ukrainians and Belorussians.
568 Dmitry Gorenburg
of the population in favour of a gradual increase in either the Bashkir
proportion or in both the Bashkir and Tatar percentages.
The following analysis focuses exclusively on rural areas because wide-
spread migration to the cities makes it difŽcult to distinguish demo-
graphic changes due to re-identiŽcation from changes due to migration.
Throughout this period, there was virtually no migration between rural
areas. Although Tatars and Bashkirs left rural areas at similar rates,
different cities served as each group’s favoured destination .
Table 7 shows that the majority of northwestern districts experience d
increases in the Bashkir proportion of the population at the expense of
the Tatars between 1939 and 1959, and again between 1970 and 1979.
Between 1959 and 1970, there was a slight shift back towards Tatar gains.
Finally, there was an overwhelming shift towards Tatar identiŽcation
between 1979 and 1989.15 Korostelev’s database on village ethnic identity
in the northwest shows that 40 per cent of the villages that registered as
Tatar in 1959 had become Bashkir by 1979. Of these, 77 per cent switched
back to Tatar identity in the 1989 Census. Another 19 per cent switched
to mixed Tatar/Bashkir identity and only 3 per cent remained predomi-
nantly Bashkir. In tracing village ethnic identity from 1870 to 1989,
Korostelev (1994, pp. 78 and 91) shows that by 1989, the Bashkir ethnic
group was reduced to its core – 95 per cent of villages identiŽed as Bashkir
in 1989 had identiŽed as Bashkir in all of the previous censuses. 16
Of the villages identiŽed in 1989 as Tatar, 48 per cent had previously
always identiŽ ed as Tatar, Mishar, or Teptiar. The rest of the Tatar vil-
lages had at some point identiŽed themselves as Bashkir. These villages
were divided into two groups. The Žrst, comprising 15 per cent of the
total number of Tatar villages, had always identiŽed as Bashkir prior to
the 1989 Census. This group mostly consisted of villages inhabited by
people who had identiŽed as Tatar-speaking Bashkirs in 1926 and by
their descendants. The second group, comprising the remaining 37 per
cent of the Tatar villages, had identiŽ ed as Tatar at some previous time,
then switched to Bashkir during the Soviet period. This group primarily

Table 7. County level population changes between censuses in northwest

Bashkortostan (number of counties)

1939 –1959 1959 –1970 1970 –1979 1979 –1989

Bashkir+, Tatar – 11 6 16 2
Bashkir+, Tatar+ 10 10 6 3
Bashkir–, Tatar+ 4 10 4 20
Bashkir–, Tatar – 1 0 0 1

Sources: Calculated from Yuldashbaev 1995, Tables 18 –20 and 1979 Census data provided
by Ildar GabdraŽ kov.
‘–’ represents decline in population. ‘+’ represents increase in population.
 Identity change in Bashkortostan 569
included Tatars switching to Bashkir identity for instrumental reasons in
the 1950s or later (Korostelev 1994, p. 91).
An example can demonstrate the extent of the demograph ic shifts that
occurred between 1939 and 1989. In 1939 Baltachevsky district was 53
per cent Tatar and 24 per cent Bashkir. By 1959, Tatars comprised only
17 per cent of the population, whereas Bashkirs had increased to 62 per
cent, essentially reversing the balance from twenty years earlier. Ensuing
censuses showed a continuing decline, so that by 1979 Tatars comprised
only 6 per cent of the district population, while Bashkirs made up 74 per
cent. Yet by 1989, the 1939 balance was essentially restored, with the
Bashkir proportion dropping to 22 per cent as the Tatars increased to 59
per cent (Figure 3). Other districts saw similar variation in ethnic identiŽ-
cation over time.
While the proportion of Tatars and Bashkirs in the total population
oscillated in the post-war period, the percentage of Bashkirs who con-
sidered Tatar their native language uniformly declined. The decline in the
Žrst thirty years was gradual, dropping from 46 per cent of the total
Bashkir population in 1926, to 41 per cent in 1959. Between 1959 and
1989, the decline was far more rapid, with only 21 per cent of the Bashkir
population declaring itself Tatar speaking in the most recent census (see
Table 1). District-level data on language choice are only available for 1959
and 1989. It shows that in eighteen of the twenty-six districts where the
percentage of Tatar-speaking Bashkirs declined between these years, the
percentage of Tatars in the total population increased. In the remaining
districts, Tatar-speaking Bashkirs either declared Bashkir to be their

Figure 3. Baltachevsky district ethnic distribution

570 Dmitry Gorenburg
native language or divided evenly between Tatar and Bashkir/Bashkir
identiŽcation. In an additional Žve districts which had sizeable numbers
of Tatar-speaking Bashkirs, their percentage of the Bashkir population
did not decline. While the data cannot prove conclusively the direction of
Tatar-speaking Bashkir identity change, they imply that about three-
quarters of the Tatar-speaking Bashkirs who changed their identity
switched their ethnic identiŽcation to correspond with their Tatar lan-
guage while the other one-quarter switched their native language to cor-
respond with their Bashkir ethnic identity.

Causes of ethnic and linguistic re-identiŽcation: 1959–1989

Unlike the previous period, the meaning of ethnic labels did not change
during this period. People who declared different ethnic identities in
different censuses were not changing their understanding of the labels’
meanings – they were changing their ethnic identiŽcation. What caused
large numbers of people to change their declared ethnic identities repeat-
edly over a few decades? And why did the number of Tatar-speakin g
Bashkirs decline so precipitously by 1989? The answers to these questions
can be found by examining government policies on the census and on
native language education, as well as in the effects of Soviet institutions.
Census policy and government pressure. The swings towards Bashkir
identiŽcation in the 1959 and 1979 Censuses can be explained to a large
extent by government tactics in explaining census questions, expressed as
administrative pressure for respondents to identify as Bashkir. Prior to
the 1959 Census, the Bashkir government conducted a propaganda cam-
paign arguing that the census question about ethnic identity was con-
cerned with ethnic self-identiŽcation, which was independent of native
language. If the two did not coincide, respondents were instructed to
declare their ethnic identity based on the self-identiŽcation (Kuzeev 1994,
p. 122). This policy went against the Stalinist conception of ethnicity,
which argues that an ethnic group can have but a single native language17
(Stalin 1994, p. 19). This campaign, repeated in later censuses, was
designed to prevent Tatar-speaking Bashkirs from re-identifying en masse
as Tatars. Furthermore, the central government required local adminis-
trators to produce a ‘quota’ of Bashkirs. Results that did not achieve this
quota were returned to the local administration for revision, leading one
administrator to note that he was trying his best but there were no more
Bashkirs to be found in his district (Khalim 1991, p. 166). Such a quota
was not required in 1970, leading to a partial swing back in favour of Tatar
identity according to that year’s census results. Finally, on the basis of his
analysis of village ethnic identiŽcation, Korostelev (1994, p. 78) notes that
pressure by census ofŽcials and even falsiŽcation at the margin were also
behind the sharp swing towards Bashkir identity in the 1979 Census,
although he does not provide any proof of such actions.
 Identity change in Bashkortostan 571
Administrative pressure during the census was certainly responsible
for the sharp swings towards Bashkir identity in the results of the 1959
and 1979 Censuses. But it cannot explain the persistence of the Tatar-
speaking Bashkir identity or the extent of the shift back to Tatar iden-
tity. To explain these phenomena, we must return to the Soviet
institutional structure and also to Bashkir language policy.
Institutions. Soviet institutions contributed to the trend towards Bashkir
identiŽ cation in the 1959 –1979 period. The advantages of being a
member of the titular ethnic group were reinforced throughout this
period. Tatar writers were required to publish their books in Bashkir;
members of the artistic community were told that they could only
succeed if they identiŽed as Bashkir painters or musicians. It was difŽ-
cult for a Tatar to become a government ofŽcial or a manager in an enter-
prise. 18 In this environment, not only was it irrational for Bashkirs to
switch to a Tatar identity, thousands of Tatars began to list themselves as
Bashkirs in ofŽcial documents (Kulchik 1992, p. 37). The extent of this
effect, even after the mass switch to Tatar identity in the 1989 Census, is
shown by a survey conducted in Buraevsky district in 1990. This survey
showed that 43 per cent of the population declared themselves Bashkir
in ofŽcial documents. An equal number declared themselves Tatar. In
response to a question about their ethnic self-identiŽcation, more than
half of the ‘passport’ Bashkirs declared that they actually perceived
themselves as Tatar despite their ofŽcial designation . Thus, 24 per cent
of the population were ofŽcially designated as Bashkir even though their
self-identity was Tatar19 (Vecherniia Ufa, 24 August 1990).
How was this change accomplished in a society where people’s ethnic
identity seemed to be rigidly deŽned on the basis of their parents’ eth-
nicity? First of all, local administrators had a great deal of leeway in
enforcing the requirement that newborns’ ethnic identity corresponded
to that of their parents. In a region like Bashkortostan, where the
administration sought to increase the Bashkir proportion of the popu-
lation, administrators at the birth registry often turned a blind eye
towards cases where Tatar parents declared their baby to be of Bashkir
ethnicity. 20 Furthermore, Tatars were encouraged to declare themselves
Bashkirs when joining the Communist Party. Party leaders were only too
glad to modify the necessary identity documents to allow such a change
(Sovetskaia Bashkiria, 29 November 1989). Despite the formal Soviet
Union-wide policy that an individual ’s ethnic identity match the identity
of at least one of his or her parents, republic leaders created conditions
where this requirement could be widely ignored by local administrators.
The creation of a sizeable group of ‘passport’ Bashkirs was seen as a
mixed blessing by many Bashkir intellectuals. On the one hand, they
desired to increase the size of their ethnic group. They found it particu-
larly galling that they were outnumbered by Tatars on territory that they
believed they ‘owned’. At the same time, they saw Bashkirs from the
572 Dmitry Gorenburg
northwest, especially those who considered Tatar their native language ,
as less truly Bashkir than those from the southern and eastern regions,
who spoke the Bashkir literary language and whose culture was more
similar to what they perceived as traditional Bashkir culture.21 The result
of this duality was a tendency for the Bashkir-controlled government to
try to force or persuade people in the northwest to identify as Bashkir
during the census campaign, followed by an equally strong tendency to
enact policies that favoured areas where ‘real’ Bashkirs lived (Kuzeev
1990, p. 41). This dual policy had the effect of increasing the amount of
revenue that could be directed towards Bashkir needs by in ating the
total number of Bashkirs, while at the same time restricting the sphere
of legitimate recipients of this largesse, cutting out ‘fake’ Bashkirs and
leaving more for the ‘real’ Bashkirs who controlled the government. I call
supporters of this policy, who controlled government ethnic policy until
