Andrew Radford - Linguistics An Introduction [Second Edition]

the same student (the second interpretation from exercise 5) is strongly preferred. Use the examples in (b) and (c) below and any others you consider relevant to assess the generality of this ‘surface’ account

of scope:

(b) Some worker in every factory likes ice cream

(c) Every voter in some constituency voted for Pratt

24

Children’s sentences

The Principles and Parameters Theory (PPT) outlined at the end of section 22 has interesting implications for the development of a theory of language acquisition, and in particular raises the question of what it is that children have to learn about the syntax of their native language? Clearly, a major part of the task of acquiring a first language involves lexical learning (i.e. learning words and their idiosyncratic properties, see section 13). However, the question we shall focus on here is what structural learning is involved in first language acquisition – i.e. what children have to learn about the structure of sentences in the language they are acquiring. (Note that we shall be concerned here only with how children acquire their native languages, not with the very different question of how children or adults acquire foreign languages.) Within the PPT model, certain aspects of sentence structure are assumed to be determined by UG principles (i.e. principles of Universal Grammar) and hence are invariant across languages. If we further assume that principles of UG are part of the child’s innately endowed language faculty, it follows that universal aspects of sentence structure will not have to be learned (see the Introduction, pp. 7–8). For example, if clauses are universally CP+TP+VP structures and this is part of the child’s innate linguistic competence at birth, it will not have to be learned.

Similarly, if noun/pronoun expressions are universally D-projections (and hence comprise either a pronominal determiner, or a prenominal determiner with a noun or noun phrase complement) and this is also part of the child’s innate knowledge, this too will not have to be learned. In other words, the child does not have to learn those aspects of sentence structure which are universal by virtue of being determined by innately endowed UG principles.

So what do children have to learn about sentence structure in their native

language? The answer is that they have to learn those aspects of structure which vary in a parametric fashion from one language to another. A key assumption of the PPT model is that all structural variation between languages can be characterised in terms of a set of parameters, each of which is binary and hence has two possible values (e.g. the Head Position Parameter, which specifies that a particular type of phrase has head-first or head-last word order, the T Strength Parameter, which indicates whether T is strong or weak, the Null Subject

Parameter, which states that finite verbs do or do not license null subjects). It follows from this that the only structural learning which children face in acquiring their native language is the task of determining the appropriate value of each of the structural parameters along which languages vary.

349

350

senten ces

If our reasoning here is along the right lines, it leads us to the following view of the language acquisition process. The central task which the child faces in

acquiring the structural properties of a language is to construct a grammar of the language. The child’s language faculty incorporates a theory of Universal

Grammar which includes (i) a set of universal principles of grammatical structure, and (ii) a set of structural parameters which impose severe constraints on the range of structural variation permitted in natural languages (perhaps limiting the range of variation to a series of binary choices). Since universal principles of grammatical structure don’t have to be learned, the child’s structural learning task is limited to that of parameter-setting (i.e. determining an appropriate setting for each of the relevant structural parameters).

The assumption that acquiring the syntactic structure of a language involves the relatively simple task of setting a number of structural parameters at their appropriate value provides a natural way of accounting for the fact that structural learning is a remarkably rapid and error-free process in young children.

Setting parameters: an example

A good example to illustrate the approach we have just outlined is

provided by examining the acquisition of word order. Young children acquiring English as their native language show evidence from the very earliest two- and three-word sentences they produce of knowing (tacitly, not explicitly, of course) that phrases in English uniformly have head-first word order. Accordingly, the earliest verb phrases and prepositional phrases produced by English children consistently show verbs and prepositions positioned before their complements, as structures such as the following illustrate (produced by a young boy called Jem at age one year, eight months; head verbs or prepositions are italicised):

(442) a.

Touch heads. Cuddle book. Want crayons. Want malteser. Open door.

Want biscuit. Bang bottom. See cats. Sit down

b.

On mummy. To lady. Without shoe. With potty. In keyhole. In school.

On carpet. On box. With crayons. To mummy.

So, children acquiring English set the Head Position Parameter at the

head-first setting appropriate to all types of phrases in English from the very earliest multi-word utterances that they produce. They do not use different orders for different words of the same type (e.g. they don’t position the verb see after its complement but the verb want before its complement), or for different types of words (e.g. they don’t position verbs before and prepositions after their

complements).

A natural question to ask at this juncture is how we can account for the fact that from the very outset of multi-word speech, we find English children correctly positioning heads before their complements. The Principles and Parameters model enables us to provide a principled explanation for how children manage to learn

Children’s sentences

351

word-order properties like these in such a rapid and error-free fashion. The answer provided by the model is that learning these aspects of word order involves the comparatively simple task of setting a binary parameter at its appropriate value.

This task will be a relatively straightforward one if the Head Position Parameter determines that the only possible choice is for a given type of phrase in a given language to be uniformly head-first or uniformly head-last. Given such an

assumption, once a child hears (and can parse) a verb phrase such as help

daddy, the child will immediately be able to infer that English is a head-first language. So, child structures like those in (442) are consistent with the parameter-setting model of acquisition outlined above. However, there is what at first sight appears to be some puzzling counter-evidence to the claim that children set parameters at their appropriate value at a very early age.

Null subjects in early Child English

In influential research carried out in the early 1980s, Nina Hyams

observed that children acquiring English at around two years of age frequently omit sentence subjects and produce sentences such as those in (443): (443)

Play it. Eating cereal. Shake hands. See window. Want more apple. No go in.

Hyams maintained that sentences like these have an implicit (i.e. ‘understood’) subject, a claim which is made more plausible by the fact that when children produce a seemingly subjectless sentence, they sometimes produce an expanded variant of the sentence immediately afterwards in which the ‘understood’ subject is made explicit – as in the following examples (collected by Martin Braine) produced by Stevie between the ages of two years, one month and two years, two months: