grammaticality judgments

[Karin Stromswold]

I elicited children's grammaticality judgments for my dissertation
(Stromswold, 1990, Learnability and the acquisition of auxiliaries,
MIT.  Available through MITWPL), and my students and I have continued
to use the technique with moderate success with children as young as 3.

In our studies, we have a dog puppy that the children feed either a
bone (for grammatical sentences) or a rock (for ungrammatical
sentences).  The key to making sure that children are judging
grammaticality and not the meaning conveyed by the sentence is to make
sure that, during the course of the experiment, each child judges both
grammatical and ungrammatical versions of each sentence  (e.g., What
does Kermit eat?  and What Kermit eats?).   By doing this, any effects
of semantics is a wash when you do statistical analyses of your data.

Of course, one has to worry that the experimenter's delivery of the
ungrammatical sentences is less natural than the experimenter's
delivery of the grammatical sentences, and that children might be
judging this and not grammaticality per se.  We have dealt with this in
2 ways.  First, we have the experimenter practice saying the
ungrammatical sentences, and we record the experimental session and
check to make sure that the experimenter did a reasonable job.  A
second way we have attempted to deal with this is by using pre-recorded
sentences rather than 'live' sentences.  The problem with doing this is
that children don't pay as much attention when the sentences are not
presented live.

Good luck,

Karin Stromswold
Rutgers - New Brunswick

On Dec 16, 2004, at 9:35 AM, James Russell wrote:

> Hello,
>
> By this question I'm sure to betray much ignorance. . . but can
> anybody offer advice on trying to evoke grammatically judgements from
> pre-school children. For example, a puppet 'speaks' a sentence (e.g.,
> 'The apple in on the table' or *The apple on is the table') and the
> child must judge if what the puppet  said was "silly" or "OK".  I do
> recall the silly/OK procedure being used by somebody. . .
> I know of Stephen Crain's use of a truth-value judgment procedure in a
> similar way; but I (quixotically?) want to evoke judgments of
> well-formedness.
>
> James Russell
> Experimental Psychology
> Cambridge, UK
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 2285 bytes
Desc: not available
URL: <http://listserv.linguistlist.org/pipermail/info-childes/attachments/20041216/d69943a1/attachment.bin>