the 1970s, the ‘self-consciousness’ faction.
Language policy and 1989. The dual track policy described in the
previous section produced the desired results for several decades. It
failed in 1989 because of a combination of factors. First, democratization
and the Gorbachev reforms, which emphasized individual merit, made
ethnically-based privileges morally suspect – many such privileges were
eliminated. Without privileges, many Tatars who had changed their iden-
tity to Bashkir for instrumental reasons switched back to their previous
Tatar identity. Second, democratization also made government pressure
and campaigns to persuade Tatars to declare themselves Bashkir more
difŽcult. This led many Tatars who had identiŽed as Bashkir because of
pressure from local administrators or census workers to return to their
previous ethnic identiŽcation (Iskhakov 1993, p. 36). Both of these
groups were people who had never truly changed their ethnic identity,
merely changing their public ethnic label.
Finally, many people who considered themselves ethnically Bashkir
but who spoke the Tatar language were persuaded to switch because of
Bashkir education policies in the late 1970s and early 1980s. In the late
1970s, those Bashkir leaders who advocated the ‘one nation – one lan-
guage’ conception of Bashkir identity took control of the education min-
istry and other cultural policy-making government bodies from the
previously dominant ‘self-consciousness’ faction, which had pursued the
policy of maximizing Bashkir numbers described above. The one-
language faction used their control over ethnic policy to remove Tatar
from the list of the republic’s ofŽcial languages in the 1978 constitution
(Khalim 1991, p. 171). Following this constitutional change, the govern-
ment sought to institute the ‘one nation – one language ’ policy by switch-
ing schools in Tatar-speaking Bashkir villages to Bashkir language
instruction. This policy was fostered by the belief that Bashkirs in this
area were former Bashkir-speakers who had been assimilated into the
Tatar cultural and linguistic milieu. Bashkir leaders justiŽ ed the policy by
 Identity change in Bashkortostan 573
arguing that Bashkirs had an obligation to be  uent in their ‘native’ lan-
guage and that the best way to achieve this  uency was through Bashkir-
language education in the schools.22 The Bashkir language that was used
in this instruction was the literary language that was entirely foreign to
northwestern Bashkirs (Kuzeev 1994, p. 122). As a result, the number of
students receiving Tatar language instruction decreased from 110,000 to
21,000 between 1978 and 1987 (Khalim 1991, p. 175).
The introduction of Bashkir-language schooling fomented widespread
resistance among the inhabitants of the region. While a minority of the
Tatar-speaking Bashkirs accepted the ‘one language – one nation’ policy,
changing their native language to correspond with their ethnic identiŽ-
cation, the vast majority refused to accept Bashkir as their native lan-
guage and began to re-identify as Tatar.23 Prior to the Bashkirization of
the schools, local Tatar-speaking Bashkirs had arrived at a modus vivendi
with the Bashkir state. Since the 1920s they considered themselves to be
ethnically Bashkir, while speaking the Tatar language and subscribing to
much of Tatar culture. This situation was made possible by the lack of
impact of Bashkir public identity on private life.
Locals who spoke Tatar but considered themselves ethnically Bashkir
gained the advantages of a titular ethnic identity while preserving the
freedom to speak Tatar and follow Tatar cultural practices at home.
Bashkirization of the schools sought to force local inhabitants to make
their cultural and linguistic identities coincide with their public ethnic
identities, thus violating the implicit compact. Local Tatar-speakin g
Bashkirs responded with a mass campaign to restore Tatar education.
This resistance was spearheaded by the schoolteachers, who made up a
signiŽcant part of the rural intelligentsi a and played a key role in mould-
ing local ethnic identity (Korostelev 1994, p. 82). The campaign was
eventually successful, but only after the removal of the Bashkir Com-
munist Party chief by Gorbachev in 1987.
The abortive Bashkirization campaign had a powerful effect on ethnic
identiŽ cation in the 1989 Census. By attempting to force Tatar-speakin g
Bashkirs to switch to the Bashkir language, the government signiŽcantly
increased the cost of maintaining Bashkir ethnic identity for non-Bashkir
speakers. Tatar-speaking Bashkirs knew that Bashkirization had
occurred only in Bashkir villages. Neighbouring Tatar villages kept their
Tatar schools. In this context, many Tatar-speaking Bashkirs came to
believe that only by changing their ethnic identity to match their linguis-
tic identity could they be safe from future linguistic Bashkirization
campaigns.
It is important to note that available statistical data cannot prove that
the identity change that occurred among Tatar-speaking Bashkirs in the
1980s was substantively different from previous instances of purely
public identity change. I have reached this conclusion on the basis of
interviews and other non-quantitative data. For example, Marat
574 Dmitry Gorenburg
Ramazanov, one of the leaders of the Tatar nationalist movement,
described himself as a Tatar-speaker of Bashkir ethnicity who re-identi-
Žed as a Tatar and joined the Tatar nationalist movement because
Bashkirs insisted that he change his language while Tatars were willing
to accept him as he was.24 This sort of claim is substantively different
from the argument made by some Tatars in earlier periods who noted
that they always considered themselves Tatar despite their identity docu-
ments, which showed them to be Bashkir (Khalim 1990, p. 7). Based on
this difference, we can hypothesize that two separate processes of iden-
tity change, one public and one private, were occurring simultaneously
in northwestern Bashkortostan in the 1980s.
Statistical proof for this claim may come from the coming 1999 Census.
With the restoration of pro-Bashkir favouritism in the 1990s, we should
expect a return to Tatar-speaking Bashkir identity if the shift away from
it in the 1980s was simply a reaction to the end of incentives for Bashkir
identiŽcation by people who always considered themselves Tatar. If, on
the other hand, Tatar-speakin g Bashkir identity continues to decline
despite the re-emergence of previously existing incentives to identify as
Bashkir, this will show that the 1980s saw a change in private ethnic
identiŽcation that cannot be reversed simply through a return to the old
incentives structure.

