Links to pages: 671, 672, 673, 674, 676, 677

Comments Welcome

George W. Grace

University of Hawaii

20 May 1993

ETHNOLINGUISTIC NOTES

Series 3, Number 44

WHAT IS THE LANGUAGE FACULTY?

I understand "the language faculty" to refer to whatever abilities underlie and are required by the ability to use human language as humans do. But what are these abilities really?

It's customary in linguistics to distinguish the inherent faculty--what humans are born with (and no other species possesses in its entirety) that permits them to learn to use a particular language--from what is subsequently learned (if that's the proper term) in "language acquisition". That is, it's customary to distinguish the innate capacity to learn from that which is learned.

The current standard view in linguistics is that the innate capacity takes the form of a universal grammar which, when applied to the linguistic data which are made available by the child's environment, is sufficient to permit the child to construct a grammar for a particular language. (To this universal grammar itself presumably must be added some innate motivation that impels the child actually to carry out the construction of that grammar). Thus, what is acquired in the process of "language acquisition" is (according to this view) a "grammar" of a particular language system. This grammar consists essentially of a syntax and a lexicon.

In summary, according to the standard view in linguistics, the immediate reason why we human beings are able to use language is that we have a grammar of our language installed in our minds. And the reason why we (in contrast to all other animals) are able to have this grammar is that we're born with the ability (and motivation) to construct a grammar.

But it's immediately apparent that this isn't enough. If we took a computer and installed a syntactic description and a dictionary of, say, English, we could not expect the computer automatically to begin speaking. And that would remain true no matter if the syntactic description and dictionary were adequate beyond the wildest dreams of any linguist living today. Some people might want to argue that the syntax and dictionary would give the computer the ability to generate grammatical sentences at random, but whatever the validity of such a claim, the ability to generate sentences at random seems to have little or nothing in common with the ability to use language as humans do--which is to say, purposively.

A language-using being would seem to have to be one that (1) had purposes which it spontaneously strove to achieve, (2) was able to arrive at strategies directed at achieving those purposes, and (3) was able to understand and evaluate the possibilities afforded by the available language (that is, by the being's linguistic knowledge) to contribute to such strategies.

I assume that all animals have purposes and (at least in the simple sense that striving [conation] implies a strategy) strategies for achieving them. And at least other primates surely have, in some measure, the ability to design their strategies (that is, their strategies are not entirely instinctual in origin). Thus, the first two prerequisites had already been supplied in pre-human stages of our evolutionary history.

How do we humans go about incorporating the use of language into our strategies for achieving our purposes? For that matter, how do we go about shaping our strategies in general? I'm not aware that anyone has studied such questions. However, in cases where the strategic planning proceeds slowly enough to be accessible to introspection, that introspection suggests that we use a procedure analogous to hypothesis-testing. That is, we consider alternative courses of action, attempt to imagine the effects of each, and then perhaps decide on one (which, of course, may be revised again at any stage). Where a contemplated course of action would involve speaking, it seems that we must be able to contemplate various possible utterances and make some evaluation of what effect each might have.

Note here that one would presumably never be considering the simple question, "should I, or should I not, speak?" (with the question of what one might possibly say if one did speak left entirely open). Only slightly more likely, I think, would be the question, "should I, or should I not, say this particular thing?" (where "thing" refers to some particular "propositional content"). Generally it seems, the question is much closer to "should I, or should I not, make this particular utterance?" (where the utterance is envisaged in full detail).

But, in fact, there's another question that precedes "Should I make this particular utterance?". It's, "what would it be like if I made that utterance? What would that sound like to the audience? What would it appear that I was doing?" What I'm saying is that the decision whether or not to utter a particular utterance is generally dependent on a prior estimate of the effect that it's likely to have on the audience.

In sum, I think that a general prerequisite for using language is the ability to imagine potential courses of action and to imagine the likely effects of each, and on this basis to arrive at a strategy to be pursued. But the ability to use human language requires also the more specialized ability to imagine potential utterances--potential speech acts--and their likely effects and to incorporate these into one's strategic planning.

The discussion so far assumes that what we're considering from the beginning is the complete utterance in all of its detail from wording to delivery. However, I think that goes too far. My feeling is that the options that we weigh don't necessarily have exact wording and intonation, (and accompanying gestures and facial expressions) fully specified--that when we decide on a particular plan of utterance, there are likely to be some details left to be filled in later. Certainly all of us are aware of the experience of "tacking on" something at the end of an utterance "as an afterthought". What I would like particularly to insist on, however, is that it is not a matter of the initial decision involving "propositional" ("referential") content, with the filling-in consisting of an "encoding" into English or some other language. It's probably more accurate to say that we think in terms of an utterance of a certain kind--one having certain kinds of characteristics. To put it another way, we think in terms of hypothetical utterances which may be only partially specified.

The kinds of characteristics that would be considered significant probably vary greatly with the situation. In some situations, the delivery--forceful, say, or tender, or laughing--might be more significant than anything else in the utterance. The question is: "How would it sound (what would be the likely effect) if I spoke like this--produced this kind of utterance?"

What I'm proposing then is that the language faculty, which I defined as "whatever abilities underlie and are required by the ability to use human language as humans do" must involve an ability to imagine oneself producing a particular utterance or an utterance conforming to certain specifications, to imagine "how it would sound" to the particular audience in the particular situation, and to evaluate the effect of such an utterance for the purposes one is pursuing. The point to be emphasized here is that this implies the ability to conceive of utterances, or at least kinds of utterances, as unitary phenomena, to call them up as wholes and to imagine the audience's reaction to each as a whole.

It should be particularly emphasized that, if any "encoding" by means of syntax and dictionary really does occur, it surely must come later than the kind of decisions that I've attempted to describe here. That is, it would have to come after the decisions that specified what it was that was to be encoded.

However, I can't see any reason to think that encoding by means of syntax and dictionary occurs at all in everyday language use. The syntax-dictionary model looks as though it was inspired by former language teaching methods in which exercises involving translation into the target language using grammars and dictionaries were emphasized (and where language-learning was conceived of as the internalization of grammar and lexicon--I've always imagined this conception to go back to the Latin curriculum when Latin was the language of learning in Europe, but I haven't tried seriously to investigate that possibility).

In any case, this procedure makes no sense at all as part of a model of language use unless a "message" to be encoded has somehow previously been supplied. Any account of the language faculty that doesn't attempt to explain where the "message" comes from simply misses the point.

I hope that this will explain why I don't believe that knowing a language is knowing a syntactic system and a lexicon, or that "acquiring" a language should be thought of as learning or otherwise installing in oneself a syntactic system and lexicon. I hope it will also make clear why I don't think the innate abilities that underlie our ability to use language have anything to do with constructing grammars, but rather have a lot to do with awareness of the resources made available by language and of how they may be used.

One last point. I've talked about what is involved in speaking (or more particularly, in preparing to speak), but I've said nothing about the role of hearer--about what is involved in understanding what others say. I won't try to discuss that here except to say (1) that I assume it also to involve interpreting utterances--in this case those which have been heard--in terms of one's purposes, (2) that such interpretation depends on the ability to recall previously heard utterances and/or utterance partials, (3) that a thorough discussion of it would probably be more complicated than the discussion of the role of the speaker, and (4) that I see no reason to think that it, either, has any central role for syntactic and lexical knowledge.

Last updated 20 January 1996