! " !

#$
%! & # '
( ) * !
' & ' & + , " *
-./ --0
1 2 34 4560 /7- .
88 &8 &8& * 8
" ! 8
9 : ;..7

Language and identity in post-Soviet Moldova1

Matthew H. Ciscel
Central Connecticut State University, New Britain, CT, USA
ciscelm@[Link]

Abstract
The multilingual Republic of Moldova emerged from the Soviet Union between 1989 and 1991 as an example of the
linguistic complications that can result from imperial domination and the mobility associated with political change.
The study draws on historical, survey, and ethnographic data to illustrate the complexity of the accessible language
and social identities in the emergent country. Among the issues discussed are the status of Russian, the argument
over the status of Moldovan as an independent language or a dialect of Romanian, and the role of English as an
international language in Moldova's globalizing culture and economy. Trends in the survey data are both reinforced
and challenged by the ethnographic data. In conclusion, I argue that the linguistic identity crisis in Moldova seems
far from resolution, even though some reasons for hope can be identified.

Introduction

This study investigates the crisis in linguistic and national identity that affects the multilingual
Republic of Moldova. About two-thirds of Moldovans speak a dialect of Romanian that has been
highly politicized in recent history. Focusing on the contentious arguments over their linguistic
identity both during and after the Soviet period, I argue that the persistent notion of a separate
Moldovan language is rooted both in an ancient Moldovanism that predates nineteenth-century
Romanian nation-building and in Moscow’s exploitation of this identity during the Soviet era.
Survey data that illustrates the balanced status of Romanian/Moldovan and Russian in the
country today provide a connection between Moldova’s troubled past and its present identity
crisis.

The connection between linguistic and national identities has been a common theme in the
sociolinguistic investigation of nation-building and nationalism. Joshua Fishman’s (1973) essays
on this topic serve as the field’s seminal works. However, ideas and theories related to this topic
have been carried off in many directions: polemical, philosophical, empirical, ethnographic,

1
The fieldwork that led to this study was partially supported by the Title VIII Combined Research and Language
Training Program, which is funded by the State Department of the United States of America and administered by the
American Council for International Education. However, any opinions and mistakes that appear herein are solely the
author’s.

106
among others. For instance, Safran asserts that ‘a nation of purely “political” essence is a
fantasy... for in order for the “political” to do its work, there must be an identitive readiness that
is based on psychological and cultural foundations’ (1999:91). He goes on to pick out linguistic
markers as the most salient form of such foundations in individual interactions. These individual
interactions are the specific object of ethnographic studies such as Miller (1999), which develops
a model of enacted identity comprised of context, language use, and group membership.
Through ethnographic data about a group of non-English-speaking, immigrant schoolchildren in
Australia, Miller argues that ‘the important link between second language use and social identity
must be seen in its relation to empowerment, being heard, and the ongoing process of self-
realisation’ (1999:163). In a multilingual society, the status of the individual depends crucially
upon that individual’s access to and proficiency in the language(s) of highest status within the
national context. In Moldova, the notions of national context and the status of competing
languages have been subjected to instability by recent and more distant historical events.
The roots of ancient Moldovanism are still productive today because of two historical
phenomena. The first involves Russian claims to the territory. Originally annexed by imperial
Russia in 1812, the Moldovan territory east of the Prut River remained isolated during the entire
process of Romanian nation-building that affected western Romanian-speaking territories (King
1999:49). Because of this, it did not participate in the consolidation and Latinization of the
standard Romanian language. The second involves the impoverished, rural character of the
Moldovans. During the Tsarist occupation, new urban and industrial areas were intentionally
populated with russophone and germanophone minorities. As a result, the few Romanian
speakers with access to education and political power were forced to adopt Russian, while the
majority remained powerless peasants. These peasants maintained the ancient Moldovan identity,
due to their forced exclusion from both Romanian and Russian identities.

Before and during the Soviet period, the peasant’s notion of ancient Moldovanism was exploited
by the irredentist policies of Moscow toward the region, which had been unified with Romania
after World War I. The propagation of a separate Moldovan language, using the Cyrillic script,
was a central policy through much of this period. Despite this policy, however, the so-called
Moldovan language never gained full functionality, because urbanized speakers of the so-called
language were still encouraged to shift to Russian. As such, the notion of ancient Moldovanism
was revived in modern, Soviet clothing.

