linguistlist.org>
Editor for this issue: Karen Milligan <karenlinguistlist.org>
fs.tufs.ac.jp>re Mike Maxwell's comments: > Children, according > to Chomskian theories, have a number of built-in concepts--things like > > grammatical categories, phrase structure, argument structure, traces > etc.; > they attach those concepts to the first things they hear (or, ICO deaf > > children, see) that meet the descriptions. One reason I have come to find Chomsky's particular theory of the Language Faculty implausible is that second language learners generally have no difficulty learning the grammatical categories, phrase structure, etc. of a second language. What adults have so much difficulty with in a second language and what children learn with ease and amazing speed-- pronunciation and vocabulary of the first language-- cannot possibly be part of Universal Grammar. In other words UG, as Chomsky formulates it, just doesn't give a satisfactory explanation for the differences between first and second language acquisition. > But I think there's another response, too: we have intuitions about > things > that virtually never appear in corpora, and by and large our > intuitions > agree. (When they don't agree, it's usually places where we don't > feel > strongly about whether a construction is acceptable, or at least > that's my > experience.) So for example, I recall hearing parasitic gap > constructions > for the first time in 1984, when a linguist reported on a trip she had > made > to MIT. My intuitions on the sentences in question accorded perfectly > with > those of the other linguists around the table, but I can't imagine > ever > running into parasitic gap constructions in corpora. But this doesn't mean that our intuitions are innate. We have intuitions about learned behavior too. People who have grown up in the same culture have common intuitions about what constitutes polite behavior, for example. If we all agree in a grammaticality judgement, it simply means we are all fluent speakers of the same language, that is, that we have internalized the norms or conventions of our speech community. Generativists tend to confuse the hypothesis of mentalism (language is in the mind) with that of innatism (it's there at birth). The fact that we know something doesn't mean we were born knowing it. > In sum, the fact that corpora are repetitive is all the more > reason for thinking that my knowledge of English is *not* based only > on routine, because my knowledge goes far beyond that routine. If you have learned calculus, then you can solve a calculus problem you have never seen before, but that doesn't mean you have an innate knowledge of calculus, though presumably any intelligent human being has an innate ability to learn calculus. A theory that language is learned (rather than innate) need not be a theory that language is learned by routine. Plato's theory of knowledge-- the ability to learn anything is prior knowledge of the thing-- is, I think, not generally adopted outside of linguistics. > The "mother acquisition device" in ducklings can't be physically > proved, > either. Do you believe in it? The entire history of science is a > succession of ideas that couldn't be "proven", in the sense that one > couldn't point to instances of the concepts until much later: gravity, > > evolution, continental drift, atoms, ions, black holes, quarks... But > > scientists in the relevant disciplines came to believe in those > concepts > (with resistance in many cases!) on the basis of other kinds of > evidence. Yes, but we are no where near having a theory of language acquisition that is supported by the standards of evidence routinely required in the natural sciences. Chomsky's theory is based on logical arguments, very little evidence, and many assumptions which have never been explicitly formulated as hypotheses, much less subjected to empirical test. For example I often see statements like "The theory of the language faculty is (called) UG." This is incorrect. UG is simply A theory of the language faculty, and it's plausible only if one also accepts a Platonic theory of knowledge. UG is only one way in which we might connect the structure of language with the structure of the mind. I believe it is widely accepted by theoretical linguists simply because it's convenient-- It allows us to do what linguists have always done-- structural analysis of language data-- while claiming that we are doing neuro-science. I think that scientists should avoid believing things, by the way. Theories are disposable-- falsifiable, otherwise they don't belong to science. > The existence of atoms was accepted by all > scientists long before the invention of microscopes (in the 1970s, if > I > recall) that were capable of "seeing" them. Why should our standards > for > evidence in linguistics be different? The atomic theory was around for a long time-- Democritus, Lucretius. It was the orthodox position adopted by Islamic theology in the 10th century.But the mainstream (the Aristotelian stream) rejected it, for the reason that atoms are logically impossible. This is obvious and undeniable, at least if you accept Aristotle's assumptions. The atomic hypothesis as formulated by Dalton was eventually accepted because it explains certain facts about nature-- for example that when a compound is decomposed into its elements the relative weights of the elements is always in a constant