reading 1 - DRC Flashcards
mental representation for speech and reading comprehension (input to output)
input = speech
activate –> existing mental representations of sound/written words
link to –> meaning
output = comprehension
building blocks for speech comprehension
speech input –>
form (phonology) –>
syntax + morphology –>
semantics
building blocks for written word comprehension
written word input –>
form (phonology + orthography) –>
syntax + morphology –>
semantics
form - orthography - 3 system types
alphabet - graphemes represent phonemes e.g. english
logographic system - characters represent words - e.g. mandarin
alphabetic language that looks like a logographic system - letters represent words but are grouped like characters - e.g. korean
form - phonology - is this necessary in reading
much debate over this
most agree activation of form occurs when reading
helps understanding
unconscious
different routes to comprehend written word
input –> output via 4 potential activations
- activate letters
- activate phonemes (via letters)
- activate orthographic form
- activate phonological form
- letters –> phonemes –> phonological form
- letters –> orthographic form –> phonological form
- letters –> orthographic form
triangle model of reading processing
semantics - orthography - phonology
meaning - written form - word sound
testing DRC model
adjust the strength of connections computationally
provide input
process input through model
assess output of model against human performance
- computational modelling allows us to ask very specific questions
- test the results
- but rarely able to model all the variables and parameters at work
- not likely to get an absolute answer
- but excellent to test specific questions
DRC model of reading aloud
dual route cascaded model of visual word recognition and reading aloud
combination of excitatory and inhibitory connections from print to speech
from print/visual features / letters to speech output
lexical and non-lexical routes
DRC 2 routes
‘non-lexical’ route converts letters into sounds to activate a phonological representation that links to meaning
‘lexical’ route activates an orthographic representation that is linked directly to meaning
DRC - lexical route
- written word
- orthographic lexicon (whole word orthographic representations)
- semantics
DRC - non-lexical routes
- written word
- spelling to sound correspondence (grapheme-phoneme correspondence - GPC)
- phonological lexicon (whole word phonological representations)
- semantics
phonological lexicon can check for irregularities of words - have to take lexical route
DRC - interaction between lexical and non-lexical routes
phonological lexicon can activate orthographic lexicon and vice versa
spelling to sound correspondences
relationship between letters and sounds = grapheme phoneme correspondence
could be key to understanding problems with dyslexia
single grapheme = single phoneme
grapheme can be made up of many letters
1 letter grapheme = f
2 letter grapheme = th
3 letter grapheme = igh
4 letter grapheme = ough
regular and irregular words (grapheme phoneme correspondence)
regular = follow a set of rules that dictate how a grapheme should be pronounced = aligns with graphemes most frequent pronunciation
MINT is a regular word
PINT is an irregular word
both words contain the grapheme ‘i’
In MINT the pronunciation is regular /ɪ/
In PINT it is irregular /aɪ/
native english speakers learn regularities implicitly
transparent vs opaque orthographics
transparent = lots of regular correspondences e.g. dog, print, cobweb
transparent language = shallow orthography e.g. finnish, italian
opaque = few regular correspondences e.g. yacht, knight, colonel
opaque language = deep orthography e.g. English
which route of DRC has to be taken for irregular words
lexical route
word sound used to then attribute meaning to it - needs to use different rule
advantages of DRC model (4)
- 2 routes for processing written word
- orthographic lexicon and phonological lexicon
- processing of regular and irregular words
- encountering new or novel words, can be processed via grapheme-phoneme correspondence
is spelling to sound correspondence important in learning to read
“to appreciate the relationship between letters and sounds is necessary and non-negotiable when learning to read in alphabetic writing systems” - Castle et al (2018)
self-teaching hypothesis
Share (1995)
children ‘de-code’ words using an understanding of how letters correspond to sounds
existing phonological representations are a
accessed – used to access meaning
phonological representation is used to develop an orthographic lexicon of whole words
they teach themselves to read
self-teaching and the DRC model - requirements for this
Share (1995) - children use phonological representations to develop an orthographic lexicon
necessary to have:
- lots of phonological representations
- good verbal language
- lots of exposure to printed word
in order to build orthographic lexicon
self-teaching DRC (Pritchard et al., 2018)
‘non-lexical’ route is used to decode words and access an existing phonological representation
contextual cues are used to select the target word from a list of many spoken word candidates – select best match from context
exposure to print facilitates the development of an orthographic lexicon – lots of reading on a regular basis necessary
skilled readers difference in DRC
link between orthographic and phonological lexicon becomes less strong as reading becomes more skilled
e.g. reading the words “path” instantly triggers the meaning without needing to sound it out first
self-teaching hypothesis - struggles with this
de-coding words to access phonological representations is necessary for development of an orthographic lexicon
children who struggle to link graphemes with phonemes might not be able to teach themselves to read
