Lecture 7 - Learning Words as Labels

Question 1

Perceiving sounds vs. learning words

Answer 1

infants start out hearing everything and narrow in on native language by 12 months VS. produce words badly at 12 months (presumably underlying representations are good)

Question 2

Stager and Werker

Answer 2

habituate kids to two nonsense words “bih” and “dih” in conjunction with moving images and then once kids are “bored” (habituated) they then “switch” the word with the image to see if they dishabituated (show attention)

at 8 months they do dishabituate

at 14 months they DO NOT dishabituate (BUT with really dissimilar words like lif vs. neem they dishabituate)

Question 3

Stager and Werker

Why?

Answer 3

perceptual regression

maybe their attention is divided between learning the objects vs. learning the words

the unfamiliarity of the words (maybe should use something like wall or doll instead)
- they did do that and the infants did fine - something to do with whether the infants know the word or not

Question 4

Word recognition experiment

Swingley and Aslin, 2000, 2002

conclusion?

Answer 4

target fixation given correct pronunciation vs. mispronunciation

kids were confused when the words were switched “ball” vs. “gall” (look more when it’s the correct pronunciation)

conclusion: kids do have very specific representation of word sounds in familiar sounds, early in life

Question 5

kids flexibly interpret mispronunciations

why?

Answer 5

mostly interpreted as the intended word

because human beings can’t memorize every single new verbalization of a word from person to person: every-time you hear a different accent you don’t think it’s a novel word

Question 6

exception to the flexible interpretation of words:

Answer 6

when there is a similar word around: instead of fish and feesh but when you hear goat instead of coat and both objects are present it can be a problem

Question 7

for new words you have somewhat weaker representations of…

Answer 7

…the sound patterns

bih vs. dih are difficult without exposure

Question 8

Speech Output of infants pre-6 months

Answer 8

mostly vowels, some back consonants (velars)

aah, gaagaa

Question 9

speech output of infants post 6 months

Answer 9

canonical babbling starts: wider range of places of articulation

using lips, tongue to produce sounds

regardless of language, kids seem to be makig very similar sounds (until a little while before words appear)

english: mostly “stop” consonants: ptk, bdg, mn, s, h

Question 10

Babbling to words

Answer 10

kids who prefer certain babbling patterns end up using those patterns later in life

babbling frequency predicts # and accuracy of words produced at 2-3 years

Question 11

late in prespeech, kids start to drop out sounds that aren’t ….. and adopt rhythms of ……

Answer 11

in child’s language

native language

Question 12

First words

Answer 12

some resemble adult words in sound and social context: hi, byebye

Question 13

protowords

Answer 13

words made up by kid but used for particular function

• controls vocal apparatus
• knows words have specific meaning
• but they don’t recognize that language has to be a “common code”