Phoneme /i/ in Turkish and Kazakh

Introduction

This study applies to the /i/ status in Turkish and Kazakh in the field of phonology and phonetics. Turkish and Kazakh are both Turkic languages with a common origin. They arise from Old Turkic. The Old Turkic phonemes, arguably, prolong to exist in the modern Turkic languages such as Turkish and Kazakh. Also, their acoustic realizations began to differ from one another. What is the reason of this difference? This study endeavors to explain the phonological nature of /i/ in Turkish and Kazakh and to describe reason of the acoustic difference in both languages.

It is necessary to specified the phoneme and speech sound notions. Carr insists on the existence of such type of relations between phoneme and speech sound that are typical for a type and its token. He says ^{[1, p. 404]}: “Phonemes are often taken to be types, and specific speech sounds uttered on specific occasions to be their tokens”. According to Trask, “type: a single abstract linguistic object, such as a phoneme; token: a single pronunciation of a linguistic form by a particular individual on a particular occasion” ^{[2, pp. 355, 364 (reference by [1, p. 404]})].

This study deals with the /i/ phoneme. Determination of the /i/ status is befogged by the fact that according to the observation made by me in Turkistan (a city in Kazakhstan), Turks can understand the Kazakh words and Kazakhs can get sense of the Turkish words including [i] (such as bir‘one’, biz ‘we’ etc.) in spite of the different pronunciations of the Turkish [i] and Kazakh [i]. This makes the idea of the common phonemes’ existence in Turkish and Kazakh sensible. It is possible thanks to the common origin of Turkish and Kazakh arisen from the Old Turkic language. The phonological structure of Old Turkic has been explored by Radloff ^[3], Clauson ^[4], Tekin ^[5], Róna-Tas ^[6], Guzev ^[7], Avrutina ^[8] and other researchers. According to some researchers the vowel system of Old Turkic includes 8 phonemes: /a/, /ɨ/, /o/, /u/, /ɛ/, /i/, /œ/, /y/ ^{[9, p. 23; 10, p. 11]}. According to another point of view there are 9 phonemes in the vowel system of Old Turkic: /a/, /ɨ/, /o/, /u/, /æ/, /ɛ/, /i/, /œ/, /y/ ^{[11, p. 45; 8, p. 81]}. The letter used for /i/ in Old Turkic also represents /ɨ/; some of researches ^{[12, p. 22; 13, p. 89 (references by [8, p. 72]})] insist on the condition of /i/ of being prior over /ɨ/; Pritsak even considers that /ɨ/ phoneme did not exist ^{[13, p. 89 (references by [8, p. 72]})]. This illustrates the strength assurance of the researches in the existence of /i/.

Phonetic and phonological systems of modern Turkish and Kazakh have been explored by Marchand ^[14], Lees ^[15], Selen ^[16], Demircan ^[17], Yavuz ^[18], Dzhunisbekov ^[19], Kenesbaiev ^[20], Kutalmış ^[21]. The biggest part of the most important researches on the Turkish and Kazakh phonological systems are done in 1960-1980’s by Russian, Turkish, Kazakh and German linguists. The vowel system of Turkish includes 8 phonemes: /a/, /ɛ/, /ɨ/, /i/, /o/, /œ/, /u/, /y/ ^{[16, p. 21; 22, p. 37]}. The acoustic realization of /i/ in Turkish is described as high, close front and unrounded vowel ^{[18, p. 42]}. The Kazakh vowel system covers 9 phonemes: /a/, /æ/, /e/, /ɯ/, /i/, /o/, /œ/, /u/, /y/ ^[19]. The realization of /i/ in Kazakh is described as high, close, central and unrounded vowel ^{[19, pp. 75-84]}. However, the origin of the modern /i/ phoneme in Turkish and Kazakh is still questionable. According to my observations (done in Turkistan, which is a city of Kazakhstan) different acoustic and articulation realizations of /i/ in Turkish and Kazakh do not bring any troubles to native speakers and Turks not knowing Kazakh understand the Kazakh words arisen from Old Turkic and per contra Kazakhs easily understand Turkish pronunciation of common words. This fact makes the idea of the /i/ proto-phoneme’s existence acceptable. This is the reason why Turks can get sense of the Kazakh words and per contra.

