Classifiers, lexical tone, determinatives: a look at their emergence and diachrony in Old Chinese and beyond

Abstract: The emergence of first, non-obligatory “sortal” noun classifiers, as documented in Ancient Chinese inscriptional documents (Huang Zaijun 1964, Ma Guoquan 1979 , Wen Qianxi 1992, Xu Lili 1997, Zhao Peng 2006, Xu Li 2007 etc.), coincides temporally with the attrition and, eventually, complete loss of a productive Old Chinese derivational morphology in non-peripheral Chinese dialects (Sagart 1999, Jin Lixin 2006, Schuessler 2007), the rise of lexical tone (Mei Tsu-lin 1970, Sagart 1986, 1993, Wang Feng 2006) in Late Old Chinese and several adjacent language families (Luo Meizhen 2007), as well as the proliferation of semantic determinatives in phonosemantic compounds, attested in the locally diverse epigraphical writing systems of the so-called Warring States (mid 5th c.-221 BCE), Qin (221-206 BCE) and Han (206 BCE-220 CE) periods. It has been repeatedly suggested that the gradual obsolescence of affixal morphology, including morphemes for the marking of certain fairly specific noun classes (e.g. countable items, kinship terms, insect names etc.), and the slow rise of monosyllabic noun classifiers might have been related (Sagart 1999, Behr 2002). Substantial parts of the morphology encoding noun properties were expressed in affix positions which underwent functional coalescence and subsequent loss of phonological material, a process which was then compensated by the rise of contour tones during the Late Old Chinese period. Once fully phonologized, these lexical tones continued, albeit very unsystematically, to encode certain noun properties, formerly expressed segmentally. At the same time, the sets of spoken classifiers, formerly by and large limited to ritual objects of high value or prestige, expanded greatly in number and came to include other semantic fields, such as fauna, flaura, everyday utensils, body parts etc. In the writing system, the same period witnessed a fairly sudden accretion and diversification of non-pronounced script classifiers (a.k.a, if wrongly, as “radicals” in the European sinological tradition), and structural parallels between the graphic and linguistic classification types have begun to be documented (Wiebusch & Peyraube 1993, Wiebusch 1998, 1999 and subsequent work). Although such written classifiers are first attested already almost a millenium earlier in Shang oracle bone inscriptions (13th-11th c. BCE), their percentage among the attested character structures is then fairly low. It is only during the Warring States, Qin, and Han periods, that we encounter a greater variety of graphs constructed with written classifiers, the first recognition and standardization of written classifiers in lexicographic lists, and their obligatory use with phonophorics, which could formerly be written without them, or were combined with different classifiers to denote the same lexical referent. I will therefore argue that the processes of spoken and written classifier diversification in this period are not decoupled, and that the shift from a rudimentary agglutinating noun class morphology to its analytical expression by (still non-obligatory) spoken classifier-noun collocations is mediated by tonogenesis. Apart from summarizing and exemplifying the processes described above, and giving some statistical overview, my paper will explore symmetries and asymmetries in the development of linguistic and graphic classifier systems in Old Chinese, review claims about the early emergence of a universal classifier ge in Early Archaic Chinese (Wang Lianqing 1996), and about the diffusion of the first noun classifiers between Old Chinese and Tai-Kadaiic languages. It will also address the role of early dialectal diversity in the development of written classifiers in Warring States Chinese (cf. Wichmann & Davletshin 2006).