Post-Soviet Identities

Due to this colonial heritage, the national identities of Moldovans at independence in 1991 took
many forms: staunchly Romanian, staunchly Russian, and several notions of moderate Moldovan
identity. This section of the study contains data from a survey of 124 students and young
professionals in the capital, Chi in u. The data presented here relate primarily to social identity
and its relation to language use. I will begin with a brief overview of the recent ethno-linguistic
context.

107
 Table I.
1989 Soviet census data on nationality groups in Moldova and their L1s and L2s (%)

Nationality Population % L1 L1 L1 L2 L2
Self Moldovan Russian Moldovan Russian

Moldovans 2,794,749 64.5 95.4 (95.4) 3.3 1.7 53.1

Ukrainians 600,366 13.8 61.6 1.6 36.7 12.8 43.0

Russians 562,069 13.0 99.1 0.6 (99.1) 11.7 0.6

Gagauz 153,458 3.5 91.2 1.1 7.4 4.4 72.8

Bulgarians 88,419 2.0 78.7 2.4 18.1 6.9 68.3

Jews 65,672 1.5 25.9 0.8 72.9 15.2 23.1

Roma 11,571 0.3 82.0 13.5 3.6 30.6 41.9

Other 56,579 1.4 - - - - -

TOTAL 4,335,360 100 88.9 62.0 23.2 3.9 44.6

The Soviet census of 1989 provides the most recent demographic statistics for the country, some
of which are presented in Table I (compiled from Karasik 1992, Gordon 1993 and King 1999).
The census collected self-reported data on nationality and linguistic fluency. The notion of
nationality in the Soviet Union was explicit and official, appearing in each citizen’s identity card
and passport. In addition, following official policy, Romanian and Moldovan were included as
two different national identities. As such, there were only about two thousand Romanians
counted in Moldova in 1989. Any local Romanians were counted as Moldovan. Similarly, many
of the smaller minorities such as urbanized Ukrainians and Jews were likely counted as Russians,
this being a more prestigious category to belong to, particularly for fully assimilated urbanites
and those from multiethnic families. Because of these and other limitations, census data are
often incomplete or inaccurate in unpredictable ways. These limitations of the data
notwithstanding, the trends represented here remain relevant as a foundation for understanding
the complexity of the country’s ethno-linguistic identity dynamics.

As can be seen in Table I, self-identifying Moldovans made up almost two-thirds of the

108
language (L1) with the same name as their nationality. For instance, the ‘L1 Self’ for Bulgarian
nationals would be Bulgarian, while for Jews it would be Yiddish, etc. In 1989, almost all
Moldovan nationals (95.4%) identified their L1 as Moldovan, with only a few (3.3%) who had
shifted to Russian. However, more than half of Moldovans did claim fluency in L2 Russian.
Russian nationals, on the other hand, very rarely shifted to Moldovan (0.6%) and also rarely
spoke L2 Moldovan fluently (11.7%). Similarly, the other national minorities, aside from the
Roma, tended to shift to or adopt L2 Russian rather than Romanian. For instance, among the
Gagauz, only 5.5% had some fluency in Moldovan while 80.2% had developed fluency in
Russian, either as L1 (7.4%) or L2 (72.8%).

The preference to learn Russian among minorities reflected the symbolic and institutional status
of Russian during the Soviet period. The urban-rural split was also significant, with Russian
spoken in the cities and national languages spoken in rural areas. For instance, although 3.3%
and 53.1% of all Moldovan nationals had shifted to L1 Russian and acquired it as an L2,
respectively, the percentages of assimilated Moldovans in the capital city were much higher,
with 11% shifted and 74% fluent as an L2. Moreover, Gordon (1993:135) shows that the
linguistic russification of Moldovan nationals increased dramatically in the two decades before
independence, with a 13.6% increase in Russian fluency between 1970 and 1979 followed by a
more moderate 7.2% increase between 1979 and 1989. This trend has abated significantly since
independence, because Romanian now enjoys a status almost equal to Russian in many areas,
particularly institutional. In contrast, the ethnic Russian adoption of the Romanian language
remains low, although it has certainly increased somewhat. As a result of the linguistic
russification of both Moldovan and other non-Russian nationals in the country, Russian had by
1989 become somewhat more widely spoken than Moldovan (Romanian). Looking at the totals
in Table I, one can see that Moldovan was spoken by 65.9% of the population as either an L1
(62%) or an L2 (3.9%), while Russian was spoken by 67.8% (with 23.2% L1 and 44.6% L2). In
addition, Russian was still much more highly valued, based on its prestige as an international,
cultured language. As mentioned, it was also the language of the urban elites.