1. Characterisctics of /i/ in Turkish and Kazakh

The /i/ phoneme takes place both in Turkish and Kazakh. In both languages it is shown by the < i > letter but it has different acoustic and articulation realizations in the languages. The discussion of the nature of /i/ arises from the field of phonology and phonetics. To separate a phoneme from a sound is necessarily to analyze the topic clearly. Thus, the situation we have consists of two levels: 1) on the first level the Old Turkic /i/ phoneme is realized in Turkic languages in different ways; and 2) on the second level each of these languages prolongs to deal with the proto-phoneme which is realized according to the language acoustic peculiarities. Thus, the realization of /i/ in Kazakh differs from the one in Turkish. The problem which is necessary to be solved is ‘what makes phonemes of a common origin such as /i/ in Turkish and Kazakh to be realized different’.

The difference between the Kazakh and Turkish realization of /i/ is easy to hear. The Kazakh [i] is shorter than the Turkish one and is closer to the pronunciation of the /ɯ/ phoneme. For [i] in Turkish the 1st formant is 400 Hz, the 2nd formant is 2100-2400 Hz ^{[16, p. 45]}. For [i] in Kazakh the 1st formant is 300-500 Hz and the 2nd formant is 1700-2000 Hz ^{[19, pp. 62-63]}. According to the Turkish and Kazakh linguists, vocal organs move different during the production of the [i] allophones in Turkish and Kazakh. For Turkish it is described as [i] “is a high, unrounded and front vowel; during the production of the sound lips are opened narrowly, the front part of the tongue moves to the front, it touches teeth by both sides”.^[1]

The moment of the production of the Kazakh [i] is described as “almost the whole tongue body stays in the center of the oral cavity and moves front a little; the tip of the tongue leans on the low incisors, the middle part of the tongue rises to the back part of the hard palate, the back resonator is extended because of sides of the back and the root of the tongue are prominent; thus, the [i] vowel should be described as a central vowel moved front”.^[2]

Thus, there is an articulation difference between the realizations of the Turkish and the Kazakh phonemes: in Turkish “the front part of the tongue moves to the front”, in Kazakh “the center part of the tongue rises to the back part of the hard palate”. Tongue movement brings changes into the resonator characteristics: “… being gathered in the middle of the oral cavity tongue can divide it into two more or less identical parts; moved front tongue divides the resonator into a small part in front of and a larger behind; when tongue rises back the smallest part of the resonator will be behind and the biggest one in front; the formant data depends on this, /…/ F1 and F2 frequencies differ according to the tongue movements”.^[3] That is why this difference between the Turkish [i] and the Kazakh [i] is possible to be heard.

2. Are Turkish /i/ and Kazakh /i/ of Common Origin?

The /i/ phoneme in Turkish and in Kazakh is shown by the < i > letter in both languages. There is the letter that, arguably, is to represent /i/ in Old Turkic. As it is said above, some of the researches discuss on the issue of the /ɨ/ status, but there is no doubt in the existence of the /i/ phoneme in Old Turkish. Some examples of words arguably included the /i/ phoneme are shown in table N 1.

Some Old Turkic words still exist in the Turkish and the Kazakh vocabularies. If the Turkish version of a word includes /i/ phoneme we find /i/ phoneme in the Kazakh version and vice versa. Moreover, the /i/ phoneme is used in the Old Turkic versions: such is the illustration of the origin of the Turkic /i/ phoneme. Examples of words which are common for Turkish, Kazakh and Old Turkic and include the /i/ phoneme are shown in table N 2.

Table 1. Some of Old Turkic Words Including /i/ Phoneme

It is hard to estimate the date of the division of the proto-phoneme in the Turkish and the Kazakh versions. But it seems that at least 150 years ago this division had been done. We can find confirmation of this in some works of the 19th century. Some experts seem to be confused by the acoustical nature of the Kazakh [i] treating it as [ɯ] that shows existence of difference between the Kazakh [i] and the Russian [i] the pronunciation of which is close to the Turkish [i]. Laptev in its graduate thesis published in 1900 ^[24] besides the /ɛ/ (< e >), /i/ (< i >) and /ɯ/ (< ы >) phonemes detected the existence of their short versions shown by letters < ë >, < ï > and < ӹ >. According to the Kazakh phonetical system, all vowels are divided into front /ɛ/, central /æ/, /i/, /œ/, /y/, back /a/, /o/, /u/ and mixed^[4] /ɯ/ groups ^{[19, p. 56]}. According to the rule of palatal harmony in the Kazakh a word may either contain front and central vowels or back and mixed vowels ^{[19, p. 10]}. That is why it is possible to conclude that phonologically only two groups (the front-central vowels and the back-mixed vowels) make sense. Front-central vowels are not to be used in the same word with the back-mixed vowels. However, it is interesting that in Laptev’s work the < ӹ > which expresses the short mixed [ɯ] allophone takes place in words which include front vowels. Examples are shown at table N 3.

Table 2. Examples of Common Words in Turkish and Kazakh Including < i > and Comparison Them with Old Turkic Versions

Table 3. Words Examples Which Include <ӹ> According to Laptev

In spite of the rule of palatal harmony in the Turkic languages according to which the vowels in a word agree in palatality ^{[25, p. 64]} Laptev treats [i] sound as [ɯ]. Moreover, the existence of the palatal [k] which needs the palatal vowel ([ɛ], [i], [y], [œ] or [æ]) does not make situation clearer and we can find the kim word (“who”) in the text written as < кӹм > with < ӹ > expressing the postpalatal /ɯ/. Existence of doubt is clear because it is used the < кïм > form (with short [i]) in the same text. The same situation exists with the bir word (“one”), which is found as < бïр > [bir], < бӹр > [bɯr] and < бëр > [bɛr] in the text. This shows confusion of Laptev by treating /i/ - /ɯ/ phonemes in Kazakh. It is important to stress that Kazakh was not the native language of the researcher and he made his field study without any special equipment, only relaying on his hearing. It seems the reason of confusion rises from the acoustic peculiarities: F2 of the [ɯ] allophone in Russian is 1500 Hz ^{[27, p. 48]}; closeness of the Kazakh [i] (F2 of which is 1700-2000 Hz) and the Russian [ɯ] parameters made Laptev (the native language of whom was Russian) to use the new < ӹ > letter.

Laptev made his study in 1900 and had a problem with separation of the [i] - [ɯ] sounds. The difference between the [i] and [ɯ] allophones in Kazakh still is hard to be heard nowadays. The acoustic realizations of the Kazakh /i/ and /ɯ/ phonemes should be pronounced very rapid, that’s why to define the difference between them is a hard work for a person who does not speak Kazakh as a native language. Laptev’s native language was Russian. Thus, his confusion shows that at the time of his research the difference between acoustic realizations of the Russian [i] and the Kazakh [i] existed. For the Russian [i] values of the 1st and 2nd formants are almost the same as for the Turkish one: the 1st formant is 400 Hz and the 2nd formant is 2000-2400 Hz ^{[28, p. 135]}. Laptev’s treating the Kazakh [i] as [ɯ] shows that he treated the Kazakh [i] different from the Russian [i] which was his native language. Similarity between formant values of the Russian [i] and the Turkish [i] lets us suspect the existence of a difference between the Kazakh [i] and the Turkish [i] either.

3. Nature of Difference Between /i/ and /ɨ/

The feature that makes acoustic realizations of /i/ and /ɨ/ different, arises from the tongue movement. It refers to where in the oral tract (in the front or in the back of it) the tongue has the highest point ^{[18, p. 36]}. According to this the Turkish vowels are divided into front (/ɛ/, /i/, /œ/, /y/) and back (/a/, /ɨ/, /o/, /u/) vowels ^{[17, p. 62]}; articulately some researchers as Yavuz also single out central vowels in Turkish: /a/ ^{[18, p. 36-37]}. It seems that the right classification belongs to Selen ^{[16, p. 21]} that made complex analysis of the phonetical system of Turkish. According to her, the Turkish vowels divided into front [ɛ], [i], [œ], [y], central [ɨ] and back [a], [o], [u] ones. As it was mentioned above, Dzhunisbekov also singled out such a vowel group as ‘central’; his classification of the Kazakh vowels includes front [ɛ], central [æ], [i], [œ], [y], back [a], [o], [u] and mixed [ɯ] groups ^{[19, p. 56]}. As we see, the Turkish /i/ and the Kazakh /i/ are classified different: front vowel in Turkish but central vowel in Kazakh.

Pattern 1. Difference Between Movement of Tongue During Production of ‘Front’ Sounds in Turkish and Kazakh^[7]

Note: The tongue position in Turkish is indicated by a solid line, in Kazakh by a dotted line.

