Sounds of letters in Indian alphabets

Discussion in 'Etymology, History of languages, and Linguistics (EHL)' started by hadronic, Jul 7, 2013.

  1. hadronic Senior Member

    New York
    French - France

    I was wondering : how come all letters all have the same corresponding sounds in all Indian alphabets, across space and across time ?
    In opposition to that, the letters of the Latin alphabet have seen their associated sound change dramatically over the time and the spread of its use to denote different languages.
    Is it that Indian languages evolved very little (*kh remained kh in all languages, etc...), or was it that every time one phoneme changed, it got rewritten with the letter corresponding to that new sound (let's say [*kh] lost its aspiration, all kh letters got replaced by k letters)? But then, what if it morphed into a sound that wasn't existing before ? Like [*kh] > [x] ? Did this never happen ?
  2. asanga Member

    The phonological changes from Sanskrit to Hindi have been about as numerous as from Latin to Italian or Spanish. There has been simplification of consonant clusters, lengthening of preceding vowels to compensate, loss of final vowels, etc.

    Old Indo-Aryan karman > MIA kamma > NIA kām "action, work"
    hasta > hattha > hāth "hand"
    akṣi > akkhi > āñkh "eye"
    śata > saya > sau "one hundred"
    yāti > jāai > jāe "he goes"
    pṛcchati > pucchai > pucche "he asks"

    Almost all of these sound changes simplified Sanskrit phonology, rather than adding new sounds to the inventory, so Indic scripts already had the characters to represent them. There are only a handful of truly new consonants. MIA intervocalic ḍ and l became ṛ & ḷ in some NIA languages, and ph sometimes becomes f.

    phala > phala > phaḷ (Marathi) "fruit"
    phala > phala > fal (Hindi dialect)
    paṭhati > paḍhai > paṛhe "he reads"

    The consonants x, z, ɣ, f, q are also found in Arabic-Persian loanwords, but other than z and f these have not been universally assimilated. Furthermore, they entered the subcontinent with the Persian alphabet, which provided a ready model for the use of diacritics to represent new sounds.

    The vowels are a different story. Even in Hindi, which superficially has the same set of vowels as Sanskrit, the quality of the short vowels and the dipthongs have changed. Other NIA languages have completely new vowels, but other than the Sinhala signs for short e and o, influenced by Dravidian scripts, I don't think any NIA script represents them in writing.

    Other than the vowels, Indo-Aryan languages therefore had few problems accommodating phonetic spelling despite the many sound changes. India also had no equivalent to the prestige of the Latin alphabet and the associated prevalence of etymologically conservative spelling. Early Indian scripts like Brahmi and Kharoshti were first used to write MIA languages, and Sanskrit was only grudgingly written down centuries later. From the beginning the spelling reflected contemporary phonology rather than Sanskrit etymology. In his rock inscriptions, Ashoka spells his title Piyadasi "He who looks with kindness", rather than Sanskritized *P(r)iyada(r)śī. Modern Sanskrit loanwords do follow Sansrkit orthography, and in most NIA languages the phonetic values of the letters ṛ ṇ ṣ ḥ have changed to ri n ś h, respectively.

    Another reason for the prevalence of phonetic rather than etymological spelling could be that Sanskrit grammarians had a sophisticated understanding of articulatory phonetics long before writing became commonplace, and the varṇa (phoneme) has always taken precedence over the akṣara (grapheme).

    Śikṣā (which in modern terms includes phonetics, phonology, and morphophonology) was a required subsidiary science to study the Vedas. The Vedic pronunciation manuals accurately describe the organs of speech, place of articulation, manner of articulation, and phonation of every varṇa found in Vedic Sanskrit.

    For example, in Sanskrit ka-kāra ("the letter ka") doesn't refer to any particular grapheme, but to the phoneme which is kaṇṭhya (velar), spṛṣṭa (obstruent), śvāsa (pulmonic), vivāra (with open glottis), aghoṣa (voiceless), and alpa-prāṇa (without aspiration). Whether it's written as क, ক, ਕ, ક, etc. is irrelevant. In contrast, the Latin name refers to the grapheme "c", and not a phonetic value. Whether it's pronounced [k], , [tʃ], etc. in modern languages is irrelevant.

    So, the shape of the characters in Indian scripts has been free to change, while the phonetic value has not. In contrast, the shapes of the Latin letters has been very conservative (esp. the capitals), while the phonetic values have been free to change.
    Last edited: Jul 10, 2013
  3. Wolverine9 Senior Member

    American English
    ^ That was a very good explanation. Well done! One correction, though. The words for "fruit" are with short a rather than ā; so, it should be phala instead of phāla.
  4. hadronic Senior Member

    New York
    French - France
    Thanks a lot for this very complete answer !

Share This Page