Andrew Radford - Linguistics An Introduction [Second Edition]

one?’). In terms of the analysis outlined here, this means that the two sentences have the respective structures in (463) below (with want being a finite form in

463a and a non-finite infinitive form in 463b): (463) a.

[

φ

φ

CP [C ] [TP I

[T +FIN] [VP [V want ] one]]]

b.

[

φ

φ

CP [C ] [TP me [T –FIN] [VP [V want] one]]]

On this view, children alternate between using finite verbs and infinitives in contexts where adults use finite verbs: for this reason, Ken Wexler and his

co-researchers have dubbed the relevant stage the Optional Infinitive stage

(sometimes abbreviated to OI stage). This stage typically lasts until around the child’s fourth birthday (with the use of non-finite clauses in finite contexts gradually becoming less and less frequent as the child gets older).

A related phenomenon which we find during the relevant stage is that children alternate between using tensed and untensed verb forms in contexts where adults require tensed verbs (i.e. verbs inflected for present/past tense). This pattern is illustrated by the sentences in (464) below (produced by Claire at ages two years to two years, one month):

(464) a.

David did it. Claire did it. Bear did it. Claire fell down. Claire woke up.

Happened the hammer? Look I found. That goes little one. That one goes

another one. There goes another one. Goes there. Goes here.

b.

Pixie eat dinner. Jane help dinner time. Bunny stand up. Cow fall down. Claire close it. Claire do puzzle. Jane do it. That go there. Chair go there. That one fit.

Daddy sit in chair. Raggedy Ann sit down. Raggedy Ann lie down. Porcupine

lie down. Raggedy Ann stay there. Raggedy Ann to wake up. Jane see

Mommy. Pig say oink (reply to ‘What does the pig say?’)

The sentences in (464a) are finite clauses containing a finite verb like happened/

goes overtly inflected for tense, but those in (464b) appear to be non-finite clauses containing an untensed verb like eat/go. In terms of the framework we are using here, Claire fell down and Claire fall down will have the respective simplified structures (465a, b):

358

senten ces

(465) a.

[

φ

φ

CP [C ] [TP Claire [T +FIN] [VP [V fell ] down]]]

b.

[

φ

φ

CP [C ] [TP Claire [T –FIN] [VP [V fall] down]]]

Both clauses are CP+TP+VP structures, but they differ in that T is finite in (465a) and so the verb fell is overtly inflected for past tense, whereas T is non-finite in

(465b) and so the verb remains in the uninflected form fall (i.e. the same form as we find in infinitives). Once again, we see the familiar pattern of children alternating between finite and non-finite forms in finite contexts.

Children’s nominals

Up to this point, we have concentrated on the clause structures pro-

duced by young children, noting that they sometimes produce non-finite clauses in finite contexts, and so, for example, omit auxiliaries (or finite verb inflections like present/past tense -s/-d) where adults require them. We find a similar pattern of development in relation to children’s nominal structures. From around two years of age, children start to produce adult-like DP structures of the form determiner +

noun, using both definite determiners like the/this/that and indefinites such as a/

another/some. However, alongside determinate nominals containing overt defi-

nite or indefinite determiners, we also find children producing bare nominals which contain a noun but no determiner (in contexts where adults would require a determiner), as illustrated by the following sentences produced by Claire at ages two years to two years, one month:

(466) a.

There’s the hat. Piggie see the water. Baby drink the coffee. Daddy sitting in the chair. Horsie swimming in the pool. Do the green one. Put that mommy in

the carriage.

b.

It’s a baby. It’s a dolly. It’s a girl. There’s a spider. There’s a bunny. There’s another one. There goes another one. Put another fence.

c.

Daddy sit in chair. Girl sleeping. Baby eating dinner. Baby eating juice. Claire do puzzle. Pig say oink. Read book. Ring bell. See flower.

The italicised nominals in (466a) are DPs headed by the definite determiners the/that, and likewise those in (466b) are DPs headed by the indefinite determiners a/another. Since Claire is clearly able to form DPs at this stage, it seems reasonable to assume that all her nominals are DPs (as indeed must be the case if

principles of UG specify that all nominals are D-projections). But this in turn means that bare nominals such as those italicised in (466c) must also be DPs; and since they contain no overt determiner, they must be headed by a null determiner.

Note that Claire doesn’t just use null determiners in contexts where adults do (e.g. with proper names like Claire), but also in contexts where adults require an overt determiner (e.g. modifying a singular count noun like chair/girl/puzzle, etc.). In other words, just as she omits auxiliaries in obligatory contexts, so too she omits determiners in obligatory contexts (i.e. in contexts where adults would require an overt determiner).

Children’s sentences

359

Research conducted by Teun Hoekstra and Nina Hyams has suggested that there

are systematic parallels between the role of T in clauses and the role of D in nominals. They note that just as the tense specification of T serves to anchor a clause in time, so too the definiteness specification of D serves to anchor a nominal in space. On this basis, they argue that definiteness and tense are two different manifestations of a single common property, which they refer to as finiteness. In the terminology of Hoekstra and Hyams, nominals which contain an overt determiner are finite, whereas those which lack an overt determiner in a context where adults would require one are non-finite. This means that a sentence such as Pig say oink (which Claire used in reply to ‘What does the pig say?’) will have the structure (467) (simplified by not showing the internal structure of the VP say oink): (467)

CP

TP

DP

T'

C D

N T

VP

ϕ

ϕ –FIN

pig

ϕ –FIN

say oink

In the same way as the overall clause is non-finite by virtue of containing a TP

headed by a non-finite T, so too the subject DP is non-finite by virtue of being headed by a non-finite D.

Hoekstra and Hyams argue that children’s finite clauses always have finite

subjects (so that they say The doggy is barking but not *Doggy is barking), but that their non-finite clauses can have either finite or non-finite subjects (so that they say both The doggy barking and Doggy barking). They maintain that this

follows from UG principles – more specifically from the specifier-head agreement relation which holds (universally) between a finite T and its subject. Since there is no agreement relation between a non-finite T and its subject, there are no finiteness restrictions on the choice of subject in a non-finite clause.

Hoekstra and Hyams’s claim that children sometimes use a non-finite D

in contexts where adults use a finite D can be extended in interesting ways

to account for the fact that young children often alternate between producing genitive and accusative possessors – as the following examples produced by a three-year-old boy called Nicholas illustrate (the data kindly being provided by Joseph Galasso):

(468) a.

I want my key

b.

I want me duck

(469) a.

What’s his name?

b.

What’s him name?

(470) a.

Where’s Zoe’s bottle?

b.

Where Daddy car?

360

senten ces

In the (a) examples in (468–70), the italicised possessor my/his/Zoe’s has genitive case, whereas in the (b) examples the possessor me/him/Daddy is accusative. Why should young children alternate between genitive and accusative possessors?