Visual and Linguistic Cues

Question 1

What visual cues can be used?

Answer 1

global info; speech reading, gender, age, source, direction, expression, segmental (VMP), accent, familiarity
eg- Nygaard and pisoni 98

Question 2

What is a ‘viseme’?

Answer 2

when several phonemic contrasts look the same, eg bilabial viseme (apa, aba, ama). a lot of variation because of coarticulation

Question 3

segmental reading patterns?

Answer 3

PLACE: visible if tongue is near front but not far into mouth
VOICE: invisible (vocal cords vibrate)
MANNER: partially visible eg timing difference of lip protrusion; can’t see nasality

Question 4

what is ‘speech reading’ and who?

Answer 4

Everdel 07: look at eyes and jaw movements and expressions

Question 5

ability to speech read depends on? and who

Answer 5

Irwin, Thomas and Piling 0: listener ability, speaker character, predictability, accent of speaker, familiarity with accent and context – not very reliable realistically

Question 6

what is known in acoustic vs visual ONLY settings?

Answer 6

visual: only place really obvious, acoustic: only voicing really obvious

Question 7

what is superadditivity

Answer 7

do better than predicted with two than just one (of v or a)

Question 8

McGurk effect?

Answer 8

bottom up process; visual input overrides acoustic when different despite not fitting eg /g/ sound + aud /b/= perceived /d/

Question 9

McGurk study

Answer 9

Hazan and Chen 07: in japanese vs chinese; more susceptible in different language (eg chinese more than english)

Question 10

what lingustic cues can be used?

Answer 10

knowledge of own language, phonotactic constraints, lexical info (neighbours, frequency), syntactic constraints, semantic info, pragmatic information (expression, code for goals of speaker, eg “tomorrow”)

Question 11

Phoneme information demonstration?

Answer 11

Phoneme restoration effect (warren 70; kashino 06)- fill in or replace phonemes when absent unconsciously and know licit and illicit combinations of phonemes in own lang

Question 12

cross modal priming experiment?

Answer 12

Zwitzerlood- ambiguous words depend on activation of neighbouring/ rhyming words (Marslen-wilson cohort model)

Question 13

Network activation model

Answer 13

Luce and Pisoni-98: network activation; more competitive neighbours means more reliance on context- ortega and hazan 2000: ease of activation depends on neighbourhood density and frequency– serial search?

Question 14

how do pragmatics help?

Answer 14

rules out meanings irrelevant of situation if ambiguous (expression and interpretation of context/ situation)