Notes on the history of ligatures and genitive marking
Yüklə 28 Kb.
|
|
R. Blust June 21-24 Santa Barbara Workshop Reactions to Constable and Simons, ‘Language identification and IT’ In general I find this a very programmatic piece. Most statements are hedged around with qualifications which suggest that there is no clear single answer to problems of language name coding. Everything depends on the purpose for which the coding is done. Much of the discussion seems to me to be aimed at librarians, the traditional ‘guardians of information’, who are now forced to confront an entirely new range of problems in connection with the need to communicate with machines that do not think like human beings. It is clear that linguists must also take notice of these types of problems, but unless we have a specific set of goals that we wish to achieve there is no point in discussing how to code language names (or language data more generally) in order to achieve the best results in retrieving data. So, let us begin with the goals! Since these haven’t been spelled out at all clearly for the workshop I really have to wonder what we are going to be discussing in La Casa de Maria. I must also chafe a bit at writers who use sentences such as ‘This occurs, for example, in ISO 639-2’ when no previous identification of the referent has been given, and little that is subsequently said helps us to achieve a clearer picture. To some extent, then, I cannot escape the impression that the authors are writing for themselves. Another thing that comes through in reading this piece is the authors’ view of the centrality of the Ethnologue as a source of general information about the languages of the world. Since the Ethnologue is a publication of the Summer Institute of Linguistics, and since the piece we have been asked to read is an SIL Electronic Working Paper this view is perhaps not surprising. I myself regard the Ethnologue as the single best compilation of general information on the languages of the world that is currently available. But this does not mean that it cannot be improved. Since the Constable and Simon paper clearly aspires to a higher degree of accuracy and consistency in representing general facts about the world’s languages than is true of many extant sources, one is genuinely taken aback at a statement such as the following on page 10, para. 1: ‘The Ethnologue catalogs over 6,800 living languages spoken in the world, based on a primary criterion of non-mutual intelligibility.’ Is this statement even approximately accurate? The Ethnologue is organized by 1. continent, 2. country, and 3. in some cases subsections of a country (e.g. Malaysia: Peninsular, Sabah, Sarawak) . All languages spoken in a given country are listed alphabetically. At the end of this alphabetical listing is a formulaic statement ‘The number of languages listed for (country) is (number). Of those, (number) are living languages and (number) are extinct.’ But consider the 78 listings for English, which is cited (and separately counted) in each of the following country listings:
The Grenadines 44. Zambia 71. Pitcairn
Ocean Territory Emirates
What is the basis for listing English 78 times in the Ethnologue? Clearly intellibility is not the deciding factor here. What seems to matter more is whether there is a significant community of first- or second-language speakers of the language in question in a given country. But if this is true, one must wonder why English is not also listed under e.g. Argentina, where there is a large expatriate community of English speakers, Holland or the Scandinavian countries, where English is a de facto second language for most urban dwellers under 50, or Kuwait or other Gulf states, where English is a de facto second language at least within the oil business and general commercial communities. While English may show an exceptional pattern of inflation in counting the numbers, similar inflations are found for French, which is listed in virtually every African and Caribbean country which once was a colonial possession, and for Spanish, which is listed separately throughout Latin America and the Caribbean with the marginal exceptions of Brazil and a few other places. For many of the less well-known languages of the world that I am familiar with it is also clear that the figures are inflated by as much as 15-20% through counting dialectal varieties as distinct languages. Barbara Grimes, the editor of Ethnologue co-authored a listing of Austronesian languages in the so-called ‘Comparative Austronesian Dictionary’ (Barbara F. Grimes, Joseph E. Grimes, Malcolm D. Ross, Charles E. Grimes and Darrell T. Tryon. 1995. Listing of Austronesian languages. In Darrell T. Tryon, ed., Comparative Austronesian Dictionary: An introduction to Austronesian studies, Part 1, Fascicle 1:121-279. Berlin and New York: Mouton de Gruyter). The authors of this piece reached a figure of 1,202 Austronesian languages, by far the largest number ever estimated for this family. Some of this increase is due to the actual discovery of new languages (either through initial recognition, or through separation from other communities which were previously thought to speak the same language), primarily by workers in the Summer Institute of Linguistics. Much of it, however, is due to over-differentiation. Why, for example, are Javanese of Java, Javanese of Surinam and Javanese of New Caledonia counted as three distinct languages, when the latter two are spoken by the descendants of immigrant laborers who left Java within the past 200 years? Many similar cases of over-differentiation could be added to this list. My point is that Constable and Simons have not taken care to make the same kinds of careful distinctions in defining ‘language’ that they are advocating others to make. This undermines their credibility and weakens a set of proposals that to my way of thinking are not at all clearly formulated (or at least communicated to others) in the first place. Yüklə 28 Kb. Dostları ilə paylaş:
|