* * *

We have seen from census data that a signiŽcant number of Bashkirs

re-identiŽed as Tatars between the 1979 and 1989 Censuses. Many of
these people had not actually changed their ethnic identity, having pre-
viously switched ethnic labels for instrumental reasons. Their re-identiŽ-
cation was similar to the mass re-identiŽcation from Bashkir to Tatar
identity that occurred at the beginning of the twentieth century.
What made the 1989 identity change different from previous mass
changes of identity in this area was the presence of a second group of
people, comprised primarily of Tatar-speaking Bashkirs, who actually
changed their personal identity, rather than merely taking on a different
public identity that would improve their economic or political status.
Unlike the Ž rst group, these people had never previously considered
themselves Tatar, having maintained Bashkir as their ethnic identity con-
tinuously since before the communist revolution. By attempting to make
individuals ’ linguistic identities match their ethnic identities, Bashkir
leaders raised the costs of maintaining the ethnic identity for these
people to unacceptable levels. This change, especially in combination
with a decline in the economic and political beneŽts of Bashkir ethnic
identity, led over 100,000 Tatar-speakin g Bashkirs to change their ethnic
identity from Bashkir to Tatar. By trying to force linguistic identity to
match ethnic identity, Bashkir leaders unwittingly brought about the
 Identity change in Bashkortostan 575
opposite outcome, wherein local inhabitants changed their ethnic iden-
tity to match their linguistic identity.
Such an identity change was possible partially because the cultural dis-
tance between Tatar and Bashkir ethnic identities was small. It was also
possible because local inhabitants had become used to neighbours chang-
ing their public ethnic identity in order to fulŽl government targets for
ethnic distribution or to maximize their utility. Throughou t the twenti-
eth century many inhabitants of northwestern Bashkortostan frequently
changed their ethnic label without actually changing their behaviour or
social interactions in any way. However, the frequency of this change in
label made ethnic identity as a whole seem more mutable even for those
who did not change their public identity, making it signiŽcantly easier for
them to change their ethnic identity when conditions warranted.

