Child language

acquisition
General Linguistics
Candide Simard
 Child language acquisition

1. main research methods

2. milestones of child language
acquisition
3. Cross-linguistic and cross-cultural
aspects of CLA
4. Major theories to explain CLA
 main research methods
2 main objectives in collecting data:
naturalness: similar to language child uses everyday, with everyday partners
representativeness:
a) what child usually says
b) how general population speaks

1. parental diaries (from 18th C onwards): one observer/one child

– risk of errors and omissions
– focus on out-of-ordinary samples

2. observational studies (1960s onwards see Brown, A first language, the early
stages): audio or video recordings, small number of children, natural contexts,
regularly over longer period of time
– Longitudinal: same participants over long period of time
– Cross-sectional observational studies: groups of similar age compared
Transcription and Annotation: time –extensive so usually small numbers of
child participants
http://www.media.mit.edu/cogmac/projects/hsp.html
https://www.youtube.com/watch?v=TgrQlhVPBjc

researcher share their data: CHILDES: Child Language Data Exchange System
http://childes.talkbank.org/
3. experimental studies: from more or less naturalistic in design to tightly
controlled tests, narrowly designed research questions. Data elicited under
carefully designed techniques. Less overall data from each participant, but
usually larger number of participants.
• e.g. HASP (high amplitude sucking paradigm): infants suck at a
higher rate when presented with novel stimuli.
• e.g. production and comprehension of syntax…Wug test
• e.g. Truth-value judgment task… story is told then s=children
asked if statement made about it is correct or not
 HASP (high amplitude sucking paradigm)
Intra-uterine Learning
Fetuses appear to be sensitive to
prosody, the characteristic rhythm,
tempo, cadence, melody, intonational
patterns with which a language is
spoken
What do newborns “know” about
language?
• Discriminate consonants and vowels
• Count syllables
• Prefer infant directed speech
• Categorically discriminate content
vs. function words
• Prefer listening to their mother’s
voice
• Prefer stories and songs heard
prenatally
• Prefer listening to the native
language
See Janet Werker, and William Fifer
http://www.uottawa.ca/vr-recherche-research/frontier/pdf/Janet_Werker.pdf
MILESTONES OF CHILD
LANGUAGE ACQUISITION
All children of the world:
•similar path of language development
•reach milestones in same order.
 first sounds

Speech perception during first year:

• First few months: can discriminate between sounds such
as /p/, /pha/ and /b/ . As they acquire phonological
system of their mother tongue, they become less able
to discriminate sounds from other languages
– e.g. Werker on Hindi-English, English learning infants
(6-8 months) and Hindi speaking and English speaking
adults. Sounds /da/ and /Da/ (dental retroflex)
• Young infants discriminate native and non-native consonant
distinctions
• Adults have difficulty discriminating some of the same non-
native consonants

 children have ability to learn any language, but capacity

to hear like a native quickly dissipates.
 first sounds
• Babies first sound produced: cry! Not intentional attempt to
communicate

• 2nd to 5th month: cooing, vowel like sounds indicating pleasure or

playfulness

• 4th to 6th month (until 1 year): babble, vowel and consonant sounds.
Often palatal and labial. Start to conform to sound patterns of
adults between 6th and 10th month.
– Adult speakers can discriminate between babble of Chinese, Arabic,
English, or French infants

– Innate and unconscious, also interactive and social.

– Children encouraged to babble by caregivers smiles and touches, give

and take interactions called proto-conversations (from 5 month some
infants can imitate simple sound sequences)

– http://www.youtube.com/watch?v=37JxkDNJ0Aw&feature=related
 first words
• Holophrastic stage (around 1 year): child associates meaning
to sounds they produce  characterised by one-word
sentence.
• Use one-word phrase to communicate variety of complex
functions.
• Parents understand various meanings with contextual cues
(rich interpretation).
• No link between when first words are uttered and later
intelligence!
• Concrete objects grounded in and central to everyday
experiences and interactions.
• content words and not function words
– English children: nouns
– Korean children: verbs

Overextension: meaning of one word extended to all

words of similar category, sometimes reverse process:
underextension.
 2-word stage

Around age 2: 2-word stage.

