Transcript PSY 369: Psycholinguistics - the Department of Psychology at

PSY 369: Psycholinguistics
Language Acquisition II
Language Sponges
How do they do it (and what are they doing)?
Learning words

12 ms
2 yrs
3 yrs
6 yrs


first words
200 words
1,000 words
15,000 words
About 3,000 new words per year, especially in the primary
grades
As many as 8 new words per day
Language Sponges

Learning words


General patterns and observations
Proposed Strategies





Fast mapping
Whole object
Mutual exclusivity
Learning Syntax
Learning Morphology
Early word learning

First words (Around 10-15 months)

Emergence of systematic, repeated productions of
phonologically consistent forms

Idiomorphs - personalized words




Developed in systematic ways
Not simply imitation, rather are creative
Learned importance of consistency of names
Typically context bound (relevant to the immediate
environment)


Important people, Objects that move, Objects that can
be acted upon, Familiar actions
Nouns typically appear before verbs
Semantic Development

Naming “Explosion”

1-general names


2- specific names



“red”
5-personal/social


“mommy”
3-action words
4-modifiers


“dog”
“yes, no, please”
6-functional

“what”
Semantic Development

Word Invention








to broom (to sweep)
to fire (to burn)
to scale (to weigh)
a fix-man (a mechanic)
a tooth-guy (a dentist)
a locker (a lock)
bum wiper (bathroom tissue)
yester-minute (a minute ago)
Semantic Development

Applying the words to referents

Extension


Finding the appropriate limits of the meaning of
words
Underextension


Applying a word too narrowly
Overextension

Applying a word too broadly
Semantic Development

Later words:

Children come to use words in more adult-like ways


Words start to be used in wider range of contexts
Children use wider range of word types:
 referential words (ball, doggie, chair)
 proper names (Mummy, Spot)
 actions (open, wash, tickle)
 properties, states, qualities (more, gone, up, on, dirty)
 social-pragmatic words (no, please)
 few ‘frozen’ phrases (all gone, what’s that)
Extensions of meaning
“tee”
Extensions of meaning
“tee”
1:9,11
Extensions of meaning
“tee”
1:9,11
1:10,18
Extensions of meaning
“tee”
1:9,11
1:10,18
1:11,1
“googie”
Extensions of meaning
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
Extensions of meaning
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
1:11,24
“googie”
Extensions of meaning
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
1:11,24
1:11,25
“tee/hosh”
Extensions of meaning
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
1:11,24
1:11,25
1:11,26
“tee/hosh”
“hosh”
Extensions of meaning
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
1:11,24
1:11,25
1:11,26
1:11,27
“tee/hosh”
“hosh”
“pushi”
Extensions of meaning
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
1:11,24
1:11,25
1:11,26
1:11,27
2:0,10
“tee/hosh”
“hosh”
“moo-ka”
“pushi”
“hosh”
Extensions of meaning
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
1:11,24
1:11,25
1:11,26
1:11,27
2:0,10
2:0,20
“tee/hosh”
“hosh”
“moo-ka”
“pushi”
“hosh”
“biggie googie”

One-word-per-referent heuristic
Extensions of meaning


If a new word comes in for a referent that is already named, replace it
Exception to that was “horse,” but it only lasted a day here
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
1:11,24
1:11,25
1:11,26
1:11,27
2:0,10
2:0,20
“tee/hosh”
“hosh”
“moo-ka”
“pushi”
“hosh”
“biggie googie”
Strategies for learning

Expansion and contraction can occur at the same time
“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
1:11,24
1:11,25
1:11,26
1:11,27
2:0,10
2:0,20
“tee/hosh”
“hosh”
“moo-ka”
“pushi”
“hosh”
“biggie googie”
Strategies for learning
Child tries different things, if a word doesn’t work then
try something else

e.g., hosh didn’t for for the large dog, switched to
biggie doggie

“tee”
1:9,11
1:10,18
1:11,1
1:11,2
“googie”
1:11,24
1:11,25
1:11,26
1:11,27
2:0,10
2:0,20
“tee/hosh”
“hosh”
“moo-ka”
“pushi”
“hosh”
“biggie googie”
Indeterminacy: Frog
Frog
Frog?
Green?
Ugly?
Jumping?
Quine’s gavagai problem

The problem of reference:


a word may refer to a number of referents (real
world objects)
a single object or event has many objects, parts
and features that can be referred to
Frog
Frog?
Green?
Ugly?
Jumping?
Learning word meanings

Learning words

Fast mapping

Using the context to guess the meaning of a word
Please give me the chromium tray. Not
the blue one, the chromium one.


