Title: "When I was a baby I kept a diary. Recently I was rereading it. It said: Day One: Still tired from the move. Day Two: Everybody talks to me like I'm an idiot." Steven Wright
1"When I was a baby I kept a diary. Recently I
was rereading it. It said Day One Still tired
from the move. Day Two Everybody talks to me
like I'm an idiot." Steven Wright
2- LING1000/LING7100First Language
AcquisitionLecture A - Readings
- Chapter 10. Language acquisition. In Fromkin,
V., Rodman, R., Collins, P. (1997) An
introduction to language. Sydney Harcourt Brace.
pp 409-451. - Chapter 4. The acquisition of language. In
Cairns, H. S. (1999). Psycholinguistics An
introduction. PRO-ED Studies in Communicative
Disorders. Austin, TX PRO-ED. Pp 77-89.
37A.1 Biological preparedness for language
- Human infants are specially tuned to hear human
language from birth. - New born infants can hear sound distinctions
that are not distinctive, or phonemic, in their
native language. - Part of this biological preparedness is
Universal Grammar (UG), which is a
language-making capacity that specifies the
general form of human language and guides the
child in the acquisition of the target language.
4- .
- Part of this biological preparedness is
Universal Grammar (UG), which is a
language-making capacity that specifies the
general form of human language and guides the
child in the acquisition of the target language.
57A.2 Early language acquisition from sounds to
babbling
- In the first six months the infant's
vocalisations are primarily soft coos and gurgles
and do not resemble genuine language. - At about 7 months true babbling begins. The
infant starts to make sounds that exhibit
language-like features. These sounds are single
syllables consisting of stop consonant (p, b, t,
d, etc) and a vowel like "ah".
67A.2 Early language acquisition from babbling
to language
- At first the babbles will be a string of
similar syllables, ba, ba, ba. Eventually this
will become more varied, baga, bada, and will
start to sound like phonemic segments. This is
called segmental babbling. - The vocalisations will also have sentence
intonation, with the infant appearing to engage
in a "conversation" that has no semantic content.
77A.2 The one-word stage
- The first words are not produced until 12-18
months, although infants as young as 9 months can
recognise individual words from a string of
speech - The first words often sound like babble,
although they are consistently used to refer to
one thing. - This is called the one-word, or holophrastic,
stage because each word conveys as much meaning
as an entire phrase.
8Processes at the one-word stage
- During the early one-word stage that child may
underextend the meaning of a word, using it too
narrowly. For example, a child might learn the
word birdie in reference to the family budgie,
and not use it to refer to other birds. - The child will also overextend the meaning of a
word by applying it too widely, as in the case of
a child using the word doggie to refer to all
four-legged animals. Overextension appears to be
limited to production.
97A.3 The two-word stage
- When the child has learned about 50 words, a
vocabulary spurt occurs and the infant starts
producing two-word utterances from which the
grammar of the language is developed. - Fast-mapping. A child will hear a word once or
twice, learn its grammatical class, but will have
only a vague idea of what it means.
10Lexical learning principles
- At the two-word stage the child is aided by a
number of lexical learning principles. - The whole object principle.
- The only one name principle.
- The principle of extendability.
11Lexical learning principles
- The whole object principle. The child assumes
that the word used to name an object refers to
the whole object. For example, doggie will refer
to the whole dog, and not to the dog's tail,
ears, etc. -
12Lexical learning principles
- The only one name principle. The child assumes
that there is only one name for each object. When
someone points to a dog (a word a child already
knows) and says tail, the child will assume that
it is a word for a part of the dog, or may refers
to the type of dog it is, etc.
13Lexical learning principles
- The principle of extendability. The child has
the expectancy that individual words will refer
to categories of similar things.
147A.3.2 The two-word stage and the emergence of
grammar
- At the two-word stage the infant starts
producing utterances with very basic grammatical
structure. - a chair baby cookie
- a rice that Daddy
- more apple a feet
- more milk want juice
- doggie allgone pull car
15Features of two-word stage grammar
- English-speaking children obey English word
order restrictions, with subjects preceding verbs
(doggie allgone) and verbs preceding objects
(pull car). - Few function words are used.
- No bound morphemes are used by English
children, but use of inflections by children
learning highly inflected languages (e.g.,
Italian). - Subjects are often omitted.
