H860 Reading Difficulties
Week 3
Developmental Dyslexia I: The
Phonological Representations
Hypothesis
3 times exceptional!!
Today’s session
1. The theory bit2. Break3. The practice bit
Recap from last week
• For a child to learn to read, a lot of precursors have to be in place…
• Early language development appears to be important, though there is no one obvious predictor of reading success
Dyslexia: A Phonological Difficulty
What is Phonological Processing?
• Refers to the child’s ability to use phonological information – in particular, the sound structure of oral language – to process oral and written language
• Predicts reading acquisition• Most widely recognized locus of deficit
in developmental dyslexia• Three kinds of phonological processing
Phonological Awareness
• Refers to a child’s awareness of and access to the sound structure of oral language
• Progresses from awareness of words to syllables to onsets/rimes to phonemes
• Required for systematic application of the alphabetic code
How Is Your Phonological Awareness?
• Say flimrix without /flim/• Say jopfrad without /frad/• Say sornch without /ch/• Say stipnir without /t/• Say sflummy without /l/• Say tropunwax without /pun/• Say belaxidrous without /k/• Say shalopiniwen without /pi/
• Refers to the child’s ability to code information phonologically for temporary storage in working memory.
• Relies on the phonological loop – the phonological store containing the most recent 2 seconds of phonological information, as well as information in long-term memory (see work by Sue Gathercole)
Phonological Memory
How Is Your Phonological Memory?
• zep dax• nich lav gup• steck ruv poll nost• poll zep gup ruv nich• lav dax nost ruv steck zep• gup ruv lav dax nost steck poll• nich zep steck ruv dax gup nost gup
Rapid Naming as Phonological Processing
There is some disagreement about what tests of rapid naming assess, questioning the construct validity of these measures.
According to the CTOPP manual, “rapid naming of objects, colors, digits, or letters requires efficient retrieval of phonological information from long-term or permanent memory” (p. 6).
Young readers must efficiently retrieve sounds associated with letters, common letter segments, and whole words. If phonological codes cannot be retrieved efficiently, decoding will be slow and effortful.
How Is Your Rapid Naming?
Medicine Law Design
Government Public Health
Where to begin (generally)
• Working with real words is easier than working with nonsense words
• Dealing with bigger chunks is easier than dealing with isolated sounds
• Choosing is easier than producing• Beginnings are easier than endings/middles• Deleting sounds is easier than adding or
substituting• Blending is easier than segmenting
What is the common denominator?
phonological representations of lexical items areunder-specified, “fuzzy”
phonologicalrepresentati
on
semantic representation
motorprogramme
phonological representati
on
Phonological Representation Hypothesis of Dyslexia
OK, but…
1. How does a child develop a well-specified phonological representation?
2. What is the role of auditory processing?
3. Different languages have different phonological systems – what are the implications for dyslexia?
4. How has the phonological representations hypothesis informed remediation?
• Infants have early-emerging perceptual biases e.g. to mothers’ voice, rhythm of language, categorical perception
• Early input allows statistical learning and enhanced sensitivity to phonological contrasts of language
• Infants begin to segment words from speech stream around 7-8 months (Jusczyk & Aslin)
How does a child develop a well-specified representation?
But it takes a while for word-learning to be able to fully utilize the infant’s perceptual sensitivities…
8 months - infants can perceptually differentiate bih vs. dih, but fail in ‘switch procedure’
17 months – success at switch task
Infant should look longer/besurprised
Stager and Werker
Important factors
1. Developmental pressure for segmental specificityas acquire vocabulary (present in all languages)
2. Phonological factors such as phonological neighborhood density and its characteristics in different languages
(after e.g., Walley & Metsala, 1998)
Phase 1: Spoken words have holistic representations
Phase 2: Children represent syllabic segments
Phase 3: Children represent onsets and rimes
Phase 4: Full phonemic representation develops
Developmental pressure for segmental specificityas acquire vocabulary (present in all languages)
Phonological factors such as phonological neighborhood density and its characteristics in
different languages
Logically, there are 3 types of neighbor for CVC words:
cot - cop: onset-vowel neighbor (OVN)cot - cat: consonant neighbor (CN)cot - hot: rime neighbor (RN)
>> lots of RN = rime well-specified?
>> lots of OVN = OV well-specified?
