Topics in the phonology and morphosyntax of Balinese based on the dialect of Singaraja, North Bali

Abstract

This study describes features of the phonology, verbal morphology, and morphosyntax of the Balinese of Singaraja, in Buleleng district, north Bali. In chapter two I describe the phoneme repertoire, and their realisations. I give evidence for meaningful postlexical rules, of the kind described by Woodbury (1987). In chapter 3 I show that there is a regular interaction between semantics and phonology at a prelexical level, contra the double articulation hypothesis, and in support of the findings of Fudge (1970). I propose an (ultimately pre-linguistic) explanation for this interaction. The findings support and enrich optimality theory, which seeks to formalise the gradient nature of wellformedness (Prince & Smolensky 1993). The semantic evidence helps explain why non-optimality is tolerated in the phonology. While the phoneme is an important unit of organisation at an intermediate level, it is neither a phonological primitive, nor the largest recurrent phonological unit. I give evidence in chapter 4, and in chapter 7, for complex phonological formants more complex than the phoneme. These can be 'segmental', or involve processes such as reduplication (pace the treatment of reduplication in Prince 1987). In chapter 5 various levels of prosodic organisation are described. The syllable, the foot and the phonological word are each relevant to word-formation processes (though no evidence was found for the mora). I give evidence in this chapter, and in chapters 6 and 8, that all morphemes must be well-formed at foot level during the lexical phonology. These data support, and can be accounted for in terms of, the Prosodic Morphology theories developed by McCarthy and Prince (eg 1990). In chapter 6 constraints on the structure of morphs and morphemes are described. Balinese shows similar cooccurrence restrictions to Arabic (Greenberg 1960, McCarthy 1986) and Javanese (Uhlenbeck 1949, Yip 1989), dispreferring the occurrence of more than one consonant of a given place of articulation, underlyingly. Morphemes showing exceptions to this 'preferred rule' predictably belong to the expressive semantic classes defined in chapter 3. Non-prosodic phonological processes involved in morpheme- and word-formation (including concatenation and morphophonemic alternations) are described in chapter 7. Chapter 8 catalogues the various formal types of reduplication. Two previously undescribed surface reduplication types in Balinese are identified: Foot-reduplication and internal reduplication. As with the restrictions on consonant cooccurrence morpheme-internally (chapter 6), reduplication processes involving vowel overwrite give evidence for a non-linear organisation of phonological structure. A variety of evidence is presented to show that phonological or 'inherent' reduplication is asynchronic process in Balinese, (despite previous claims that there is no evidence for this). Finally, I describe theoretically problematical reduplication processes where nondistinctive material is transferred (cf Steriade 1988), and offer an explanation for them in terms of the Prosodic Morphology theory of McCarthy and Prince. The second part of the thesis describes aspects of the morphosyntax, in particular word formation processes applying to derive verbs in Balinese, and related aspects of clause syntax. After an overview of basic features of morphosyntax and terminology in chapters 9 and 10, I describe in detail the functioning of verbal affixes, in chapters 11 to 13. In chapter 14 I give evidence that Balinese shows 'undergoer primacy' in a variety of ways, at the same time arguing against an ergative analysis of the syntax (pace Blake 1993). My major conclusions relate to the function of verbal morphology, and the way meaning is assigned to lexemes and clauses. The meaning of a given root-affix combination is broadly predictable, once the semantics of the root mopheme is known. To that extent, morphology signals primarily semantic information in Balinese. The Actor/Undergoer distinction which informs the morphological and syntactic patterning of other Austronesian languages (Foley & Van Valin 1984, Durie 1984, 1987) is an important one in Balinese. Prefixes (or their absence) signal whether the subject of both transitive and intransitive verbs is an Actor or an Undergoer (chapter 11). The semantic type of the Undergoer is further specified by verbal suffixing (chapter 12). Although formally Balinese is close to the ideal of an agglutinating language, the meaning of a lexeme is not simply the sum of the meanings of its component morphemes (cfKoch 1990). Meaning assignment is best treated using a model such as those in the word & paradigm tradition (Beard 1988, Aronoff 1994). Affixes are not in themselves meaningful: they serve primarily to signal the presence of meaning elements inherent in the lexemes of which they form part. As such some lexical bases always occur with affixes, even though they have not undergone derivational processes. (I argue against the 'precategorial' analysis of such lexical bases in chapter 10). Where affixes participate in derivational processes, the meaning of the resulting lexeme is assigned by the derivational process as a whole. Indeed, I show that in some cases an adequate account of the meaning conveyed by a given lexeme can only be achieved if discourse and other contextual factors are also taken into account (Chapter 13). The evidence from Balinese supports the view of (for example) Wierzbicka and Givon that transitivity is fundamentally a semantic notion. The primary function of affixing on verbs in Balinese is to signal the presence of entitities in the semantic structure of the lexeme; these may or may not be represented in the syntax, according to contextual and other factors. Semantically and formally transitive verbs thus often behave as though they are syntactically intransitive, and affixes often described as 'transitivising' are so as long as one accepts this semantic interpretation of the term. This approach enables a wider explanatory coverage than previous descriptions of Balinese verb morphology, or of similar facts in related languages (Chung 1976, Kana 1986, Alsagoff 1992). Show more