Hindustani phonology

Schwa is a short vowel which vanishes to nothing at unstressed position. /ə/ is often realized more open than mid [ə], i.e. as near-open [ɐ].[2]

The open back vowel is transcribed in IPA by either [aː] or [ɑː].

Among the close vowels, what in Sanskrit are thought to have been primarily distinctions of vowel length (that is /i ~ iː/ and /u ~ uː/), have become in Hindustani distinctions of quality, or length accompanied by quality (that is, /ɪ ~ iː/ and /ʊ ~ uː/).[3] The historical opposition of length in the close vowels has been neutralized in word-final position, for example Sanskrit loans śakti (शक्ति – شَکتی 'energy') and vastu (वस्तु – وَستُو 'item') are /ʃəkti/ and /ʋəstu/, not */ʃəktɪ/ and */ʋəstʊ/.[4]

The vowel represented graphically as ऐ – اَے (romanized as ai) has been variously transcribed as [ɛː] or [æː].[5] Among sources for this article, Ohala (1999), pictured to the right, uses [ɛː], while Shapiro (2003:258) and Masica (1991:110) use [æː]. Furthermore, an eleventh vowel /æː/ is found in English loanwords, such as /bæːʈ/ ('bat').[6] Hereafter, ऐ – اَے (romanized as ai) will be represented as [ɛː] to distinguish it from /æː/, the latter. Despite this, the Hindustani vowel system is quite similar to that of English, in contrast to the consonants.

In addition, [ɛ] occurs as a conditioned allophone of /ə/ (schwa) in proximity to /ɦ/, if and only if the /ɦ/ is surrounded on both sides by two schwas.[4] and is realised as separate vowel. For example, in kahanā /kəɦ(ə)naː/ (कहना – کَہنا 'to say'), the /ɦ/ is surrounded on both sides by schwa, hence both the schwas will become fronted to short [ɛ], giving the pronunciation [kɛɦɛnaː]. Syncopation of phonemic middle schwa can further occur to give [kɛɦ.naː]. The fronting also occurs in word-final /ɦ/, presumably because a lone consonant carries an unpronounced schwa. Hence, kaha! /kəɦ(ə)/ (कह – کَہہ 'say!') becomes [kɛh] in actual pronunciation. However, the fronting of schwa does not occur in words with a schwa only on one side of the /ɦ/ such as kahānī /kəɦaːniː/ (कहानी – کَہانی 'a story') or bāhar /baːɦər/ (बाहर – باہَر 'outside').

The vowel [ɔ] occurs in proximity to /ɦ/ if the /ɦ/ is surrounded by one of the sides by a schwa and on other side by a round vowel. It differs from the vowel [ɔː] in that it is a short vowel. For example, in bahut /bəɦʊt/ the /ɦ/ is surrounded on one side by a schwa and a round vowel on the other side. One or both of the schwas will become [ɔ] giving the pronunciation [bɔɦɔt].

As in French and Portuguese, there are nasalized vowels in Hindustani. There is disagreement over the issue of the nature of nasalization (barring English-loaned /æ/ which is never nasalized[6]). Masica (1991:117) presents four differing viewpoints:

1. there are no *[ẽ] and *[õ], possibly because of the effect of nasalization on vowel quality;
2. there is phonemic nasalization of all vowels;
3. all vowel nasalization is predictable (i.e. allophonic);
4. Nasalized long vowel phonemes (/ɑ̃ː ĩː ũː ẽː ɛ̃ː õː ɔ̃ː/) occur word-finally and before voiceless stops; instances of nasalized short vowels ([ə̃ ɪ̃ ʊ̃]) and of nasalized long vowels before voiced stops (the latter, presumably because of a deleted nasal consonant) are allophonic.

Masica[12] supports this last view.