However, after 1989 the national and linguistic identities in the country began to shift radically.
Perestroika led to the revival of the Romanian identity and demands by both Romanian and
Moldovan identity groups for more power in the political, economic, and cultural markets of the
country. The language laws of August 31, 1989, recognized the identical linguistic character of
Romanian and Moldovan, returned official use of the language in the country to the Latin
alphabet, and made it the sole official language. The apparent radicalism of this pro-Romanian
movement led to an even more radical, violent reaction from the Russian-speaking minorities in
the eastern region of Transnistria and the southern area of Gagauzia after the independence of
1991. Democratization of politics starting in 1994 led to moderation of the pro-Romanian
ideology among leaders in Chi in u, but the break-away status of Transnistria, where the Russian
army remains in a nominally peace-keeping role, has remained unresolved. In 2003, Moldova is
ruled by a democratically-elected, revived Communist party that propagates the old Soviet
policies of russification and Moldovan distinctiveness. But, more than a decade of non-
Communist rule have established the pro-Romanian opposition as a permanent alternative in
politics and identity choices. Although many still argue that Russian should be a second official
language (because ‘everyone’ speaks it, as the ideology maintains), ‘Moldovan’ is still the only
official language in the country’s Constitution. At the same time, most educated, urban
Moldovans will acknowledge that Romanian and Moldovan are the same language, even if they
still refer to it from time to time as Moldovan. As I will show in greater detail below, the

109
national identity of Moldovans as separate from Romanians or Russians has been largely
accepted, but cultural, particularly linguistic, identities have been much more contentious.
Article 13 of the 1994 Constitution concerns the national language and the use of other
languages in the country. The official translation into English reads:

(1) The national language of the Republic of Moldova is Moldovan, and its writing is based on
the Latin alphabet.
(2) The Moldovan State acknowledges and protects the right to preserve, develop and use the
Russian language and other languages spoken within the national territory of the country.
(3) The State will encourage and promote studies of foreign languages enjoying widespread
international usage.
(4) The use of languages in the territory of the Republic of Moldova will be established by
organic law.

Noteworthy here is that the term ‘Moldovan’ is used to the exclusion of ‘Romanian’ and that the
Russian language is the only other language named explicitly. This draws out both the dominant
roles of these two languages in the country and the contentiousness over which label to use for
the national language. In order to clarify these issues, I will consider how languages are
distinguished and identified in a national context like Moldova.

A language can be identified based on three criteria: structural differences, distinctions in social
or national group membership, and differences in the value associated with them. The Moldovan
dialect of Romanian does have several structural features that mark it as potentially distinct from
standard Romanian. Many common words with labial consonants in initial position in standard
Romanian are pronounced with non-labial equivalents in the Moldovan dialect. As a result, rural
Moldovans will say /gine/ for standard /bine/ (in English, ‘well’), /ki ware/ for /pit ware/
(‘legs’), and /njere/ for /mjere/ (‘honey’). In addition, unstressed vowels are more commonly
centralized toward schwa than in the standard variety. The lexicon of Moldovan also varies
somewhat from standard Romanian. For example, rural Moldovans often say curec instead of
the standard varz (‘cabbage’). Some other examples are pepeni for castrave i (‘cucumbers’) and
nic for nimic (‘nothing’). These words are all simple regional variants. However, there are a
number of other lexical differences based on borrowings from Russian. These include the
Moldovan use of cresl for standard Romanian fotoliu (‘armchair’), butilc for sticl (‘bottle’),
and cran for robinet (‘faucet’). However, few of these variations were actually integrated into the
standard Moldovan promoted by the Soviet authorities, because they were associated too closely
with the powerless spoken varieties of the rural peasants. Therefore, the standard Moldovan
promoted throughout the Soviet period was essentially identical to standard Romanian with the
exception of the alphabet and the use of Russian, rather than French, borrowings for technical
terminology. As such, the structural criterion for distinguishing the languages disappeared with
the introduction of the Latin alphabet in 1989 and the opening of the border with Romania in
1990 (Dyer 1999).

It is worth noting that a last remnant of the Cyrillic alphabet remains in some written forms of
Romanian in Moldova. This remnant concerns the choice between the graphemes <î> and <â>,
which both represent the high, central, unrounded vowel [ ], a sound common to Slavic but not to
other Romance languages. In standard Romanian, both graphemes appear in a distribution that
allows the Latin roots of the language to be apparent. For instance, <î> is used in words like în

110
(‘in’), where an <i> or <e> appears in French, Italian, and Spanish cognates. But, <â> is used in
words like pâine (‘bread’), where the <a> would appear in western Romance languages. In
Moldova, on the other hand, many street signs and written forms of Romanian avoid the <â>
altogether. A typical example is the spelling pîine on bread shops. The <î> more closely
resembles the Russian grapheme for the same sound, < >. This avoidance of <â> reflects an
anti-Romanian and pro-Russian ideology, because it preserves an element of the Slavic
appearance and eschews the Latinate.

