prehistoric derivation
The separation of Proto-Indo-Iranian language into Avestan and Vedic Sanskrit is estimated, on linguistic grounds, to have occurred around or before 1800 BCE. The date of composition of the oldest hymns of the Rigveda is vague at best, generally estimated to roughly 1500 BCE.
Both Asko Parpola (1988) and J. P. Mallory (1998) place the locus of the division of Indo-Aryan from Iranian in the Bronze Age culture of the Bactria–Margiana Archaeological Complex (BMAC).
Parpola (1999) elaborates on the model and has "Proto-Rigvedic" Indo-Aryans intrude the BMAC around 1700 BCE. He assumes early Indo-Aryan presence in the Late Harappan horizon from about 1900 BCE and "Proto-Rigvedic" (Proto-Dardic) intrusion to Punjab as corresponding to the Gandhara grave culture from about 1700 BCE. According to this model, Rigvedic within the larger Indo-Aryan group is the direct ancestor of the Dardic languages.
The early Vedic Sanskrit language was far less homogenous, and it evolved over time into a more structured and homogeneous language. The language in the early Upanishads of Hinduism and the late Vedic literature approaches Classical Sanskrit, states Gombrich. The formalization of the late Vedic Sanskrit language into the Classical Sanskrit form is credited to Pāṇini, along with Patanjali's Mahabhasya and Katyayana's commentary that preceded Patanjali's work.
Chronology
According to Michael Witzel, five chronologically distinct strata can be identified within the Vedic language:
- Rigvedic - Many words in the Vedic Sanskrit of the Rigveda have cognates or direct correspondences with the ancient Avestan language, but these do not appear in post-Rigvedic Indian texts. The Rigveda must have been essentially complete by around the 12th century BCE. The pre-1200 BCE layers mark a gradual change in Vedic Sanskrit, but there is a disappearance of these archaic correspondences and linguistics in the post-Rigvedic period.
- Mantra language - This period includes both the mantra and prose language of the Atharvaveda (Paippalada and Shaunakiya), the Rigveda Khilani, the Samaveda Samhita, and the mantras of the Yajurveda. These texts are largely derived from the Rigveda but have undergone certain changes, both by linguistic change and by reinterpretation. For example, the more ancient injunctive verb system is no longer in use.
- Samhita prose - An important linguistic change is the disappearance of the injunctive, subjunctive, optative, imperative (the modi of the aorist). The new innovation in Vedic Sanskrit appears such as the development of periphrastic aorist forms. This must-have occurred before the time of Pāṇini because Panini makes a list of those from the northwestern region of India who knew these older rules of Vedic Sanskrit.
- Brahmana prose - In this layer of Vedic literature, the archaic Vedic Sanskrit verb system has been abandoned, and a prototype of pre-Panini Vedic Sanskrit structure emerges. The Yajñagāthās texts provide a probable link between Vedic Sanskrit, Classical Sanskrit, and languages of the Epics. Complex meters such as Anuṣṭubh and rules of Sanskrit prosody had been or were being innovated by this time, but parts of the Brahmana layers show the language is still close to Vedic Sanskrit.
- Sutra language - This is the last stratum of Vedic literature, comprising the bulk of the Śrautasūtras and Gṛhyasūtras and some Upanishads such as the Katha Upanishad and Maitrayaniya Upanishad.
Phonology
Vedic differs from Classical Sanskrit to an extent comparable to the difference between Homeric Greek and Classical Greek. Tiwari ([1955] 2005) lists the following principal differences between the two:
- Vedic had a voiceless bilabial fricative (ᳶ [ɸ], called upadhmānīya) and a voiceless velar fricative (ᳵ [x], called jihvāmūlīya)—which used to occur as allophones of visarga (अः [h]) appeared before voiceless labial and velar consonants respectively. Both of them were lost in Classical Sanskrit to give way to the simple visarga. Upadhmānīya occurs before p and ph, jihvāmūlīya before k and kh.