Indeed, there is an important issue making these phonemes different: the issue is the movement of tongue. Positions of tongue create different vowels. Thus, Shcherba’s study on the French vowels demonstrates the existence of not only the front and back vowels but also “back vowels moved front” ^{[29, p. 53]}. However, not only the way (front or back) of tongue movement but also the part of tongue that produces such a movement makes serious difference to acoustic peculiarity of a sound. In regards the /i/ phonemes in Turkish and Kazakh, it is said that at the moment of the [i] production, as it was mentioned above, in the Turkish “the front part of tongue moves front” ^{[16, p. 44]}; at the moment of the [i] production in the Kazakh “the middle part of tongue rises to the back part of the hard palate” ^{[19, p. 52]}. The more the position of tongue is in the front, the higher is the F2 frequency ^{[23, p. 176]}. This is illustrated by the Turkish and the Kazakh parameters (see: table N 4): the Turkish F2 is 2100-2400 Hz ^{[16, p. 45]}, the Kazakh F2 is 1700-2000 Hz ^{[19, pp. 62-63]}.

The divergence of the tongue movements during sounds production in Turkish and in Kazakh is easy to be define during the articulation of allophones of /k/. In Turkish the /k/ phoneme is shown by the < k > letter. Selen describes the acoustic characteristics of the [k] allophones as follows: “There are two production points of [k]. The first point is situated between the back part of tongue and the velum… The second point takes place between the front part of tongue and the palate. /…/ Palatal consonant is used with front vowels and velar consonant is used with back vowels.”^[8] This different movement of tongue is possible to be mentioned during articulation, for example, kel- (to come) – kal- (to stay); büyük (big) – ılık (warm); yelek (vest) – yalak (windbag). In Kazakh there are two letters which represent allophones of the /k/ phoneme: < к > for [k] (used with front-central vowels) and < қ > for [q] (used with back-mixed vowels), see Kenesbaiev’s statement: “…[k] is produced by coupling of the middle part of the tongue and the velar /…/ During the [q] production the back part of the tongue couples with the back of the velum”.^[9] For example, кел- [kɛl] (to come) – қал- [qal](to stay);тоқ[toq] (full) – төк[tœk] (goldfield). So in Turkish there is a couple of palatal / velar plosive allophones (for more information see ^{[16, p. 85-88]}) and in Kazakh we have a pair of velar / uvular stops (for more information see ^{[30, pp. 94-95]}). The front vowels and palatal consonants in Turkish are produced by the front part of the tongue; the front-central vowels and palatal-velar consonants in Kazakh are pronounced with the middle part of the tongue touching the back part of the palate or the velar. I showed the situation on pattern N 1.

Pattern 2. Difference Between Movement of Tongue During Production of ‘Back’ Sounds in Turkish and Kazakh

Note: The tongue position in Turkish is indicated by a solid line, in Kazakh by a dotted line.

The same divergence between tongue positions in the languages exists during the production of ‘back’ sounds. Thus, the central-back vowels and velar consonants in Turkish are pronounced with the middle part of tongue touching the back part of the palate or the velar; the mixed-back vowels and uvular consonants in Kazakh are produced by the back part of the tongue (schematically I showed this on pattern N 2).

Table 4. F1 and F2 Values of Turkish and Kazakh Wovels^[10]

It is clear that articulately system of ‘front’ sounds in Turkish is situated more front in the oral cavity then the system of the Kazakh ones. And on the contrary, points of production of the ‘back’ sounds in Kazakh are situated more back in the oral cavity then the Turkish ones. As Thomson supposes, there is such an issue as ‘articulate habits which determine development of particular muscles of speech organs ^{[31, p. 216 (reference by [23, p. 80]})]. Regarding this issue Zinder says: “From the anatomical and physiological point of view, there is no fundamental difference in the structure and functioning of the pronunciation organs in representatives of different races or nationalities. /…/ …foreign “accent” does not depend on inborn or race peculiarities of speech (to be more precise, pronunciation) organs but on the pronunciation habits.”^[11] Thus, we can say that there are special ‘pronunciation habits’ of Turkish and Kazakh peoples which lead to different acoustic realizations of the /i/ phoneme in these languages.