The second criterion used to identify a language, the one involving social group distinctions, is
clearly present for Moldovan and Romanian. However, language is an aspect of national
identity that tends to need deeper roots than other aspects. Because of this, North Americans still
refer to their language as English rather than as American or Canadian, even though their
national identity is clearly American or Canadian. Similarly, Austrians call their language
German and many Belgians call theirs French. Just as the Belgians have with regard to French,
Moldovans may have developed a national identity separate from Romania in the two centuries
of isolation, but their language goes back much further. For that reason, many Moldovans who
call themselves Moldovan nationals will still call their language Romanian (Crowther 1996).
However, for the same reason, the notion of ancient Moldovan that was preserved during the two
century isolation continues to support a separate Moldovan linguistic identity, one which has
been reinforced by Soviet propaganda. As a result, the social criterion produces two competing
identities for the indigenous people of Moldova: ancient Moldovanism, valorized by Soviet
ideology, or modern Romanianism.

Finally, the criterion of status, or value, also draws a clear distinction between Romanian and
Moldovan. Standard Romanian is a language rich in literary and scientific traditions. Of course,
it has not had this status as long as many other European languages and is not widely adopted as
an L2, but it is certainly multi-functional and thriving. Moldovan, on the other hand, borrowed
the Romanian classics for its literature and never achieved much use as a scientific and technical
language, despite Soviet efforts and claims (Korletianu 1979:5). On the contrary, Moldovan
remained fairly limited functionally (Bruchis 1988). I have discussed the patterns of shift to
Russian, particularly in the urban areas during Soviet times. This shift was facilitated by the
unequal status of Russian and Moldovan. Russian was promoted by the Soviets as the language
of international communication and the language of the revolutionary vanguard. As a symbol of
imperial power, Russian was widely adopted in the Soviet empire by minority groups like the
Moldovans, whose own language was far less prestigious and valuable in economic and political
markets. Because of this, languages like Moldovan, although promoted as national languages,
lost significant ground, both functionally and in sheer numbers of speakers, to Russian.

In essence, Moldovan was always the basilect, i.e. the low-status, intimate language, in a
diglossic relationship with dominant Russian. This low prestige is also apparent in the attitudes
of those who identify with a standard Romanian linguistic identity toward the Moldovan dialect
and Russian borrowings. In conclusion, the notion of a separate Moldovan language is
sustainable only based on the criterion of social group distinctions, since low status is not a
justification for language maintenance in a contemporary, democratic society.

111
A Survey of Identity and Language Use

In this study, I examine three self-identified L1 groups: Romanian, Moldovan, and Russian. The
other linguistic minorities, such as Ukrainian, Gagauz, and Bulgarian, exist in much smaller
numbers in Moldova and primarily in the rural areas. The L1 Romanian and Russian groups are
distinguished by all three criteria related to language identification. They are structurally
different, associated with distinct ethno-national social groups, and highly valued in their
respective linguistic markets. The inclusion of Moldovan as a separate L1 is based on the
association of many Moldovans with this linguistic identity, presumably based on social group
distinctions. I will also consider these three languages as L2s. Although many Moldovans are
bilingual, their bilingualism varies in degree. Most, when pressed, will associate more strongly
with one national language than with another. For this reason, even when a bilingual learned
both languages from childhood and is fairly balanced in competence, I will consider the language
with strongest group associations the L1 and the other an L2.

Finally, I will be studying the role of L2 English in Moldova. The English language is widely
seen as a source of economic and cultural development for the struggling peoples of eastern
Europe. This belief has been criticized as imperialistic and deceptive, benefitting the interests of
Western powers (Phillipson 1992, Phillipson and Skutnabb-Kangas 1997:39-40). Even given
this role, belief in the power and status of English is apparent in Moldova, driven by the
perceived and, thereby, actual necessity of English competence for access to the meager
economic opportunities in the small private sector (Ciscel 2002b). Many urban Moldovans are
more motivated to learn L2 English than the local L2 of their neighbors. In other words, caught
between the old empire of the East and the new empire of the West, Moldovans scramble to
determine not only their own identity, but how outside identities will influence them in the
process. Ironically, the East-West struggle in Moldova has been a stalemate, producing
disruptively balanced numbers of Russian, Moldovan, and Romanian social and linguistic
identities.