167A.4 Beyond the two word stage
- Around the age of 3 the child will gradually
start to produce longer sentences. - This development is measured in the mean length
of utterance (MLU), the average number of
morphemes in the child's utterances. - The steady increase in MLU is related to the
increasing use by the child of bound morphemes
and function words. - The increase in MLU is also due to increasing
working memory capacity in the child, which
allows the production of longer utterances.
17Acquisition of bound morphemes in English
- Earliest to appear is the -ing marker on verbs
signalling the present progressive form and the
plural /s/ and the possessive /s/. - The function words in and on also appear, as
do the, a, and a little later, an. - Acquired much later are auxiliaries and the
copula, as well the third person singular /s/.
These forms carry little communicative meaning.
18Other features of early grammar development
- The overgeneralization (in English) of the
regular past tense verb (went ? goed) and plurals
(foots and feets). This reflects the tendency of
all children to make irregular processes in their
language more regular. - Auxiliaries and copulas also start to appear,
and with them the child begins to use subjects
consistently. - Modals and negative modals such as can't and
don't also appear. contractions like wanna and
hafta are learned as fixed forms - Passive forms are rarely used but evidence for
the beginning of movement rules (e.g.
subject-auxiliary inversion) in question
formation. Aux inversion first appears in yes/no
questions ( Is doggie sleeping?) and later in
Wh-questions (Why can't I go outside?)
19Toward more complex grammar
- At about 3 years the child will begin to
produce complex sentences. A complex sentence is
one with two verbs (e.g., Watch me draw circles).
- I want Mommy ? I want Mommy get it.
- I see you. ? I see you sit down.
- I don't know who ? I don't know who it is.
- More complex material is always placed at the
right, or the end, of the utterance in English.
207A.5 Theories of child language acquisition
Imitation
- A traditional view of first language
acquisition held that the infant learned the
language by imitating the speech of those around
him. There is strong evidence indicating that
this is not case - The child often produces forms that do not
occur in the language, as in He goed to the
store, and thus could not be imitated.
21- Even when the child tries to imitate adult
speech, they are unable to do so if the speech is
beyond his current grammar. - Adult He's going out. Adult That's
an old-time train. - Child He go out. Child
Old-time train - Adult Adam, say what I say Where can I put
them? - Child Where I can put them?
- Thus imitation alone cannot account for the
rapid development of complex grammar knowledge.
227A.5 Theories of child language acquisition
Reinforcement
- Reinforcement has also been proposed as a means
by which the child learns the grammar. The child
learns to produce correct sentences because he is
positively reinforced when he says something
right and negatively reinforced when he says
something wrong. - There is evidence that reinforcement has little
or no effect on learning. - Caretakers seldom provide positive or negative
reinforcement. - When they do explicitly correct a child's
utterance, it is usually concerns the factual
nature of the statement and not the grammatical
structure used.
23- There is little evidence that reinforcement has
much of an effect on learning. - Caretakers seldom provide positive or negative
reinforcement. - When they do explicitly correct a child's
utterance, it is usually concerns the factual
nature of the statement and not the grammatical
structure used.
24- Even when explicit grammatical correction is
attempted, it is often unsuccessful - Child Nobody don't like me.
- Mother No, say 'Nobody likes me'.
- Child Nobody don't like me.
- (dialogue repeated eight times)
- Mother Now, listen carefully, say 'Nobody likes
me'. - Child Oh, nobody don't likes me.
25- Neither imitation or reinforcement alone are
enough to account for child language acquisition.
26First language acquisition
27Lecture B First Language Acquisition
287B.1 Language acquisition as rule-making
- A theory of first language acquisition must
account for the fact that - children make non-random errors
- acquisition happens in a regular manner across
diverse languages and settings - the basic grammar is learned quickly
- acquisition comes about without formal
instruction - acquisition happen regardless of the kind of
interaction the child is exposed to when
learning.
29Child grammar development is systematic
- The systematic nature is evident in the stages
that many children go through in the mastery of
negation (also over regularising the past tense
ed) - Declarative I want some food.
- Stage 1 No want some food.
- No added to the beginning of utterance.
- Stage 2 I no/don't want some food.
- Unanalysed negative element inserted.
- Stage 3 I don't want no food.
- Negative element inserted negation spread, some
becomes no - Stage4 I don't want any food.
- Negative element inserted correctly some changed
to any - Although not all children show exactly the same
development as above, all show similar regular
changes. Note that the child has never been
exposed to these forms and thus could not imitate
them. This rule-making must thus be driven by
other processes. - Sign language also follows the same
developmental pattern as spoken language.
