week 6 Flashcards
(15 cards)
Why it is important to read
Poor literacy associated with:
Lower income
Higher rates of unemployment
Increased risk of debt
Fewer adults in a married/cohabiting relationship
Higher rates of teenage pregnancy
Less likely to be socially active
Poorer living conditions
Higher prevalence of depression
Written and spoken differences
Similarities
Many of the same skills used. E.g.:
Language skills
Memory
Executive function
Attention
Theory of Mind
Differences
Differ in physical form AND the social function of the message
Additional skills needed for both
Complexity and speed
Number of rare words (per 1000 words; Haynes & Ahrens, 1988)
Conversation among graduates: 17.4
Expert witness testimony: 28.4
Books for pre-schoolers: 16.3
Books for children: 30.9
Words per minute
We speak at a rate of ~ 150 WPM ~ 2.5 words/sec (Tauroza & Allison, 1990)
We read (silently) at a rate of ~ 200-300 WPM ~ 3-5 words/sec (Brysbaert, 2019)
Visual word recognition
Is written word recognition a 2-stage process: letter recognition first, then word recognition?
Word superiority effect (Reicher, 1969) suggests not.
Participants perform more accurately when stimulus is a word (e.g. poster) than when it is a random letter string (e.g. gdasch)
Word knowledge influences in visual word recognition.
Phonemes and Graphemes
Phoneme: The smallest unit of sound that can alter meaning
E.g. change /b/ in ‘big’ to /d/ = dig
Grapheme: Basic unit of written language, a letter or combination of letters corresponding to a phoneme
E.g. change /b/ in ‘bank’ to /th/ = thank
Regularity and consistency
Regularity
Degree to which the particular graphemes of a word correspond to particular phonemes
Consistency
Is a particular spelling typically pronounced in the same way?
Heteronyms and Homophones
Heteronym
A type of homograph (words with same spelling)
Both the meaning AND pronunciation of a word is different despite using the same spelling
Homophone
Words which are pronounced the same but spelled differently
How do we measure word recognition
Fluency/efficiency of reading
Whole sentences/paragraphs
Single word tasks (e.g. speeded reading aloud)
Lexical decision tasks
Words and non-words presented individually, participants have to decide whether it is a real word or not. Reaction time measured
Semantic decision tasks
Word is presented, participants required to make a semantic judgement (e.g. “Is this a living thing”)
Eye tracking tasks
Can provide information on both reading speed (decoding) and comprehension monitoring
Neurophysiological measures
Complementary measures of neural recruitment
Homophone effect
Homophone effect (van Orden, 1987; also Lesch & Pollatsek, 1993)
semantic categorization: Does the word belong to a semantic category? Press Y for ‘yes’, N for ‘no’.
If the sound of ‘rows’ activated during reading, more Y to ‘rows’ relative to control words, e.g. ‘robs’
Phonology in reading
Strong phonological theory’ (Frost, 1998; also Lukatela & Turvey, 1994) - the core lexical representations that underpin visual word recognition are phonological, and thus, phonological processing is a mandatory part of recognition process.
Supported by evidence of ‘fast’ phonology in VWR
Major models of word recognition are ‘weak phonological theories’ (e.g. Coltheart, Rastle, Perry, Langdon, & Ziegler, 2001) - consider the printed word recognition to be driven primarily by the analysis of orthographic form.
Phonological influences on recognition are viewed as secondary and non-essential
Models of visual word recognition
Helps understand the processes underpinning language comprehension and production, and (in some cases) links to how it all works in the brain
Early models:
“box and arrow”
Modern models:
computer programs simulating
how words are recognized
Dual-route Cascaded model of reading
print-Grapheme-phoneme rule system-speech
Print + Speech-meaning
Support for DRC
Very high rates of accuracy in reading both regular and irregular words
Simulates several aspects of word reading in skilled readers:
‘Dual route’ approach also supported by fMRI data and by studies of acquired dyslexia
The triangle model of reading
Connectionist* systems suggest no need for separate mechanisms of word reading
Parallel distributed processing all sensory information and all stored knowledge used interactively to read all words.
Unlike Dual Route, can model learning how to read
‘Rules’ are not pre-specified
Used to model different levels of success
Comparison of models
Dual Route
Computation
two mechanisms: rules and lexical ‘look-up’
Influence of regularity
Learning
not implemented: adult-state model
Cross-linguistic application
Intervention studies
Triangle
Computation
single mechanism:
spreading activation across simple neuron-like processing units
Influence of consistency
Learning
crucial aspect of the model
Cross-linguistic applications
e.g. Chinese (Yang, Zevin, Shu, McCandliss, & Li, 2008, 2009)
Intervention studies
Harm, McCandliss, & Seidenberg (2003)
