Fundamental Principles for Chinese Wordhood: Case Studies on Negation and "Separable" Words

漢語成詞基本原則:否定詞及「離合詞」個案研究

Student thesis: Doctoral Thesis

View graph of relations

Author(s)

  • Hio Tong CHAN

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date7 Apr 2016

Abstract

Wordhood has long been a controversial issue in Chinese linguistics and language processing. The influential definition by Bloomfield (1933) that a word is a “minimum free form” is generally accepted by linguists and is assumed to be universally applicable to all languages. Nevertheless, the wordhood issue is exacerbated by many misconceptions and much confusion arises in applying this definition to word recognition because of the arbitrary labeling of linguistic units without a clear definition.
This study comprises an in-depth investigation of Chinese wordhood based on a comprehensive examination of each of the four terms in the definition of word as a “minimal (meaningful) free form”, another plain paraphrase of Bloomfield’s definition. The aim is to uncover the key notions underlying these four terms and to avoid intuitive and ambiguous interpretations. This study further explores a number of fundamental principles of wordhood underlying word recognition in terms of and beyond a given word definition, based on the hypothesized autonomy of individual modules in our language faculty, particularly the autonomy of morphology and syntax. The first principle in distinguishing words from larger linguistic structures such as phrases is that the former are morphologically generated structures and function as syntactic atoms in the construction of the latter, which are syntactically generated. Another principle is that the linguistic structure and semantic composition of a word are decomposable only at the morphological level, not at the syntactic level (i.e., the phrase level or higher levels).
Based on the given word definition and these principles, a number of applicable wordhood criteria and tests are examined and applied to solve controversial cases of Chinese wordhood. Two data sets, namely, Chinese negation words and “separable words”, are used to verify the soundness of the principles and the workability of these tests by determining whether each word-like form is morphologically or syntactically generated. The validity, feasibility, and practicality of applying these principles and tests are confirmed by data from various sources.