FInal Exam Cog and Lang Development

Words

- are symbols: stand for something without being part of it.
- are arbitrary: the sounds /dog/ mean "dog" but there is nothing dog-like about the sounds
- have reference: stand for their referents, not just accompany them

Context-bound words

words children only say in a specific context

Are children's first words all context-bound words? Or are some first words already referential?

- longitudinal, observational study of 4 English-learning infants
- recorded all language production
- examined whether first 10 words were context-bound or not
- results: children's first 10 words included both context-bound and referential words
- role of input: the "context" of the context-bound words was the mother's most frequent usage

The first 50 words

- acquired around 18 months
- of first 50 words, 45% are nouns and only 3% are verbs

Natural Partitions Hypothesis

- meanings linked to verbs are harder to learn because
1. nouns typically refer to objects, which are easy to perceive and are stable in the environment
2. verbs typically refer to actions that are transient and relational

Theory: different learning mechanisms in acquisition of nouns and verbs

- the meanings of nouns can be learned by mapping sounds onto the nonlinguistic context
- the meanings of verbs often require a syntactic frame that identifies the nature of the relationship involved

Do all children show the noun bias, cross-linguistically and cross-culturally?

- 8- to 16-month-olds learning English, Cantonese, or Mandarin
- parents completed a checklist of words their child could produce
- results: object nouns and action words were more similar in frequency in Cantonese and Mandarin than in English
- role of parental input and syntax in acquisition of nouns among English learners

Underextensions

when children use a word in fewer contexts than they should ("dog" to refer only to collies)

overextensions

when children use a word in more contexts than they should ("Daddy" to refer to all adult men)

Why do children overextend word meanings?

- experimental study with 2-year-olds
- children named three pictures (priming) and then a target object
- one picture was perceptually similar to the target object
- results: priming led to higher rates of overextension of the primed word

Word Spurt

after children acquire 50 words (around 18 months), they begin acquiring words at a much faster rate

Is the Word Spurt real?

- some children may show spurts, but these are highly individualistic and may depend on specific experiences
- word spurt may not happen at the same time for every child

Word comprehension

- children can understand much more than they can say
- children's word comprehension can begin as young as 5 months

Measuring infants' word comprehension

- tested 6-9-month-old infants
- parents wore a visor and headphones and repeated sentences that they heard (ex. "Can you find the banana?")
- results: even the youngest children (ages 6-7 months) looked longer at the labeled picture, indicating word comprehension

follow-up study with children ages 6-20 months

- results: all ages were above chance at looking at target picture, but there was a big improvement between 13-14 months

Referential style

more nouns and fewer social expressions

Expressive style

more social expressions and fewer nouns

Sources of the referential/expressive difference

- mothers who spend more time teaching labels for objects tend to have more referential children
- first-born children of college-educated parents are more likely to be referential
- children may simply differ in style

Larger vocabularies are found among children who...

- hear more speech from adults
- hear more variation in words and kinds of sentences
- hear more decontextualized language (talk about the past and future, pretend play, and definitions)
- hear words in more elaborative contexts
- hear speech that is responsive to their own behavior (especially for infants under 18 months)
- are first-born
- are of high SES household

Word segmentation

ability to break stream of speech sounds into distinct words

Transitional probabilities in speech sounds

- speech stream contained 3-syllable "words" that were only distinguished by their transitional probabilities
- 8-month-olds discriminated words versus part-words after 2 minutes of listening

Are "chunks" really words for the child? Or just a sound sequence?

8-month-olds discriminate "words" from non-words in an English-language context but not a nonsense context

Stress as a cue to word segmentation

- in English, stress is usually on the first syllable
- infants can use stress as a cue to word boundaries as early as 9 months

What happens if transitional probabilities and stress cues do not coincide?

when pitted against each other, 8-month-old English-language learners used stress cues instead of transitional probabilities

The whole-object assumption

a new word refers to a whole object, not to actions, parts of objects, or properties of objects

The mutual-exclusivity assumption

children assume each object has only one name

The taxonomic assumption

- words refer to things of the same kind

shape bias

children preferentially attend to shape to determine category membership

Training the shape bias

17-month-olds were randomly assigned to 2 conditions

shape bias training

7 weeks of training (novel objects with the same shape, but different texture and color, were labeled with the same novel word)
- control: no training
- results: shape-bias training increased learning of object names but not other words

