CONFERENCE: ICSTLL35 LOCATION: Arizona State University PRESENTER: Richard S. COOK STEDT Project Manager Linguistics Department University of California, Berkeley mailto:rscook@socrates.berkeley.edu http://stedt.berkeley.edu/ TITLE: Word Families in Sino-Tibetan ABSTRACT: This study explores theoretical and methodological issues relating to the assembly and use of intralinguistic synchronic word family data for the purpose of interlinguistic diachronic comparative analyses. The Chinese notions of "Xiesheng Series", "Yi Sheng" ('also phonetic') and "paronomasticity" are related to the Karlgrenian "word family" (Karlgren 1933, "Word Families in Chinese") in the logogenetic context of the notion of the "allofam" (Matisoff). It is well known in the mainstream Sinitic lexicographic tradition (and in its Sinological and Sino-Tibetic offshoots) that within the Chinese speech community the orthographic system encodes aspects of interlexical semantico-phonologic overlap. Such overlap is held to exist in degrees of idiosyncraticity, the result of more or less localized (spatiotemporal) subjective interpretation, innovation, and adoption. The mainstream interpretation of the semantico-phonologic basis for identification of a word family group arises in relatively local conditions, which broadens with the spread of orthodoxy. The conventions involved may warp in varying degrees with temporospatial displacement, as reimplementation and reanalysis of the existing communicative standard is part and parcel with the natural process of maintaining that standard. Proceeding from recognition of this fact in Chinese, the over-arching principle may be stated most simply as follows: Within a given language, similar sounding utterances are often held to convey highly similar or related meanings. This truth is exhibited most broadly by the fact that any two speakers of any given language are able to communicate with each other at all. If there were not consistency despite temporospatial displacement, meaningful communication would not be possible. This principle is certainly a cognito-linguistic universal if ever there was one (assuming that meaningful communication is in fact possible). It remains however to be demonstrated just exactly what is meant by inter-utterance phonologic or semantic similarity. While it is comparatively easy to quantify intralinguistic phonological similarity in terms of various features and at various levels of complexity (e.g. phonemic, n-syllabic), quantification of semantic relation is in one respect the real stumbling block. To what extent is it possible for two utterances to be adjudged similar in meaning, and how does one know for sure? Intuitively, meaning similarities are easy to recognize, and in fact, it is argued, introducible at will. In classifying two objects together for the purposes of comparision, one necessarily introduces a shared feature among those two objects, this shared feature being most broadly class membership itself. Once a shared feature has been recognized, it is easy to imagine how new shared features might develop and take deep root with time. It is argued, therefore, that this must be recognized as a cognitive mechanism or natural process, and that only if it is possible to safeguard against it or to place it in proper perspective, might it be possible to perform meaningful linguistic analysis. Strategies for isolating suitably safeguarded items for comparison have focused on "core vocabulary" and its natural extensions, and have also focused on the relative stability of polymorphemic structures. Working first synchronically and intralinguistically, the proposed methodology is shown to proceed primarily on the basis of an intralinguistic (monolingual) glossed syllable canon metric of morphophonological similarity built computationally with recourse to a suitably defined corpus of "homogenetic" data. Only with the illumination provided by such corpora may the analysis proceed into the deepest darkest interlinguistic and diachronic depths. ---------------------------------------- Richard S. COOK STEDT Project Manager Linguistics Department University of California, Berkeley mailto:rscook@socrates.berkeley.edu http://stedt.berkeley.edu/ ----------------------------------------