LINGUIST List 3.622

Tue 11 Aug 1992

Sum: Child Language

Editor for this issue: <>

Directory

1. Geoffrey Russom, Child Language: Summary

Several people responded to my request for children's wh-sentences
in which one element was extracted from an NP. Most steered me
to Bob Wilson and Ann M. Peters, "What are you Cookin' on a Hot?:
Movement Constraints in the Speech of a Three-Year-Old Blind Child"
(LANGUAGE, vol. 64, no. 2 (1988)). My wife Jacque sends thanks for
the reference, which was useful in preparing her presentation on
differences between AI language projects and word-processing software
(grammar correctors especially). Barbara Hall Partee and Dan Finer
sent their own data, and several respondents asked for a summary of
such material. Jacque and I went through our notes after reading
the article to see how our own experience related to the authors'
conclusions. In general, it seems that children without vision
problems do produce sentences like those of the blind child Seth,
though perhaps not so persistently.
 Wilson and Peters cited Partee's example as support for their
theory that anomalous wh-questions arise when a child tries to
take the role of interrogator in pedantic question-and-answer
games. In her case the interrogator was an older brother (7 1/2
years old), who asked, "What do dogs sweat through their?" When
the 6 1/2 - year old brother failed to comprehend, this was
paraphrased as "How do dogs sweat? What part of their body do
they sweat through?" Some such predetermined answer as "Their
tongues" was being elicited. A relation is suggested (by the authors
and also by Dan Finer) between such questions and structures like
"You spilled your WHAT?" or "This is a nice ______" (where the child
is encouraged to fill in the blank with a word well known to
the adult).
 Finer's child produced some anomalous wh-questions around
the age of 3 years 10 months. One was "What should we play with
some ____?" A particularly interesting example with immediate
self-correction (in a space of less than 15 seconds, Finer says)
was "What do diesels put diesel fuel in their____?", corrected to
"What do diesels put diesel fuel in that place?", then to
"What do diesels put diesel fuel in?", the adult form.
Instant self-correction seems to indicate that the child identifies
such structures as unsatisfactory without adult feedback (no doubt
because a universal is being violated) but also demonstrates
that the anomalous structure is related by the child to
normal cases of wh-movement. At the time when the wh-questions were
produced, Finer says, his child avoided pied piping generally and
produced relative clauses with resumptive pronouns ("Smokey is an
engine that he pulls a train") or with neither gap nor pronoun
("And we'll get the video that the engines don't have names").
 So far as I know, my wife and I didn't do much pedantic
interrogation of our children. We were already well-socialized
Chomskeyites before they were born, and doubted that any special
type of interaction would affect their language development.
Both of us tend to treat children like adults and remember that
we preferred being treated that way when we were that age. My
only hard-and-fast rule of parenting was not to do anything
consistently weird to the child over a long period of time.
Nevertheless, both Gavin and Gillian produced anomalous wh-questions.
 Gavin started producing multiword utterances at around 15 months
(e.g. "Ta-ta hot" = "The pancake is hot"). By 18 months we were
hearing things like "Daddy sit down bed watch TV, OK? OK?" and
"Daddy take zoo see camel llama?". At 23 months inverted questions
arose, at first with "be" as the verb of choice ("Are you have
a knife?", "What are you see?", "What are you do?", "Is daddy
paint with this one?" -- indicating a paintbrush). In the
same month Gavin had adult-style inversion in a triplet of
questions ("Mommy, can I go in the attic? Where is the attic?
Are attics dangerous?" -- the last two admittedly cases in which
the proper inverted verb is his first choice "be").
During this period Jacque noticed a tendency for Gavin to repeat
wh-questions even after answers had been given by us and apparently
understood by him. Gavin produced his anomalous wh-questions at
about 2 years 0 months. They took the form "Who is it ball?" and
"Whose is it bicycle?". By 2 yrs. 6 mo. we get adult-style utterances
("Whose orange juice are you making?"). We get clauses at this point
like the ones mentioned by Finer ("That one looks like the cat
that I pinched his tail and he scratched me"). Another ambitious
sentence from 2 yrs. 6 mo., one that incorporates a question, is
"I want a banana, and when I eat it part of the way, will you eat
it all up?"
 Gillian produced anomalous wh-questions at 27.5 months.
Some were like Gavin's ("Who is it birthday?" "Who is it book?").
But she also extracted the noun from such possessive NP's ("Who is
my daddy your___?", addressed to her mother. Answer: husband).
I wrongly attributed this last one to Gavin in my original posting.
Gillian persisted in the use of an anomalous wh-structure one time
when she was inspecting features on the head of her brother's
"Star Wars" figure: "These are his eyes. This is his nose.
Mommy, what're these two black things his? What're these things
called his?" Note that the anomaly persists during repetition for
emphasis, and that Gillian is not carrying out a pedantic
interrogation, since she does not know the answers to the questions.
An apparent self-correction occurred at 28 months, when Gillian was
becoming aware of what it was like (rubesco dicens) to share a
bathroom with an older brother. Peering into the toilet, which had
inconsiderately been left unflushed, she asked "Who is it pee-pee?".
Jacque said that Gavin was the responsible party. Gillian then flushed
the toilet, and asked "Whose pee-pee was that?", even though she
seemed to have taken in Jacque's answer.
 These are simply the observations of linguists who happen to
be parents. Neither one of us studies child language acquisition
professionally, and both of us are up to our ears in other kinds
of projects (I'm doing Germanic metrics and morphophonemics at
present). I would simply suggest that children do seem to
violate A-over-A-type constraints with some persistence even though
they also show a kind of dissatisfaction with the violations not
wholly explainable in terms of adult feedback. Perhaps children's
experimentation with word orders is sometimes deliberate rather than
intuitive. It would be surprising if they wholly lacked the ability
to make up sentences that were weird but somehow "logical". Why should
such an ability be restricted to adult linguists? A final
question: Why do NP's with possessives figure so frequently among
these anomalies?

 -- Rick Russom