lecture 8 - reading process

Question 1

word identification

Answer 1

automatic process, evidence = stroop task

need to use eye movements to study it

Question 2

moving window =

Answer 2

text is mutilated apart from experimenter-defined area surrounding fixation point

Question 3

EZ reader model

Answer 3

next eye movement is programmed after only partial processing of currently fixated word
can attend to 2 words during single fixation, but only one word is processed at a time = serial processing model
if next word is predictable from context its skipped
common, short or predictable words are not fixated

Question 4

reaction time techniques: naming task (reading aloud)

Answer 4

string of letters - emphasises links between orthography and phonology

Question 5

reaction time techniques: lexical decision task

Answer 5

string of letters: decide whether spell out word or not - emphasises links between orthography & semantics

Question 6

double-route cascaded model (Coltheart et al)

Answer 6

2 main routes from printed words to sound/speech (both start with orthographic analysis - identifying and grouping letters in words)
it is a cascaded model = activation at one level is passed onto next level BEFORE processing at first level is complete (differ from threshold models)

Question 7

double-route cascaded model: route 1

Answer 7

indirect non-lexical route
operates serially from left to right
converts letters into sounds one at a time (grapheme-phoneme conversion) = converting speeling into sound

Question 8

double-route cascaded model: route 2

Answer 8

direct lexical route - operates in parallel
divided into 2 sub-routes: one involves only orthographic lexicon & other involves semantics as well as orthographics (stores knowledge of language i.e. words you know)
thousands of similar words are stored in an orthographic input lexicon - visual presentation of a word produces activation of this lexicon followed by obtaining its meaning from semantic system
the sound pattern is then generated by the phonological output lexicon

route 2 = lexicon + semantic knowledge
route 3 = lexicon only

Question 9

which routes are used when?

Answer 9

healthy pps use both routes in parallel when reading aloud bu naming visually presented words = mostly lexical route because it is faster

Question 10

surface dyslexia

Answer 10

only use route 1
can’t link surface info (visual form of word) to meaning
symptoms:
accurate pronunciation of regular words but not irregular not conforming to conversion rules (these rules can also provide pronunciation of nonwords)
confuse homophones (same pronunciation but different meaning)
spelling errors that are phonologically correct
patient KT: read 100% of nonwords, 81% regular words & 41% irregular words

Question 11

phonological dyslexia

Answer 11

uses route 2 & 3 but not route 1
no problem reading previously learned words (intact orthographic input lexicon) if they’re regular or irregular
can’t use grapheme-phoneme rules to pronounce unfamiliar words & nonwords
can extract meaning from visual form

Question 12

deep dyslexia

Answer 12

use completely different systems in rigth hemisphere

problems reading unfamiliar words and nonwords & semantic reading errors

Question 13

connectionist triangle model (Plaut et al)

Answer 13

double-route model = pronouncing regular and irregular words involves different routes
BUT this model = highly interactive
all the systems are used in pronouncing all types of letter strings

3 sides of triangle: orthography, phonology & semantics

Question 14

connectionist triangle model: 2 routes from spelling to sound

Answer 14

1) directly from orthography to phonology

2) indirectly from orthography to phonology via word meanings

Question 15

connectionist triangle model: consistency

Answer 15

words and non-words vary in consistency (extent to which pronunciation agrees with similarly spelled words - neighbours)
consistent words and nonwords should be said faster than inconsistent ones

Question 16

connectionist triangle model & dyslexia

Answer 16

surface dyslexia = damage in semantic system
phonological dyselxia = general impairment of phonological processing
deep dyslexia = severely impaired phonological processing leads to reliance on semantic processing

Question 17

connectionist triangle model: evaluation

Answer 17

+ mechanism for how we learn to pronounce words (unlike others)
- explanation of phonological dyslexia oversimplified (some don’t show general phon impairment)

Question 18

although children may learn to read faster in more __________ orthographies, there may still be universal cognitive prerequisites for learning to read

Answer 18

consistent

Question 19

phonological neighbourhood

Answer 19

2 words are phonological neigbours if they only differ in one phoneme - when phon neighbours are closely related this slows down word recognition as they receive activation from each other & so compete with target word

Question 20

Word processing: Interactive activation model

Answer 20

parallel processing
bottom-up and top-down processes interact
recognition units at 3 levels: feature e.g. vertical line in top corner (bottom), letter (middle), word (top)
when a feature in a letter is detected acitivation goes to all letter units containing the feature & inhibition to all others not containing feature
when letter is identified, activation is sent to all 4-letter word units containing that letter in that position

Question 21

2 priming effects:

Answer 21

1) rapid short-lived facilitatory effect based on semantic relatedness
2) slower, long-lasting effect based on expectations

Question 22

spillover effect

Answer 22

fixation time on a word preceded by a rare word is longer