This article is about the phonology and phonetics of the Kyrgyz language.

Vowels

A formant chart showing the stem vowel space of Kyrgyz. From Washington (2007:10).

• Notes on vowel quality:
  • Kyrgyz vowel space is different in affixes and stems. Washington (2007) describes the former as more typical and more condensed.^[2]
  • In stem vowel space, the main difference between /e/ and /i/ is that the latter is more back. In affix vowel space, they can have the same backness, and differ by height.^[2]
• /a/ appears only in borrowings from Persian and is excluded from normal vowel harmony rules. In most dialects, its status as a vowel distinct from /ɑ/ is questionable. There is also a phonetic [a] which appears as a result of regressive assimilation of /ɑ/ before syllables with phonological front vowels, e.g. "àydöş" [ajdøʃ] 'sloping'.^[3][4]
• /i, y, u, e, ø, o/ are sometimes transcribed /ɪ, ʏ, ʊ, ɛ, œ, ɔ/.^[5]
• The sequence of any vowel and the consonant /z/ is pronounced as a long vowel with falling pitch.^[6]
• In colloquial speech, word-final vowels are dropped when the next word begins with a vowel.^[7]
• All vowels but /i/ may be both short and long. Long vowels are the result of historical elisions (e.g. compensatory lengthening) and contractions. For example, já "rain" < *yağ; bé "mare" (cf. Kazakh biye); too "mountain" < *tağ; dőlöt "wealth" < Arabic daulat; ulú "great" < *uluğ; elű"fifty" < *eļļiğ.

Consonants

• /n, l, r/ are alveolar, whereas /t, d, t͡s, s, z/ are dental.^[8]
  • the liquid /l/ is velarized [ɫ] in back vowel contexts.
• /ŋ, k, ɡ, x/ are velar, whereas /j/ is palatal.^[8]
  • /k, ɡ/ are palatal [c, ɟ] in words with front vowels, and uvular [q, ʁ] in words with back vowels.^[9]
    • Word-initial [c] is often voiced [ɟ].^[10]
    • In loanwords from Persian and Arabic, palatal [c, ɟ] are always followed by front vowels, whereas velar [k, ɡ] are always followed by back vowels, regardless of the vowel harmony.^[9]
    • Word-final and word-initial /k/ is voiced to [ɡ] when it is surrounded by vowels or the consonants /m, n, ŋ, l, r, j/.^[7]
• /f, v, t͡s, x/ occur only in foreign borrowings, mostly from Indo-European and Semitic.^[8]
• In colloquial speech:
  • /b/ is lenited to [w] after /l, r, j/ or between vowels.^[7]
  • /t͡ʃ/ is deaffricated to [ʃ] before voiceless consonants.^[7]
  • Intervocalic /s/ can be voiced to [z].^[7]
  • Word-final /z/ is often devoiced to [s].^[7]

Stress

Recent loanwords often retain their original stress.^[11]

Desonorisation and devoicing

In Kyrgyz, suffixes beginning with /n/ show desonorisation of the /n/ to [d] after consonants (including /j/), and devoicing to [t] after voiceless consonants; e.g. the definite accusative suffix -NI patterns like this: ķemeņi ('the boat'), aydı('the month'), tordu ('the net'), koldu ('the hand'), tañdı ('the dawn'), ķözdü ('the eye'), baştı ('the head').

Suffixes beginning with /l/ also show desonorisation and devoicing, though only after consonants of equal or lower sonority than /l/, e.g. the plural suffix -LAr patterns like this: ķemeļer ('boats'), aylar ('months'), torlor ('nets'), koldor ('hands'), tañdar ('dawns'), ķözdör ('eyes'), baştar ('heads'). Other /l/-initial suffixes, such as -LA, a denominal verbal suffix, and -LÚ, a denominal adjectival suffix, may surface either with /l/ or /d/ after /r/; e.g. тордо-/торло- ('to net/weave'), түрдүү/түрлүү ('various').

See Kyrgyz language#Case for more examples.

References

Bibliography

Further reading