Conclusions
Throughout the twentieth century, changes in ethnic identity in north-
western Bashkortostan followed changes in state policies towards ethnic
groups. Before the 1917 revolution the elimination of estate-based
restrictions on land-ownership led many Tatars and Mishars who had
identiŽ ed as Bashkirs in order to gain the right to own land to return to
their previous identities. In the 1920s the establishment of ethnic
republics which were allowed to favour titular ethnic groups led many
Bashkirs who found themselves in Tatarstan after the division of Ufa
province to re-identify as Tatars. At the same time, government policies
on the creation of a Bashkir literary language led many Bashkirs to
declare Tatar as their native language. Throughout the post-war period,
preferences for members of the titular ethnic group led many Tatars in
Bashkortostan to declare themselves Bashkir in ofŽcial documents. The
end of such privileges in the mid-1980s led many of these individuals to
reclaim the Tatar identity in the 1989 Census. Finally, the forced con-
version of Tatar-language schools in Bashkir villages to literary Bashkir
persuaded many Tatar-speaking Bashkirs to re-identify as Tatars in order
to allow their children to attend Tatar-language schools. The instability
of ethnic identity in this region is thus the result of multiple changes in
government policy towards minority ethnic groups.
These ethnic identity shifts demonstrate the crucial role of the state in
provoking ethnic change. In fact, changes in state policy and the creation
of new state institutions are the link between accounts of ethnic change
that focus on identity change for instrumental reasons and accounts of
ethnic change that focus on changes in the cultural content of ethnic iden-
tity. Government action can modify the cost-beneŽt calculus that leads
to identity shift for instrumental reasons by introducing new policies that
beneŽt or penalize the members of a particular group. The state can also
channel the cultural content of ethnic identity in new directions by such
576 Dmitry Gorenburg
actions as the creation of a new literary language or the merging of
several ethnic groups into a new, compound identity. By focusing on the
immediate causes of ethnic identity change, much of the theoretical
literature has missed the role of state action as the crucial precursor to
such change.

Acknowledgements
I wish to thank Pauline Jones Luong, Stephen ShenŽeld, Henry Hale,
Timothy Colton, Ida Gorenburg, and an anonymous reviewer for helpful
comments on this article. Particular thanks must also go to the members
of the Sawyer Seminar on Democratic Performance and the Post-
Communist Politics Workshop at Harvard University, who assessed an
earlier version of this work, and the International Research Exchange
Board (IREX), which provided funding for the Želd research on which
the article is based.

Appendix 1. Population and native language by County

Northwestern Bashkortostan

Population percentage Native language %

—————————————————————— ————————
Bashkir Tatar Bashkir Tatar
—————————————————————— ————– ———
County 1939 1959 1970 1979 1989 1939 1959 1970 1979 1989 1959 1989 1989

Alsheevsky 19 20 23 26 27 33 34 38 39 41 85 95 4
Askinsky 42 50 58 61 63 17 19 20 21 24 13 41 59
Bakalinsky 12 13 13 17 8 42 49 54 52 62 2 9 90
Baltachevsk y 24 62 75 74 22 53 17 6 6 59 0 7 93
Belebeevsky 4 3 3 4 4 15 17 21 22 24 15 46 46
Bizhbuliaksky 3 7 8 11 10 14 29 31 33 35 86 91 6
Birsky 4 3 5 13 4 10 11 12 5 14 13 39 56
Blagovarsk y 10 35 46 45 8 47 28 21 23 60 2 35 60
Buzdiaksky 21 45 35 33 26 61 42 53 59 65 13 29 71
Buraevsky 49 59 62 73 40 35 27 26 16 51 1 4 96
Davlekanovs ky 18 21 25 27 27 21 21 23 23 23 97 95 1
Diurtiulinsky 30 35 34 39 24 46 41 44 42 59 2 26 73
Ermekeevsky 21 27 24 40 11 30 31 38 25 54 2 15 84
Ilishevsky 54 63 71 79 63 33 27 21 13 30 5 3 96
Kaltasinsky 11 8 12 5 3 12 18 17 19 21 6 47 45
Karaidelsky/
Baikibashevsky 21 23 30 33 35 32 36 36 37 37 55 78 20
Krasnokamsky 39 40 31 32 24 9 13 18 23 39 2 24 73
Kushanrenkovsky 14 23 28 38 6 52 50 51 44 78 2 35 62
Mishkinsky 6 6 5 8 5 24 22 23 19 19 5 48 50
Miakinsky 27 29 27 32 26 36 40 46 44 55 65 90 9
Tatyshlinsky 57 64 61 67 54 13 8 13 7 21 N/A 12 88
Tuimazinsk y 32 28 51 62 31 38 36 34 25 56 5 13 86
Chekmagushevsky 30 37 30 34 19 62 54 62 56 76 3 7 92
Sharansky 13 13 14 27 6 34 37 39 28 51 2 13 86
Yanaulsky 38 39 42 45 17 21 20 23 19 29 37 22 77
 Identity change in Bashkortostan 577
Other Counties of Bashkortostan

Population percentage Native language %

—————————————————————— ————————
Bashkir Tatar Bashkir Tatar
—————————————————————— ————– ———
County 1939 1959 1970 1979 1989 1939 1959 1970 1979 1989 1959 1989 1989

