Word and Morpheme Segmentation (5)

why doesn’t phonetics help with word segmentation

speech is continous, we don’t speak with spaces

what is a problem with morpheme segmenetation

the distinction between words and morphemes are fuzzy

why is word segmentation a bottle neck tasks

it was to be be done before other phonological process, word meaning, and everything else 

at what age can children recognize some words in continous input 

0;6-0;8

at what age can children recognize syll boundaries

0;6-0;8

name 2 reasons why syll boundaries are useful

word boundaries only appear before syll boundaries

there are fewer syll bounds than phones in any utterance

is finding syll bounds still helpful if u speak a CV language

yes, it cuts the search space in half

what 4 things cue word segmentation (include age for 2nd+3rd one)

words child already knows, especially at beginning/end of utt

stress patterns, especially if lang has consistent stress (clear from 0;9)

phonological processes, ifants show sensitive to word-level patterns like vowel harmoney 

prosody can indicate phrase boundaries 

give an examples of undersegmentation and oversegmentation

it’sa

behave —> i am have

what are transitional probabilities

the changes a phoneme will be followed by the one after it

does the TP approuch work on it’s own, and what kind of learning is it

no

distributional learning 

what is the idea of TP

children track every sequence of phones

when would you find high and low TPs

high - reoccur because they’re inside words

low - indicative of word/morph boundaries

what kind of paradigm was the TP experiemnt w 3 nonsence words

head-turning paradigm

what is a syll unigram freq

how many times a syll is in the data set

what is a syll bigram freq

how often a syll occurs after another one

how is (B|A) read

b given a (b coming after a) 

what the equation for transitional probabilities

prob(B|A) = count(B|A) / count(A)

what is a local minima

TPs that are lower than the ones immediantly before and after it