The Sinitic Languages: A Contribution to Sinological Linguistics by Mieczysław Jerzy Künstler (review)

The title of this book already signals its perspective: a linguist looking at 'Chinese' from outside China, and from Europe in particular, perceives not a language but a family of languages, comparable in their diversity to the Romanic or Germanic language families (see Chappell 2016 for recent work from this perspective). Indeed, in a new twist on an old comparison, the author asserts that 'Pekinese and Cantonese are more distant than Icelandic and Southern German' (p. 40).

Having published the original book in Polish in 2000, the author was working on a revised English version at the time of his death in 2007. Künstler's text breaks off at the end of chapter 8, at which point Alfred Majewicz takes over the translation, without attempting the updating that Künstler was pursuing in his English version. The book is based on lectures given at University of Warsaw by Künstler, who is described as a passionate lecturer as well as a 'severe and demanding professor' (p. xiii). Both the passion and the severity come across in the book, which presents a relentless critique of Chinese linguistics, especially as practised in China. Linguists are taken to task for numerous errors of method and argumentation. In reconstructing proto-Sino-Tibetan, for example, numerals are commonly used as evidence when they are more likely to be early borrowings than true cognates (p. 63).

The alternative approach presented here begins with a shock announcement: no characters are to be used! The reasoning is that 'The links between language and writing are often overestimated' (p. 11), 'Chinese writing was never a good tool to note the language' (p. 40) and 'the Chinese cannot liberate themselves from looking at the language through Chinese characters' (p. 82). These arguments overlap with the hanzi filter as proposed by Wang (1996)—the effect whereby those aspects of Chinese which are not represented in characters tend to become invisible or distorted. Given this policy, even the book's intensive discussions of Archaic and Ancient Chinese present examples in modern pinyin readings (and without tone markers). The editor has relented to the extent of providing characters in an index. Yet the exclusion of characters becomes absurd at times, such as when the editor notes that the Cantonese mathai 'water chestnut' is written with two characters meaning 'horse hoof', which remain unprinted due to the ban (p. 207).

The author begins with a discussion of typological, genetic and areal affinities. Having adopted the deconstruction of Sino-Tibetan (ST) by Benedict (1972), he proceeds to split Sinitic from Tibeto-Burman (TB). The relationships between Sinitic and neighbouring families, including TB, are thus seen as areal and typological rather than genetic. The author often delves into the areal links between Tai languages and Sinitic, but the Mongolian and Manchurian languages 'lie outside of our scope of interest' (apparently on typological grounds: p. 13). Consequently, he does not consider the Altaicization hypothesis of Hashimoto (1976), whereas Hashimoto's hypothesis of a Miao-Yao substrate in Min is discussed (p. 216). Indeed, Künstler shows a particular interest in the southern minorities and their impact on southern varieties. The Cantonese term mathai 马蹄'water chestnut', for example, is analysed as containing the Tai prefix ma- 'fruit' (p. 31), while the interpretation of 马 蹄 as 'horse hoof' is treated as a folk etymology (p. 236). Tai-Kadai languages are rather misleadingly referred to as Thai in the book; similarly, yuè 越 is rendered 'Vietnamese' when the text makes better sense when understood as referring to the languages of the băi yuè 百越 peoples of south China (p. 216).

The organization of the book follows the history of Chinese from the oldest texts to the 20^th Century, before returning to the theme of diversity within the Sinitic family. This section begins with a chapter surveying dialect groups, entitled...