Persian phonology

The status of diphthongs in Persian is disputed.[4][5] Some authors list /ei̯, ou̯, ɒi̯, oi̯, ui̯/,[4] others list only /ei̯/ and /ou̯/, but some do not recognize diphthongs in Persian at all.[4][5] A major factor that complicates the matter is the change of two classical and pre-classical Persian diphthongs: /ai̯/ > /ei̯/ and /au̯/ > /ou̯/. This shift occurred in Iran but not in some modern varieties (particularly of Afghanistan).[4] Morphological analysis also supports the view that the alleged Persian diphthongs are combinations of the vowels with /j/ and /w/.[5]

The Persian orthography does not distinguish between the diphthongs and the consonants /j/ and /w/; that is, they both are written with ی and و respectively.

/ou̯/ becomes [oː] in the colloquial Tehran dialect but is preserved in other Western dialects and standard Iranian Persian.

For Western Persian:

The variety of Afghanistan has also preserved these two Classic Persian vowels:

In the modern Persian alphabet, the short vowels /e/, /o/, /æ/ are usually left unwritten, as is normally done in the Arabic alphabet. (See Arabic phonology § Vowels.)

Early New Persian inherited from Middle Persian eight vowels: three short i, a, u and five long ī, ē, ā, ō, ū (in IPA: /i a u/ and /iː eː aː oː uː/). It is likely that this system passed into the common Persian era from a purely quantitative system into one where the short vowels differed from their long counterparts also in quality: i > [ɪ]; u > [ʊ]; ā > [ɑː]. These quality contrasts have in modern Persian varieties become the main distinction between the two sets of vowels.[6]

The inherited eight-vowel inventory is retained without major upheaval in Dari, which also preserves quantitative distinctions.[7]

In Western Persian, two of the vowel contrasts have been lost: those between the tense mid and close vowels. Thus ē, ī have merged as [iː], while ō, ū have merged as [uː]. In addition, the lax close vowels have been lowered: i > [e], u > [o]; this vowel change also happened in Dari. The lax open vowel has become fronted: a > [æ], and in word-final position further raised to [e]. Modern Iranian Persian does not feature distinctive vowel length.[8]

In both varieties ā is more or less labialized, as well as raised in Dari. Dari ō is also somewhat fronted.[7]

Tajiki has also lost two of the vowel contrasts, but differently from Western Persian: here the tense/lax contrast among the close vowels has been eliminated. That is, i, ī have merged as /ɪ/, and u, ū have merged as /ʊ/. The other tense back vowels have shifted as well. Mid ō has shifted front: /ɵ/ or /ʉ/, a vowel usually romanized as ů. Open ā has been labialized and raised to an open-mid vowel /ɔ/.

Loanwords from Arabic generally undergo these changes as well.

The following chart summarizes the later shifts into modern Tajik, Dari, and Western Persian.[7][9]

Early New Persian	Dari	Tajiki	Western Persian	Example	Tajik	Romanization	English
/a/	/a/	/a/	/æ/	شب	шаб	šab	night
/aː/	/ɒ̝ː/	/ɔ/	/ɒː/	باد	бод	bâd	wind
/i/	/e/	/ɪ/	/e/	دل	дил	dil	heart
/iː/	/iː/		/iː/	شیر	шир	šîr	milk
/eː/	/e̞ː/	/e̞/		شیر	шер	šer	lion
/aj/	/aj/	/aj/	/ej/	کی	кай	kay	when
/u/	/o/	/ʊ/	/o/	گل	гул	gul	flower
/uː/	/uː/		/uː/	نور	нур	nûr	light
/oː/	/ö̞ː/	/ɵ/		روز	рӯз	roz	day
/aw/	/aw/	/av/	/ow/	نو	нав	naw	new

Alveolar stops /t/ and /d/ are either apical alveolar or laminal denti-alveolar. The voiceless obstruents /p, t, t͡ʃ, k/ are aspirated much like their English counterparts: they become aspirated when they begin a syllable, though aspiration is not contrastive.[13] The Persian language does not have syllable-initial consonant clusters (see below), so unlike in English, /p, t, k/ are aspirated even following /s/, as in هستم /ˈhæstæm/ ('I exist').[14] They are also aspirated at the end of syllables, although not as strongly.

The velar stops /k, ɡ/ are palatalized before front vowels or at the end of a syllable.

In Classical Persian, the uvular consonants غ and ق denoted the original Arabic phonemes, the fricative [ʁ] and the plosive [q], respectively. In modern Tehrani Persian (which is used in the Iranian mass media, both colloquial and standard), there is no difference in the pronunciation of غ and ق. The actual realisation is usually that of a voiced stop [ɢ], but a voiced fricative [ɣ]~[ʁ] is common intervocalically. The classic pronunciations of غ and ق are preserved in the eastern varieties, Dari and Tajiki, as well as in the southern varieties (e.g. Zoroastrian Dari language and other Central / Central Plateau or Kermanic languages).

Some Iranian speakers show a similar merger of ج and ژ, such that [d͡ʒ] alternates with [ʒ], with the latter being restricted to intervocalic position.

Some speakers front /h/ to a voiceless palatal fricative [ç] in the vicinity of /i/, especially in syllable-final position. The velar/uvular fricatives are never fronted in such a way.

The /r/ is more heavily trilled at the beginning of a word than in the middle or at the end;[13] a geminate trill [rː] occurs word-medially, especially in loanwords of Arabic origin. An approximant [ɹ] also occurs as an allophone of /r/ before /t, d, s, z, ʃ, l, ʒ/; [ɹ] is sometimes in free variation with [r] in these and other positions, such that فارسی ('Persian') is pronounced [fɒːɹˈsiː] or [fɒːr̆ˈsiː] and سقرلات ('scarlet') [sæɣeɹˈlɒːt] or [sæɣerˈlɒːt]. /r/ is sometimes realized as a long approximant [ɹː].

The velar nasal [ŋ] is an allophone of /n/ before /k, ɡ/, and the uvular nasal [ɴ] before /q/.

/f, k, s, ʃ, x/ may be voiced to, respectively, [v, ɡ, z, ʒ, ɣ] before voiced consonants; /n/ may be bilabial [m] before bilabial consonants. Also /b/ may in some cases change into [β], or even [v]; for example باز ('open') may be pronounced [bɒːz] as well as [βɒːz] or [vɒːz] and/or [vɒː], colloquially.

The pronunciation of و [w] in Classical Persian shifted to [v] in Iranian Persian and Tajik, but is retained in Dari. In modern Persian [w] may be lost if preceded by a consonant and followed by a vowel in one whole syllable, e.g. خواب /xwɒb/ ~ [xɒb] 'sleep', as Persian has no syllable-initial consonant clusters (see below).

In standard Iranian Persian, the consonants /ʁ/ and /ɢ/ are pronounced identically.

Consonants, including /ʔ/ and /h/, can be geminated, often in words from Arabic. This is represented in the IPA either by doubling the consonant, سیّد саййид [sejˈjed], or with the length marker ⟨ː⟩, [seˈjːed].[15][16]

Syllables may be structured as (C)(S)V(S)(C(C)).[13][17]

Persian syllable structure consists of an optional syllable onset, consisting of one consonant; an obligatory syllable nucleus, consisting of a vowel optionally preceded by and/or followed by a semivowel; and an optional syllable coda, consisting of one or two consonants. The following restrictions apply:

• Onset
  • Consonant (C): Can be any consonant. (Onset is composed only of one consonant; consonant clusters are only found in loanwords, sometimes an epenthetic /æ/ is inserted between consonants.)
• Nucleus
  • Semivowel (S)
  • Vowel (V)
  • Semivowel (S)
• Coda
  • First consonant (C): Can be any consonant.
  • Second consonant (C): Can also be any consonant (mostly /d/, /k/, /s/, /t/, & /z/).