Typological variation across Sinitic languages : contact and convergence

Abstract

Decades of works dedicated to the description of (previously) lesser-known Sinitic languages have effectively dispelled the common myth that these languages share a single “universal Chinese grammar”. Yet, the underlying cause of their grammatical variation is still a matter for debate. This thesis focuses on the typological variation across Sinitic varieties. Through comparing the typological profiles of various Sinitic languages with those of their non-Sinitic neighbors, we discuss to what extent the variation within the Sinitic branch can be attributed to areal diffusion. Variation across Sinitic is often explained from the perspective of language contact – sandwiched between Altaic languages to its north and Mainland Southeast Asian (MSEA) languages to its south, Sinitic can be considered typologically intermediate between these two groups of languages, where Northern Sinitic shows signs of convergence towards Altaic languages and Southern Sinitic towards MSEA languages. For example, the northern varieties tend to have a smaller number of classifiers, tones and codas, as well as a stronger tendency to disyllabicity and head-final constructions. However, the notion of “Altaicization” (Hashimoto 1976) is a moot point. Despite the typological differences between Northern Sinitic and Southern Sinitic, as Bennet (1979) argues, there is little evidence for “Altaicization” as many of such differences can hardly be put down to Altaic influence; instead, they are more likely due to the typological convergence between Southern Sinitic and MSEA languages. Moreover, there is evidence that the typological variation across Sinitic cannot be amply explained by areal influence from non-Sinitic languages. Some Sinitic varieties are known to exhibit certain distinct typological characteristics. For instance, analyzing the disposal, passive, and comparative constructions across the Sinitic branch, Chappell (2015b) argues that there are no fewer than five principal linguistic areas in China. Taking into account over 350 language varieties of seven different genetic affiliations (Sinitic, Turkic, Mongolic, Tungusic, Hmong-Mien, Tai-Kadai, Austroasiatic) and 30 linguistic features, we conduct a typological survey with the aid of the phylogenetic program NeighborNet (Bryant & Moulton 2004). Our results suggest that convergence towards their non-Sinitic neighbors has indeed played a pivotal role in the typological diversity of Sinitic languages. Based primarily on their degree of Altaic/MSEA influence, the Sinitic varieties in our database are classified into four areal groups, namely 1) Northern, 2) Transitional, 3) Central Southeastern, 4) Far Southern. This classification scheme reflects the intricate interplay between areal convergence, regional innovations, and retention of archaic features. The findings suggest that contact-induced typological change can occur rather rapidly, especially if given the appropriate sociolinguistic conditions. Furthermore, this thesis highlights the interdependence between the meticulous analysis of qualitative linguistic data and the proper application of quantitative tools in typological studies. Although this study is chiefly concerned with Sinitic typology, the quantitative approach adopted herein can potentially help shed new light on the challenge of typological comparison in other areas.