The Northwest of South Asia and beyond: The issue of Indo-Aryan retroflexion yet again

Abstract

Abstract The Vedic Sanskrit contrast dental : retroflex is widely attributed to Dravidian substratum influence (starting with Pott 1833, 1836). An objection is that retroflexion is explainable by internal developments and that the phonological systems of Sanskrit and reconstructed Dravidian differ considerably (e.g. Bloch 1925; Hock 1975, 1982). Tikkanen (1987, 1988) raises the alternative possibility of a northwestern substrate. I reexamine the evidence and arguments offered so far. The trigger for the internal development of Sanskrit retroflexion is the retroflex sibilant ṣ and its variant ẓ(h), segments absent in Proto-Dravidian. The Northwest, by contrast, has retroflex sibilants; in fact, a triple contrast s : ṣ : ś is widespread and occurs as early as the Prakrits, Eastern Middle Iranian Saka, and even Avestan (with s : š : ṣ̌ : š́). Significantly, it also occurs in Vedic. I argue that the retroflex sibilant, the trigger for Vedic retroflexion, arose in the Northwest, not by contact with Dravidian. The development of retroflex stops can be explained in terms of later convergent developments in peninsular South Asia, developments which in Dravidian also introduced alveolar stops. In the concluding section I discuss the relevance of this account for convergence studies in general.