Last revised 12 December 1995.
Comments Welcome
George W. Grace
University of Hawaii
H.G. Widdowson wrote (1989: 135), "(...) communicative competence is not a matter of knowing rules for the composition of sentences and being able to employ such rules to assemble expressions from scratch as and when occasion requires. It is much more a matter of knowing a stock of partially pre-assembled patterns, formulaic frameworks, and a kit of rules, so to speak, and being able to apply the rules to make whatever adjustments are necessary according to contextual demands. Communicative competence in this view is essentially a matter of adaptation, and rules are not generative but regulative and subservient."
If this is considered as a description of the kind of knowledge of language that people actually have, it seems about right to me. I see no need to complicate this by adding a "linguistic competence", consisting centrally of a systematic grammar. While I suspect that most people probably have some knowledge that could be described as grammatical (the more glib or articulate individuals more than the less articulate), I don't imagine the grammatical knowledge as ever being united into a single all-encompassing system. I think of the grammatical knowledge as more like a collection of know-hows to deal with various contingencies--what has been called recipe knowledge.
I also like Paul Hopper's (1987) idea that grammar is a by-product of speech, rather than a prerequisite, and also that a grammar is always a movement toward structure rather than structure achieved. (Cf. also Murat Roberts's (1944: 300) description of grammar as "fossil idiom", and his statement (1944: 299) "Yesterday's discourse is today's language; today's discourse is tomorrow's language").
I became convinced some time ago that the kinds of language use that we regularly observe--that is, use of language by humans to express their thoughts and feelings and to understand the speech of others--cannot be accounted for by the hypothesis of an internal grammar, but must be based essentially in the learning of usable utterances.
I think this idea first reached the level of conviction because of the difficulties some languages seemed to present to any potential learner. For example, the first Algonkian language that I read about struck me as virtually unlearnable. Athabaskan languages made the same impression, and I remember that Harry Hoijer was supposed to have said that he never expected to be able to learn to speak Navaho (or any other Apachean language).
When I say they seemed "unlearnable", what I mean is that it was difficult for me to imagine how anyone could master the complex morphology and particularly the great variety of morphologically-conditioned allomorphs sufficiently ever to be able to assemble or analyze sentences in real time. It seemed inescapable that larger syntagmatic combinations (whole, or very nearly whole, sentences) must be learned as units for anyone ever to achieve fluency, and that the children who learn those languages must pursue pretty much that strategy. Of course, although it was Algonkian and Athabaskan languages that (if I am remembering rightly) first impressed me with the magnitude of the problem, they are not unique. There are many languages around the world that could serve to illustrate the problem. I think particularly of many Papuan languages and many other Native American languages (e.g., Mohawk). (1)
However, after I'd thought about it for a while, I began to see that English-speakers use much more of that strategy than I'd been imagining.
If we think in terms of two possible strategies--a holistic strategy, which aims at learning appropriate utterances, and an analytic strategy, which aims at constructing a grammar--my idea would be that the holistic strategy is universal, whether or not it's accompanied by some analytic strategy. But I wouldn't be surprised if most people become aware of little (or not so little) grammatical (sub)systems here and there in their languages and analyze them as systems.
This idea, if it's true, implies (or if it doesn't automatically imply, it suggests and I hereby propose) that our knowledge of our language should be thought of as essentially taking the form of a repertoire of linguistic expressions rather than as what I've been calling a "governing system" (i.e., a system [traditionally supposed to consist of a syntactic system and lexicon] which is supposed to authorize the sentences of the language and warrant the meanings of each sentence so authorized (2)). This proposal seems to help explain several things.
The following list of matters that I think this hypothesis--i.e., that our knowledge of our languages takes the form of a repertoire rather than of a governing system--helps to explain includes seven items:
(1) How dialect chains work. It's possible for repertoires to overlap in various ways and degrees. We don't have to worry about which people share the same governing system, about how degrees of sameness are to be conceptualized, and about how partial understanding works for people who (must, according to the governing-system hypothesis, be supposed to) have different, but similar, governing systems.