Abzelilovsk y 69 71 76 82 85 4 4 4 3 3 100 100 0

Arkhangelsky 24 29 35 38 41 8 9 10 10 11 99 97 1
Aurgazinsky 9 8 9 11 11 17 46 47 46 48 94 93 6
Baimaksky 50 61 71 76 80 8 7 6 5 5 99 99 0
Belokataisky 28 32 38 39 40 3 5 5 5 5 99 98 0
Beloretsky 23 31 48 54 58 3 4 4 4 4 99 98 0
Blagoveshchensky/
Pokrovsky 4 5 6 7 8 10 11 15 16 18 46 76 17
Burziansky 83 81 89 92 95 3 6 3 2 2 100 100 0
Gafuriisky 30 36 37 42 42 21 19 24 24 24 87 97 2
Duvansky 2 11 14 16 18 9 11 13 14 14 96 92 5
Zianchurinsky 41 51 57 61 66 10 11 13 13 13 99 99 0
Zilairsky/
Matraevsky 22 26 36 42 48 3 5 5 4 4 99 99 0
Iglinsky 6 13 15 17 18 14 16 18 19 19 95 92 2
Ishimbaisky/
Makarovsky 33 60 64 69 70 17 10 7 6 7 84 97 2
Karmaskalinsky/
Buzoviazovsky 14 19 21 22 23 48 42 45 46 48 95 94 5
Kiginsky 33 29 34 37 36 42 53 55 54 57 93 93 6
Kugarchinsky/
Yumaguzinsky 30 38 42 46 48 12 14 14 15 16 99 99 0
Kumertaunsky/
Kuiurgazinsky 21 24 29 32 36 12 13 14 15 17 99 98 1
Meleuzovsky/
Voskresensky 9 11 29 32 35 6 9 10 12 15 97 98 1
Mechetlinsky 49 48 49 51 51 27 29 31 29 30 87 95 4
Nurimanovsky 21 24 28 32 29 31 33 37 36 38 83 91 7
Salavatsky/
Maloiazovsk y 43 45 52 55 59 30 31 30 30 29 99 97 1
Sterlibashevsky 18 22 24 27 27 49 56 60 61 62 87 94 5
Sterlitamaksky 4 5 8 12 14 13 23 27 23 27 92 91 4
UŽ msky 2 2 3 6 8 8 16 20 23 29 73 74 17
Uchalinsky 40 55 73 78 75 31 18 16 12 15 96 99 1
Fedorovsky 6 9 11 13 14 20 27 33 34 35 97 95 4
Khaibullinsky 42 44 53 59 66 3 4 5 6 4 99 98 1
Chishminsky 9 10 11 12 13 53 55 58 60 61 76 86 12

Sources: Yuldashbaev, table 18 (1939), calculated from table 19 (1959), table 20 (1970, 1989), table 21
(1989 native language). 1979 data provided by Ildar GabdraŽkov. Some counties changed names or
merged during this period. Both names are provided .