- Vedic had a retroflex lateral approximant ([ɭ]) (ळ) as well as its breathy-voiced counterpart [ɭʱ] (ळ्ह), which were lost in Classical Sanskrit, to be replaced with the corresponding plosives [ɖ] (ड) and [ɖʱ] (ढ). (Varies by region; Vedic pronunciations are still in common use in some regions, e.g. southern India, including Maharashtra; it was also metrically a cluster, suggesting original Rigvedic pronunciations of [ʐɖ] and [ʐɖʱ] (see Mitanni-Aryan) before the loss of voiced sibilants, which occurred after the split of Proto-Indo-Iranian.
- Vedic also had a separate symbol ळ for retroflex l, an intervocalic allophone of ḍ, transliterated as ḷ or ḷh. In order to disambiguate vocalic l from retroflex l, vocalic l is sometimes transliterated with a ring below the letter, l̥; when this is done, vocalic r is also represented with a ring, r̥, for consistency (c.f. ISO 15919).
- The pronunciations of syllabic [r̩] (ऋ) and [l̩] (ऌ) and their long counterparts [r̩ː] (ॠ) and [l̩ː] (ॡ) no longer retained their pure pronunciations but had started to be pronounced as short and long [ri] (रि) and [li] (ल्रि) during the decline of Sanskrit.[dubious – discuss] It partly varies by region like in southern India, including Maharashtra, people will pronounce ऋ as ru (रु)and ऌ as lru (ल्रु), this is pronounced according to Taittiriya Pratishakya of the Taittiriya Shakha or branch of the Yajur Veda. Virtually no region pronounces as would in Vedic Sanskrit generally. Only while reciting Vedas is it normally heard. But if the Pratishakhya, i.e. the pronunciation manual, mandates it otherwise, the latter is followed.
- The vowels e (ए) and o (ओ) were actually realized in Vedic as diphthongs [ai̯] and [au̯], but they became pure monophthongs [eː] and [oː] in Classical Sanskrit, such as daivá > devá.
- The vowels ai (ऐ) and au (औ) were actually realized in Vedic as long diphthongs [aːi̯] (आइ) and [aːu̯] (आउ), but they became short diphthongs [ai̯] (अइ) and [au̯] (अउ) in Classical Sanskrit, such as dyā́uḥ > dyáuḥ.
- The Prātishākhyas claim that the "dental" consonants were articulated from the root of the teeth (dantamūlīya, alveolar), but they became pure dentals later. This included [r], which later became a retroflex consonant.
- Vedic had a pitch accent which could even change the meaning of the words, and was still in use in Pāṇini's time, as we can infer by his use of devices to indicate its position. At some later time, this was replaced by a stress accent limited to the second to fourth syllables from the end. Today, the pitch accent can be heard only in the traditional Vedic chantings.
- Since a small number of words in the late pronunciation of Vedic carry the so-called "independent svarita" on a short vowel, one can argue that late Vedic was marginally a tonal language. Note however that in the metrically-restored versions of the Rig Veda almost all of the syllables carrying an independent svarita must revert to a sequence of two syllables, the first of which carries an udātta and the second a so-called dependent svarita. Early Vedic was thus definitely not a tone language like Chinese but a pitch accent language like Japanese, which was inherited from the Proto-Indo-European accent. See Vedic accent.
- Pitch accent was not restricted to Vedic: early Sanskrit grammarian Pāṇini gives both accent rules for the spoken language of his (post-Vedic) time as well as the differences of Vedic accent. We have, however, no extant post-Vedic text with accents.
- The pluti (trimoraic) vowels were on the verge of becoming phonemicized during middle Vedic but disappeared again.
- Vedic often allowed two like vowels to come together in hiatus without merger during sandhi, possibly through a laryngeal /H/, such as bháas ~ bháHas > bhā́s.