(2) How language mixing is possible. How, under the governing-system assumption, explain the fact that people can understand mixed utterances--utterances such that it's not possible to assign the whole utterance to one or another governing system--that is, to say which language's syntax it's in. (It's because of the governing system presupposition, of course, that there has been so much effort expended on attempting to discover grammatical constraints on code shifting/mixing).
(3) The extent to which formulaic speech permeates English conversation (and presumably that of all other languages). On this, see particularly the work of Andrew Pawley (esp. Pawley 1985, 1994).
(4) The fact, originally pointed out by Ann Peters, that (even in the environment of modern civilization [I would add]) some children begin their language "acquisition" not with single words, but with holistically-known sentence-like units. See especially Peters 1977, 1983.
(5) Reports that the language teaching methods spontaneously employed by people in one place in New Guinea and another in New Britain conformed to what has been called the phrase-book model rather than the grammar-lexicon model. (3)
(6) C-J. Bailey (1982) has proposed a distinction between the kinds of development that occur "when languages are left alone, i.e., when they have no contact with other systems" and those that occur "as the result of contact with other systems" (1982: 10). Languages that are left alone develop in the direction of "synthesis (the amalgamation of functional particles as flectional bundles)"; contact with other languages leads the development in the direction of "analysis (the breaking down of inflections into nothing or, more specifically, into functor words)"
Thurston divided the nine languages spoken in the area of Northwestern New Britain that he studied into those which had important roles as lingue franche and those which didn't. He noted that the former "must be easy for neophytes to acquire" (1987: 38), and that they are relatively analytic. Those not involved in lingua-franca roles are characterized by phonological efficiency at the expense of morphological transparency; replacement of regular derivational forms with suppletives; and augmentation in the number of opaque idioms.
Peter Trudgill (1989) proposes that the distinguishing factor is whether or not a language has post-adolescent learners. Later learners have need of transparent, analytic structures. Younger learners do not. The hypothesis that language learners (at least, first-language learners) learn expressions holistically offers what to me is the most plausible explanation of why, when left alone, languages should tend to become increasingly opaque and irregular in structure.
(7) What goes on in a speaker (to-be) prior to speech. I won't say that the hypothesis of holistic learning fully explains the form of the original intention and the process by which it is realized in speech, but it provides a much better clue than the governing-system hypothesis. The governing-system hypothesis, as far as I've been able to figure out, assumes that a speech act normally consists of an attempt to convey a proposition (which presumably is delivered ready-made by some module of the brain unrelated to language).
It now seems apparent that most speech acts don't have the conveying of a proposition as their main objective. It also seems apparent that the decision to perform a particular speech act must be based on some kind of assessment of the overall effect on the potential audience of the act as a whole. That is, we must somehow be able to contemplate potential acts as wholes, including whole linguistic expressions. (4)
1. I find the following quotations from Marianne Mithun (1989: 311-312) on the acquisition of polysynthesis in Mohawk especially instructive:
"Although derivational suffixes and noun incorporation are pervasive in adult speech, there was no evidence of their acquisition as processes by these children. The older children used forms containing them, but showed no overgeneralization or innovation. These facts are probably more indicative of the nature of this linguistic knowledge in Mohawk than of the linguistic immaturity of the children. Most derived forms are learned and stored as lexical units, although the learning and storage are undoubtedly facilitated by the pervasive regularities running through such a vast lexicon. It is likely that much innovative derivation by adults is more a process of analogizing on the basis of sets of acquired lexical items than the application of general rules. Speakers are very conscious of which derivationally complex words already exist in the language and which could but do not. Many excellent adult Mohawk speakers have great difficulty deriving new forms or incorporating creatively on demand, although they have little trouble interpreting neologisms. These same speakers can easily switch the tense, person or number of a verb when asked, even with invented verb stems."