Notes
1. Banks (1996) provides a helpful summary and analysis of these and other debates
among anthropologists who study ethnicity. I am indebted to this journal’s anonymous
reviewer for this reference.
2. The data on Bashkir and Tatar identity are drawn primarily from Russian and
Soviet censuses. The Russian empire conducted a full census in 1897 and a rural census in
1913. The Soviet government conducted censuses in 1920, 1926, 1939, 1959, 1970, 1979, and
578 Dmitry Gorenburg
1989. From 1897 to 1920, the census did not ask a direct question about ethnic identity,
deŽ ning ethnicity purely on the basis of native language. In the context of an undifferenti-
ated linguistic environment where Bashkir and Tatar blended together in an intermediate
dialect, peoples’ perceptions of their linguistic identity depended to a large degree on their
ethnic identity. After 1926 the census asked separate questions about ethnic identity and
native language .
3. For purposes of clarity, I focus exclusively on shifts between Bashkir and Tatar iden-
tities Throughout the Soviet period, some Bashkirs and Tatars became RussiŽ ed and
declared Russian to be their native language. By 1989 5 per cent of Bashkirs considered
Russian to be their native language (Yuldashbaev 1995, p. 25). Furthermore, the children
of mixed marriages between Russians and minorities tended to identify themselves as
Russian. However, this study focuses on identity change among rural inhabitants while the
vast majority of both mixed marriages and cases of linguistic RussiŽcation occur in urban
areas. Most non-Russian inhabitants of rural areas do not have the possibility of switching
to Russian identity. For this reason, in this study I do not discuss changes of identity from
Bashkir or Tatar to Russian.
4. This section provides a schematic account of identity change in northwestern
Bashkortostan during this period in order to assist in the explanatory process. For Ž ne
detailed accounts of the processes described here see Ganeeva (1992); Korostelev (1994);
Rodnov (1995); Yakupov (1992).
5. Throughout this article, references to Tatars should be understood to include
Mishars unless explicitly stated otherwise. The groups were not distinguished by Soviet
censuses after 1926.
6. The Cossacks are the best known example of an estate in Imperial Russia.
7. In the Birsk region, for example, Bashkirs owned 61 per cent of all land in 1917
(Kuzeev, Moiseeva and Babenko 1987, p. 37; Rodnov 1995, pp. 69 and 77).
8. Calculated from Khismatullin (1992, Table 1); Korostelev (1994, p. 80); and
Yuldashbaev (1995, Tables 6–8).
9. A distinct Bashkir culture and language did exist, but was concentrated in areas to
the south and east.
10. The Tatar migration did not reach the eastern regions (Ganeeva 1992, pp. 60 –1).
11. Like the 1897 Census, both these censuses asked respondents about their native
language, rather than their ethnicity.
12. Out of 114,000 Teptiars in 1926, 60,000 were reassigned as Tatars, 29,000 as Bashkirs
and 23,000 insisted on Teptiar (Bikbulatov 1992, p. 43).
13. Unfortunately, canton-level data for the 1920 Census are unavailable .
14. Local ethnographers note that this re-identiŽcation cannot be explained by
migration or other natural causes. The 1921 famine, while causing great loss of life, affected
Tatars and Bashkirs equally (Kuzeev 1990, p. 38; Iskhakov 1993, p. 122).
15. Appendix 1 shows the changes in the proportion of Tatar and Bashkir population
in each district from 1939 to 1989.
16. It is important to note that both this database and district-level census results are
aggregate data and cannot show individual identity change. The village database, for
example, shows the predominant ethnic identity of each village, within which identity
change in multiple directions may be occurring. This article seeks to analyse broad trends
in individual identity change on the basis of personal incentives, rather than attempting to
explain speciŽ c cases of identity change. The latter task cannot be performed with
aggregate data because of the ecological inference problem (but see King 1997).
17. Stalin deŽned nationality as a society of people based on common language ,
territory, economic life, and culture (Stalin 1994, p. 20).
18. Vasil Babenko, Researcher at Bashkortostan State Assembly, interview, November
1995.
19. 18 per cent of the respondents declared their self-identiŽ cation as Bashkir; 66 per
cent stated that they were Tatar.
 Identity change in Bashkortostan 579
20. Zufar Enikeev, Member of Bashkortostan State Assembly, interview, November
1995.
21. Sergei Kabashov, Deputy Department Administrator, Bashkortostan Cabinet of
Ministers, interview, December 1995.
22. For an example of Bashkir nationalist rhetoric on this issue, see ‘How it happened’,
Zamandash , 28 December 1990. For a Tatar response, see Khalim (1991, pp. 168 –77).
23. Ildar GabdraŽkov, Assistant Director, Department of Ural Peoples at Ufa Science
Centre, Russian Academy of Science, interview, November 1995.
24. Interview, November 1995.