COT:49 neighborsrime = 24
cop cat hot gotcockkit lot notcog kite dot whatcod curt rot yachtcon.. cute..jot pot
shotscot..
COOK: 17 neighborsrime = 9
could lookkick.. book..
Phonological Neighborhoods
25
35
45
55
65
75
Dense RN Sparse RN
% c
orr
ec
t
Dys CA RL
Will children show better performance in rhyme tasks when words are from dense RN?
Further questions
1. How does a child develop a well-specified phonological representation?
2. What is the role of auditory processing?
3. Different languages have different phonological systems – what are the implications for dyslexia?
4. How has the phonological representations hypothesis informed remediation?
phonological representations of lexical items areunder-specified, “fuzzy”
phonologicalrepresentati
on
semantic representation
motorprogramme
phonological representati
on
Phonological Representation Hypothesis of Dyslexia
1. Why does it matter in the classroom whether the trouble-spot is auditory or phonological?
Rosen, S. (2003). Auditory processing in dyslexia and specific language impairment: is there a deficit? What is its nature? Does it explain anything? Journal of Phonetics, 31, 509-527
Note the auditory processing jury is still out…
Further questions
1. How does a child develop a well-specified phonological representation?
2. What is the role of auditory processing?3. Different languages have different
phonological systems – what are the implications for dyslexia?
4. How has the phonological representations hypothesis informed remediation?
syllable - onset/rime - phoneme
window win dow
street str eet
street /s/ /t/ /r/ /i/ /t/
Development of Phonological Awareness: Language Universal?
Tapping Task: English vs Italian
0
20
40
60
80
100
Eng 1 Eng 2 Ital 1 Ital 2
% c
orr
ec
t
phoneme syllable
PHONOLOGICALREPRESENTATIONS
OF WORDS
Vocabulary size andrate of expansion
(language universal?)
Speech processingskills (input and
output)(language universal?) Phonological
neighbourhooddensity
(language specific?)
Linguistic factorseg, sonority profile
(language specific?)
Word frequency/familiarity/
ageof acquisition
(language universal?)
Speech-readingeg, lip shape(language universal?)
Reading and spelling acquisition
(phoneme level)(language specific?)
1. When children learn to read, they need to learn howsounds are represented by letters
2. The most accessible phonological units to the beginnerare ‘large’ ones - syllables, onsets, rimes
3. The most accessible orthographic units are letters,which typically correspond to phonemes (‘small’ units)
>> this creates a “mapping dilemma”
Mapping Sounds to Letters:Universal Dilemma, Language Specific Solutions?
1. Languages like Italian and Spanish have many simpleor open syllables (consonant-vowel structure)
2. Here, children with vocabularies segmented at theonset-rime level can easily solve the mapping problem
3. For many words, onsets and rimes = phonemes
e.g. gatto (cat) /g/ /a/ /t/ /o/casa (house) /c/ /a/ /s/ /a/
Mapping Sounds to Letters:Consistent Alphabetic Orthographies
4. Coupled with consistent spelling systems, in whichthe same letters always map to the same phonemes, thisreduces the mapping problem considerably
e.g. gatto (cat) /g/ /a/ /t/ /o/ G-A-TT-Ocasa (house) /c/ /a/ /s/ /a/ C-A-S-A
Mapping Sounds to Letters:Consistent Alphabetic Orthographies
5. For languages with more complex syllable structures, like German, onset-rime segmentation does not maponto phonemes
e.g. Ball /b/ /al/ B - ALLHand /h/ /and/ H - ANDGarten /g/ /ar/ /t/ /en/ G - AR - T - EN
However, spelling-sound consistency still eases themapping problem. Letters consistently predict phonemes.
Mapping Sounds to Letters:Consistent Alphabetic Orthographies
Mapping Sounds to Letters:Inconsistent Alphabetic Orthographies
• Inconsistent spelling systems, like English, increase the mapping problem, particularly for langs with complex syllables:
- onset-rime segmentation does not map to phonemes - letters map to different phonemes
e.g. ball /b/ /al/ B - ALLhand /h/ /and/ H - ANDgarden /g/ /ar/ /d/ /en/ G - AR - D - EN
Languages vary in the degree to which letters have a 1:1mapping to sound
GreekFinnishGermanItalianSpanish
SwedishDutchIcelandicNorwegian
FrenchPortugueseDanish
English
Orthographic Consistency
Implications for learning foreign languages?