30- Although not all children show exactly the same
development as above, all show similar regular
changes. Note that the child has never been
exposed to these forms and thus could not imitate
them. This rule-making must thus be driven by
other processes. - Sign language also follows the same
developmental pattern as spoken language.
317B.2 Universal grammar
- The evidence is strong that the infant is born
with a innate capacity to learn the first
language which is unique to humans
(species-specific). Language acquisition is
assumed to obey a biological timetable that
specifies a period in which it can take place,
that is, when the child must be exposed to
language. - During this critical period, language learning
is swift and relatively effortless. After this
period the acquisition of grammar is difficult if
not impossible.
32- This innate capacity is called Universal
grammar (UG). UG provides the blueprint for the
general form of language, specifying what human
languages can look like. - UG also specifies how individual languages can
vary. For example, a basic difference is whether
a language marks grammatical information by word
order (like English) or by inflection (like
Spanish or Italian). - The UG parameters provide the child with
possible alternatives to choose from when
attempting to make sense of the target input.
33The role of the environment
- The linguistics environment thus serves as a
trigger for the emergence of the innate UG
knowledge by providing positive evidence for the
learner. The child does not need to discover
either the basic organizational features or the
parameters through experience. It is already
provided by UG. - Speech in the child's Universal Grammar Grammar
- environment ? Acquisition Principles
? and - Lexicon
347B.2.1 Linguistic competence and language
performance
- The child's underlying linguistic competence
exceeds his productive ability. - Data from naturalistic production
underestimates the child's competence, so
researchers try to elicit evidence of this
underlying knowledge through experimental tasks
357B.3 The innateness hypothesis
- The child is assumed to be born with a
language-making capacity that specifies the form
that human languages can take and greatly
facilitates the child's acquisition of the target
language. - This biological capacity is different from the
general cognitive processes that are used for
learning and problem-solving in other domains
acquiring a first language is fundamentally
different from learning how to play chess or hit
a cricket ball. - The view that the child is born with an innate
language-making capacity is strongly supported by
the fact that there is a critical period for
first language learning. Estimates what that
critical age is range from 7 - 15 years.
36Evidence for a critical period
- The study of other species, particularly the
learning of birdsongs - Human children raised in early isolation from
human contact and then later unable to master the
language. These include the French case of
Victor, 'the wild child of Aveyron' in the late
1800s, and Genie, a more recent case in the US. - Empirical studies showing adult language
learners are unable to develop the native-like
intuitions of child learners.
377B.5 First language acquisition and social
interaction.
- Social interaction also plays an important role
in first language acquisition. - Researchers like Halliday approach the
development of language as the mastery of the
language devices and functions needed for social
interaction. - The first four functions, which develop in the
first year. - instrumental ('I want' function) object-oriente
d used to satisfy material needs - regulatory ( 'do as I tell you')
person-oriented used to regulate
behavior - interactional ('me and you') used for getting
along with others - personal ('here I come') used to express
individuality
38- Soon after these four functions are established
three more functions appear. - heuristic ('tell me why') used to learn about
reality - imaginative ('let's pretend) used to create
imaginary worlds - informative ('I've got something to tell you')
- used to communicate content
397B.7 Can chimps learn human language?
- The claim that human language is an innate,
species-specific capacity has been challenged by
researchers who have attempted to teach human
language non human primates like chimps and
gorillas. They have had limited success - Non-human primates are not able to make the
range of sounds needed to express human language,
so sign language and symbolic languages using,
e.g. plastic tokens, have been used.
407B.7 Non-human primate language learning
All-Stars
- Washoe was raised like a deaf child, and by the
time she was 4 years old had learned eighty five
signs. She was also able to put together simple
'sentences' like baby mine, gimme flower, more
fruit. - Kanzi was taught a language using plastic
lexigrams that represent morphemes. Researchers
reported that Kanzi's was able to create a
grammatical expression ("dog go"). Little
subsequent grammatical development was reported. - Nim Chimpsky learned about 125 signs in the
four years of the study. Nim used only a basic
grammar, showed no tendency to be more creative
as he learned more, and little evidence of
spontaneous speech. - Koko, a gorilla, learned several hundred signs
and could string together sentences and even coin
new words like finger bracelet for ring.
41- The chimps and gorillas showed great intelligence
in mastering the communication systems taught.
However, unlike human children, they showed
little creativity or spontaneity in their
language use, and were able to master only a
basic grammar.
42- First language acquisition Week 7
- End of Lecture B.