Hindi is not Sanskrit: Phonetics and Phonology

There is a weird developing cultural trend in some circles of treating Hindi as a sole inheritor of Sanskrit. Obviously, this is tied to the political centralisation around certain religious ideologies, which seem to have an obsession for a singular national language (invariably Hindi). But a lot of the arguments for this special status of Hindi are grounded in linguistic nonsense. I responded to one instance of this on Twitter, but it was a long thread that isn’t really nice to view on there so this blog post is a form of that.

First of all, I will say that I love Hindi. All of my research work is tied to it right now and I expect much of it will continue to be. I had a phase like many children of immigrants do where I was too shy/embarrassed to speak in Hindi and so my language skills atrophied. But I basically reversed that process in high school. So, I don’t have any agenda against Hindi, just annoyed by the misinformation.

Second, I’m only touching on phonetic and phonological differences. This is one tiny facet of a language and its speech community; a full overview of the differences is probably necessary but I thought to star with sounds since those are basic and also subject to a lot of misinformation (“Hindi has the most scientific alphabet”, “Hindi is written exactly like it sounds”, “Hindi sounds are identical to Sanskrit”, etc.) Now to the actual discussion.

Vowels

First off, the vowels. Hindi has gotten rid of Sanskrit’s phonemic length and replaced it with quality distinctions in monophthongs. Here’s the basic vowels compared between Hindi and Sanskrit.

Letter Hindi Sanskrit
/ɐ/ /ɐ/
/ä/ /äː/
/ʊ/ /u/
/u/ /uː/
/ɪ/ /i/
/i/ /iː/

Also note schwa deletion in Hindi. Where we Hindi-speakers pronounce भारत, आनंद, etc. as bʰārat, ānand, in Sanskrit all of the orthographic schwas were always spoken (=> bʰārata, ānanda) unless there was a virāma (=halant). Eastern IA is more conservative in keeping these.

The guṇa vowels ए and ओ were diphthongs in Vedic Sanskrit /ɐi̯ ɐu̯/, but were monophthongised by Classical Sanskrit /eː oː/. The latter (minus length distinction!) is how they are in Hindi. Hardly very conservative though, not like the Vedic system at all. The vr̥ddʰi vowels ऐ and औ were always diphthongs /äi̯ äu̯/ in Sanskrit. In Hindi we have reduced them to new monophthongs /ɛ ɔ/ that never existed in Sanskrit. The diphthong realisation is only retained before glides in MSH (e.g. भैया, कौवा). Bhojpuri keeps the diphthongs!

Now to the “problem” vowels: ऋ ॠ ऌ ॡ. Only the first is common in Sanskrit, but due to the historical r ~ l alternation a couple verb roots have ऌ in their paradigms. In Sanskrit these were syllabic consonants /r̩ r̩ː l̩ l̩ː/. Think of how Americans say their “r”s in “butter”, but without the tongue curled so far back. Eastern European languages also have similar sounds. Basically no Indian language retains this sound; Hindi calls them /ɾɪ, ɾi, lɾɪ, lɾi/, Gujarati has /ɾu …/, etc.

To sum up, the vowel system is quite changed from Sanskrit to Hindi, even if the way the vowel system is structured is the same in writing. Always keep in mind that writing != pronunciation! For Sanskrit, Devanagari corresponded 1-to-1 to speech, but it does not in Hindi.

Consonants

Now for consonants. There are fewer differences here. First, the lost nasal consonants ङ ञ ण. In Sanskrit these first two were nasal sounds at the place of articulation of the corresponding row in Devanagari, /ŋ ɲ/. We don’t do that in Hindi. In Hindi, we do not have the first one really; a few words like वाङमय ‘literature’ do have it though, and we invariably insert the /g/ there: /ʋäŋgmɐj/.1 Skt would call it /ʋäːŋmɐjɐ/. The second one we call /nj/. In Skt it’s a single nasal, touching the roof of the mouth. Finally, the third one is usually just the same as न. Some people do have a different sound (especially priests when chanting Sanskrit etc.) which is like ड़ but more nasal, transcribed /ɽ̃/. In Sanskrit the flapping quality of ड़ was not there, it was a curled nasal /ɳ/.