In a survey conducted in 2001, 124 respondents at four educational institutions and two private
companies, all in the capital city, provided information about their language uses and attitudes.
The ages of respondents ranged from 15 to 46 with a mean of 21.3. The lower-end skew
apparent in these age figures is attributed to the fact that most respondents were university
students. The sample is hardly representative of Moldovans as a whole, even though it does
capture the character of the well-educated, professionally oriented urbanites. The data appear in
Table II.

112
 Table II.
Cross-tabulation of self-reported national identity and L1 in survey subjects

Nat ID \ L1 Moldovan Romanian Russian TOTAL

Moldovan 29 (4mx) 39 (2mx) 14 (11mx) 82 (17mx)

Romanian 17 17

Russian 1 (1mx) 1 (1mx) 13 (7mx) 15 (9mx)

Ukrainian 7 (4mx) 7 (4mx)

Gagauz 1 (1mx) 1 (1mx)

None 1 1 2

TOTAL 30 (5mx) 58 (3mx) 36 (23mx) 124 (31mx)

About two-thirds of respondents chose a Moldovan national identity. The remainder are
concentrated in groups of Romanians with L1 Romanian and Russians or other minorities with
L1 Russian. But, exceptional outliers also appear. These include two self-described Russians,
one with L1 Moldovan and the other with L1 Romanian. Similarly, they are both from mixed
families, as indicated by the number in parentheses with ‘mx.’ However, the majority of mixed
families, by almost three to one, produced L1 Russian speakers. Also, two respondents chose not
to identify with any nationality at all.

As seen above, self-reported national identity categories do not necessarily reflect the complexity
of multiple social identity categories available to the individual (for further discussion see Safran
1999, McNamara 1997). For this reason, I developed an algorithm to determine an identity score
based on responses to several items on the survey questionnaire, including self-reported location,
kinship, and attitude factors. Each factor was placed on a scale from -2 (very Russian) to 2 (very
Romanian) and then integrated into the composite identity score as follows: 30% location of
birth and childhood, 20% linguistic kinship (L1s of parents), 40% political attitude related to a
series of issues in Moldova, and 10% by the language the questionnaire was filled out in.
Further details of this calculation can be found in Ciscel (2002a).

The overall social identity score for each respondent fell in the range between -2 and 2. In order
to establish categories, this range is divided into four equal parts: scores 1 to 2 are extremely pro-
Romanian (XRO), scores 0 to 1 are moderately pro-Romanian (MRO), scores -1 to 0 are
moderately pro-Russian (MRU), and scores -2 to -1 extremely pro-Russian (XRU). This
approach neatly separates the more extreme identity stances from the centralized, multi-cultural
population in the Moldovan context. It further separates the middle into Slavic and Latinate
oriented branches. The counts of social identity score and L1 appear in Table III.

Although artificial and, to a certain extent, arbitrary, the identity scores and categories both

113
reflect the complexity of the phenomenon and more convincingly represent the range of political
ideologies available in the national context. The extremely pro-Romanian group (>1 XRO)
primarily contains L1 Romanian speakers. Furthermore, all members are from unmixed families
and tend to hold more extreme political views. The moderately pro-Romanian group (0>1 MRO)
contains a balance of L1 Moldovan and Romanian, including many of those from mixed families
in these L1 groups. The moderately pro-Russian group (-1<0 MRU) is more problematic. It is
made up of 23 L1 Russian speakers, many from mixed families, and seven L1 Moldovan
speakers. To reflect the L1 difference, this group is further divided into two groups for analysis
in this study. The seven L1 Moldovan speakers are categorized as assimilated, or russified,
Romanians (ARO), while the 23 L1 Russian speakers remain in the MRU category. Finally, the
thirteen respondents who fall into the extremely pro-Russian group (<-1 XRU) are less likely to
come from a mixed family and tend to hold more extreme political views. A separate group of
L1 Russian speakers with an identity score greater than zero does not occur in the data. In sum, I
have proposed five categories of social identity (the four in Table III, plus the extra ARO group
discussed above) that both reflect the backgrounds and attitudes of the respondents and divide
them into salient identity groups. The categories, from most pro-Romanian to most pro-Russian,
are (1) XRO, (2) MRO, (3) ARO, (4) MRU, and (5) XRU. The categories also subsume L1
identities, since a chi-square of the data in Table III indicates significant correlation between
identity and L1 groups (176.193, dF=12, p=0.000).