All got the olive tray
Several weeks later still had some of the meaning
Constraints on Word Learning

Markman (1989)

Perhaps children are biased to entertain certain
hypotheses about word meanings over others


These first guesses save them from logical ambiguity
Get them started out on the right track



Object-scope (whole object) constraint
Taxonomic constraint
Mutual exclusivity constraint
Strategies for learning
Object-scope (whole object) constraint

Words refer to whole objects rather than to parts of
objects
Dog
Strategies for learning
Taxonomic constraint


Words refer to categories of similar objects
Taxonomies rather than thematically related obejcts
‘Here is a lux’
‘Show me another lux’
Strategies for learning


But in ‘no-word’ conditions, they would be
shown the first picture
See this? Can you find another one?
Strategies for learning
% Theme / Category
4 and 5 year olds' choice of theme vs. category
100
Theme
Category
70
40
10
-20
No word condition
Novel word condition
Strategies for learning
Mutual exclusivity constraint (Markam and Watchel 1988)


Each object has one label & different words refer to
separate, non-overlapping categories of objects
An object can have only one label
‘Show me a dax’:

they choose the corkscrew

because it is a less well known object for which they
don’t have a label yet.
Problem with constraints

Most of the constraints proposed apply only to object
names.


There have been cases where children have been
observed violating these constraints


What about verbs? (Nelson 1988)
Using for example the word ‘car’ only to refer to ‘cars moving
on the street from a certain location’ (Bloom 1973)
The mutual exclusivity constraint would prevent
children from learning subordinate and superordinate
information (animal < dog < poodle)
Language explosion continues

The language explosion is not just the result of simple
semantic development; the child is not just adding
more words to his/her vocabulary.

Child is mastering basic syntactic and morphological
rules.
Language explosion continues

Proto-syntax (?)

Holophrases

Single-word utterances used to express more than the meaning
usually attributed to that single word by adults
“dog”
might refer to the dog is drinking water


May reflect a developing sense of syntax, but not yet
knowing how to use it
Controversial claim (e.g., see Bloom, 1973)
Language explosion continues

Syntax

Basic child grammar (Slobin, 1985)


Similarities across all languages
Mean length of utterance (MLU) in morphemes

Take 100 utterances and count the number of morphemes
per utterance
Daddy coming. Hi, car. Daddy car comed. Two car outside. It
getting dark. Allgone outside. Bye-bye outside.
# morphemes: 3, 2, 4, 3, 4, 2, 2
‘-ing’ and ‘-ed’ separate morphemes
‘allgone’ treated as a single word
MLU = morphemes/utterances
= 20/7 = 2.86
Language explosion continues
6
5
MLU
4
3
2
1
0
0
10
20
30
age (months)
40
50
60
Language explosion continues

Syntax

Roger Brown proposed 5 stages

Stage 1: Telegraphic speech (MLU ~ 1.75; around 24 months)



One and two word utterances
Debate: learning semantic relations or syntactic (position rules)
Children in telegraphic speech stage are said to leave out the ‘little
words’ and inflections:
 e.g. Mummy shoe NOT Mummy’s shoe
 Two cat NOT two cats
Language explosion continues

Syntax

Roger Brown proposed 5 stages

More than two words

Stages 2 through 5
 Stage 2 (MLU ~2.25) begin to modulate meaning using word
order (syntax)
 Later stages reflect generally more complex use of syntax
(e.g., questions, negatives)
How do kids learn the syntax?

Innateness account

Pinker (1984, 1989)

Semantic bootstrapping
Child has innate
knowledge of
syntactic categories
Child learns the
and linking rules
meanings of
some content words Child constructs some
semantic representations
Child makes guesses
of simple sentences
about syntactic structure
based on surface form
and semantic meaning
How do kids learn the syntax?

“It is in the stimulus” accounts

Children do not need innate knowledge to learn grammar

Speech to children is not impoverished (Snow, 1977)
 Children learn grammar by mapping semantic roles (agent, action,
patient) onto grammatical categories (subject, verb, object) (e.g.
Bates, 1979)