week 4 Flashcards
Language acquisition
Between birth and 5 years of age, children transform into linguistically competent individuals who have mastered the basic structure of their native language
The Process of Language Acquisition
Acquiring a language involves both comprehending what other people communicate to you and producing language of your own
Infants know a great deal about language long before their first linguistic productions
Language comprehension
Understanding speech sounds, words and sentences produced by others
Language production
Articulating sounds and words; combining words into sentences that others can understand
areas of language
Phonology: learning the sound and combinations of sound used in native language
Semantics: learning the meaning encoded in word/sentences
Syntax: learning the rule about how words can be produced
Pragmatics: learning how language is used in different context
Phonological development
Learning the sounds and combinations of sounds that constitute words in your language
Infants are born able to perceive and distinguish speech sounds from all the world’s languages
With experience, the zero in on the sounds of their native language
Over time, they develop enhanced processing of their native language, and reduced ability to distinguish sounds of foreign languages
This is an example of perceptual narrowing
The challenge of word segmentation
One of the earliest linguistic challenges children face is segmenting the speech stream
Statistical learning
Ability to perceive and track regularities (patterns) in language
Allows infants to identify candidate words in speech streams
Powerful, domain-general mechanism (i.e., not specific to language learning)
Statistical learning: Saffran et al. (1996)
8 month old infants were familiarized to 2 minutes of this artificial language
Some speech sounds always occurred next to each other (“ki-bu”; others rarely/never did (e.g, “ti-ku”)
Then presented infants with familiar “words” (ki-bu) or non-words (ti-ku)
Infants listened longer to nonwords than words, showing a novelty preference for these new nonwords
Suggests that 8 month olds were sensitive to the statistical regularities with which speech sounds co-occurred
semantic Development
Comprehension:
Around 4.5 months: own name
Around 6 months: first nouns (common objects, body parts, people)
10-13 months: verbs
Production:
- First words around 10 to 15 months
- Large individual differences. Median vocabulary size:
16 months: produce 55 words 23 months: produce 225 words 30 months: produce 573 words 6 years: produce 6000 words
By 6 years, comprehend approx. 14,000 words
Comprehension
Associating words with correct person/object/action signals true understanding of word’s meaning
2 images presented side-by side
Hear word/phrase that matches one of the images
Comprehension look longer at image that matches word
Production
Vocabulary spurt around 18 months
rate of growth of productive vocabulary accelerates
add 8-24 new words per week
Objects – especially things infants can manipulate
Relational words (refer to state/location of objects – e.g., under, here)
Verbs – conceptually more difficult – refer to relations between objects/people (e.g., bump)
Emotions/internal states – unobservable – require more advanced cognitive abilities (e.g., feel, sad)
How do children map meanings of words
Based on the input, essentially an infinite number of hypotheses about word meaning = UNDERSPECIFICATION
Fast Mapping
One solution to underspecification of word meanings
Make an initial ‘fast mapping’ between new word and likely meaning
Modify this guess as they receive further input
Learning by exclusion
Under- and overextensions
Underextension
Children map words to an overly narrow class of referent
E.g., use truck only for a toy truck
Haven’t yet generalised the word to all ducks
(earlier in development)
Overextension
Overgeneralise words to overly broad class of referents
E.g., daddy for all men
(later in development)
Cognitive Biases
A number of assumptions or cognitive biases guide children’s acquisitions of word meanings – also known as lexical constraints
Whole object assumption
New word refers to whole object, rather than part of it
Mutual Exclusivity assumption
New word meaning does not overlap with known word (can overcome whole-object assumption)
Nature vs nurture
Empiricism (nurture)
Knowledge is gained through experience
Perceptual abilities, concepts are learned
There are no innate structures
Nativism (nature)
Knowledge, ability to perceive, concepts are present at birth
Innately ‘programmed’ to unfold via maturation
Empiricism
Language is learned/derived via exposure e.g. Skinner (1957), Saffran et al. (1996)
Learning process driven by domain-general perceptual and cognitive processes
association, imitation, reinforcement
statistical learning – word segmentation
Nativism
Infants are born with the capacity to acquire language (learning is not required) e.g. Chomsky (1959); cognitive biases for word learning
Innate language acquisition device (LAD): humans are born with a specific innate component in the brain for language
Poverty of the Stimulus: The speech children hear is unclear/contains errors, e.g. hesitations, slips of the tongue and contains insufficient information for them to be able to learn the grammar
Children make predictable grammatical errors would not be predicted by learning theory as adults do not make them
Theories of language acquisition
Interactivist view – biologically prepared to acquire language through experience
Learning language is social!
Requires children to be part of a community in which people interact and talk with each other
Infant Directed speech
Characteristic way in which adults modify their speech when addressing young children
Exaggerated intonation, higher pitch, elongated vowel sounds (Fernald & Kuhl, 1987)
Produced spontaneously and consistently across languages and cultures (Kuhl et al., 1997)
Analogous behaviour observed in deaf mothers using sign-language (Masataka, 1992)
Benefits of infant directed speech
Measured amount of speech directed to infants in low-SES Spanish-speaking families
All-day recordings to parent-infant interactions in the home
29 infants at 19 months; language skills tested at 24 months
Large variation in amount of IDS between families
Children who heard more IDS at 19 months had larger expressive vocabularies & more efficient basic language processing at 24 months
Speech not directed at infant unrelated to child outcomes
Social Contingency
Role of social contingency in word learning
38 mother-child dyads; children = 2 years
Mothers taught children 2 novel words in a within-participants design
Uninterrupted teaching
Teaching interrupted by mobile call
Children heard novel word the same number of times in each condition
Children learned the novel words in uninterrupted teaching condition only
Interruptions to socially contingent interactions disrupt learning
Multilingual contexts
Multilinguals > Monolinguals – more than half the world’s population is bilingual
2/3 of children in the world are raised as bilingual
Immigration – e.g., to UK, US
Multi-ethnic societies – e.g., Singapore: Chinese, Malay, Tamil; English as common 2nd language
Simultaneous bilinguals (roughly from birth)
Parent speaks language different from community
Parents speak two different languages
Two language used by community
Sequential bilinguals (L1 then L2)
Changing family circumstances
New language when entering school
Bilingual vocabulary growth
Does the language development of children learning 2 languages differ from that of children exposed to just 1 language?
When total vocabulary for both languages is considered, BL and ML vocab growth rates match, though growth in each language is slower