Phoneme counting task at end of Grade 1
% correct
Greek (Porpodas) 100Italian (Cossu et al.) 97Turkish (Durgunoglu & Oney) 94German (Wimmer et al.) 92Norwegian (Hoien et al.) 83French (Demont & Gombert) 61English (Perfetti et al.) 65
Effect on phonemic awareness
0
20
40
60
80
100
E Da Fr Du Ge Sw N Ic It Sp Po Gk Fi
% c
orr
ect
NW Word
COST A8: Word and Nonword Reading (cvc), Grade 1
Phonological deficit:
usually speed rather than accuracy
Poor phonological recoding:
nonword reading deficit, extremely slow
Poor reading and spelling
effortful decoding and poor spelling
Profile of Dyslexia across Languages
0
20
40
60
80
100
Greek German
% c
orr
ec
t
Dyslexic CA
Nonword Reading: German & Greek“At Risk” beginners (Porpodas, Wimmer)
0
1
2
3
4
Greek German
tim
e i
n s
ec
s
Dyslexic CA
Speed of Nonword Reading in secondsGerman (G2) and Greek (G1)
E. Paulesu, J.-F. Démonet, F. Fazio, E. McCrory, V. Chanoine, N. Brunswick, S. F. Cappa, G. Cossu, M., Habib, C. D. Frith, U. FrithScience, 2001
Dyslexia: Cultural Diversity and Biological Unity
E. Paulesu, J.-F. Démonet, F. Fazio, E. McCrory, V. Chanoine, N. Brunswick, S. F. Cappa, G. Cossu, M., Habib, C. D. Frith, U. FrithScience, 2001
Dyslexia: Cultural Diversity and Biological Unity
• in the six groups of normal controls (A)
• and in the six groups of dyslexics (B)
• (C) The figure shows the brain areas that were significantly more active in all normal compared to all dyslexic readers.
Further questions
1. How does a child develop a well-specified phonological representation?
2. What is the role of auditory processing?
3. Different languages have different phonological systems – what are the implications for dyslexia?
4. How has the phonological representations hypothesis informed remediation?
Great resources
http://dww.ed.gov/Doing What Works – translational site
http://ies.ed.gov/ncee/wwc/What Works Clearing House – your one-stop shop for evidence-based practice
Question…
“There ain't one single piece of scientific evidence that says that phonics helps dyslexics. In fact scientific research suggests otherwise. If anyone can show me one single research article published in a respected journal that validates the use of phonics to teach dyslexics, I will shave my head.”
One answer…
• 2-6% of all children remain poor readers in spite of the application of intensive programs
Torgesen…
• “We know that approaches featuring systematic, explicit instruction in phonemic awareness and phonetic decoding skills produce stronger reading growth in children with phonological weaknesses than do those that do not teach these skills explicitly (Brown & Felton, 1990; Felton, 1993; Foorman et al., 1998; Hatcher, Hulme & Ellis, 1994…..)”
Torgesen…
• “It also seems clear that these phonemically explicit approaches should include careful instruction to help children apply their phonetic decoding skills to real words and they should provide many opportunities to read connected text for fluency and meaning”
Torgesen…
“HOWEVER, to know what kind of instruction is most effective is not the same thing as knowing how much of that instruction, delivered under what conditions, will lead to adequate development of word reading and passage comprehension skills in children with phonological processing weaknesses…”
Torgesen concludes…
• We must examine the intensity and duration of instruction required to eliminate reading failure in children with the most severe phonological disabilities and most disabling environmental backgrounds”
Different phonic approaches
• Given rise to successful use of systematic phonic approaches in early grades
• Synthetic phonics – explicitly teaches readers to segment words into individual letters/letter combinations and to select phonemes for the letters
• Analogy phonics – teach students to pronounce/spell a new word by making an analogy to a known word (word families)
• Analytic phonics – takes analogy phonics a step further. Teach students to analyze letter-sound relationships by comparing unknown words to known words (avoiding saying sounds in isolation e.g. bar – begins like bus and ends like car)
• These studies report overall gains and a one-size fits all strategy:
A) If you have carried out phonics work with children, what determines your approach?
B) Do you think better matching to programs to children could result in less ‘treatment resisters’ or do you think something other than phonics is needed in these cases
(and if so, what?)
Resulting questions