• For English: noun and verb baby sleep, or verb and
modifier eat now.
– Ordering is not fixed and morphological marking not
used.
• Comprehension far outweighs production.
– Around 1 year, children understand 70 words but can
produce 6.
• Vocabulary spurt (from the end of 2nd year):
children production of vocabulary develops
rapidly. ≈200 words a month…
 first sentences

• ‘Many diary, observational and experimental

studies concerned with how children become
competent users of their language’s system of
morphology and syntax’.
• Seems to involve the formation of internal ‘rules’.
• Comprehension skills still outpace production.
Children react to correct grammatical
representation of meaning in their native
language.
• Morphoplogical and syntactic development is
predictable.
• Many patterns and processes are constant across languages
and cultural groups.

• https://www.youtube.com/watch?v=wxtLhgzntg8
 first sentences in English
From 2 to 5 years: multiple-word stage.
Inflectional morphology (widely studied by Brown, 1973)

Order Morpheme Example

1 Present Progressive I driving
2-3 Prepositions in, on
4 Plural balls
5 Irregular Past Tense broke, fell
6 Possessive Daddy's chair
7 Uncontractible Copula This is hot
8 Articles a, the
9 Regular past tense She walked
10 3rd person present tense, regular He works
11 3rd preson present tense, irregular She does
12 Uncontractible auxilliary Ross is winning
13 Contractible copula He's a clown
14 Contractible auxiliary She's drinking
l
 Mean Length of Utterance
MLU: widely used measurements of the complexity of
children’s language.
• Calculated from average number of morphemes per
utterance.
1. order of acquisition is similar
2. age at which they are acquired varies from child to child.
3. MLU good index of level of development (better than age).
– But: more recent studies have stressed the importance of
vocabulary as predictor of grammatical development.

More conclusions from Brown:

• the order of morpheme acquisition does not depend on
frequency of exposure (in parental speech).
• the morphemes are acquired in order of syntactic and
semantic complexity.
 predictable stages
• Phase 1: The child uses the correct past tense of go, for instance, but does
not relate this past tense went to present-tense go. Rather, went is treated
as a separate lexical item.

• Phase 2: The child constructs a rule for forming the past tense and begins
to overgeneralize this rule to irregular forms such as go (resulting in forms
such as goed).

• Phase 3: The child learns that there are many exceptions to this rule and
acquires the ability to apply this rule selectively.

• Maybe from parents’ perspectives, development is U-shaped –children can

appear to be decreasing rather than increasing in their accuracy of past-
tense use as they enter phase 2.
– But this apparent back-sliding is an important sign of linguistic development.
• Developmental sequences in other areas of grammar.
• ‘Tendency for researcher to focus on ‘mistakes, but given many
opportunities for incorrect and unconventional language use, grammatical
erros are in fact uncommon’… (p. 219)
 wugs
Jean Berko’s wug study
• Children understand
the rule of plural
formation before
they are 5 years old.
In developing rules
children go through
predictable stages
http://www.youtube.com/watch?v=ElabA5YICs
A

http://www.youtube.com/watch?v=IeF33vu2K_
M&feature=related
Cross-cultural perspectives
Onset times for both word comprehension (8 to
10months) and word production (11 to 13 months);
Wide individual variation exists within each language
concerning the pace and size of vocabulary
growth.
In languages, where verbs are learned before nouns
(i.e. Korean), the children tend to perform better
on tasks which are related to verbs, while English-
speaking children perform better on noun-related
tasks. Why???

→In general, one can say that structural

differences do influence the nature of the
developmental sequence for each language.
CROSS-CULTURAL
PERSPECTIVES
Cultural differences in child rearing, different ideologies concerning
language use and what is meant to be a ‘good’ child, → resulting in
different interactional patterns with and around the infant.

Samoan: children are not believed to have individual personalities or

control over their behavior; very young children are not expected
to initiate talk, early vocalizations are not interpreted as
meaningful attempts to communicate.

Central American homes, infants’ interactions are mostly with multiple

social partners, European-American homes, infants’ interactions
are mostly with one adult person at a time.

In the United States

European-American mothers tend to actively participate in verbal
exchanges with their children and place greater attention on task-
specific goals (organizing a narrative chronologically) than on social
conversational goals (like including all those present in the
conversation).
In the highly social environments of Latino homes, mothers’ roles
might be to support children’s conversations with others.
Researchers have reported that Mexican-American family
members explicitly instructed preverbal infants to participate in
social conversations through the use of ‘tell him/her’. Through
these early instructions, children learn to participate in multiparty
conversations from a very young age.
Another important difference in
language acquisition concerns not
which languages are being learned,
but rather the number of languages
acquired by a child
 bilingualism
Bilinguals demonstrate greater metalinguistic awareness
(knowledge and awareness about language as a
system), greater mental flexibility, and the ability to
think more abstractly, still bilingual education
remains a controversial topic.
• bilingual children go through a period of code-mixing:
move back and forth between their two languages,
seemingly without discrimination.
→ universal among bilingual children and is apparent
even at the babbling stage.
→ could be early code-switching, even very young
children have the social or strategic competence to
move between two languages depending on the
conversational context.
Studying bilingual children helps in understanding the nature and
organization of the two languages in the brain.

• Does a bilingual child begin with just one grammar and lexical
system that later becomes differentiated as the child learns to
distinguish between the two languages?
the unitary system
– evidence is that all young children seem to go through a period of
mixing their two languages, particularly at the lexical level.

• Or does a bilingual child have two grammatical and lexical

systems from the outset?
the separate systems hypothesis.
– evidence point to the fact that even very young children can
differentiate between their languages prior to entering the two-
word phase and are often sensitive to their interlocutors’ language
competences.
• Different kinds of interactions between child and
other speakers of different languages.
• Amount of exposure to each language directly
contributes to competence levels i.e. amount of
interactive exposure and size of the active vocabulary
in each language.
– most children are stronger in one language,
– and nearly all children tend to associate each language with
particular contexts, skills, and activities.
– So in fact very few balanced bilinguals → tend to use each
of their languages for different functions.
• Bilinguals are not 2 monolinguals in one person but
individuals whose competences reflect their
particular learning experiences and
patterns of language use.
THEORIES: EXPLAINING
HOW CHILDREN ACQUIRE
LANGUAGE.
Theories must explain central facts of child
language development:
• the rapid rate of development
• the systematic regularities concerning
which items are early or late acquired, as
well as the routine errors children make in
this process; and
• how language learning can be accounted for
given children’s varied environment.
– each of the explanations is embedded in
different schools of thought or paradigms
within the field of linguistics and psychology.
 behaviourism
One of the earlier theories, held that language is essentially a habit, a
behaviour like any other, which is mastered through general learning
principles.
– imitation,
– reinforcement, and
– punishment.
• The child’s more target-like utterances are rewarded (and thus tend to
be repeated) while the non-target like utterances are ‘punished’ (and
thus generally disappear over time). As parents’ expectations for their
child’s language change as the child grows, they alter their
reinforcement strategies.
Stimulus → response → reinforcement
• Behaviourists tend to focus on observable behaviours rather than
internal or innate processes.
– they assume that children are essentially “interested bystanders,” bringing
no special abilities or innate mechanisms to bear on the language acquisition
process.
• Commonsense appeal, behaviourism is no longer the dominant research
paradigm in the study of child language, one its main proponents B.F.
Skinner.
 nativists
Language is not the result of general learning mechanisms, but rather is an innate
capacity. This special capacity is limited to humans and differs in important ways
from any type of animal communication.

Arguments based on several observations:

All children acquire language easily and rapidly, most adults typically struggle for
decades to master the complexities of a second or foreign language, children
reach near mastery of their native tongue in just a few short years, without
instruction or apparent effort.
All children, regardless of the language they are learning or the quantity or the
quality of input they receive from their caregivers, acquire their first language
at the same rate and by progressing through the same developmental stages.
Adult speech young children hear is a poor model – filled with incomplete sentences,
false starts, slips of the tongue, etc. Nevertheless, children take this
fragmentary and degenerate input and are able to construct a complex grammar
– far more complex than they could have ever learned from reinforcement or
general learning mechanisms.

But:
poverty-of-stimulus argument: input alone is inadequate to support children’s
language learning.
 nativists
Children rarely receive specific feedback, also negative evidence,
on the grammaticality of their utterances, as adults typically
focus on the content of a child’s utterance rather than its
linguistic accuracy.
→ when adults do provide feedback or explicit language
instruction to their children, the children are by and large
oblivious to it.

Children create or generate a rule-based system

• If children were truly relying upon imitation and reinforcement
to learn their native language, why then would they produce such
utterances as he goed to the store or we saw mouses today?
– Never been uttered by an adult.

• overregularization errors are evidence that the child is in the

process of creating language, testing hypotheses about language,
and, in general, acquiring the rules of grammar.
The only possible explanation for the uniformity of the language acquisition process,
the complexity of the linguistic knowledge children possess at such young ages
despite the scarcity of the feedback they receive, is that language must be
innate.

Language is:
• functional ‘organ’, like vision and hearing capacities.
• language is a species-specific or uniquely human cognitive capacity which is the
result of
• an language faculty.

There is, a time limit, also known as a critical or sensitive period (Lenneberg,
1969), for this process to take place, and some evidence suggests that after
this period has ended (typically around puberty), complete acquisition of a first
or second language becomes difficult, if not impossible.
i.e. Genie, wild boy of Aveyron

• Now, even if the nativists approach to language acquisition is still a major force
within and beyond the field of linguistics, in recent years it has come under
increasing attack from several directions on both theoretical and empirical
grounds.

– One of its main proponents: Chomski →LAD (Language Acquisition Device), blueprint in
the brain that enables them to recognise and manipulate the structure of language
 Connectionism
Nativists argue that the linguistic input children receive is not rich enough
to support the extraction of complex linguistic generalizations;
therefore, children must by necessity be endowed with an innate
knowledge of linguistic rules that guide the language acquisition
process.
• In Chomsky’s most recent account, the concept of the linguistic rule has
been largely abandoned in favor of some very general principles he has
called universal grammar.
• Two such general principles reflect, first, that language structures can
be recursively embedded, and second, that two structures can merge to
form a new entity.

Cognitive scientists challenge the poverty-of-stimulus argument. Another

approach, connectionism or ‘parallel distributed processing’, proposes
that general learning processes, such as our brain’s capacity to process
information that is distributed across the whole brain, are sufficient
for at least some aspects of language learning.
Major concept: the network → consists of a large number of
nodes that may connect to other nodes by pathways that vary in
their strength.
• Key characteristic of networks is that they need not have
recourse to rules. Rather, they have the ability to make
associations based on regularities they detect in the input

• Children can learn the regularities of language through an

inductive process based on exposure to many examples, that is
to frequency of occurrence.

• Connectionists have created computer models (also called neural

networks) that are trained to learn by exposure to many
examples.

• Connectionist models are, however, criticized for what is called

task veridicality (i.e., conditions under which networks are fed
input differ from those in which children are exposed to
language).
→ the increasing sophistication of connectionist models will,
if nothing else, motivate linguists to reconsider the nativist
approach.
 Interactionism and usage-
based approaches
• Nativists: emphasize the importance of innate linguistic knowledge
• Connectionists: stress the role of general learning mechanisms.
• Social interactionists point to the importance of child-caregiver interactions in
the language acquisition process.