The actual reason for the difference between the Turkish and the Kazakh languages consists in the field of the issue that makes such ‘pronunciation habits’ existent. In other words, Turkish and Kazakh phonological systems are very close to one to another. In spite of the existence of common phonemes in Turkish and Kazakh, there are particular acoustic and articulation differences between them. It means that not only phonological peculiarities make language systems different, but languages with similar phonological systems can differ from one another by their articulation peculiarities. The actual problem is to explore the nature of the production of such articulation peculiarities, which experts name ‘pronunciation habits’. According to neuroplasticity each habit is reflected in the brain system ^[32]. History of Kazakh and Turks shows that these peoples had a particular ‘pronunciation habit’ and then they changed it. Our aim is to understand what makes peoples change their ‘pronunciation habits’. Only the brain knowledge can help linguists to understand it.

4. Conclusion

The realizations of the /i/ phoneme in Turkish and Kazakh differ from each other by the tongue movement: the Turkish [i] is pronounced by the front part of tongue moving front and the Kazakh [i] is realized in the speech with the middle part of tongue rising to the back of the palate. This movement makes the realizations of /i/ to sound different: the F1 of the Turkish [i] is 400 Hz and the F2 is 2100-2400 Hz; the F1 of the Kazakh [i] is 300-500 Hz and the F2 is 1700-2000 Hz. The analysis of the texts showed that this difference was existed at least 150 years ago. Moreover, this is typical not only for the /i/ realizations but for all acoustic system of both languages: in Turkish tongue moves further front then in Kazakh. However, this difference makes no sense to the Turkic-native speak people and they can easily understand the realization of /i/ with unwonted tongue movement. This is possible thanks to their common background: /i/ is one of their proto-phonemes.

^[1] “Dar, düz bir ön dil ünlüsüdür; sesin çıkışında dudaklar dar bir biçimde açıktır, dilin ön kısmı öne doğru kabarır iki yanı üst sıra dişlerin kenarlarına değer” ^{[16, p. 44]}.

^[2] “… основная масса языка занимает срединное положение в полости рта, несколько продвинута вперед. Кончик языка упирается в нижние резцы, средняя часть спинки языка приподнята по направлению к задней части твердого неба, края задней части и корня языка сильно выпуклы, что несколько расширяет задний резонатор. В общем, гласный [i] должен определяться как центральнорядный, продвинутый вперед” ^{[19, p. 52]}.

^[3] “… будучи сконцентрированным в середине полости рта, язык может разделить ее на две более или менее одинаковые части; продвинутый вперед, он разделяет резонатор на небольшую по объему часть впереди и бóльшую сзади; при оттягивании назад меньшая часть резонатора будет сзади, а бóльшая впереди; с этим связана и формантная картина, в которой /…/ частотное положение FI и FII зависит от того или иного положения языка” ^{[23, p. 196]}.

^[4] The ’mixed vowel’ term belongs to Trubetzkoi ^{[26, p. 113]} and means a vowel which occurs with the tongue rising not in the front and not in the back but in the middle of the mouth (more back than during a central vowel).

^[5] The Old Turkic examples are taken from ^{[5, pp. 129–166]}.

^[6] Transcription is given according to: ^{[18, pp. 22, 37]}.

^[7] Differences between movements of tongue showen on parttens N 1 and N 2 are noticed by myself by exploring the movements of oral organs during the articulation of assorted words.

^[8] “< k > ünsüzünün iki çıkarılma noktası vardır. Birinci nokta dil sırtının arka tarafıyla artdamak arasındadır… İkinci nokta, dil sırtının önüyle öndamak arasındadır. /…/ Öndamak ünsüzü ön dil, artdamak ünsüzüyse arka dil ünlülerinin eşliğinde gelir.” ^{[16, p. 85]}.

^[9] “...к образуется посредством плотной смычки средней части языка и неба /…/ При образовании қ задняя часть языка полностью смыкается с задней частью неба” ^{[20, pp. 294, 296]}.

^[10] Values of Turkish vowels are taken from: ^[16]. Values of Kazakh vowels are taken from: ^{[19, p. 66]}.

^[11] “С анатомо-физиологической точки зрения принципиальной разницы в строении и функционировании произносительных органов у представителей разных рас или национальностей нет. /…/ ...«акцент» иноязычных обуславливается их речевыми (точнее – произносительными) привычками, а не врожденными или расовыми особенностями органов произношения.” ^{[23, p. 80]}.