The latest BP Podcast is up. You can listen on Libsyn, iTunes and Stitcher. Probably the easiest way to keep up the podcast since we don’t have a regular schedule is to subscribe at one of the links above.
The conversation was between Brown Pundits’ contributors Razib, Zach and Slapstik on the evolution of Sanskrit and Indo-Aryan more generally within the Indian subcontinent. The conversation started off from thoughts on the origin of Sanskrit from Proto-Indo-Aryan, the language of the feudal elite of the Bactria-Margiana Archaeological Complex (or BMAC).
We spoke of the broad swathe of time from (roughly) 1500-500 BCE wherein this elite planted themselves across the breadth of the Indo-Gangetic plains south of the Himalayas (cf. hima-vanta > himavata). The peculiarities of the obstinately oral culture led to phenomenal developments in grammar (vyAkaraNa) to preserve the fidelity of speech. We also briefly covered some peculiarities of development of prkRta-s which show influence of Dravidian speech, the Indo-Aryan nature of Dardic languages and some comments on Iranic languages.
One of the important questions raised by Zach was on the parallel development of Mitanni-Aryan. Specifically why doesn’t its existence prove an outward diffusion of Indo-Aryan dialects? While I did not want to de-rail the conversation in the podcast with a technical argument (based on the RUKI rule), I think this point does deserve some supplementary explanation here. The basic argument can be set out as follows:
- Retroflexes in earliest Vedic were almost always phonemic (pUrNa), and it is well established within IA linguistics that retroflexion existed before the loss of voiced sibilants /z/, /ź/ and /Z/ in proto-Vedic. So, we have PIE *misdhom (salary, reward) > *mizdhom (IIr, voiced sibilant) > *miźdhom (IIr, via satem RUKI sound law) > *miZDham (Proto-Vedic, using sibilant-dental consonant saMdhi)
- We also know that Proto-Dravidian lacked sibilants and replaced the phonemic voiced-sibilants by their approximants /y/ or /w/, which later merged with the preceding vowel. So, the effect of Dravidian substratum on proto-Vedic *miZDham is given by: *miZDham > *mi(y)Dham > mIDham, and mIDham is attested in Rg Veda as meaning reward or prize of a contest.
- The same word (meaning payment) is also attested in Mitanni-Aryan as miśta, which seems to be easily derivable from IIr *miźdhom via simple de-voicing of /ź/ and de-aspiration of /dh/. There’s no known rule or precedent of deriving miśta from IA mIDham. Even using the latter-day example of Romany languages, the aspirated retroflex /Dh/ should be approximated by /r/, which is clearly not the case here.
Therefore, the simplest explanation of miśta in Mitanni-Aryan and its cognate mIDha in Vedic is that both terms are derivable using known sound laws from the older Indo-Iranian version of the word, as opposed to one from the other. The same argument also shows how retroflexion existed in oldest Vedic and that the simplest explanation of lack of voiced sibilants in Vedic is the substrate effect of Proto-Dravidian.
For readers more interested in this topic, I would suggest The Horse, The Wheel and Language by David Anthony on the archaeological evidence of steppe Indo-European culture. The book is very strong on archaeology, but it gets some of the linguistics’ arguments wrong. On linguistics itself, Cardona and Jain’s Indo-Aryan Linguistics remains the go-to text. Note that this book is more technical, but very rewarding.