"Perhaps most intriguing of all is the rarity of errors in the speech of the children observed. The sheer numbers of productive affixes in Mohawk, in addition to the pervasive morphophonemic alternation, would seem to provide pitfalls to the language learner at every step. Once pronominal prefixes were acquired, however, no unusual forms or usages were observed at all. This suggests that the lexicon, the memory of existing words, may play a much greater role in this language than the highly productive morphology might lead a casual observer to suspect." Back up
2. The term "governing system" is my own, of course. If you should want to see what I've said about the concept, you may jump to references in Ethnolinguistic Notes of Series 3. (However, you will have to find your own way back). Click where indicated to go to Series 3, no. 43, site 1, site 2, no. 45, site 1, site 2. Back up
3. Don Laycock, who reported the case from the upper Sepik basin in Papua New Guinea, commented (1979: 92), "But it seems likely that this method of teaching by whole sentences of potential use--the phrase-book method--is the normal one in Papua New Guinea; my own informants commonly adopted this method during eliciting."
In the other case, William R. Thurston (1987) reports that when the people of Northwestern New Britain undertake to teach their languages to outsiders, speakers of all of the nine languages of the region conduct the instruction according to a single, unvarying plan. And, although Thurston was not aware of Laycock's report, the plan that he describes sounds no different from Laycock's. Back up
4. I've talked a bit about the problem of the antecedents of acts of speaking in Ethnolinguistic Note, Series 3, number 44. You can click here to go to the approximate beginning of that discussion. Again, you will have to find your own way back. Back up
Bailey, Charles-James N., On the yin and yang nature of language, Ann Arbor, Karoma Publishers, 1982. Back up
Hopper, Paul, Emergent grammar, in: Proceedings of the Thirteenth Annual Meeting of the Berkeley Linguistics Society, 1987, pp. 139-57. Back up
Laycock, Don. 1979. Multilingualism: Linguistic boundaries and unsolved problem in Papua New Guinea, in: New Guinea and neighboring areas: A sociolinguistic laboratory (edited by Stephen A. Wurm), The Hague, Paris, New York, Mouton. pp. 81-99. Back up
Mithun, Marianne. 1989. The acquisition of polysynthesis. Journal of Child Language 16: 285-312. Back up
Pawley, Andrew. 1985. On speech formulas and linguistic competence. Lenguas Modernas 12: 84-104.
Pawley, Andrew. 1994. Formulaic speech. In William Bright (Editor in Chief), International Encyclopedia of Linguistics, Vol. 2, pp. 22-25. Oxford/New York: Oxford University Press. Back up
Peters, Ann M. 1983. The units of language acquisition. Cambridge (etc.): Cambridge University Press. Back up
Roberts, Murat H. 1944. The science of idiom: A method of inquiry into the cognitive design of language. Publications of the Modern Language Association 59: 291-306. Back up
Thurston, William R. 1987. Processes of change in the languages of north-western New Britain (Pacific Linguistics B99), Canberra, The Australian National University. Back up
Trudgill, Peter. 1989. Contact and isolation in linguistic change, in: Language change: Contributions to the study of its causes (edited by Leiv Egil Breivik and Ernst Håkon Jahr) (Trends in linguistics, studies and monographs, 43), Berlin, New York, Amsterdam, Mouton de Gruyter, pp. 227-37. Back up
Widdowson, H.G. 1989. Knowledge of language and ability for use. Applied Linguistics 10: 128-37. Back up
| Home Page | The Ethnolinguistic Notes | The Ethnolinguistic Notes, Series 1 and 2 | Ethnolinguistic Notes, Series 3 | The Ethnolinguistic Notes, Series 4 | Reflections: Language Evolution |
| Reflections: Knowledge of Language | Personal Page | The Human Predicament | Why Write Unpublishable Things? | Modest Proposals | Odds and Ends | Pictures |
First put on the Web on 10 December 1995
Last revised 12 December 1995.