Language

Acquisition
Studies in Linguistics
Language Acquisition
→How do we learn a language?
→How do we acquire the rules of language?
→Do we memorize things like sentence
structures?
Mechanisms of Language
Acquisition
Skinner (1957) language as a kind of verbal
behavior
Children learn language through imitation,
reinforcement, analogy, etc.
Chomsky (1959) language is a complex
cognitive system
It cannot be acquired by behaviorist principles
Do Children Learn through
Imitation?
Do Children Learn through
Correction and Reinforcement?
Do Children Learn through
Correction and Reinforcement?
Parents generally tend to correct “content”
(truth value) rather than “grammar”
Recasting ?
Do Children Learn through
Correction and Reinforcement?
Do Children Learn through
Analogy?
The child hears: “I painted a red barn.”
Through analogy produces: “I painted a blue barn.”
From a model you extend it to all of the new cases
by similarity
“I painted a barn red.” (we can switch the order?)
What about a different verb?
Heard: “I saw a red barn.”
Through analogy: “I saw a barn red.” But
something’s gone wrong…
Do Children Learn through
Structured Input?
Simplified language? Motherese, child-directed speech (CDS), baby talk
(emphasis is on the environment)
We tend to talk to young children in a special way: more slowly and more
clearly.
We may speak in a higher pitch and exaggerate our intonation, and
sentences are generally grammatical.
Motherese is not syntactically simpler:
•syntactically complex sentences such as questions (Do you want your juice
now?);
•embedded sentences (Mommy thinks you should sleep now);
•imperatives (Pat the dog gently!);
•negatives with tag questions (We don’t want to hurt him, do we?)
Analogy, Imitation,
Reinforcement?
Analogy, imitation, and reinforcement cannot account
for language development because they are based on
the (implicit or explicit) assumption that what the child
acquires is a set of sentences or forms rather than a set
of grammatical rules.
Theories that assume that acquisition depends on a
specially structured input also place too much emphasis
on the environment rather than on the grammar-making
abilities of the child.
These proposals do not explain the creativity that
children show in acquiring language, why they go
through stages, or why they make some kinds of
Children Construct Grammars

Language acquisition is a creative process.

Children are not given explicit information about
the rules, by either instruction or correction.
They extract the rules of the grammar from the
language they hear around them.
Children Construct Grammars
Observations of children acquiring different
languages under different cultural and social
circumstances reveal that the developmental
stages are similar, possibly universal.
Children are equipped with an innate template or
blueprint for language=Universal Grammar (UG).
This blueprint aids the child in the task of
constructing a grammar for her language=the
innateness hypothesis.
The Innateness Hypothesis
Poverty of the Stimulus
Although children hear many utterances, the
language they hear is incomplete, noisy, and
unstructured.
Child-directed speech is largely well formed, but
children are also exposed to adult–adult
interactions.
These utterances include slips of the tongue, false
starts, ungrammatical and incomplete sentences,
and no consistent information as to which
The Innateness Hypothesis
But somehow children come to know aspects
of the grammar about which they receive no
information.
In this sense, the data they are exposed to is
impoverished.
It does not account for the richness and
complexity of the grammar they attain.
The Innateness Hypothesis
According to the innateness hypothesis, the child
extracts from the linguistic environment those
rules of grammar that are language specific, such
as word order and movement rules.
But he does not need to learn universal principles
like structure dependency, or general principles of
sentence formation.
All these principles are part of the innate blueprint
for language that children use to construct the
grammar of their language.
The Innateness Hypothesis
What accounts for the ease, rapidity, and uniformity of
language acquisition in the face of impoverished data?
The answer is that children acquire a complex grammar
quickly and easily without any particular help beyond
exposure to the language because they do not start from
scratch.
UG provides them with a significant head start. It helps
them to extract the rules of their language and to avoid
many grammatical errors.
Because the child constructs his grammar according to an
innate blueprint, all children proceed through similar
developmental stages.
Learning & Development of
Language

