Overview
Central questions about first language acquisition
Methods of studying first language acquisition
First language acquisition as an innate behavior
Theories of first language acquisition
Milestones in first language acquisition
First language acquisition
i.e. developmental (as opposed to experimental) psycholinguistics
- How is it that by age 5 children know their language?
- What they do along the way and why?
- Is language development independent of intelligence, other cognitive skills?
- How is it that children are able to completely master a first language, whereas adults rarely can completely master a second language?
Methods of studying first language acquisition
Production studies
Productions are spontaneous or elicited
spontaneous production studies: typically diary of subject is kept from ages 1-3; this type of study is good for studying the age of acquisition of phonology, lexicon
elicited production studies: e.g. which doll should he pick up? kid might answer: the one that's jumping or the jumping one; this type of study is good for eliciting knowledge of syntax at a particular age
Comprehension studies
act-out tasks
e.g.: experimenter: "make the hippo jump over the rhino. then make bullwinkle jump over him." kid: moves the toys
these don't work with really young children
judgement tasks
e.g. experimenter: "X happened. is that right?" kid: yes/no
these can be used with younger children
perception tasks
can be used even with infants: measure pacifier sucking rate, heart beat
habituation to task (e.g. listening to [pa]): measured by slower sucking rate
new sound presented (e.g. [ba]): sucking rate increases
introspection
not well explored as experimental technique
probably works well with older kids only
Production lags behind comprehension
How to determine if child has acquired a particular characteristic?
Assumption: correctly produced --> acquired; if mistakes are made has not acquired that feature; but language learners can correctly comprehend much more than they can produce
- Recognition of polite forms precedes the ability to produce them.
Puppets requesting candy used direct forms like:
‘Give me candy.’
Or indirect forms like:
‘I would like some candy.’ Or: ‘May I have some candy?’
Indirect forms were judged more polite.
Direct vs. indirect commands
direct commands produced before indirect commands, but even before indirect commands are produced, their politeness function is acquired:
“The children listened to two different puppets requesting candy, and had to decide who asked most nicely for it. For example, one puppet might say:
Give me candy.
And the other:
I would like some candy.
Or:
May I have some candy?
Bates (1976) found that children three years old thought the puppet that asked indirectly was more deserving of the candy, but they could not explain this choice. If pressed for an explanation, they often said “this puppet said ‘please’,” although none of the puppets did! So it seems as though recognition of polite forms precedes the ability to produce them.”
- Recognition of sounds precedes the ability to produce them.
“fis” example in O’Grady p. 444: ‘One of us, for instance, spoke to a child who called his inflated plastic fish a fis. In imitation of the child’s pronunciation, the observer said: “This is your fis?” “No,” said the child, “my fis”. He continued to reject the adult’s imitation until he was told, “That is your fish.” “Yes,” he said, “my fis.”
(Child can recognize sh vs. s before able to produce both sounds.)
Theories of first language acquisition
competence: list-like and rule-like
acquisition of list-like competence (knowing which words/morphemes are part of your language):
by the time a child is 6, typically has a vocabulary of about 14,000 words. children may acquire much of this kind of competence through something like memorization (but probably involves much more structure--semantic decomposition/componential analysis).
Theories of acquisition of rule-based competence:
acquisition of rule-like competence (rules for phonology, morphology, syntax). different hypotheses:
- Imitation hypothesis: children learn solely by imitating what they hear
- Reinforcement hypothesis: children learn by being positively or negatively reinforced for certain kinds of behavior
- Active construction of grammar hypothesis: children are actively constructing and refining a grammar of the language of their environment (much like linguists).
Evidence for Active construction of grammar hypothesis:
- Children don't get a lot of corrections
they do get some lexical/content corrections
they don't get a lot of grammatical corrections
(problem for reinforcement hypothesis)
- Children don't absorb a lot of the corrections they do hear
Child: / My teacher holded the baby rabbits and we patted them.
Adult: / Did you say your teacher held the baby rabbits?
Child: / Yes.
Adult: / What did you say she did?
Child: / She holded the baby rabbits and we patted them.
Adult: / Did you say she held them tightly?
Child: / No, she holded them loosely.
adult: held, said twice, ignored both times
Child: / Nobody don’t like me.Mother: / No. Say ‘nobody likes me’.
