Child Language Acquisition
Language Acquisition
First Language Acquisition:
So far we have looked at various ways of accounting for our
phonetic, phonological, morphological, syntactic, semantic and
pragmatic competence in our language(s).
This gives us some idea of what exactly constitutes our
linguistic knowledge.
Now maybe we can begin to appreciate just what a child does when
s/he learns language:
What exactly does a child do?
Whatever it is, it's universal
-it is acquired regardless of culture, language, class, etc.
(e.g. Polish case system is easily learned, Hanunoo color
terms, Japanese topic- comment structure, etc.)
-it's effortless.
Language acquisition occurs in stages:
1) BABBLING (+/- 6 months)
-child produces the full range of possible speech sounds-even
those which do not occur in speech heard in immediate
environment, (the TARGET LANGUAGE) and which s/he may later find
"impossible" to reproduce when learning a foreign language.
2) HOLOPHRASTIC / ONE-WORD STAGE: (+/- 12-14 months)
-the words produced in holophrastic speech are not just any
words.
For example you get:
cookie
drink
bad
fast
go
yes/no
But never:
*in
*the
*and
Remember the distinction? It looks again as though the
distinction between OPEN CLASS and CLOSED CLASS words comes into
play. This, then is further evidence to the PSYCHOLOGICAL REALITY
of that division.
These single words may even function as illocutionary acts:
May ASSERT/COMMAND/QUESTION.
3) TWO-WORD STAGE (+/- 24 months)
-still virtually no closed class words
-some pronouns, especially ME/YOU.
How can you tell the difference between a two-word utterance and
two one-word utterances?
-intonation
-structure (often N V)
4) TELEGRAPHIC SPEECH (e.g. for English)
-no 3-word stage
-basically English sentences, but still without closed class
items.
-some affixes (past tense marker, plural)
-SVO word order (almost invariable)
-constant changing/adding of rules
e.g. Labov and Labov studied their daughter Jessie's
acquisition of inversion in Wh-questions:
Adult rule of inversion: What do you want?
Where have you been?
Why are you crying?
Who did you see?
How will you do that?
That is, MODAL/HAVE/BE inverts with subject.
Child speech is more likely to contain: What you want?
Where you have been?
Why you are crying?
Who you saw/see?
How you will do that?
JESSIEat age 3.4: NO inversion
3.5 70% How
30% Where
30% What
10% Why
3.7 80% How
50% Where
40% What
5% Why
3.9 95% How
85% Where
70% What
10% Why
4.5 95% How
95% Where
100% What
50% Why
4.9 100% How
100% Where
100% What
80% Why
-Looks as though, initially, there is NO RULE OF INVERSION.
-Then a rule is hypothesized, but its domain is limited.
-Domain of rule gradually enlarges until it is the same as
adults'.
In the light of these data, let's consider some hypotheses which
have been proposed to account for how a child learns language:
-Child memorizes sentences (SILLY)
-Child learns by uttering random strings and having only the
correct ones reinforced. (The 'behaviorist'/Skinnerian view)
-Innate programming
Let's look at the Reinforcement Hypothesis:
-What ways of reinforcement might be used?
-parent DOESN'T say: "I'll give you a bowl of ice cream if
you invert subjects with auxiliaries in all your WH-
questions."
-rather, some parents may CORRECT their children by telling
them what they should have said: OVERT POSITIVE
REINFORCEMENT.
-but, since it's pretty clear that children do not memorize
sentences, then we can't suppose that the new "correct" version
is just being stored away somewhere.
-Rather the child has to hypothesize some kind of rule to
account for the new sentence, in order for the correction to
have any effect.
-Otherwise the correction would hold ONLY FOR THAT TOKEN.
-Other types of reinforcement:
COVERT POSITIVE: child hears only grammatical strings
OVERT NEGATIVE: child is told certain strings are wrong: (flip-
side of O.P.)
COVERT NEGATIVE: child doesn't hear ungrammatical strings.
So we get a picture of the child as RULE HYPOTHESIZER:
-INPUT = linguistic data i.e. speech heard around child's
environment
-OUTPUT = rules in child's mind.
What must the kid start with to do this? Preprogramming?
The observable chain of events is:
1) Child is born: no language.
2) Child receives input/stimuli.
3) Voila! Child acquires language.
How much is acquired and how much is preprogrammed?
Well, what if the stimuli are varied? Then, if language learning
depends solely on environmental input, the learning process
should also differ:
Method: Find kids who, for some reason or other, lack certain
types of input:
1) Blind child: Kelli (Lila Gleitman and Barbara Landau)
-Original claim: Blind children are 6 months later (on the
average) than seeing children in acquiring language.
Kelli: Not late at all, above average, in fact.
-only deviation was found in her acquisition of visual
terms:
COLOR TERMS: she knew that color is a property of
concrete objects.
e.g. that you can say "red ball" but not "red birthday
party"
LOOK: has a different semantic range for Kelli: e.g.
LOOK UP
(reaches up with her hands)
vs. SEE
Turns out that the reason blind children are on the average
6 months later than sighted children in acquiring language is
that the vast majority of blind children at that time were
blinded immediately following birth by oxygen administered
because they were on the average 3 months premature. If children
who were born 3 months prematurely but were not blinded are
studied, it turns out that they too are about 6 months late in
acquiring language! Thus the lateness seems to be due to the fact
that the nervous system is not quite 'done' when the child is
born and has nothing to do with blindness.
2) Deaf child, in a signing environment (Laura Petitto)
-Child goes through PRECISELY the same stages as a hearing
child, except the modality is manual/visual instead of
oral/aural.
3) Deaf child, in an oral language environment: severe deficit,
irremediable if allowed to persist too long
4) Two deaf children (twins) in an oral environment: invent their
own sign language! (Lila Gleitman)
5) Seeing/hearing child deprived of linguistic stimuli: Genie
(Vicki Fromkin and her students) Severe deficit, irremediable.