proportion-- more than that, it makes explicit quantifiable predictions that can be and were tested by experiment. Yes, the standards of evidence in linguistics should be the same as in natural science, but in the present state of things they are not. The universal grammar hypothesis has never, as far as I know, even been stated in explicit falsifiable terms, and it has certainly never been subjected to empirical test. It does not, and should not, enjoy the same status as the atomic theory. -- ----------------------------------------------------------- Robert R. Ratcliffe Associate Professor, Arabic and Linguistics, Dept. of Linguistics and Information Science Tokyo University of Foreign Studies Asahi-machi 3-11-1, Fuchu-shi, Tokyo 183-8534 JapanMail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue
andrew.ac.jp>J.Mukherjee wrote: >To begin with, the concept of native competence seems to play a central >role. In other words: is the difference between a native speaker of English >and someone who has learned English as a foreign language a fundamental one? >This automatically raises the question if a L2 learner of English can ever >achieve nativelike competence. From the Chomskyan point of view (and/or on >the basis of regarding L1, but not L2, as a matter of "prewired >imprinting"), the answer to the first question would be yes and to the >second one no. ***The 'Chomskyan' view--that is, the view of generative theories of grammar, so far as I know--is agnostic on the issue of L2 competence. Indeed, there's a fairly lively debate going on among SLA researchers on precisely this issue (see, e.g. Eubank 1991; Epstein, Flynn, & Martohardjono 1996 & commentary therein). >However, Pawley/Syder (1983: 193) address the problem "that native speakers >do not exercise the creative potential of syntactic rules to anything like >their full extent, and that, indeed, if they did so they would not be >accepted as exhibiting nativelike control of the language". ****But who said they do? And what follows from the claim that they don't? The fact (I devoutly hope it's a fact) that none of us walk like John Cleese doesn't prejudice the claim that our more humdrum ways of walking are an expression of a (largely innate) competence. Furthermore, these lexical, semantic and/or >grammatical patterns elude a merely intuition-based observation. >One theoretical implication, among others, is that the L1 speaker of English >has a nativelike competence not because he/she uses whatever is possible, >but rather because he/she uses what is probable, i.e. what is frequent. What >challenges the generative approach is corpus linguistic evidence of >patterned routine (in the widest sense) which turns out to be all-pervading >in native language use. It does not come as too much a surprise that such >empirical evidence is rejected by generativists. ***'rejected' is ambiguous here: I doubt that many generativists reject the findings of corpus linguists as to how we behave verbally. What is rejected, and rightly so, is the sort of conclusions from that evidence that Mukherjee, among many, seem to draw. It most certainly does *not* follow from these corpus studies that one's competence is the effect of using our sense of probabilities to choose what we say. I know, for instance, that 'John ate an apple and Mary a pear' is grammtical; I also know it stinks, and I can safely bet big bucks that in all my depressingly numerous years as a native speaker I have never used that sort of sentence. Actually, I would expect that, if anything, generativists would even *predict* that >these lexical, semantic and/or >grammatical patterns elude a merely intuition-based observation These patterns are patterns of behavior, and generative theory has nothing to say about behavior. But if competence simply *were* these patterns of behavior, then wouldn't we be able, *ex hypothesi*, to predict them? One last point about competence: What corpus studies cannot ever tell us, and what is central to the idea of competence, is what it is as native speakers (or non-native speakers, for that matter) that we know to be *ungrammatical*. The frequency with which I say things like 'John ate an apple and Mary a pear', after all, is exactly the same as the frequency with which I say things like 'John may been have eating an apple'. The causes of those two zero frequencies, however, are quite different. And could somebody tell me why it is that the peculiar canard that the 'Chomskyan' concept of competence is based on written language keeps popping up over and over again, and always without any evidence in support of it? This is a definite candidate for inclusion in any Dictionary of Received Ideas. One more comment and then I'll shut up: The idea of a native speaker of a language is often rejected, I suspect, out of a sense that it functions to support invidious and unjustifiable distinctions between 'us' and 'them'. This suspicion is understandable, but misplaced insofar as it leads to a rejection of the idea of competence: There is nothing in generative theories of linguistic competence that would justify, for instance, refusing to hire someone because that person is not a 'native speaker'. Kevin R. Gregg Momoyama Gakuin University (St. Andrew's University) 1-1 Manabino, Izumi Osaka 594-1198 Japan tel.no. 0725-54-3131 (ext. 3622) fax. 0725-54-3202Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue