Leung, Helen Hue Lam (2016). The semantics of utterance particles in informal Hong Kong Cantonese (Natural Semantic Metalanguage approach). PhD thesis, Griffith University, Brisbane. PDF (open access)
This study identifies the semantic invariants of some commonly-used Cantonese utterance particles in Hong Kong Cantonese. The particles are a distinctive and ubiquitous feature of informal, everyday Cantonese, occurring every 1.5 seconds on average. The particles are necessary for expressing speakers’ transitory attitudes, assumptions, or feelings connected with an utterance. Although they are not grammatically obligatory, conversation sounds unnatural when they are omitted. There are approximately 30 ‘basic’ particles, which can combine with each other to form ‘clusters’, resulting in roughly 100 variations. This number easily surpasses that of comparable particles in Mandarin, and is matched by very few, if any, other languages. Semantic analysis of Cantonese utterance particles is challenging because their meanings are extremely elusive, even to native speakers. The range of use of each particle is so varied and wide-ranging that some Cantonese speakers and scholars have concluded that the particles have no stable semantic content. Prior research on the particles has produced contradictory, vague, obscure or inaccurate descriptions.
This study demonstrates that particles have meaning, by using the Natural Semantic Metalanguage (NSM) approach to identify the semantic invariants, or ‘core’ meanings, of a selection of commonly-used utterance particles, namely laa1, wo3, gaa3, laa3, and zaa3. NSM expresses the meanings of words and concepts in reductive paraphrases called explications, where the language used is limited to a set of semantic primes. Using this method, each particle’s meaning is identified and stated in versatile explications which are clear, accurate, translatable, and testable. The explications reliably explain each particle’s range of use in the Hong Kong Cantonese Corpus, which comprises 180 000 words of naturally-occurring Cantonese. One of the most significant findings is that explications for Cantonese utterance particles are typically short and simple. The results prove that the particles have stable and identifiable meanings.
In addition, the explications reveal the role of semantics in determining why particles can or cannot combine in particular ways. The particles selected for analysis occur in many common clusters, e.g. gaa3-laa1, gaa3-zaa3-wo3, while other clusters are unacceptable, e.g. *laa1-wo3. The meanings of particle clusters are widely claimed to be the combined meanings of the particles of which they are made up, but there have been no serious attempts to verify this. To do so would first require accurate definitions of the individual particles. The explications proposed in this study shed light on this neglected area. It is found that where particle clusters are acceptable in speech, the combined explications reveal the meanings of the clusters. A semantic critique of sub-morphemic analyses of monosyllabic particles is also presented.
This study also considers the complexities of using NSM for Hong Kong Cantonese. If basic NSM assumptions are correct, any explication should be able to be expressed in simple and natural Cantonese, giving the same meaning as in any other language. This thesis identifies and evaluates Cantonese exponents of all the 65 proposed semantic primes, and explores some Cantonese-specific issues. Each particle explication is presented in English and Cantonese.