Table III.
Cross-tabulation of identity ranges and L1

ID \ L1 Romanian Moldovan Russian Ukrainian TOTAL

>1 XRO 34 1 35

0<1 MRO 24 (3mx) 22 (3mx) 46 (6mx)

-1<0 MRU 7 (2mx) 23 (19mx) 30 (21mx)

<-1 XRU 12 (4mx) 1 13 (4mx)

TOTAL 58 (3mx) 30 (5mx) 35 (23mx) 1 124 (31mx)

This categorization has both weaknesses and strengths. The greatest weakness is that the
categories are artificial, in that they do not reflect any one particular membership choice actually
made by the respondents. Insofar as measurement of such an actual choice is even possible,
however, the artificial score is composed of several different scores that do reflect the
respondents’ actual positions, at least as reported in the context of the survey. By balancing
several related measures of attitude and identity, the score, although artificial, is able to
triangulate and estimate a closer approximation of an actual, holistic identity than a single item,
such as nationality. Therefore, the weakness of artificiality can also be interpreted as a strength
in that the score is a composite.

Even so, the particular weighting of the aspects and the division of final scores into categories

114
can be criticized as arbitrary. Certainly, these processes were arbitrary. However, the resulting
categories appear to represent well the salient categories of group membership on the ground in
Moldova. For example, as discussed, extremist views at each end of the Romanian-Russian
identity spectrum tend to be occupied by small, powerful groups of relative elites. This dynamic
is captured in the scores. Also, the L1 and mixed family groups are neatly divided along a
continuum by the scoring algorithm. Finally, the projection of identity categories along this
continuum captures the individual variation and apparent ease of shifting among groups by
adjusting a score up or down a few points in one direction or the other. In sum, although
imperfect and likely, at times, inaccurate for particular individuals, the composite identity scores
and categories proposed here are more powerful variables than national or linguistic identity
alone. As such, their use in the study is justified.

The pattern of language use reported across the 124 subjects in the survey appears in Figure 1.
The data presented here involve a composite score based on responses to a series of questions
about frequency of language use in various social contexts and with different interlocutors. The
points on the graph represent the degree of use of one language compared to another across the
established social identity groups. For instance, the XRO group uses its L1 Romanian just more
than 50% (1 degree) more often than L2 Russian. The same group uses L2 Russian and English
equally often (0). Also, the XRU group uses L2 Romanian less often than English (-0.25
degree). The differences among identity groups for both L1 to local L2 and local L2 to English
are highly significant (p=0.000 for each, based on one-way ANOVAs).

Figure 1: Differences in frequency of use of L1 to local L2

115
The general patterns reveal that members of the extremist XRO and XRU categories use their
respective L1s much more often than the local L2, as would seem logical. In addition, they tend
to use the local L2 and English equally often, despite the non-local status of English. Notably,
the XRU members are more extreme than those in XRO, whose use is actually very similar to
that of the MRU category on the L1 Russian side. The members of the MRO category use their
L2 Russian fairly often, although not as often as their L1 Romanian. Correspondingly, they use
L2 Russian much more often than English. This tendency is even more extreme in the members
of the small ARO category, where Romanian and Russian use are almost balanced and English is
used much less often than L2 Russian.

During the fieldwork, I had greater access to three groups of students at the State University of
Moldova than to other subjects, because I had volunteered to teach an essay writing class once a
week to each group during the Fall Semester of 2001. Because this was an optional class without
a grade, attendance was low and sporadic. However, a few students in each of the three groups
attended regularly, allowing me to get to know them well. Six of them also volunteered to do an
individual interview and to fill out an additional questionnaire about their language proficiency.
The stories of these six students of English translation flesh out the skeletal statistics of the
broader survey. One of these ethnography subjects will be presented here in detail in order to
illustrate the complexity of individual experience suggested by the above quantified data.

Lidia (a pseudonym) was categorized as belonging to the MRO group in this study. Her identity
score was 0.467, based primarily on her claims to Moldovan rather than Romanian national and
linguistic identity. She was raised in the capital city of Moldova, but with parents who she
reports to be L1 Moldovan speakers. Her access to and acquisition of L2 Russian were earlier
and more complete than for many of the XRO subjects. In the second year of school, at age 8,
she began formal lessons in Russian, which continued throughout the remaining ten years in
school, averaging three hours per week. She also reported using some Russian from an early age
with neighbors and speaking it often with schoolmates after the age of 8. Even so, she reported
using Russian regularly with only about ten percent of friends and family, primarily with
neighbors and a few distant relatives. Lidia’s use of L2 Russian in the contexts reported on the
first questionnaire is more frequent than that of the other L1 Romanian subjects in the
ethnography. In addition, her Russian was reported and tested at the highest proficiency of any
L2 among the six subjects.

