PART I
Preliminaries
CHAPTER ONE
Introduction
ACROSS THE PENINSULA of mainland Southeast Asia, languages again and again display a complex grammatical pattern involving a word which we may label âACQUIREâ. Here are the features of the pattern, in schematic terms: a verb meaning âcome to haveâ (as in (1)) is used as a postverbal modal element (as in (2)), a marker introducing a postverbal adverbial phrase (as in (3)), and a preverbal aspectual marker of âattainmentâ (4):
(1) He ACQUIRE fish. âHe got fish.â
(2) He fry fish ACQUIRE. âHe can fry the fish.â; âHe managed to fry the fish.â
(3) He fry fish ACQUIRE fast. âHe fried the fish fast.â
(4) He ACQUIRE-fry fish. âHe did fry the fish.â; âHe got to fry the fish.â
This book presents a detailed survey and analysis of the semantics and grammar of this pattern in five languages of mainland Southeast AsiaâLao, Khmer, Vietnamese, Hmong, and Kmhmu Cwangâalong with supplementary data from over a dozen more languages of the area. The study concentrates on semantic, pragmatic, and grammatical structure, showing the extensive and detailed parallels in grammatical organisation in these many (often unrelated) languages. Phenomena of this kind are familiar from studies in areal linguistics in other regions of the world, and when the degree of âgeneticâ relatedness between the languages is uncertain, a fundamental question that arises is whether parallelisms between the languages are due to a genetic relationship between them, to a historical relationship of contact between speakers of the languages, or to independent yet parallel innovation. The structural parallelisms investigated in this book have emerged over many centuries by a combination of structural diffusion and parallel development.
Two major thematic conclusions emerge. First, the puzzles which âlinguistic areaâ phenomena presentâin particular, the most general question of how it is that complex patterns actually come to be shared by languages in geographic proximityâmust be approached adopting a non-metaphorical view of the true processes of linguistic change and âlanguage contactâ. Second, in comparing semantic and grammatical patterns cross-linguistically, and in considering how they change over time, it is necessary to insist on special care in semantic description, and to pay close attention to the distinction between semantics and pragmatics, and how these two aspects of linguistic meaning interact.
This introductory chapter begins by outlining a theoretical view of language (including linguistic structure, âlanguagesâ, and language contact) which serves as a background for the mostly descriptive content of later chapters. Section 2 treats theoretical and descriptive issues in the study of semantics and pragmatics, followed in Section 3 by a brief discussion of âgrammaticalisationâ. Section 4 then turns to some conceptual preliminaries in anticipation of specific semantic problems which arise in Chapters 3â7, and Section 5 raises two methodological issues. The chapter finishes with a preview of the bookâs overall structure.
1. LINGUISTIC EPIDEMIOLOGY: A NON-METAPHORICAL VIEW OF LANGUAGE
My aim in this first section is to establish a non-metaphorical theoretical position on the nature of language, to serve as a framework for the more traditional grammatical study presented in the body of the book.
1.1 The ontology of linguistic signs and âsystemsâ
Humans make order of the world by abstracting conceptual categories from real instances (e.g. of words). What we normally refer to as âwordsâ and âgrammatical patternsâ (or ârulesâ) are conceived entities based on patterns of thought (accessed by introspection) or behaviour (observed). Our only evidence beyond introspection for what we call words and grammatical categories is behavioural, in speakersâ production of linguistic signifiers with consistent communicative intentions, and in listenersâ consistent responses. We never actually encounter âwordsâ and âgrammatical patternsââthey are inferred (by speakers and linguists alike) from empirically observable effects of peopleâs behaviour (Fraser 1992, 1996). This is entailed by the âno telepathy assumptionâ (Hutchins and Hazlehurst 1995)âthat as person and as researcher one can only theorise about what is in the heads of oneâs social associates. People are isolated from one another by the physical skin, and to âbridgeâ individual minds we rely on semioticsâi.e. any of the many means to get others to think or know things by exposing them to some kind of artefactual signal. Let us consider how, as separate and mobile individuals, we can coordinate collective social practice on such a massive scale and to such a fine degree of detail as we do with language.
1.1.1 Methodological individualism, and the one-to-many problem
An individual speaker has the ability to produce and interpret highly complex patterns of linguistic behaviour, consistent with his being identifiable (by others and by himself) as a âspeaker ofâ some language L. Whatever is required for that ability is transported with him wherever he goes. One may conclude from this that whatever is fundamental for an individualâs linguistic ability at a given moment is contained âwithin his skinâ at that moment.1 The doctrine of methodological individualism states that the fundamental unit or locus of any social process is the individual, and thus all explanations must be phrased in such terms (cf. Lukes 1968, Nettle 1997, Hedström and Swedberg 1998). There is no magic scaffolding of âsocial structureâ into which individuals are placed. While this view locates meaning essentially in the individual, this does not entail that meaning is merely idiosyncratic (Strauss and Quinn 1997:16). Clearly, the linguistic habits of people in the same community are highly convergent, and there is robust social commonality to patterns of meaning (e.g. of words). Individuals in social association fall into patterns of convention and consensus in their use of linguistic categories, by a natural tendency to conform, in the interests of solving recurrent âcoordination problemsâ as efficiently as possible (Schelling 1960, Lewis 1969, Schiffer 1972, Clark 1996). Thereby, large bodies of linguistic signs cohere and show system properties (i.e. become grammars) in peopleâs minds, and the grammars of individuals share a lot. The individual linguistic signs (i.e. meanings associated with phonological sequences and associated combinatorial properties) making up these grammars are best understood as theories, constructed by individual speakers over time by a process of trial and error (cf. Keller 1994:150, Hutchins and Hazlehurst 1995). It is sets of signs or âlinguistic itemsâ which we regard as âlanguagesâ (Nettle 1999:8), and the high degree of convergence of individualsâ theories of sign meanings is what leads us to imagine that signs exist independently of people.2
The sheer volume of linguistic material (e.g. speech sounds) we introduce into each otherâs environment presses us into a collective holding pattern with regard to our respective hypotheses of linguistic sign meaning. We collectively orbit a single âmassâ of conceptual categories, the gravity of social association continually drawing our idiolects towards convergence. One of the strongest centripetal forces is the significant commitment of time required to maintain the large set of linguistic items which identifies one as part of a social group (Dunbar 1992, 1996, 1999, Nettle 1999:8). While there are many similarities between speakersâ idiolects, the non-metaphorical view is that linguistic âsystemsâ remain ultimately housed in individuals (not implying the simple-minded view of âidiolectâ attacked by Weinreich, Labov, and Herzog 1968). As Le Page and Tabouret-Keller (1985:4) remark, âgroups or communities and the linguistic attributes of such groups have no existential locus other than in the minds of individualsâ. âA languageâ (e.g. âEnglishâ) in the everyday sense is primarily an idea, the idea of âthe languageâ in the minds of speakers of a specific group of people who speak in an identifiable way.
1.1.2 Historical continuity
Consider the paradox of the following statement: âThe villagers at Na Meo say they have been living in their present location for 286 yearsâ (Chamberlain 2000:100). Like the 50-year old axe which has had three new blades and four new handles, an idea has endured, transcending physical facts of continuity. Similarly, no single empirically definable linguistic system literally endures across generationsâthere is only successive and overlapping embodiment by individuals of highly convergent and often effectively identical complex systems of linguistic signs. The standardly implied historical continuity of âa languageâ over centuries assumes that âtransmissionâ of the system from generation to generation is essen-tially faithful. We turn, then, to the notion of âtransmissionâ.
1.1.3 Transmission, ânormalâ and otherwise
While some of the cognitive wherewithal for language acquisition is no doubt prewired, structural attributes of specific languages are not simply transmitted to children by parents in an ovular/seminal bundle. Specific ideas for ways of saying things in a given language must be learned by the child, painstakingly, constructed and maintained through ceaseless practice. The passage of this transmission is through air, over days, weeks, months, years, with great interference and noise. It is not âthe languageâ that is inheritedârather, its parts are inherited, one by one. Thus, while the long standard genetic metaphor in models of language change may be applicable to individual signs (Nettle 1999:6), it does not apply to âlanguagesâ as wholes.
Thomason and Kaufman (1988) define ânormal transmissionâ as applying when âwhat is transmitted is an entire languageâthat is, a complex set of interrelated lexical, phonological, morphosyntactic, and semantic structuresâ (Thomason and Kaufman 1988:11). This notion of ânormal transmissionâ risks conveying the impression that language learners are normally not exposed to linguistic habits of people other than their parents and other members of their own group, implying that linguistic communities, like cultural groups, are normally...