Andrew Radford - Linguistics An Introduction [Second Edition]

ing sets of items from a small number of superordinate categories (e.g.

sport, fruit, vegetable). Then ask native speakers to rate each of the

items on a scale of 1–7 for their ‘goodness’ of category membership.

The instructions you should use are:

I am going to read out the names of a number of items each of which is an X

(sport, fruit, vegetable, etc.). Using a number between 1 and 7, you must

indicate how good a member of X you consider each item to be. For example,

suppose X is sport and the item I read is tennis. If you think that tennis is a particularly good member of this category, you should give it 7, if you think it is a particularly bad member, you should give it 1, if you think that it is

intermediate, you should give it 4, and so on. Are there any questions?

Summarise your results in a systematic way and, where possible,

pool them with those of others in a class so that the total sample is as

large as possible. Discuss the significance of your results.

13

Children and words

In the previous sections of this part of the book, we have introduced a large number of the tools used by linguists when they examine words and their structure in a range of languages. From now on, we seek to apply some of these tools,

beginning with the child’s acquisition of words. Like most aspects of first

language acquisition, this process, once started, is something that parents and other adults take very much for granted. The very first strings of sounds produced by the child which are recognised as words are greeted with great acclaim, but from then on sight is often lost of the child’s massive achievement.

In considerations of first language acquisition, it is customary to be concerned with questions of order. For example, if we suppose that part of what is involved in acquiring a language is the establishment of appropriate word classes and assigning specific words to those classes, we can immediately ask whether there is evidence that children acquire word classes in a particular order. Assuming a positive answer to this question immediately gives rise to a second, more difficult question: why? Pursuit of the first question is a largely descriptive enterprise, which could be viewed as a prerequisite to seriously posing the second; answers to the second question will, if adequate, provide us with an explanatory account of some aspect of acquisition. In this section, we shall see that there is considerable evidence for small children controlling remarkably sophisticated systems of

linguistic representation from a very early age. Of course, in a general sense, this is what we might expect if the child comes to acquisition innately equipped to achieve linguistic competence.

Early words – a few facts

It has been estimated that small children acquire on average about ten

new words each day. While they sometimes make what adults regard as errors in their use of words, some of which we shall discuss below, in many respects

children’s early words are used with remarkable linguistic accuracy.

The linguistic concepts which have been introduced earlier enable us to raise a number of questions about order of acquisition. As far as major lexical categories go, children’s early production vocabularies exhibit a preponderance of nouns, typically used to refer to objects in the child’s immediate environment (e.g.

mummy, daddy, dolly, car). Alongside these, children are often quick to develop 186

Children and words

187

a small number of ‘general purpose’ verbs. The sort of thing we have in mind will be familiar to parents and is illustrated by the following interaction:

(179)

parent: (puts hat on doll)

child:

(removes hat, gives it to parent) Do it.

parent: (puts hat on doll)

child:

(removes hat, gives it to parent) Do it.

parent: (hides hat behind back)

child:

(finds hat, gives it to parent) Do it.

parent: (indicates behind back) Put it here?

child:

(nods) Do it.

Here we see the verb do (or possibly the sequence do it, if this is the only context in which it appears) being used to cover a range of actions, and reliance on one or more verbs of this type is characteristic of the early stages.

Small numbers of adjectives (e.g. nice, big) and prepositions (e.g. up, down) also occur in transcripts of early child speech. Now, it is important to be clear that in making this sort of claim, we are viewing things from the perspective of adults.

At the earliest stages, children do not string words together into phrases and sentences, nor do they systematically inflect words, so the morphosyntactic

criteria for recognising lexical classes, which were introduced in section 9, cannot be applied to the very beginnings of language production. However, when these criteria do become applicable, evidence for lexical categories is readily available (see section 24).

A different, and in many ways more interesting, question arises if we contrast the acquisition of lexical categories with that of functional categories (see section 9).

While the evidence that lexical categories are present from a very early stage is overwhelming, the same cannot be said for functional categories. A typical utterance from a two-year-old is (180):

(180)

Car go innere (as child places car in toy garage)

Setting aside the phonological characteristics of the phrase innere, there are two observations to make about this utterance. Firstly, car, a singular count noun, requires a determiner in English (a car, the car, this car, etc.); secondly, since car is a third person singular subject, the agreement inflection -s should appear on the verb (car goes innere). Both of these items are missing from the child’s utterance, and such apparent omission of members of functional categories (in this case a member of D) along with certain inflections is a characteristic of early child English. Indeed, the extent of such omissions and their implications for theories of the child’s morphosyntactic development have been, and continue to be, major research questions. We shall return to these matters in detail in section 24.

Suppose, for present purposes, that members of functional categories are indeed absent in the English-speaking child’s early language. Various possibilities might account for this including a lack of perceptual salience (typically, functional category items do not carry stress) and semantic opaqueness – coming to terms

188

words

with the semantics of determiners (the, a, this, that) or complementisers (that, if, for) looks like a rather forbidding task, and while nouns which refer to concrete objects and verbs which denote activities bear some relation to the child’s

non-linguistic experience, it is not clear that this is true of for in I’m anxious for you to eat this. We can surely understand a child systematically ignoring such items. More intriguingly, it has been suggested that the early absence of functional categories (if, indeed, they are absent) may be explicable in terms of an unfolding genetic programme. After all, to say that language is part of human genetic

endowment is not to say that all aspects of language are simultaneously available to the child. Indeed, if this latter were the case, we might expect first language acquisition to be an even faster process than it is. In fact, the account of early sentences which we present in section 24 does not suggest that functional categories are completely absent in the early stages; rather, it proposes that they are

‘deficient’ in certain respects. Whatever view turns out to be correct here, the suggestion that the course of acquisition is at least partly determined by genetic mechanisms remains a live option (exercise 1).