Child: / Nobody don’t like me.
(preceding dialogue repeated 8 times)
Mother: / Now listen carefully. Say ‘nobody LIKES me’.Child: / Oh...Nobody don’t LIKES me.
(problem for reinforcement hypothesis)
- Children produce novel utterances
children can come up with novel sentences: children produce sentences they have never heard before--such as sentences which are ungrammatical by adult standards:
new syntax:
‘other one spoon’
new morphology:
causative:
'you're fedding me up'
(wants mother to change sister’s diaper before feeding her) ‘Don’t eat her yet. She’s smelly!’
‘These flowers are sneezing me!’
nouns used as verb:
‘Put me that broom. Let’s get brooming.’
‘Why you didn’t jam my bread?’
‘I hate you and I’ll never unhate you or nothing!’
(problems for Imitation hypothesis: these are not obviously not imitations of the adult language)
- Children make systematic, not random, errors
there is a systematicity to their errors which suggest that they are learning rules; i.e. errors are rules
children appear to be equipped with some knowledge of what a possible human lg is
Phonological
Natural, not random, classes of consonants
Inventory of English consonants (age 2)
voiced and voiceless stops, voiceless fricatives, nasals, labial glide
p b / t d / k gf / s / h
m / n
w
Inventory of English consonants (age 4) (adult system, minus interdental fricatives, [])
p b / t d / k gt d
f v / s z / / h
m / n /
l
w / r / j
Novel phonological rules
child / target / rule:“[gu] here” / glue / no consonant clusters are allowed
“it not [lu] off” / flew
“no me [lip]” / sleep
“[kak] ticking” / clock
“daddy [kk]” / stick
“allgone [t]” / twig
“eat [ol]” / granola
“more [brd]” / bread
“mummy [gb]” / give / syllable-final consonants are stops
“me got [æpm]” / asthma
“me [ll]” / little / only vowels can be syllable peak
“bus [ltu] no” / little
“it [btu] me” / bitten
“take [mnæn]” / banana / all consonants in a word must be either oral or nasal
Morphological
Regularizing and overgeneralizing errors tend to decrease after age 2:
Regularization of plurals: gooses
Regularization of past tense forms of verbs: heared, hitted, goed, bringed, comed; I tooked it smaller
Regularization of comparative forms of adjectives: He hitted me. He’s a puncher he is. He’s being badder and badder.
Semantic
Overextension/broadening:
child’s word / first referent / extensions‘mooi’ / moon / cakes, postmarks, round marks on window, the letter O, round shapes in books
‘bird’ / sparrows / cow, dog, cats, any moving animal
‘fly’ / fly / specks of dirt, dust, all small insects, child’s own toes, crumbs, small toad
‘koko’ / rooster crowing / piano, phonograph, tunes played on violin, accordian, all music, merry-go-round
‘wau-wau’ / dog / toy dog, soft slippers, picture of old man in furs, all animals
note that overgeneralization makes the most of a small vocabulary
Underextension/narrowing:
semantic underextension also happens but is less frequent than overgeneralization
child’s word / first referent / extensions‘car’ / the family’s Pontiac / none
‘plant’ / the fern in the kitchen / none
‘mow-mow’ / the family’s cat / none
‘dish’ / the child’s dish / none
Syntactic
Three stages in the acquisition of negative sentences
stage / productions / rule1 / No...wipe finger. / Attach ‘no’ or ‘not’ to the beginning of a sentence.
No a boy bed.
No singing song.
No the sun shining.
No money.
No sit there.
No play that.
No fall!
Not...fit.
Not a teddy bear.
More...no.
Wear mitten no.
2 / No square is...clown.
I can’t catch you. / ‘no’, ‘not’, ‘can’t’, and ‘don’t’ appear after the subject and before the verb
I can’t see you.
We can’t talk.
I don’t want it.
Don’t bite me yet.
No pinch me.
He no bite you.
He not little, he big.
That no Mommy.
There no squirrels.
Touch the snow no. / ‘no’ can appear at the end of a sentence
This a radiator no.
3 / We can’t make another broom. / no’, ‘not’, ‘can’t’, ‘don’t’ and ‘won’t appear after the subject and before the verb
I don’t want cover on it.
I gave him some so he won’t cry.
I didn’t see something.
I not hurt him.
