Lecture 13

Language acquisition
 Content

• Language acquisition
• Child Language Acquisition
• Bilingualism
• Adult Language acquisition
• Second Language acquisition
 Language acquisition

Language acquisition refers to the process of acquiring a language, usually due to immersion (i.e. hearing the
language often and in everyday contexts). Most of us acquire our native language just from being around
others such as our parents.
The study of child language acquisition is the study of the processes by which children learn a language. At
a very young age, children begin to understand, and gradually use, the language spoken by their caregivers.
The study of language acquisition involves three main areas:
• First-language acquisition (your native language i.e. child language acquisition).
• Bilingual language acquisition (learning two native languages).
• Second-language acquisition (learning a foreign language). Fun fact - There's a reason why French
lessons were so di cult - babies' brains are much more primed for language learning than our adult brains
are!
ffi
 Child language acquisitioin
First Language acquisition

1. There are four main stages in child language acquisition:

1. The babbling stage (3-8 months)
2. Children rst start to recognise and produce sounds eg 'bababa'. They don't yet produce any recognisable words but
they are experimenting with their newfound voice!
2. The one-word stage (9-18 months)
1. The one-word stage is when babies start to say their rst recognisable words, eg using the word 'dog' to describe all
u y animals.
3. The two-word stage (18-24 months)
1. The two-word stage is when children start communicating using two-word phrases. For example, 'dog woof ', meaning
'the dog is barking', or 'mummy home', meaning mummy is home.
4. The multi-word stage (telegraphic stage) (24-30 months)
1. The multi-word stage is when children start to use longer sentences, more complex sentences. For example, 'mummy
and Chloe go school now'.
fl
ff
fi
fi
 Pre-birth: Preparation of the human brain for language acquisition

Language acquisition begins well before a child is born. Babies are initially familiarized with speech
and language in the womb. The human ear begins to function at the 3 trimester or the 7-month
rd

mark of pregnancy (Saxton, 2017). During this period, unborn infants respond to all types of sounds.
What is more fascinating is that they can discriminate speech sounds from non-speech sounds. In
the womb, babies are more sensitive to their own native languages than foreign ones, and also make
a well-de ned distinction between their mother’s voice and others (Saxton, 2017). This
phenomenon is one of many natural occurrences that supports the claim that language is hardwired
in the human brain.
This stage of language perception/hearing in the womb is essential for language acquisition after
birth. Children who are born with congenital auditory impediments or hearing loss are deprived of
this stage. As a result, language acquisition is generally delayed by four to ve months after birth
(DeLuzio, 2020).
fi
fi
 Perception and production of speech sounds

• Infants display an ability to discriminate and recognize speech sounds

• – They will even respond to linguistic contrasts when those contrasts are not present in
the language(s) spoken around them
• – They can perceive di erences in voicing, place of articulation, manner of articulation
• – But they do not react to nonlinguistic aspects of speech (loudness, gender-based
pitch di erences, etc.)
• Infants appear to be born with the ability to perceive and focus on the sounds that are important
for language, so they can learn any human language
– But by 6 months babies begin to lose to ability to discriminate between sounds that are not
phonemic in the language(s) they are acquiring
•
ff
ff
 Babbling

At this stage, infants begin to explore the properties of sounds through production. The sounds of early
babbling are universal. However, by the time a child reaches the age of 8 months, a drift occurs in the
characteristics of babbling (Helms-Park, 2018). Babbling becomes more distinctive. Infants begin to make
sounds that would only occur in their own native languages.
A French baby and an American baby will not babble the same, as there are certain sounds that English and
French do not share. At this stage, parents should expect their children to go through three phases. They follow:
• One consonant and a vowel at a time. Example: “ma,” “da,” “du,” “bi”
• Reduplicative babbling. Example: “da-da-da-da”
• Non-reduplicative babbling. Example: “ba-du-ba-du.”
In these phases, the most frequent vowels and consonant sounds that are produced by English babies are: ‘i’, ‘u’,
‘a,’ and ‘b’, ‘d’, ‘g’, and ‘m’. Vowels are generally easier to produce than consonants, as they do not require many
obstructions in the vocal or nasal tract.
 Babbling

• Babbling begins at about 6 months and is considered the earliest stage of language
acquisition
• Babies may babble phonemes that do not occur in the language(s) they are
acquiring
• 95% of babble is composed of the 12 most common consonants around the world

• Early babbles mostly consist of CV sequences but become more varied later on
• By 1 year babbles are composed only of the phonemes used in the language(s)
they hear
• Deaf babies babble with their hands like hearing babies babble using sounds
•
Babbling initially contains a wide repertory of speech sounds, and nally zeroes in on the speech sounds of the target
language.
• Around 8-10 months, adults can recognize the target language of a babbling child (French, Cantonese, Arabic).
By 6 months, Chinese children's babbling contains more tonal variation than the babbling of an American child.
• Deaf children babble vocally as well, though naturally there's no gradual approximation to a target language.
• Also: gestures analogous to babbling are noted in deaf children learning ASL. Conclusion: inner biological
clock governs onset of babbling; this clock is unaware of child's deafness.
• Just as spoken language babbling shows properties of the language being acquired, so does sign language
babbling. Children with tracheostomy surgery that eliminates vocal production do not hear themselves babble. When
the tracheostomy hole is closed, there are some delays in speech, but they quickly jump to age-appropriate behavior,
skipping various stages.
Conclusion: Babbling is linguistic (not motoric) — but it is not "practice" essential for language acquisition. It is a by-product of
the process of language acquisition, not an essential building block.
fi
 One word stage

During this stage, children begin to acquire and produce real words of their native languages. A child in
this stage will use single-word constructions to communicate. The use of single-word items is meant to
convey full sentences provided the context. Usually, children go through a holophrastic stage, where their
one-word utterances may convey more meaning
– up used to indicate something in the sky or to mean “pick me up”

As expected, word productions are extremely redundant. The one-word productions have three functions:
• Express emotions: “uh-oh,” “bad”
• Request a desired action: “up,” “down,” “gimme”
• Name objects: “baba” (bottle), “truck,” “doggie,” “mine”
At this stage, children know the meanings of words that they say, as speaking is often accompanied by
nger-pointing (Helms-Park, 2018). This suggests that children know more language than they can express
fi
 Lexicon
Discerning word-boundaries:
Kids already know more than they say. Work of (the late) Peter Jucszyk and others:
• In 11-month old, selective looking reveals preference for pauses that coincide
with word boundaries over pauses inserted between syllables of words.
• In 9-month old, no preference.
So what happens between 9 and 11? A bootstrapping problem.

• How do you segment speech into words if you don't know the words? No magic bullet, it seems!
two possibilities:
• Prosodic bootstrapping:
Some common, but not universal phonological phenomena respect word boundaries.
• For example, in French, the nal a of panorama is longer than the rst a of matimaticien. 3-day old infants can distinguish ma-ti from panorama typique
from ma-ti out of matematricien. So the information, if relevant to guessing word boundaries, is perceived... [Mehler et al -- this is a sucking/habituation experiment]
• Also, typical stress patterns get noticed. Czech - initial / French - nal / Polish - penultimate. English: 90% of nouns have stressed initial syllables. If a
child anticipates this sort of regularity, it at least narrows hypotheses... In a 1993 study by Juszyk, Cutler and Redanz, 9-month old American infants listened longer
to words with initial stress than to other words. No such di erence among 6-month olds.
• Maybe these features are exaggerated in production by parents (motherese)?
• Statistical bootstrapping
Sequences of syllables that overlap word boundaries are often less likely than sequences of syllables within words, simply because it's words that are being learned.
• pretty baby: [tibej] (pret-ty ba-by) less common because it's not a word. If kids are sensitive to these probabilities, they can form hypotheses about
word/non- word on the basis of these probabilities.
• Computer corpus analysis does fairly well (emphasize "fairly") nding word boundaries this way. Do babies do this too? Perhaps.
fi
ff
fi
fi
fi
 Two word stage

Children usually enter this stage when they have acquired about 50 words. They begin to demonstrate
their knowledge of the word order that occurs in their language. English has a SUBJECT VERB
OBJECT (SVO) word order. It is very common for production to take the following structures:
• Doer + action: “Tommy play”
• Action + a ected object: “Kick ball”
• Possessor + possessed object: “Daddy key”
• Negation/Refusal/ Cessation of existence: “No veggie,” “all gone milk”
At this stage, lots of verbs, nouns, and adjectives emerge, as well as intonation! Intonation is a clear
indication that children do not perceive speech production as one chunk, but rather as individual
words coming together to convey a more complex meaning.
ff
 After the age of one, children gure out that sounds are related to meanings and start to produce their rst words
• Usually children go through a holophrastic stage, where their one-word utterances may convey more meaning
– up used to indicate something in the sky or to mean “pick me up”

• This suggests that children know more language than they can express
A major obstacle that babies must over come is to be able to identify where word boundaries are
• English-speaking children may be able to use stress as a cue for word boundaries (prosodic bootstrapping)
– Every content word in English has stress
• If a word has two syllables, the stress either falls on the rst syllable (trochaic stress) or the second syllable (iambic stress), but
the vast majority of English words have trochaic stress
– Experiments have shown that children do use stress as a cue for word boundaries since most English words have stress on the
rst syllable
• Buthowdochildrenknowthestresspatternofthe language they are acquiring?
– Babies may use statistical frequency of syllable sequences to determine word boundaries
• In one experiment, babies were able to recognize the nonsense words pabiku, tutibu, golabu, and babupu out of
strings of nonsense syllables because those strings of syllables in the ‘words’ occurred more frequently than the random strings of
syllables
• Children may use statistical strategies to determine a few words, and from there may be able to determine the
rhythmic, allophonic, and phonotactic properties of the language and then can determine even more words from this knowledge
fi
fi
fi
fi
 Morphology
Kid`s task is to

• determine the morphemes/words of the language

• determine the rules that combine morphemes into words
Example:
learn that walked contains two morphemes

• On the one hand...

Some evidence that language acquisition device is "pretuned" to pick up morphological patterns.
Those of you who read the paper by Safran, Aslin & Newport on statistical learning by infants (one of the choices for the rst paper) might be interested in a follow-up by Marcus et al
(1999) in Science:
In a clever twist on the Sa ran et al. experiment, Gary Marcus of NYU and three colleagues did a similar experiment with 7-month old (actually, slightly younger) infants. (Science, January 1,
1999). Once again, a two-minute stream of synthesized speech was played to the infants. Once again, the speech stream was composed of three-syllable nonsense words, made from
synthesized speech, with no cues identifying the word boundaries. Once again, the children were played di erent recordings in a later presentation which they could control by gazing at
or looking away from a blinking light.
This time, however, the nonsense words in the rst presentation were not randomly chosen sequences of syllables, but conformed to simple rules, or templates. One group of infants, for
example, heard a string of nonsense words in which the second two syllables were identical (ga ti ti, li na na). Another group heard words in which the rst and third syllables were identical
(ti ga ti, na li na).
The key recording in the second presentation did not consist this time of the same words rearranged, but consisted of entirely new words that followed the same rule as the rst set of
words. Could infants distinguish new words that followed the rule from new words that did not? Strikingly, they could. The infants preferred novel second presentations over familiar
second presentations. But "familiar" here meant -- not containing the same "words" -- but containing new words that followed the old rule. In their words: "Infants [can] extract abstract
algebra-like rules that represent relationships between placeholders (variables), such as “the rst item X is the same as the third item Y,” or more generally, that “item I is the same as item
J...In addition to having the capacity to represent such rules, our results appear to show that infants have the ability to extract those rules rapidly from small amounts of input and to
generalize those rules to novel instances."
ff
fi
fi
ff
fi
fi
fi
Wug test
On the other hand
The acquisition of morphology clearly demonstrates the rule-governed nature of language acquisition
– Children typically learn a morphological rule and then overgeneralize
– Children go through three stages in the acquisition of an irregular form:
• • In phase 1 they use the standard irregular past tense forms because they have learned these irregulars as separate lexical items (broke, brought)
• • In phase 2 the child has learned the rule for past tense and therefore attaches the regular past tense morpheme to the irregular verb (breaked, bringed)
• • In phase 3 the child realizes that there are exceptions to the morphological rule and bring the standard irregular forms back into their vocabulary (broke, brought)
• The“wugtest” demonstrates correct plural allomorph to nouns they have never heard before
– Which shows they have an understanding of natural classes of phonemes and are not just imitating words they have heard before
•• Children acquiring languages other than English learn subject-verb agreement very early
•• Children also demonstrate their knowledge of derivational rules and can create new words
– E.g. broomed (“swept”)
• At about two years of age, children start to put words together to form two-word utterances
– The intonation contour extends over the two words as a unit, and the two-word utterances can convey a range of meanings:
• mommy sock = subject + object or possessive
• Chronological age is not a good measure of linguistic development due to individual differences, so instead
linguists use the child’s mean length of utterance (MLU) to measure development
 Telegraphic stage

At this stage, children experience a vocabulary spurt or “explosion.” Production is pidgin-like, as

grammatical/function words (little words) such as ‘the,’ ‘a,’ ‘is,’ ‘will,’ ‘of,’ ‘by,’ pluralization, tense (past
-ed), verb endings/person agreements (she eat ‘s’) are omitted.
During this stage, a child will primarily communicate with content words (verbs, nouns, adjectives) and
some pronouns such as ‘me’ and ‘mine.’
Although many function words are missing at this stage, children obey the word order and structure of
their language. Considering the English word order (SVO), English speaking children are expected to
produce phrases similar to the following sentences:
• “Want more ice-cream.”
• “Mommy go bye-bye.”
• “He go play.”
• The telegraphic stage describes a phase when children tend to omit function morphemes such as articles, subject pronouns, auxiliaries and
verbal in ection. For example: He play little tune or Andrew want that.
• However, while function morphemes are absent, these sentences have hierarchial constituent structure like adult sentences:
• Telegraphic utterance are not just words strung together and reveal the child`s knowledge of syntactic rules.
• A child must know syntactic categories of words in order to apply syntactic rules
• Semantic bootstrapping: the notion that children rst use the meaning of a word to gure out its syntactic category
– Word frames may also help children determine the syntactic categories for words
• Some frames such as you__it and the___one occur frequently enough that kids may be able to identify which words can occur in each frame
(verbs for the former and adjectives for the latter)
• Between 2;6 and3;6 a language explosion occurs and children undergo rapid development
– By the age of 3, most children consistently use function morphemes and can produce complex syntactic structures:
• He was stuck and I got him out • It’s too early for us to eat
• • After 3;6 children can produce wh-questions,and relative pronouns
• • Sometimeafter4;0childrenhaveacquiredmostofthe adult syntactic competence
fl
fi
fi
Syntax
 After telegraphic stage

Early in this stage, complete sentences begin to emerge. Imperative and declarative sentences appear rst (Helms-Park,
2018):
• Imperative (command): “Give me my toy.”
• Declarative (Relay information): “That’s my toy truck.”
Coordination (acquired rst) and subordination sentences also appear early in this stage, however, they are mastered by
4-5-years-old (Helms-Park, 2018):
• Coordination (Two clauses joined in a symmetrical relation): “I am tall, but she is short.”
• Subordination (Two clauses joined in a non-symmetrical relation): “When I am older, I am going to be
bigger.”
Progression of more complex questions and negatives appear between 2.6- 3.0-years-old (Helms-Park, 2018). By a child’s 3 rd

birthday, they begin to use more grammatical/function words, such as regular plural ‘-s,’ endings on verbs ‘-ing,’
prepositions (of, in, by), but third person agreement comes a bit later “She loves candy.” Near the end of this stage,
almost everything is acquired between the ages of 5-6 (Helms-Park, 2018).
fi
fi
 The acquisition of pragmatics

Deixis:
– Children often have problems with the shifting reference of pronouns
• Children may refer to themselves as ‘you’
– Problems with the context-dependent nature
of deictic words
• Children often assume the hearer knows who she is talking about
 The acquisition of sign language

• Deaf babies acquire sign language in the same way that hearing babies acquire
spoken language:
– babbling, holophrastic stage, telegraphic stage
• When deaf babies are not exposed to sign language, they will create their own signs,
complete with systematic rules

