Some descriptive linguists define basic linguistic documentation as reflecting the very essence of the discipline of linguistics, which they regard as taking precedence over all other areas of linguistic activity:
 |
| sussy_What Kind of Linguistics |
The ideal apprenticeship (for a linguist) is to undertake fieldwork on some previously undescribed (or scarcely described) language-recording, transcribing and analyzing texts; observing how people use the language in the daily round; writing a grammar and phonology; compiling a dictionary; and publishing a volume of annotated texts. (Dixon 2001)
There are probably as many different ways of describing the phonology, morphology, and syntax of a language as there are linguists, though different approaches can often be gathered roughly under a variety of different theoretical labels. Linguistic descriptions that are rigidly bound to particular theoretical constructs do not tend successfully to outlast the theories which spawned them. Grammars that are written, for example, according to the strict structuralize formulae that were in vogue up to the early 1960s, or according to the transformational model which became all the rage immediately after this, are often extremely difficult for people to read today. Similarly, descriptions that are expressed exclusively in terms that derive from some of today’s theoretical traditions are often equally unreadable to those who hail from other traditions.
The most readable and arguably, therefore, the most valuable-accounts of languages seem to be those which are relatively open to theoretical eclecticism. By this, I mean that a linguistic description should set out to allow the linguistic data to govern the form of the description, rather than requiring a single theoretical model rigidly to dictate the shape of the entire grammar. Some would argue that there is a single theory-neutral model for grammatical description which can be used for any language, as implied by Dixon’s (1997: 128–38) use of the expression
Basic Linguistic Theory. While it would perhaps be nice if this were the case, any linguistic description is in reality going to exhibit certain kinds of theoretical biases, many of which may be implicit. The most important consideration in producing a good linguistic description is that the writer’s particular theoretical assumptions should be clearly recognizable to the reader, that any terminological conventions should be clearly explained, and that the grammar should be richly exemplified with natural data so that readers coming from other theoretical persuasions are able to make sense of your discussion. With new technological tools it is possible to create rich contextual data in a media corpus for which your analysis provides a gateway (e.g. Thieberger 2004) and which allows others to confirm and extend your analysis.
Linguistics is a hugely diverse discipline, and Weld linguists can contribute to our knowledge in a wide variety of ways. While most linguists are primarily interested in matters of phonology, morphology, and syntax, the ways in which language is used ‘in the daily round’ noted by Dixon seldom receive more than passing mention. Only a minority of linguists-better, perhaps, linguistic anthropologists-have shown themselves to be interested in documenting how language functions as a ‘mode of social interaction [which] provides the material out of which a group of people recognize themselves as a community’ (Duranti 1997: 99). Such accounts cannot be based on the traditional direct ‘elicitation’ of language data from ‘informants’; rather, the fieldworker must become a long-term participant observer, recording extensive collections of both audio and video data of natural language use between different individuals within the community.
At the same time, it is clear that Duranti’s (1994) discussion of the ways in which grammatical constructions in Samoan are used to achieve socio-political goals depends crucially on a prior analysis of the language in a traditional descriptive account such as that of Mosel and Hovdhaugen (1992). Even the title of Duranti’s linguistic anthropological field account-From Grammar to Politics-implies this priority. Similarly, Schieffelin and Ochs (1986) show how language acquisition by children is related to socialization into different cultures using the same kinds of basic linguistic information. Traditional descriptive accounts, then, should be promoted as an essential stepping stone for bigger-and some would no doubt argue, better-things.
The goals of fieldwork can sometimes be much more limited than a full published grammar or dictionary, or a detailed ethnographic account of the communicative strategies that are used by people in a particular society. Sometimes, previously published materials may spark the interest of specialists who return to somebody else’s previous Weld site in order to analyze some particular features of the language in more detail. For instance, in some of the languages of Vanuatu, there are phonemically contrastive linguo-labial consonants in which nasals, stops, and fricatives are produced with the tip of the tongue touching the upper lips. Ian Maddieson undertook a fieldtrip from the United States to Vanuatu in the 1980s simply to study these speech sounds in detail in a number of languages in which they are found. A guide that is specifically geared towards instrumental phonetics is beyond the scope of this volume, though Ladefoged (2003) andMaddieson (2001) provide a great deal of very readable material.
Published grammars concentrate for the most part on regular patterns, resulting in a lack of attention devoted to the study of patterns of variable data. There is absolutely no reason why the kinds of corpus-based statistical studies that have been carried out extensively on different varieties of English could not be carried out in other languages as well, e.g. Dorian’s (2001) work on Gaelic. However, in order to do this, a linguist would need to pay close attention to a much wider range of sampling and data gathering issues than is commonly done in order to ensure that statistically representative samples of different categories of speakers have been recorded. While I have also decided to exclude sociolinguistic studies from the scope of this volume, readers who are interested in an up-to-date guide to data-gathering of this kind could consult Milroy (1987).
REFERENCE
Terry Crowley. (2007). Field Linguistics; A Beginner’s Guide. New York: Oxford University Press.
Terry Crowley. (2007). Field Linguistics; A Beginner’s Guide. New York: Oxford University Press.