I not crying
That not turning.
Don’t put the two wings on.
I didn’t did it.
You didn’t caught me.
I am not a doctor. / ‘not’ appears after forms of ‘be’
It’s not cold.
This not ice cream.
I isn’t...I not sad.
3rd stage shows some remaining problems with auxiliaries ‘be’ (I not crying), forms of ‘be’ (I isn’t),
First language acquisition as an innate behavior
intense interest in first lg. acquisition by linguists because characteristics of first lg acquisition seem to support Innateness Hypothesis. i.e., children must be equipped from birth with certain lg. learning capabilities.
first language acquisition seems to be an innate behavior, with the characteristics identified by Lenneberg (discussed in File 82); cf. differences between walking, learning to play football or ballet dancing (not innate behaviors)
Characteristics of first language acquisition
(all terms used by Chomsky)
- Poverty of stimulus
`very limited data suffice for the language faculty of the mind/brain to provide a rich and complex language'. children are exposed to:
adult performance, not competence, with variety of speech errors, omissions; somehow they figure out competence underlying performance.
motherese: slowed down (more pauses); short, simple sentences; repetitions; exaggerated intonation; higher pitch overall, diminutives; motherese appears to be universal or nearly so
- Speed of learning
- Lack of instruction
- Cross-linguistic regularities in learning
- Uniformity of resulting grammars
Characteristics of innate behaviors
Innate behaviors: walking, language
Not innate behaviors: playing football, ballet dancing, etc.
1. Emerge before needed.
lg, walking emerge before child has to fend for him/herself;
ballet might never emerge without instruction/coaching
2. Not the result of a conscious decision.
learning ballet, someone has to decide to take ballet lessons
3. Not triggered by (extraordinary) external events (other than immersion in linguistic environment).
acquisition of skill in ballet is triggered by immersion in ballet environment and drilling in ballet skills
4. Not effected by explicit instruction:
child is not explicitly taught to speak or walk, in fact, resist correction;
ballet is explicitly taught, corrections are an important part of learning process
5. Normal stages of achievement, independent of environment, can be identified.
in talking, walking there are milestones, normal achievements (see file 86), which occur at typical ages independently of language environment;
mastery of certain ballet skills depends on individual; students of ballet take different amounts of time to master a pirouette (apocryphal story: it takes 1 million repetitions to nail a pirouette), and some might never master a pirouette
6. Critical age
hypothesis: there appears to be a critical age for language learning. explains: (1) why adults don't learn a second language easily, with full mastery (particularly of phonology)
(2) children learn languages easily without teaching, but if not exposed to a language, lg. ability atrophies after certain point, just as there is a critical age for acquisition of vision
case of Genie (described in Readings) (discovered age 14 in 1970, studied 1970--1979), other feral children (Isabelle mentioned in Files). Genie turned out to be more advanced cognitively than linguistically
is there a critical age for learning to walk? cf. evidence concerning critical age for development of vision
is there a critical age for learning of ballet? (some required abilities, such as extreme flexibility and ability to spin, do seem to be acquired more easily by children than adults)
lateralization and Critical Age hypothesis: thought possibly lateralization of brain related to critical age hypothesis
lateralization (for some skills) found in fetuses, newborns, infants (males?)
but possibly 2 hemispheres not so specialized for lg. or certain other functions at birth:
infants whose left hemispheres removed (eg. tumor)---young brains can recover from trauma whereas adult brains cannot: if left hemisphere at birth, right hemisphere can take over language functions so that language use in later years is almost perfect (incomplete mastery of syntax);
adult brains cannot do this. in fact, if hemisphere is damaged at adolescence or later, and damage is to left hemisphere in strongly lateralized subject, then there will be lifelong speech problems
so, when does lateralization of brain cease? originally thought puberty--when ability to learn lg. ceases, but some new evidence that lateralization possibly finished by age 5--note most of lg. acquired by age 5
Heny suggests that lateralization precedes language; perhaps in an evolutionary sense, lateralization paved the way for language and language gravitation to left because of specialization for analytical, sequential processing
Regular milestones in first language production
children learn language in stages; basically, the older they get, the better they get.
these stages found in all languages!