LANGUAGE: THE FIRST 5 YEARS OF LIFE OF LEARNING AND

DEVELOPMENT
HTTPS://WWW.YOUTUBE.COM/WATCH?V=U49ULLCULEK
The Perception and Production
of Speech Sounds
It is possible for children to learn any human language
they are exposed to.
During the first year of life, the infant’s job is to uncover
the sounds of the ambient language.
From around six months, he begins to lose the ability to
discriminate between sounds that are not phonemic in
his own language.
His linguistic environment molds the infant’s initial
perceptions.
They have begun to learn the sounds of the language of
their parents.
Babbling
At around six months, the infant begins to
babble.
This is not chaos! The 12 most frequent
consonants in the world’s languages make
up 95% of the consonants infants use in their
babbling.
Early babbles are a repetition of consonant-
vowel sequences: mama, gaga, dada.
Later babbles are more varied.
Babbling
By the end of the first year the child’s babbles come to
include only those sounds and sound combinations that
occur in the target language.
Babbles begin to sound like words, although they may not
have any specific meaning attached to them.
At this point adults can distinguish the babbles of an
English-babbling infant from those of an infant babbling in
Cantonese or Arabic.
During the first year of life, the infant’s perceptions and
productions are being fine-tuned to the surrounding
language(s).
First Words
Some time after the age of one, the child
begins to repeatedly use the same string of
sounds to mean the same thing.
At this stage children realize that sounds
are related to meanings.
They have produced their first true words.
The age of the child when this occurs varies
and has nothing to do with the child’s
intelligence.
First Words
First Words
Most children go through a stage in which their utterances
consist of only one word = the holophrastic or “whole
phrase” stage
These one-word utterances seem to convey a more
complex message.
For example, when J. P. says “down” he may be making a
request to be put down, or he may be commenting on a
toy that has fallen down from the shelf.
When he says “cheerios” he may simply be naming the
box of cereal in front of him, or he may be asking for some
Cheerios.
Segmenting the Speech Stream
How do infants discover where one word begins
and another leaves off?
Speech is a continuous stream broken only by
breath pauses.
The ability to segment the continuous speech
stream into discrete units—words—is one of the
remarkable feats of language acquisition.
Segmenting the Speech Stream
Studies show that infants are remarkably good at
extracting information from continuous speech.
They seem to know what kind of cues to look for
in the input that will help them to isolate words.
One of the cues that English-speaking children
attend to that helps them figure out word
boundaries is stress.
Segmenting the Speech Stream
Every content word in English has a stressed syllable.
(Function words such as the, a, am, can, etc. are ordinarily
unstressed.)
If the content word is monosyllabic, then that syllable is stressed
as in dóg or hám.
Bisyllabic content words can be trochaic, which means that
stress is on the first syllable, as in páper or dóctor, or iambic,
which means stress is on the second syllable, as in giráffe or
devíce.
The vast majority of English words have trochaic stress.
Segmenting the Speech Stream
Stress is very salient to infants, and they are quick to
acquire the rhythmic structure of their language.
Just a few months old infants are able to discriminate
native and non-native stress patterns.
Before the end of the first year their babbling takes on
the rhythmic pattern of the ambient language.
At about nine months old, English-speaking children
prefer to listen to bisyllabic words with initial rather than
final stress.
Studies show that infants acquiring English can indeed
use stress cues to segment words in fluent speech.
Segmenting the Speech Stream
In a series of experiments, infants who were seven
and a half months old listened to passages with
repeated instances of a trochaic word such as
púppy, and passages with iambic words such as
guitár.
They were then played lists of words, some of
which had occurred in the previous passage and
others that had not.
Experimenters measured the length of time that
they listened to the familiar versus unfamiliar
Segmenting the Speech Stream
The results showed that children listened
significantly longer (indicated by turning their
head in the direction of the loudspeaker) to words
that they had heard in the passage, but only when
the words had the trochaic pattern (pú ppy).
For words with the iambic pattern (guitár), the
children responded only to the stressed syllable
(tár), though the monosyllabic word tar had not
appeared in the passage.
These results suggest that the infants—like adults
Segmenting the Speech Stream
Following such a strategy will sometimes lead to errors
(for iambic words and unstressed function words), but it
provides the child with a way of getting started.
This is sometimes referred to as prosodic
bootstrapping.
Infants can use the stress pattern of the language as a
start to word learning.
Segmenting the Speech Stream
Infants are also sensitive to phonotactic constraints and to
the distribution of allophones in the target language.
For example, in English aspiration typically occurs at the
beginning of a stressed syllable—[pʰɪt] versus [spɪt]—and
that certain combinations of sounds are more likely to
occur at the end of a word rather than at the beginning,
for example [rt].
Studies show that nine-month-olds can use this
information to help segment speech into words in English.
Segmenting the Speech Stream
golabupabikututibubabupugolabubabup
ututibu. . .
The infants were able to distinguish the words
from the nonwords
By tracking the frequency with which the different
sequences of syllables occur
Babies are sensitive to statistical information as
well as to linguistic structure to extract words from
the input