Andrew Radford - Linguistics An Introduction [Second Edition]

Having argued that English children produce non-finite wh-questions

like (449) with a null PRO subject, let’s take a closer look at the structure of such sentences. In keeping with the assumptions underlying the Principles and Parameters model, we will assume that UG principles determine that wh-phrases must move to spec-CP, and wh-questions like (447) show that children’s grammars recognise this from the earliest stages. We can maintain, then, that the wh-pronoun what (which originates as the complement of doing) moves to spec-CP in (449a). Hence, (449a) must contain a CP-projection. If we make the standard assumption that C universally selects a TP complement, and if we also assume (as we have throughout) that

subjects occupy spec-TP, it follows that (449a) will also contain TP. Finally, since

(449a) contains the lexical verb doing (and since T selects a VP complement), it will also contain VP. So, our assumptions lead us to the conclusion that (449a) is a CP+TP+VP structure derived in the manner outlined (in simplified form) in (453): (453)

[ CP WhatC [ TP PRO T [VP doing t]]]

Its adult counterpart What are you doing? will have the derivation in (454):

(454)

[CP What[C are ] [TP you [T t ] [VP doing t]]]

An important difference between the two structures is that the adult structure (454) is a finite clause (headed by the finite auxiliary are) with a nominative you subject, whereas its child counterpart (453) is a non-finite clause with a null PRO subject.

Since T in English can be filled only by a finite auxiliary (and only finite auxiliaries can move from T to C), it is scarcely surprising that (453) contains no overt auxiliary.

There are two interesting conclusions which our discussion of children’s nonfinite questions lead us to. The first is that there is essential structural continuity

Children’s sentences

355

between adult and child grammars: this (in a fairly obvious sense) is what the Principles and Parameters model would lead us to expect. After all, if some

aspects of sentence structure are determined by innate UG principles and so do not have to be learned, and if other (language-specific) aspects of structure involve children in the comparatively simple learning task of parameter-setting, we should expect to find that the very earliest sentences children produce are similar in structure to their adult counterparts.

A second conclusion which we can draw is that children sometimes use

non-finite clauses such as What doing? in contexts where adults require a finite clause such as What are you doing? More specifically, young children tend to alternate between finite and non-finite clauses in finite contexts (i.e. in contexts where adults require a finite clause). We can illustrate this in terms of the negative sentences in (455) below, produced by a girl called Kathryn between the ages of one year, ten months and two years (the data are from a study by Lois Bloom): (455) a.

Can’t see. I can’t open it. I don’t go sleep. I don’t need pants off. I don’t want those shoes. This one don’t fit.

b.

No like celery, Mommy. No want this. No go outside. Not going away. No

going home. Man no go in there. Kathryn not go over here. Kathryn no fix

this. Kathryn no like celery. Mommy no play ‘corder. Kathryn not quite

through.

Sentences in colloquial English are usually negated by a finite negative auxiliary such as don’t, won’t, can’t, isn’t, etc., and it is clear from the examples in (455a) that Kathryn already knows this. However, alongside the finite negative sentences in (455a), she produces non-finite auxiliariless negatives like (455b), sometimes negated by no, sometimes by not (confusion between no and not being typical of young children).

So, in contexts where adults require a finite clause, young children alternate between finite and non-finite clauses. An interesting reflex of the difference between these two types of child clause is that their subjects are differentially case-marked, as examples such as those below illustrate:

(456) a.

I’m pulling this

b.

Me going make a castle (Holly; two years)

(457) a.

She’s gone

b.

Her gone school (Domenico; two years)

(458) a.

He’s kicking a beach ball

b.

Her climbing up the ladder there (Jem; two years)

(459) a.

I can mend it

b.

Me finding something (Adam; two years, two months)

(460) a.

I’m having this

b.

Me driving (Rebecca; two years, two months)

In finite clauses like the (a) examples, we find nominative subjects, whereas in non-finite clauses like the (b) examples, we find accusative subjects: for example,

356

senten ces

nominative I is used as the subject of the finite contracted auxiliary ‘m in (456a), but accusative me is used as the subject of the non-finite verb going in (456b).

Why should this be?

Interestingly, these case-marking errors turn out to be predictable if we assume that by the age of two, children have acquired the adult English case-marking system. Adult English is said to have structural case, in that the case carried by a pronoun is determined by the position it occupies in the structure containing it. In

section 19, we suggested that the grammar of English incorporates a set of case assignment conditions along the lines of those given in a simplified form below: (461)

Case assignment conditions in English

A noun or pronoun expression is assigned

a.

nominative case if the specifier of a finite T (i.e. the subject of a finite clause) b.

genitive case if a possessor (i.e. an entity possessing something)

c.

accusative case otherwise (by default, if ineligible for nominative or

genitive case)

In the light of the case conditions in (461), let’s look at how we account for the fact that children alternate between structures like I’m playing and Me playing.

Given our assumption that subjects are in spec-TP, both clauses will contain a TP

(which, given the assumption that UG requires all clauses to be CPs containing a force-indicating C, will serve as the complement of a null declarative C). Since I’m playing contains a finite auxiliary in T but Me playing does not, let’s assume that T

is finite in the first case and non-finite in the second. Using the feature [± FIN] as a convenient way of marking the difference between a finite and a non-finite T, we can say that the two have the respective (simplified) structures indicated in (462): (462)

a .

CP

TP

T'

C D T

+FIN

V

ϕ

I (a)m

playing

b. CP

TP

T'

C D T

–FIN

V

ϕ

me

ϕ

playing

Children’s sentences

357

T is filled by ’m in (462a) but is null in (462b) because only a finite T can be filled by an auxiliary, not a non-finite T (and we can assume that children leave a given position empty when they have no suitable overt lexical item which can fill it). Let’s further assume that by the age of two, children have acquired the adult English case-marking system in (461), so that (at the relevant stage) there is continuity between adult and child case systems. It follows that the subject in (462a) will have nominative case by (461a) and so appear as I; and conversely that the subject in (462b) will have default accusative case by (461c) and so appear as me.

We can extend the analysis proposed here to account for the fact that many

two-year-olds alternate between saying, for example, I want one and Me want one.

Let’s suppose that when want is used with an accusative subject such as me, it is a non-finite form (i.e. the same non-finite form that we find in adult infinitive structures such as the italicised clause in ‘Have you ever known me want