• – This demonstrates the drive humans have to communicate, and also the innate basis for
language since these children create a rudimentary language without any input
 Given this guideline of language acquisition, parents should note that kids are
di erent in their rate of language development. A child who may be behind does not
necessarily indicate a language impediment. Again, every child is unique and
di erent. Some may acquire language faster than others.
It is always important to be mindful that comprehension of language always precedes
production by some months. Language competence does equate to production. Most
children who encounter delays in language development, including auditory
impediments, usually catch up with their peers by the age of 3 (DeLuzio, 2020).
ff
ff
 Theories of language acquisition
• Cognitive theory suggests that children go through stages of language development. Theorist Jean Piaget emphasised that we
can only move through the stages of language learning as our brains and cognitive processes develop. In other words, children
have to understand certain concepts before they can produce the language to describe these concepts. Theorist Eric Lenneberg
argued that there is a critical period between two years old and puberty in which children need to learn language, otherwise, it
cannot be learned su ciently well.
• Behavioural theory, often called 'Imitation Theory', suggests that people are a product of their environment. Theorist BF
Skinner proposed that children 'imitate' their caregivers and modify their language use through a process called 'operant
conditioning'. This is where children are either rewarded for desired behaviour (correct language) or punished for undesired
behaviour (mistakes).
• Nativist theory, sometimes referred to as the 'innateness theory', was rst proposed by Noam Chomsky. It states that children
are born with an innate ability to learn language and that they already have a "language acquisition device" (LAD) in their brain
(this is a theoretical device; it doesn't really exist!). He argued that certain errors (eg 'I runned') are evidence that children actively
'construct' language rather than just imitating caregivers.
• Interactionist theory emphasises the importance of caregivers in child language acquisition. Theorist Jerome Bruner argued
that children do have an innate ability to learn language however they require lots of regular interaction with caregivers to achieve
full uency. This linguistic support from caregivers is often called 'sca olding' or a Language Acquisition Support System
(LASS). Caregivers may also use child-directed speech (CDS) that helps a child learn. For example, caregivers will often use a
higher pitch, simpli ed words, and lots of repetitive questioning when talking to a child. These aids are said to enhance
communication between the child and caregiver.
• Halliday`s function of language
fl
fi
ffi
ff
fi
 What are Halliday`s functions of Language?

Michael Halliday suggested seven stages that show how the functions of a child's language become more
complex with age. In other words, children express themselves better and better as time goes by. These stages
include:

• Stage 1- Instrumental Stage (language for basic needs eg food)

• Stage 2- Regulatory Stage (language to in uence others eg commands)
• Stage 3- Interactive Stage (language to form relationships eg 'love you')
• Stage 4- Personal Stage (language to express feelings or opinions eg 'me sad')
• Stage 5- Informative Stage (language to communicate information)
• Stage 6- Heuristic Stage (language to learn and explore eg questions)
• Stage 7- Imaginative Stage (language used to imagine things)
fl
 Bilingualism
Bilingual language acquisition, or simultaneous bilingualism refers to the acquisition of two languages simultaneously from infancy
– About half the people in the world are bilingual or multilingual
– In many parts of the world, bilingualism (or multilingualism) is the norm
Unitary system hypothesis: the idea that the child initially constructs only one lexicon and one grammar
• • Evidence for: language mixing similar to codeswitching; lexical items existing in only one language
• • Evidence against: there is a lot of overlap in the lexicon for each language, and children may have gaps because each language is
used in di erent contexts and they can only learn so many words each day
Separate systems hypothesis: the idea that the child builds a distinct lexicon and grammar for each language
– Evidence for:
• • where the two languages diverge grammatically, the child will
acquire two di erent sets of rules
• • bilingual children select which language to use based on the context
• • children bilingual in sign language and a spoken language may say a word in one language and sign it in the other
simultaneously
It’s unclear how much input in each language a child needs to become bilingual
– Une personne-une langue (one person, one language) is the strategy where one parent speaks only language A to the child and the other speaks
only language B
• Bilingual children tend to have better metalinguistic awareness than monolingual speakers, meaning they have more conscious knowledge
about language
ff
ff
 Second language acquisition

Most adult language learners never become fully pro cient in their second language
– They make errors unlike children’s errors and these errors may become fossilized
• Fundamental di erence hypothesis: learning a second language is a di erent process than learning a rst language
– Di erent principles are drawn upon in L2 learning than L1 acquisition
– However, L2 learners do demonstrate rule-governed interlanguage grammars
One obvious di erence between L1 and L2 acquisition is that in L2 acquisition a speaker already knows a language
– Learners often transfer phonological, syntactic, and morphological rules from their rst language to their second
language
• But, not everything transfers from the L1 to the L2, and many errors made by learners are not found in
their L1
• Speakers with di erent L1s go through similar stages when learning their L2s
– Which points to some possibly universal developmental principles like those in L1 acquisition
ff
ff
ff
ff
fi
ff
fi
fi
Heritage language learners constitute a unique type of adult language learner
– Someone who was raised with a strong cultural connection to a language and who then chooses to
study that language more formally
• May have no prior linguistic knowledge of the language
• May be bilingual

Most researchers would not claim that it is impossible to acquire a new language after a certain age
– But it does get harder as one gets older
– There may be “sensitive” (rather than critical) periods
for acquiring certain aspects of an L2
• The sensitive period for phonology is the smallest—it is very difficult to acquire an L2 without an accent after the childhood years; other aspects
of language have a larger window
•