Challenges to word-learning assumptions

- whole-object assumption must be overridden when a child learns names for parts or properties
- mutual-exclusivity assumption must be overridden for synonyms and reference to objects at different taxonomic levels
- taxonomic assumption must be overridden for proper names

Syntactic bootstrapping

children's ability to use syntactic cues to help them learn words

Syntax

the component of grammar that governs the order of words in sentences

Descriptive grammatical rules

implicit grammatical rules that all speakers of a language follow unconsciously

prescriptive grammatical rules

taught in classes and by parents, markers of etiquette and education

Adult syntax knowledge is a combination of...

linear order, grammatical categories, hierarchical structure

The productivity/generativity of language

we can produce an infinite number of sentences that we have never heard before

Vertical constructions

a sequence of one-word sentences that are meaningfully connected

Unanalyzed word combinations

sequences of words for which children have probably memorized the whole sequence rather than separate words

Open-class words

- content words, mostly nouns, verbs and adjectives
- open in the sense that we add to this class of words all the time
- typical of two-word utterances

Closed-class words

- fixed set of function words, including
1. auxiliaries (can, do, will)
2. prepositions (in, below, of)
3. determiners (the, those, an)
4. complementizers (that, which, who)

Early three-word speech

sounds a bit like a telegraph message, composed exclusively of open-class content words

Why don't children use more closed-class words?

- because they are not necessary to convey meaning; they reserve their limited processing space for critical elements
- they leave out elements that are phonetically less prominent
- they may not understand the functions of closed-class words

Morphology

- smallest unit that has meaning in a language
- combinatorial
- free ("truck" in trucks) or bound ("s" in trucks)
- derivational (flawLESS, drinkABLE) or inflectional (walkED, walkING)

Morphological development in children acquiring English

- takes children many months to reliably use even very common inflectional morphemes
- different morphemes are acquired at different times but in a very similar order among children

Morphological development in children acquiring languages other than English

- children acquiring rich systems of morphological inflections acquire grammatical morphemes as early/sometimes earlier than they acquire word order
- following factors help children acquire morphemes earlier
1. highly frequent use
2. highly regular systems with few exceptions
3. higher perceptual salience (ex. stressed syllables)

Expressing negation

children express negation early on before they can use adult-like syntax

Asking questions

- yes/no questions: first questions are marked only by intonation
- WH-questions: who, what when, where
- later, children begin to produce auxiliaries and subject-auxiliary inversion

Individual differences in grammatical development

- some children start combining words into 2-word utterances before 18 months, while others do not do so until after 24 months

syntactic development is often measured

by the Mean Length of Utterance (MLU)
MLU can be measured in terms of words or morphemes

Nativist theory

- the capacity for learning language is innate, including guidance about the abstract structure of all languages
- learning a language is like learning to walk

Principles and parameters theory

- nativist

Universal Grammar is like a switchboard

a set of parameters, each with a few possible settings
- experience with a specific language "triggers" the correct parameter settings for that language

Constructivist theory

- children construct their grammatical knowledge from their experiences with language
- early grammatical knowledge may not be abstract
children may be conservative in their early language production, not going beyond what they hear

Usage-based accounts

- children acquire words and constructions which they combine in limited ways
- learning is driven by children's communicative skills and ability to find patterns

The connectionist approach

- connectionist nets make associations among elements in the input
- what we see as "rules" emerge from the interaction of the input and the net

Different verbs have different argument structures

ditransitive, intransitive, and transitive

ditransitive

"give" has 3 arguments associated with it: giver, given, givee

transitive

"eat" has 2 arguments associated with it: eater, eaten -->

intransitive

"sleep" has 1 argument associated with it: sleeper -->

Evidence for item-specificity

- diary study, age 16-24 months
- child used limited sentence frames for most verbs
- new utterances were most closely related to previous sentences with that specific verb, not that verb category

elicited production study

34-month-olds and 41-month-olds- heard a new verb in passive sentences
- only 20% of younger children and 55% of older children produced an active transitive sentence

Verb island hypothesis

- each specific verb has its own set of rules
- there is no general rule that applies to all verbs

Evidence for productivity

- children commonly produce overregularizations and overgeneralizations
- at first, children produce irregular forms correctly
- then, they overregularize
- later, they return to correct usage

Is grammatical knowledge abstract?