The characteristics of the language used within these interactions, on what is called
child-directed speech (or motherese, or baby talk). Caregivers, when
interacting with children, tend to use a special form of speech – including short,
simple sentences with higher pitch and exaggerated intonation contour, as well
as sentences focused on the objects and events in the child’s environment.
• Caregiver speech may also include:
– increased use of diminutives, as well as repetition and imitation.
– recasts – more target-like reformulations of the child’s original ungrammatical
utterance – to help the child master more complex language forms.
This type of speech, while varying in shape and form and not used in all speech
communities in the same way, is believed to help attract the child’s attention to
problematic forms and to actively involve him/her in the conversation.
Another important concept that is seen to facilitate language acquisition: joint
attention, very often established by the uniquely human skill of pointing.
• the social interactionist paradigm: determine the
relationship between these interactions and children’s
language development.
• Child-directed speech and establishing joint attention are
two powerful tools geared to facilitate language
development.
• Social interactionists tend to disagree with nativists about
the nature of children’s input and specifically with the
poverty-of-stimulus argument. In particular, they point to
the fact while parents do not teach grammar in any formal
way, children receive many types of feedback about the
effectiveness of their language every time they speak, and
this effectiveness is related to the grammatical
correctness of their utterances. Researchers from this
camp have shifted their focus from rules and grammar to
the lexicon and the role played by vocabulary
development; they have documented the importance of
frequency of language input children receive and shown
a clear relationship between the language which
children hear and the language which they produce.
 Interactionism and usage-
based approaches
• ‘Interactionists’ : have produced a
constructive account of language
acquisition that starts out small, builds
on what has been achieved and
emphasizes the role of caregivers’ input
(Tomasello, Constructing a Language,
2003).
• This approach ties the study of the
emergent first language to a usage-based
approach, moving by the principle of
analogy from item-based phrases to more
complex constructions.
 Interactionism and usage-
based approaches
• So, contrary to the Chomskyan approach, it is argued that
infants do not need a language instinct to explain how
children learn language. Instead, it is argued that the
essence of language is its symbolic dimension, which rests
on the uniquely human ability to comprehend intention.
Grammar emerges as the speakers of a language create
constructions out of recurring sequences of symbols;
children pick up these patterns in the buzz of words they
hear around them.
• Children do not operate with abstract linguistic entities but
on the basis of concrete item-bases constructions… same
mechanisms that children use to acquire words are also used
to acquire grammar. ‘mastery of artifacts and conventions
that children may adapt for creative uses as their mastery
progresses’ (Tomasello, 2001)
 To sum up…
• All theories of language acquisition in a
way now assume these fundamental skills:
– intention reading (mainly attention-based)
– pattern-finding (including perception of
similarity and analogy detection).
• Yet while the nativist approach posits a
further set of acquisition processes to
connect somehow with an innate universal
grammar, the usage-based approach
argues that this is completely unnecessary.
All that we need is frequency of input plus
a few general cognitive abilities.
 Nature/nurture
• ‘In a broader framework, what we meet here is yet another
version of the long-standing debate about the relative
importance of nature and nurture in human development,
with the naturalists stressing the importance of biological
and genetic programming (‘instinct’), and the nurturists
pointing to the role of the environment. While there is
growing agreement that both nature and nurture are
critical, scientists still disagree on the relative importance
of each.
• In addition, the study of language acquisition raises many
important and interesting questions about the specifics of
brain organization and brain functioning, and about the
organization of human cognition in general.
• Thus, the study of first language acquisition is a dynamic
field with many unanswered and exciting questions, whose
importance definitely goes far beyond linguistics’.
 References
Janet Werker
– http://infantstudies.psych.ubc.ca/
• William Fifer
– http://nyspi.org/DevelopmentalPsych/secti
ons/research/wpfifer.htm
• Deb Roy’s Speechome project
• http://web.media.mit.edu/~dkroy/resea
rch/index.html
• Berko, J. (1958). The child's learning of
English morphology. Word, 14, 150-177