This paper is a study of the phonological system of Russian (R) as spoken today by Crimean Tatars (CT), a small ethnic group whose native language is from the Turkic language family, whose dialect has important social ramifications in current Crimean society. Past research published on CT has focused primarily on their history and political standing. Where language is mentioned in a contemporary context, it is to call for expansion of CT-language education or to note or bemoan the widespread use of R by younger CT, as a result of their R-language education. None of this research, however, has delved into that dialect of R unique to CT or acknowledged its significance.
I assert that CT, despite their mastery of the R lexicon, grammar, and syntax, retain a distinct accent in Russian. Further, I contend that this Crimean Tatar Russian (CT-R), not the actual CT language, is currently the linguistic marker of CT ethnic identity. Exiled from Crimea by the Soviets in 1944 for supposed collaboration with the Nazis, CT were allowed back to Crimea only in 1989. In the interim most CT literature was banned, the number of CT-language publications was sharply reduced, and the CT-language was heavily Russified. Most important in forming CT-R, however, was the Soviet regime's severe restriction of formal education in CT-language. As a result, CT-R became the de-facto lingua franca of CT.
From 1996 to 1998 I interviewed 100 Crimean Tatars in Crimea. I attempted to elicit unmonitored, spontaneous, informal speech so my informants would reveal more of the sounds and phonological rules in their representation of the L2 sound system than they would in a more formal situation. To gain a representative sample of CT-R I asked informants biographical questions and to tell me stories of their families' deportation. As a result, informants concentrated more on content than delivery, and I was thus able to generate repetition of various phrases for comparison among samples.
Since second language performance depends on a number of social and individual influences, in addition to linguistic factors, I relied on Sharypova's sociolinguistic approach, considering significant extra-linguistic factors, such as age and education, to help interpret the raw data I collected. I approached the problem of establishing the phonetic content of CT-R using the theory of "interlanguage," whose chief proponents include Selinker and Ellis. Interlanguage is an L2 learner's L2 grammar, which differs from standard L2 grammar due to L1 interference and the misapplication of L2 rules. Normally, L2 speakers' interlanguage approaches nearer and nearer to the standard L2. My research, however, shows that phonologically, this is not so with CT-R speakers.
According to my research, age plays the most important role in CT-R, with both expected and unexpected results. The older generation, as expected, retains the most non-standard features in their CT-R. Since the majority of CT who grew up in exile were exposed to CT in the home from 0-3 years of age, they had sufficient exposure and interaction to learn CT motor skills, but not enough to acquire full grammar and lexicon. This is one reason why, the middle and younger CT generations, retain several features of CT-R, rather than approaching standard R pronunciation. Some examples of regular features of CT-R pronunciation:
1) Twenty percent of informants show overcompensation for lack of CT phonemically soft consonants by softening consonants, as if the consonants were phonemically soft, e.g., Ukrainu [n'] to [t'] vyslali [l'a] umirali [l''i] xotja [t''a] dele [d''e]
2) Seventy two percent of my informants fail to distinguish R phones from CT phones whose features are similar but not the same, for example, thirty percent of my informants realize R velar /x/ as the closest CT phone, pharyngeal [x], regardless of environment, for example: xleb, nix, exat', naxoditsja.
3) CT rule interference: xodila [xLd'ilč] CT rule of [V +hi, +lax drop /__#]. While it is true that CT language lacks grammatical gender markers, this CT-R phenomenon is not limited to fem. sg. or just gender makers, e.g., s vami [s vam] vladeju [vlLdej] zakony [zLkon]. The CT-R rule is [V +lax drop /__#]. Another cause may be the fact that CT stress rules indicate end stress, so when R does not show end stress, perhaps the final vowel is no longer significant.
4) Fifty nine percent of my informants show deviation in stress placement, which is not distinctive for CT-R speakers as it is for R speakers. The CT final syllable stress pattern does not appear to transfer into CT-R, however. Instead, CT-R speakers show overcompensation, and 80% of deviations in stress consist of moving stress towards the front of the word. This stress shift has three different consequences:
a) most often, CT-R speakers apply V reduction based on the non-standard stress, e.g., jazyk [jazyk], opublikovat' [Lpub'l'Ikovčt], v Krymu [fkrymu], perevozli [p'Ir'Ivoz'l'I]
b) they sometimes apply SSR V reduction first, then reassign stress, e.g., vezli [v'Iz'l'i] vysylali [vysylal'I]
c) a third variant that occurs in CT-R is jazyk [jazyk][non-standard stress, no subsequent V reduction] takoj [takoj]
5) Forty five percent of informants apply vowel epenthesis/prothesis in accordance with CT intolerance of consonant clusters and CT processes of consonant cluster simplification applied to words borrowed into CT, although CT only allows for high-vowel epenthesis [V-hi insertion / #C__C]. Furthermore, CT-R speakers only simplify consonant clusters in the environment that CT simplification applies when words are borrowed into CT, namely to consonant clusters within a syllable and not in word-final position. This CT rule interference leads to utterances such as: trudno [tLrudnč] vybrasyvali [vyb'Irasyval'I] mne [m'In'e] klichki [kLl'ic'k'I] russkij [Lrus's'k'Ij] rabota [LrLbotč] rodina [Lrod'Inč]
6) Twenty percent of my informants sporadically substitute CT phonemes for R phonemes due to overdifferentiation of sounds that are phonemic in their L1 but non-phonemic in the L2. As a consequence of the distinction between palatal [k'] and pharyngeal [k] and palatal/non-palatal consonant harmony, CT-R speakers show the pharyngeal before back vowels or in words with a majority of back vowels, e.g., kto, kazakom [1st k = pharyngeal], and is particularly prevalent in derivatives of Krym, e.g., Krym, Krymskij, krymskotatarskij [1st k = pharyngeal]
7) Fifty seven percent of my informants show some deviation in their R intonation. CT-R speakers show two different types of intonational interference: In approximately half the deviations, CT-R speakers replace R intonation with CT intonation, while in the other half, speakers use R intonation with CT IC-center placement, for example: v konce koncov na ris priexali [CT-intonation III in place of R-1]; 1 na beregu chernogo morja [non-standard IC-center].
There is a widespread belief among CT that language is at the core of their ethnicity--indeed their nationality. Certainly, all of my informants named CT as their native language, regardless of their self-evaluated level of fluency, thus indicating their strong personal motivation to retain ethnic identity. Although the majority of CT espouse CT-language education and development as their goal, my research shows that currently CT-R functions as their linguistic in-group marker. The CT have, in practice, validated CT-R through the generations, and not CT, in this role.