Now to the affricates च छ ज झ. Sanskrit might have had these as actual stops (fully blocking airflow) /c cʰ ɟ ɟʱ/ or affricates (with air) /t͡ɕ t͡ɕʰ d͡ʑ d͡ʑʱ/. Either way, these are palatal, with the body of the tongue touching the roof of the mouth. In Hindi they are totally different. It’s more of the tip area of the tongue, touch the part of the mouth right above the alveolar ridge. I transcribe these as /c͡ʃ c͡ʃʰ ɟ͡ʒ ɟ͡ʒʱ/ following Masicaa; they are less “pointy” than the English ch, j.

Now let’s look at फ. In Modern Standard Hindi of Delhi, many people have taken to pronouncing it as /f/ (like the f of English). And indeed, this is probably due to post-independence influence of English. Go to a village, they will say /pʰ/, प with a puff of air.

Our last single consonants to discuss are श and ष. In Sanskrit the former is palatal /ɕ/ (like च, in terms of how the tongue is shaped in the mouth) and the later is retroflex /ʂ/ (like ट). In Hindi, we have merged these to be both closer to the front of the mouth: /ʃ/.

Now for the conjunct ज्ञ. The form in Sanskrit is supposed to be /ɟɲ/ or /ɟ͡ʑɲ/, or in Devanagari ज + virāma + ञ. In Hindi we somehow got /gj/ ग + virāma + य! There is a lot of variation in how Indian languages pronounce this; the closest is in Dravidian langs.

Finally, क्ष is /kʂ/ in Sanskrit (क + virāma + ष). In Hindi people tend to say just श at the start of a word and क्श in the middle. We don’t have the tongue curled back like in ट. It’s extremely different from Sanskrit.

Other letters

The anusvāra (bindī) and candrabindu. In Hindi the latter nasalised vowels. The former is a nasal that assimilated to the following consonant. This is the same case as Sanskrit, except it gets weird with sonorants (sounds that allow air through). Basically, in words like संविधान or संस्कृत (व and स let air through unlike क or ड), Sanskrit would just have a nasal vowel, no actual “n” or “m” sound. Hindi behaves differently, calling them samvidhān and sanskrit.

Finally (bear with me), the visarga in words like दुःख and सामान्यतः was /h/ in Sanskrit. (The full situation is complicated so I won’t go into it.) This is different from ह because there is no vibration of the vocal cords. But in Hindi, both are the same /ɦ/.

Conclusion

To summarise, some Skt ~ Hindi examples.

Word Sanskrit Hindi
क्षत्रिय /kʂɐt̪ɾijɐ/ /ʃɐt̪ɾij(ɐ)/
ज्ञानी /ɟɲäːniː/ /gjäni/
संस्कृत /sɐ̃skr̩t̪/ /sɐnskɾɪt̪/
दुःख /d̪uhkʰɐ/ /d̪ʊkʰ/
मेष /mɐi̯ʂɐ/ /meʃ/

I won’t get much into other aspects of Sanskrit loaning in Hindi. I’ll just say that Hindi does not follow all of the rules of Sanskrit in this regard, specifically noting that we do not do proper sandhi (words like अंतर्रष्ट्रिय are misspellings).

The important points here are:

  1. Hindi is not “identical” to Sanskrit in any way. There are so many intervening developments that have totally changed all Indian languages since the time of Sanskrit. No use of Sanskrit words or styles can turn back the clock on this.
  2. It is silly to compare X language with Sanskrit and say it is closer than Y language. It is even sillier to say that this is a reason for X language to be prioritised. Languages serve as lingua franca when they are easy to use by a wide range of people not by how “close” they are to an imagined ideal. There are features that Bhojpuri conserves that Hindi doesn’t. I do not see anyone arguing to make Bhojpuri the national language. Same goes for Bengali (way more Skt vocab, less schwa deletion), Punjabi (proper ण), etc.

Footnotes

  1. It seems that some people do pronounce it [ʋäːŋmɐjɐ]. Either way, /ŋ/ is noenxistent as a phoneme in inherited Indo-Aryan vocabulary and marginal in other parts of the Hindi lexicon.