- 2-year-olds children viewed videos on two screens
- infants were familiarized with the stimuli using familiar verbs

test phase: animation

- results: children were significantly above chance, consistent with abstract knowledge of verb argument structure
- 21-month-olds were above chance at looking at the correct video when "bunny" and "duck" were changed to "girl" and "boy"
- suggests that by 21 months, infants' knowledge of verb argument structure is abstract, not limited to specific verbs that they have learned

Evidence for abstractness

- children produce rule-based forms they have never heard before
- children comprehend new verbs based on their syntactic frames

The case for multiple systems (abstract and item-specific)

- many researchers believe that children and adults sometimes use memorized forms and sometimes use abstract rules
- regular pasts can be represented with rules (add -ed)
- irregular forms must be memorized (go --> went, eat --> ate)

The tadpole-frog problem

If children's and adults' knowledge is qualitatively different, how does developmental change occur?

Solutions to tadpole-frog problem

- Solution 1: There are no tadpoles. There is developmental continuity because children have abstract, rule-based grammatical knowledge from the start.

Solution 2:

There are no frogs. There is developmental continuity because even adults' grammatical knowledge is not abstract.

Solution 3:

There are tadpoles and frogs. There is developmental discontinuity, accounted for by learning or maturation.

In the U.S., about _________% of people speak more than one language.

20-25

Simultaneous bilingualism

child learns multiple languages from birth

Sequential bilingualism

child begins to acquire another language after starting to acquire a first language

Majority language

spoken by the majority of individuals in a society

minority language

any language not spoken by the majority of the community

Heritage language

language of the culture in which a child was born

Language differentiation

child's knowledge that they are being exposed to two languages and eventual ability to separate them

Language differentiation hypotheses

fusion, autonomous differentiation, interdependent differentiation

Fusion

Initially, children do not differentiate languages, fusing them into one. Differentiation occurs later.

Autonomous differentiation

Children separate languages, acquiring each separately. Rate and process of development are independent in each language.

Interdependent differentiation:

Children separate languages, but they influence one other during acquisition.

Phonological differentiation

- infants' perception becomes tuned to their native language from age 6-12 months
- by 12 months, bilingual infants get better at distinguishing the phonemic contrasts in both languages

Do children create two different lexicons at the same time?

- If children have a fused language, they should apply mutual exclusivity, and learn one label for each concept.
- If children have differentiated languages, they can suspend mutual exclusivity to learn two labels for the same concept.
- Result: Between 30-50% of bilingual children's vocabularies consist of translation equivalents.

Bilingual learners must have suspended mutual exclusivity assumption

suggests that their lexicons are differentiated

Syntactic differentiation

- evidence suggests that bilingual children create separate syntactic systems
- bilingual children, like bilingual adults, do mix words from both languages into the same sentences

Does bilingual language learning change the way language develops?

the overall course of development is very similar for monolingual and bilingual children

Does bilingual language learning change the rate of language development?

- simultaneous bilingual infants and monolingual infants
- observed at age 1 year 10 months, 2 years 1 month, and 2 years 6 months
- at each age, parents reported for each language separately their child's vocabulary and syntax
- results: bilingual learners lagged the monolingual learners in vocabulary in each language, and bilingual children were less likely than monolinguals to produce two-word utterances in each languag

Do bilinguals eventually catch up to monolinguals in a single language?

- vocabulary: possibly not, differences exist at all ages tested

- syntax and morphology:

yes by age 10, with consistent exposure and usage of both languages

Bilinguals' OVERALL language development is ______________________________ than monolinguals'.

the same as or better

Variable properties of bilingual environments

- societal factors
- heritage language culture
- household composition

Does amount of exposure to language matter?

percentage of language exposure is correlated with vocabulary and syntax

Vocabulary in English is positively related to...

- amount of exposure to English
- number of different sources of English
- amount of English that comes from native speakers

Recent work has found bilingual advantages for...

- theory of mind (TOM): understanding others' mental states - executive function: ability to control attention and to flexibly shift attention

Do bilingual children do better than monolinguals on language-specific TOM tasks, all TOM tasks, or all information-processing tasks?

- bilingual and monolingual children

- completed three tasks: Standard Theory of Mind, Modified Theory of Mind (language switch), control task (no mental states)

- results: bilinguals out-performed monolinguals on both theory of mind tasks (hypothesized to be due to general enhancement of executive function)

Do bilingual children have better executive function than monolinguals?

- bilingual 6-year-olds

color-shape task switching

- results: bilinguals were faster than monolinguals when switching between rules