Evidence for dialectal diversity in Imperial Times, and linguistic criteria for determining the provenance of Old Tibetan texts

This paper investigates linguistic characteristics of modern Tibetan dialects from two different regions, and discusses evidence for the same traits in Old Tibetan (OT). Classical Tibetan (ClT) has a number of verbs with a present stem in ʾG- ŋG- and a past stem in bK- pK-, but the past stems are different for labial initials: their initial ph- lacks the b- p- prefix which all other initials have, and is aspirated instead (ph-). In a vast area of modern dialects from Tsang to Ladakh, however, the same stems are shown in the Comparative Dictionary of Tibetan Dialects (CDTD) to have an unaspirated initial p-. What is more, this same p- occurs already in the Chronicle and a few other OT texts, and I propose here that it may be interpreted as suggesting that these texts originate from the center of the Empire. In Amdo, by contrast, the same verb initials appear to have been pronounced as ph- already in Imperial Times. This aspirated pronunciation predominated in OT texts from Dunhuang (most of which are from the first half of the ninth century), and it became the only possible spelling in ClT. And while present and past stems for non- labial initials continued to be marked by ŋG- and pK-, imperfective stems with labial initials are regularly found to have initial m-ph- instead of the expected m-b-. That is, a nasal prefix came to distinguish present (m-ph-) from past (ph-) stems with labial initials, and for many other initials likewise (i.e. present N-Kh- vs. past Kh-, but also ʾchi ntɕʰi < *n-ɕi die’). The fact that nasal prefixed present stems are among modern dialects found only in Amdo (and northern Kham) makes it appear likely then that OT texts in which such present stems occur also originate from that region, and that ClT was also strongly influenced by the variety spoken there in Imperial Times.