References
BANKS, MARCUS 1996 Ethnicity: Anthropological Constructions, London: Routledge
BARTH, FREDRIK 1969 ‘Introduction’, in Fredrik Barth (ed.), Ethnic Groups and
Boundaries, London: George Allen & Unwin
—— 1994 ‘Enduring and emerging issues in the analysis of ethnicity’, in Hans Vermeulen
and Cora Govers (eds), The Anthropology of Ethnicity: Beyond ‘Ethnic Groups and
Boundaries’ , Amsterdam: Het Spinhuis
BIKBULATOV, N. V. 1992 ‘Voprosy vzaimodeistv iia i sotrudnichestva narodov Bashkor-
tostana v Sovetskuiu epokhu’, in V. Ia. Babenko (ed.), Sovremennye Etnicheskie Protsessy
v Bashkortostane, Ufa: Institute of History, Language, and Literature
BROMLEY, YULIAN 1981 Sovremennye Problemy EtnograŽi, Moscow: Nauka
FRANK, ALLEN J. 1998 Islamic Historiography and ‘Bulghar’ Identity among the Tatars
and Bashkirs of Russia , Boston, MA: Brill
GANEEVA, A. R. 1992 ‘Problema etnokulturnogo vzaimodeistv iia Bashkir i Tatar v
severo-zapadnoi Bashkirii’, in R. G. Kuzeev and V. A. Tishkov (eds), Etnos i Ego
Podrazdeleniia , Moscow: Institute of Ethnology and Anthropology
GROSS, JO-ANN 1992 ‘Introduction: approaches to the problem of identity formation’,
in Jo-Ann Gross (ed.), Muslims in Central Asia: Expressions of Identity and Change,
Durham, NC: Duke University Press
ISKHAKOV, D. M. 1993 Istoricheskaia DemograŽia Tatarskogo Naroda, Kazan:
Ibragimov Institute of Language, Literature, and History
ISHBERDIN, E. F. 1989 ‘Bashkirskie dialekty i literaturnyi yazyk’, in R. G. Kuzeev (ed.),
Etnicheskie i EtnograŽcheskie Gruppy v SSSR i Ikh Rol v Sovremennykh Etnokulturnykh
Protsessakh , Ufa: Institute of History, Language, and Literature
KEYES, CHARLES F. 1981 ‘The dialectics of ethnic change’, in Charles F. Keyes (ed.),
Ethnic Change, Seattle, WA: University of Washington Press
KHALIM, AIDAR 1990 ‘Interview s Aidarom Khalimom’, Oran, no. 4, pp. 5 –10
—— 1991 Kniga Pechali, ili Zapiski Aborigena, Vilnius: Mokslas
KHISMATULLIN, A. A. 1992 Archival note on number of Bashkirs and other peoples of
Ufa province, unpublished
KING, GARY 1997 A Solution to the Ecological Inference Problem, Princeton, NJ:
Princeton University Press
KOROSTELEV, A. D. 1994 ‘Dinamika etnicheskogo sostava selskogo naseleniia Bash-
kortostana’, in L. M. Drobizheva (ed.) Koniktnaia Etnichnost i Etnicheskie Konikty,
Moscow: Institute of Ethnology and Anthropology
KULCHIK, YURI 1992 Etnopoliticheskie Protsessy v Bashkortostane, Moscow: Cultural
Initiative Foundation
KUZEEV, R. G. 1990 Mezhnatsionalnye Otnosheniia: Istoriia, Teoriia, Puti Demokratich-
eskogo Obnovleniia, Ufa: Knowledge Society of the Bashkir Republic
—— 1994 ‘Istoriko-etnologicheskii vzgliad na sovremennuiu etnolingvisticheskuiu situat-
siiu v respublike Bashkortostan’, in M. N. Guboglo (ed.), Yazyk i Natsionalizm v Postsovet-
skikh Respublikakh, Moscow: Institute of Ethnology and Anthropology
580 Dmitry Gorenburg
KUZEEV, R. G., MOISEEVA, N. N. and BABENKO, V. IA. 1987 ‘Etnicheskie protsessy
v novoe i noveishee vremia’, in R. G. Kuzeev (ed.), Etnicheskie Protsessy v Novoe i
Noveishee Vremia, Ufa: Institute of History, Language, and Literature
MURZABULATOV, M. V. 1995 Bashkortostan i Bashkiry v Zerkale Statistiki, Ufa:
Institute of History, Language, and Literature
NAGATA, JUDITH 1981 ‘In defense of ethnic boundaries: the changing myths and
charters of Malay identity’, in Charles F. Keyes (ed.), Ethnic Change, Seattle, WA:
University of Washington Press
RODNOV, M. I. 1995 ‘Etnicheskii sostav naseleniia Birskogo uezda v kontse XIX –
nachale XX Veka’, in M. I. Rodnov (ed.), Stranitsy Minuvshego, Ufa: Bashkir State
Historical Museum
ROOSENS, EUGEEN 1994 ‘The primordial nature of origins in migrant ethnicity’, in
Hans Vermeulen and Cora Govers (eds), The Anthropology of Ethnicity: Beyond ‘Ethnic
Groups and Boundaries’, Amsterdam: Het Spinhuis
STALIN, JOSEPH 1994 [1913] ‘The nation’, excerpted from Marxism and the National
Question, in John Hutchinson and Anthony D. Smith (eds), Nationalism , New York:
Oxford University Press
TISHKOV, VALERY 1997 Ethnicity, Nationalism and Conict in and After the Soviet
Union: The Mind Aame, London: Sage Publications
VAIL, LEROY 1989 ‘Introduction: ethnicity in southern African history’, in Leroy Vail
(ed.), The Creation of Tribalism in Southern Africa , Berkeley, CA: University of California
Press
YAKUPOV, R. I. 1992 ‘Teptiare: Soslovie ili etnos’, in R. G. Kuzeev and V. A. Tishkov
(eds), Etnos i Ego Podrazdeleniia , Moscow: Institute of Ethnology and Anthropology
YULDASHBAEV, B. KH. 1995 Bashkiry i Bashkortostan: Etnostatistika, Ufa: World
Bashkir Congress
YULDASHEV, A. A. 1968 ‘Severo-zapadnyi dialekt Bashkirskogo yazyka v ego
otnoshenii k literaturnomu yazyku ’, in B. M. Yunusaliev (ed.), Voprosy Dialektologii Tiurk-
skikh Yazykov, Frunze: Ilim Publishing

DMITRY GORENBURG is Research Associate in the Project on Cold

War Studies at the Davis Center for Russian Studies, Harvard University.
ADDRESS: Davis Center for Russian Studies, Coolidge Hall, Harvard
University, 1737 Cambridge Street, Cambridge, MA 02138, USA.