[v] and [w] are allophones in Hindustani. These are distinct phonemes in English, but both are allophones of the phoneme /ʋ/ in Hindustani (written ⟨व⟩ in Hindi or ⟨و⟩ in Urdu), including loanwords of Arabic and Persian origin. More specifically, they are conditional allophones, i.e. rules apply on whether ⟨व⟩ is pronounced as [v] or [w] depending on context. Native Hindi speakers pronounce ⟨व⟩ as [v] in vrat (व्रत – ورت, 'vow') and [w] in pakwān (पकवान – پکوان 'food dish'), treating them as a single phoneme and without being aware of the allophonic distinctions, though these are apparent to native English speakers. The rule is that the consonant is pronounced as semivowel [w] in onglide position, i.e. between an onset consonant and a following vowel.[21]

However, the allophone phenomenon becomes obvious when speakers switch languages. When speakers of other languages that have a distinction between [v] and [w] speak Hindustani, they might pronounce ⟨व و⟩ in vrat (व्रत – ورت) as [w], i.e. as [wrət] instead of the correct [vrət]. This results in an intelligibility problem because [wrət] can easily be confused for aurat (औरत عورت) [ˈɔːɾət], which means "woman" instead of "vow" in Hindustani. Similarly, Hindustani speakers might unconsciously apply their native allophony rules to English words, pronouncing war /wɔːɹ/ as [vɔːɹ] or advance /ədˈvɑːns/ as [ədˈwɑːns], which can result in intelligibility problems with native English speakers.[21]

In some situations, the allophony is non-conditional, i.e. the speaker can choose [v], [w] or an intermediate sound based on personal habit and preference, and still be perfectly intelligible, as long as the meaning is constant. This includes words such as advait (अद्वैत – ادویت) which can be pronounced equally correctly as [ədˈwɛːt] or [ədˈvɛːt].[21]

Loanwords from Sanskrit reintroduced /ɳ/ into formal Modern Standard Hindi. In casual speech it is usually replaced by /n/.[6] It does not occur initially and has a nasalized flap [ɽ̃] as a common allophone.[14]

Loanwords from Persian (including some words which Persian itself borrowed from Arabic or Turkish) introduced six consonants, /f, z, ʒ, q, x, ɣ/. Being Persian in origin, these are seen as a defining feature of Urdu, although these sounds officially exist in Hindi and modified Devanagari characters are available to represent them.[22][23] Among these, /f, z/, also found in English and Portuguese loanwords, are now considered well-established in Hindi; indeed, /f/ appears to be encroaching upon and replacing /pʰ/ even in native (non-Persian, non-English, non-Portuguese) Hindi words as well as many other Indian languages such as Bengali, Gujarati and Marathi, as happened in Greek with phi.[14] This /pʰ/ to /f/ shift also occasionally occurs in Urdu.[24] While [z] is a foreign sound, it is also natively found as an allophone of /s/ beside voiced consonants.

The other three Persian loans, /q, x, ɣ/, are still considered to fall under the domain of Urdu, and are also used by many Hindi speakers; however, some Hindi speakers assimilate these sounds to /k, kʰ, ɡ/ respectively.[22][25] The sibilant /ʃ/ is found in loanwords from all sources (English, Portuguese, Persian, Sanskrit) and is well-established.[6] The failure to maintain /f, z, ʃ/ by some Hindi speakers (often non-urban speakers who confuse them with /pʰ, dʒ, s/) is considered nonstandard.[22] Yet these same speakers, having a Sanskritic education, may hyperformally uphold /ɳ/ and [ʂ]. In contrast, for native speakers of Urdu, the maintenance of /f, z, ʃ/ is not commensurate with education and sophistication, but is characteristic of all social levels.[25] The sibliant /ʒ/ is very rare and is found in loanwords from Persian, Portuguese, and English and is considered to fall under the domain of Urdu and although it is officially present in Hindi, many speakers of Hindi assimilate it to /z/ or /dʒ/.

Being the main sources from which Hindustani draws its higher, learned terms– English, Sanskrit, Arabic, and to a lesser extent Persian provide loanwords with a rich array of consonant clusters. The introduction of these clusters into the language contravenes a historical tendency within its native core vocabulary to eliminate clusters through processes such as cluster reduction and epenthesis.[26] Schmidt (2003:293) lists distinctively Sanskrit/Hindi biconsonantal clusters of initial /kr, kʃ, st, sʋ, ʃr, sn, nj/ and final /tʋ, ʃʋ, nj, lj, rʋ, dʒj, rj/, and distinctively Perso-Arabic/Urdu biconsonantal clusters of final /ft, rf, mt, mr, ms, kl, tl, bl, sl, tm, lm, ɦm, ɦr/.