3>word learning Flashcards
when do infants learn their first words?>
12 months
features of infants first words>
usually relevant & frequent words in environment (i.e. greetings, objects, toys, pet names)
when does a child reach (on average) 600 words?
36 months
features of infants first 600 words> (2)
-these words as expansion to different types & more abstract words
-these words as mainly still common nouns verbs & greetings
when does comprehension of words begin?
around 8-10 months
when does production of words begin?>
around 11-13 months
word comprehension=
mapping a word to its referent
what is word comprehension indicated by?> (2)
> Retention=remembering the word-referent pair (mapping)
Generalisation= generalising the word to a new category member (e.g. extend to other types of car)
word retention=
remembering the word-referent pair (mapping)
word generalisation=
generalising the word to a new category member (e.g. extend to other types of car)
what is the reference problem?>
there are infinitely many possible referents for a word
(e.g. point & gavagi> what is referring to? obj, part of obj, environment etc)
children’s ‘intrinsic biases’ on figuring out word meaning=
how every child has innate specific expectations of how labels map onto objects, (with these limiting no of potential referents)
types of children’s ‘intrinsic biases’ on figuring out word meaning (2)
> lexical constraints
social-pragmatic cues
how do children figure out word meaning> (2)
- intrinsic biases
- extrinsic cues
what are extrinsic cues for figuring out word meaning>
co-occurences between specific objects & labels; (properties of external communicative situation which assist word learning)
Types of extrinsic cues for figuring out word meaning> (2)
statistical associative info, distributional learning (e.g. hear & see together–>pair)
Lexical constraints> 1>whole object assumption study> (3)
- children 3-4 shown image of NOVEL obj & NOVEL label
- asked which of 2 things it is (part/whole)
- children chose entire object as referent (fits with assumption)
what is the whole object assumption?>
children expect a word to refer to a WHOLE object, not a part/property of it
lexical constraints: mutual exclusivity study> (4)
- showed children novel word for ENTIRE object & told them what it is
- Now children know of word
- asked on diagram where another label is
- as don’t accept 2 labels for things then apply this principle & infer this must refer to part
what is the mutual exclusivity assumption?>
children expect object to have one and only one name
Lexical constraints> taxonomic assumption study> (4)
-children showed an object with novel label (car & ‘sud’)
-showed objects that thematically similar
-asked to find another (‘sud’) & chose another car
-thus generalised words to other members of category (when use word to refer to 1st one)
lexical constraints> taxonomic assumption ALTERNATE study condition> (4)
-children showed an object with novel label (car & ‘sud’)
-showed objects that thematically similar
-asked to find another but didnt repeat label
- children chose policeman (as thematically similar)
summary on lexical constraints>
-aims to solve complex problem by simple set of constraints
problems with idea of ‘lexical constraints’>
-children do not always adhere to these
-children do learn words that refer to parts of objects
-children do learn MULTIPLE labels for an object
Socio-pragmatic- joint attention study> (3)
- after labelling children were asked to find ‘toma’ out of 3 objects
- when both parent & child were looking ast same thing 16 months+ identified correct obj
- when no joint attention only 18 month identified correct obj
socio-pragmatic cues> communicative intentions study on mutual focus> (4)
- children (2) & 3 hidden novel objs without lang
- asked child to find ‘toma’ new word but no corresponding obj
- adult find 2 of ‘wrong’ objects & vocalises this
- once all toys out, children chose correct obj without direct lexical link (through understanding communicative intentions/goals)
socio-pragmatic cues: communicative intentions> study on monitioring adults’ intentions (‘toma’) (3)
- objects hidden in buckets
-asked “where is toma” - adult takes out 3 novel objects & frowns at each until last which smile at
- children chose correct, understanding communicsative intention of ‘happy’ as found object
socio-pragmatic cues: communicative intentions> study on monitioring adults’ intentions (widget) (4)
- 2/3 year olds taught ‘widget’
- in 1 conditon “widge it” presented as an action; in another presented as an object
- asked to show “widge it”
- children in action condition were more likely to demonstrate the action; children in no action condition were more likely to make link between obj & label
socio-pragmatic account evidence
children’s word learning starts to take off about same time as socio-cognitive skills
problems with socio-pragmatic account
- children do learn word before their joint attention & intention reading skills have developed
- children with autism who find interpreting socio-pragmatic cues difficult do learn words
extrinsic cues> statistical associative info study> (5)
- children presented with novel object & novel labels (6 overall)
- presented in pairs
- have to keep track of which object goes with which label (as not explicit)
- asked to find ‘bosa’ ; can only find by co-occurence
- both 12 &14 months were able to form associations between novel objects & labels
pros of associative learning account>
pros
- simple (no need for innate knowledge)
- powerful (shown to be in place young
cons of associative learning account>
-evidence is from very simplified lab environment (1 label, few objs)
- might not be able to scale up to learning in real world
- doesnt account for learning of abstract words
Emergentist coalition model=
a hybrid account that is sensitive to multiple strategies children use to learn words
what is ‘statistical associative info’ extrinsic cue>
idea children associate words to their referents based on the probability of co-occurence of these pairs across multiple situations
3 key ideas of emergentist coalition model>
-children are sensitive to MULTIPLE cues
-children differentially weight certain cues over others at different time points
-children can move from using immature, basic constraints to more mature, sophisticated ones as they develope
emergentist coalition model> study> (3)
- 10, 12, 18, 14 months presented with 2 objects: one as interesting looking & the other as boring
- when ‘interesting’ object is asked to find AND looked at –>all children picked correct
- when asked to find ‘boring’ object–> only after 1 1/2 years were children able to shift attention away from ‘interesting’ obj
source of quantitative differences in word production> (4)
-child gender
-birth order
-personality
-language
how is gender a source of quantitative difference in word production>
-found female advantage prior to 24 m
-rates of acquistion found to differ
how is birth order a source of quantitative difference in word production>
- found advantage in first born, prior to 20 months
how is personality a source of quantitative difference in word production>
-children high on positive sociability showed advanced lang production at 20/21 months
- shy children less likely to learn & retain names of novel objects
how is ‘language’ a source of quantitative difference in word production>
- noun bias in english-speaking children
- vs noun bias not found in mandarin-speaking children (due to mandarin often having N’s omitted>thus not exposed to more N’s)
how is curiosity/interest a source of quantitative difference in word production>
- 24 month olds more readily learn word for a new category member of categories they ALREADY know more words for
- 30 month olds more readily learn words about objs that are in categories of their interests