Lidia reported great ease with her L2 Russian in a range of functions and tested without errors on
a written cloze test in that language. In contrast, her L2 English was similar to that of the other
L1 Romanian subjects, quite proficient but not like her L1. Unlike her colleagues, Lidia began
learning English only at the university, at age 17. Like the others, she had had almost five years
of formal instruction in the L2, several hours per week. Her reported grades (around 9 out of 10)
were as good as any student who had studied English in school. She has apparently been quite
immersed in the language since entering the university. She reported beginning to use the
language informally with other students in the third year, when she was 19. As a result, her score
on the cloze test was also quite good for L2 English: 2.7/3.0. Having learned L2 Russian early
and grown up essentially bilingual, Lidia seems to take to learning L2s naturally. During the
interview, she said that learning Russian had indeed come ‘naturally,’ because all around her, in
the street and even sometimes at home, people spoke it. She emphasized that learning Russian
‘had never been a problem.’ But, she also claimed that English had been easier than Russian.
Like her classmates, she said that the grammar of English was much easier and that she wished

116
she had more opportunities to speak it, to gain fluency. In other words, she saw lack of access as
the primary obstacle to her acquisition of English.

Despite her proficiency in both L2s, Lidia made consistent statements that reinforced her strong
connection to L1 Moldovan (Romanian). She claimed to dream primarily in Moldovan and
rejected outright any possibility of Russian as a second official language in Moldova. Although
she thinks learning Russian is a good idea for Moldovans, she does not think that it should be
mandatory. When asked about the recent attempts by the Communists to make Russian official,
she predicted correctly that the Romanian nationalists and the people in general would not
tolerate it. Indeed, for several months following the interview, the pro-Romanian Christian
Democratic party held daily protests against russification in the streets and in the courts,
ultimately winning in the latter. For Lidia, this was inevitable. But unlike many nationalists, she
remained positive about L1 Russian speakers in Moldova and about the need for Russian to
maintain a de facto leading role, even in the absence of de jure official status.

Given her proficiency in and frequent use of L2 Russian, Lidia’s rejection of official status for
the language seemed peculiar. Once during class, the topic of official Russian came up2. The
one L1 Russian speaker in the group argued that Russian needed to be made official, drawing an
unfortunate comparison to the role of French in France’s former colonies in Africa. More
surprising than her willingness to make such a statement in front of a group of L1 Romanian
colleagues was the simple acceptance of the idea of official status for Russian among these
colleagues. With only one exception, everyone present that day, including an XRO group
member, agreed that giving Russian official status would be acceptable or at least possible. Lidia
was the only one of the half dozen L1 Romanian speakers who rejected the idea outright, in front
of everyone. The various contradictions and hypocrisies wrapped up in this situation can only be
explained by idiosyncratic aspects of social and linguistic identity. It is perhaps because of the
precarious status of her L1, which she identifies as Moldovan rather than Romanian, that she was
so willing to defend her position. Similarly, her XRO colleague, with the weight of Romania and
its history behind her L1, likely did not feel that it was as necessary to defend the status of her L1
so virulently. Whatever the motives, it is clear from this episode that individuals often behave in
unpredictable ways with regard to identity. From a set of unpredictable practices, one can, at
times, identify patterns that reflect the spirit of a group, community, or nation.

Upon reflection, I should not have been shocked by the way the above scene played out. What
was shaken by the discourse was more my own sense of idealism and justice than my
impressions about how people interact concerning linguistic identity issues in Moldova. In fact,
the scene reinforced many stereotypes that I have drawn from daily life there and tried, as a
researcher, to resist. L1 Russian speakers often state their opinions directly, with the arrogance
of one bestowed with privilege and advantage. In a group, L1 Romanian speakers generally do
not resist. Those who do resist are sometimes marked, ironically, as arrogant or extremist by L1
Russian speakers. These are the mechanics of domination. I can imagine them working quite
well in Soviet times. However, the illustration above is only an artifact of those times, an artifact
that continues to surface at times, but one which is also often counterbalanced by L1 Romanian
nationalists, who are often more assertive than the one present in this interaction, and, at times,
even by moderate Moldovans like Lidia, who stand on the threshold of language shift to Russian
but refuse to enter.
2
This conversation was, unfortunately, not recorded. Therefore, the details of the interchange are based purely on
my recollection and notes taken at the end of the lesson.