Babbling
0-1 months: crying, coughing
2-3 months: cooing, gooing; velar consonants
4-6 months: produce greater variety of sounds, sounds more like language
7-9 months: CV syllables, often reduplicated; e.g. [tata]
12 months: relatively long sequences of gibberish
18-20 months: babbling ceases
universal characteristics of babbling; babbling independent of what sounds are heard:
deaf children babble
hearing children of deaf parents babble.
some sounds produced may not be heard in child's house;
possibly children are learning difs. between sounds of their lg., sounds not in their lg., practicing right ones. also start producing correct intonation contours of sentences.
Categorical perception of speech sounds
said above that production lags behind comprehension
some spectacular evidence that comprehension, categorization of speech sounds begins at quite an early age
When sounds are perceived, they are assigned to categories (phonemes).
Phonetic categories have prototypical (good examples) and non-prototypical (bad examples) members.
There are differences between languages with respect to prototypes: adult Swedish speakers' prototype /i/ is different from English speakers' prototype /i/; Swedish speakers will often judge English /i/ more like Swedish /e/.
Prototypes are like “perceptual magnets'': sounds which are at a certain perceptual distance from the prototype are perceived as more like the prototype than sounds which are at the same perceptual distance from a non-prototype.
in research by Pat Kuhl, it has been shown that there are phonetic category prototypes--`exceptionally good instances of phonetic categories'. i.e., there is a range of possible pronunciations of /i/; adults judge some of these as better instances of /i/ than others.
for adults, these prototypes are like ``perceptual magnets'': sounds which are at a certain perceptual distance from the prototype are perceived as more like the prototype than sounds which are at the same perceptual distance from a non-prototype. this is perceptual magnet effect of prototype.
differences between languages with respect to prototypes: adult Swedish speakers' prototype /i/ is different from English speakers' prototype /i/; Swedish speakers will often judge English /i/ more like Swedish /e/.
Advantages of categorical perception
Best et al. (1988) Examination of perceptual reorganization for nonnative speech contrasts: Zulu click discrimination by English-speaking adults and infants. J. of Experimental Psychology: Human Perception and Performance. 14:345--360.
`the perceptual reorganization at 10-12 months that Werker found closely parallels the universal milestones of beginning word comprehension and, for many infants, the first productions of words...the prephonemic sensitivity of infants under 10-12 months of age for many nonnative contrasts is surely well suited to their ability to learn whichever language surrounds them. However, as they become attuned to the ambient language and first begin to use words, phonemic perception should presumably aid their language acquisition. If phonemic perception entails assimilation of incoming sounds to the categories employed in the native language, then it may benefit the infant by sharpening the lines of structural organization within the phonological system of their language and by helping to establish perceptual constancy among the acoustic variations of words pronounced in different contexts and by different speakers. These benefits would presumably continue to aid efficient speech perception by adults, thus accounting for their continued difficulty with discriminating nonnative sounds that are assimilated to a single native phonemic category.'
Adults
adults perceive speech sounds
Linguistic experience influences categorical perception.
E.g., Japanese speakers are not good at distinguishing [r] and [l]; English speakers are. Categorical perception thus interferes with the ability to discriminate phonetic contrasts.
Miyawaki et al. (1975) "An effect of linguistic experience: The discrimination of [r] and [l] by native speakers of Japanese and English." Perception and Psychophysics 18:331--340.
Miyawaki et al. showed that Japanese speakers are not good at distinguishing [r] and [l]; English speakers are
``13 'speech' stimuli...varied in the initial stationary frequency of the third formant (F3) and its subsequent transition into the vowel over a range sufficient to produce the perception of [ra] and [la] for American subjects and to produce [ra] (which is not in phonemic contrast to [la]) for Japanese subjects. Discrimination tests of a comparable set of stimuli consisting of the isolated F3 components provided a ``nonspeech'' control. For Americans [39 adults], the discrimination of the speech stimuli was nearly categorical, i.e. comparison pairs which were identified as different phonemes were discriminated with high accuracy, while pairs which were identified as the same phoneme were discriminated relatively poorly. In comparison, discrimination of speech stimuli by Japanese subjects was only slightly better than chance for all comparison pairs. Performance on nonspeech stimuli [the synthesized control], however, was virtually identical for Japanese and American subjects; both groups showed highly accurate discrimination of all comparison pairs. These results suggest that the effect of linguistic experience is specific to perception in the ``speech mode''.'