117
Conclusions

In this paper, I have proposed a set of social identity categories to be used in the analysis of the
Moldovan context. Although the use of fixed, rigid identity categories contradicts the potentially
fluid and volatile nature of social identities, the five categories described here are intended as
approximations that will allow some control of the analysis of identity and language use
phenomena. Despite the fact that individual subjects are discussed as members of particular
categories, based on their identity scores, the multiple and complex character of these identities
and the freedom of individuals to deviate at times from the program of their category are also
assumed. These dynamic elements are captured by the representation of the range of identities as
a continuum and by the focus on the complex social and linguistic practices of individual
subjects. In sum, the categories established here are a convenience that facilitates analysis. The
results from these analyses should be understood as occurring within the context of variability
and individual differences that become evident with parallel, qualitative analyses. As such, the
categories are proposed as soft guideposts to the stories of language and identity practices that
are suggested by the brief introduction to Lidia.

Overall, the results show a pattern of language use that is deeply rooted in historical
developments and recent social changes. The identities represent present instantiations of that
history and those changes. As such, the patterns of reported language use illustrate the stalemate
in struggles for national and cultural identity today in the Republic of Moldova. The prognosis
for resolution of these issues is unclear, depending on both internal and external politics and,
perhaps more than anything else, on the economic and social dynamics in the everyday lives of
individual Moldovans like Lidia. The relative stability and moderation of the past decade
provides some hope for an eventual resolution of the crisis in linguistic identity. However,
persistent poverty and political extremism remain dangerous barriers to resolution of the crisis.
The ultimate outcomes are relevant not only to Moldovans but to members of emergent national
communities all over Europe and the world that are grappling with multilingualism and histories
of colonization, for example the Scottish in Britain, the Corsicans in France (Jaffe 1999), the
Catalan speakers of Spain (Woolard 1989), and countless others. For members of these
communities, like for the Moldovans, language and social identity are inextricably linked in the
experiences and challenges of national self-determination and individual self-realization.

References
Bruchis, Michael 1988. The USSR: Language and Realities. Boulder, Colorado: East European Monographs
Ciscel, Matthew 2002a. Language and Identity: L2 acquisition in post-Soviet Moldova. University of South
Carolina dissertation
Ciscel, Matthew 2002b. English in Moldova: A portrait of linguistic opportunism. World Englishes 21:403-419
Crowther, William 1996. Nationalism and political transformation in Moldova. Studies in Moldovan, ed. by D.
Dyer, 31-52. Boulder: East European Monographs
Dyer, Donald 1999. The Romanian Dialect of Moldova. Lewiston: Edwin Mellon
Fishman, Joshua 1973. Language and Nationalism: Two integrative essays. Rowley: Newbury House
Gordon, Ellen 1993. Legislating Identity: Language, citizenship, and education in Lithuania. University of
Michigan dissertation
Jaffe, Alexandra 1999. Ideologies in Action: Language politics on Corsica. Berlin: Mouton de Gruyter
Karasik, Theodore, ed. 1992. USSR Facts and Figures Annual, Volume 17. New York: Academic International
Press
King, Charles 1999. The Moldovans: Romania, Russia, and the Politics of Culture. Stanford: Hoover Institution
Press

118
Korletianu, N.G. 1979. V.I. Lenin i razvitie natsionalnyx iazykov [Lenin and the development of national
languages]. Voprosy Iazykoznaniia 28:3-10
Laitin, David 1998. Identity in Formation: The Russian-speaking populations in the near abroad. Ithaca: Cornell
University Press
McNamara, Tim 1997. What do we mean by social identity? Competing frameworks, competing discourses.
TESOL Quarterly 31:561-7
Miller, Jennifer 1999. Becoming audible: Social identity and second language use. Journal of Intercultural Studies
20:149-65
Phillipson, Robert 1992. Linguistic Imperialism. Oxford: Oxford University Press
Phillipson, Robert, and Tove Skutnabb-Kangas 1997. Linguistic human rights and English in Europe. World
Englishes 16:27-43
Safran, William 1999. Identity. Handbook of Language and Ethnic Identity, ed. by J. Fishman, 77-93. Oxford:
Oxford University Press
Sonntag, Selma 1995. Elite competition and official language movements. Power and Inequality in Language
Education, ed. by James Tollefson, 91-111. Cambridge: Cambridge University Press
Woolard, Kathryn 1989. Double Talk: Bilingualism and the politics of ethnicity in Catalonia. Stanford: Stanford
University Press

119