Spelling and writing

1
Spelling and writing

• What is dysgraphia?
• Cognitive models of spelling and writing
• Types of dysgraphia
• Central dysgraphias
• Peripheral dysgraphias
• Nonlexical spelling?
• Summary

2
What is dysgraphia?

• A broad division
• Central and peripheral acquired dysgraphias (or
  agraphias) refer to disorders of spelling and/or
  writing.
• Dysgraphias that affect spelling are central and
  those that affect writing are peripheral.
• Will consider acquired disorders following brain
  damage (not developmental difficulties).

3
Cognitive model of writing and spelling

• Three routes to writing from speech input
• 1) Lexical-semantic route.
• 2) Lexical-nonsemantic route.
• 3) Nonlexical route.
• Inner speech?
• Need to posit separate spoken and written output
  lexicons in the model because of the many cases
  of patients who have anomia but not agraphia (and
  agraphia without anomia).
• See Ellis Young, Chapter 7.

4
The lexical semantic route

• A word can be recognised as familiar via the
  auditory input lexicon (PWD patients).
• The meaning of the word is accessed from the
  auditory lexicon via the semantic system.
• The orthographic form of a written word is
  retrieved via the graphemic output lexicon.
• The letter representations of a word are
  assembled via the graphemic output buffer.
• This route can also be called spelling via
  meaning" as only applicable to familiar words.

5
The lexical nonsemantic route

• A word is recognised as familiar by activation of
  representations in the auditory input lexicon.
• The spoken form of the word is accessed directly
  via the phonological output lexicon.
• The written form is retrieved via the graphemic
  output lexicon using the spoken word form.
• The letter representations of a word are
  assembled via the graphemic output buffer.
• Again, this is only applicable to familiar words.

6
The nonlexical (or sublexical) route

• This is assumed to be activated when we spell
  unfamiliar words such as invented nonwords.
• It is a metaphor for the application of knowledge
  about sound-spelling correspondences in order to
  assemble an oral or a written spelling.
• Will misspell words that have irregular
  spellings
• e.g.,CASTLE -gt "cassel TABLE -gt "taybul
• these are called regularisation errors.

7
Types of dysgraphia

• Central dysgraphias
• Phonological dysgraphia
• Surface dysgraphia
• Nonsemantic spelling
• Deep dysgraphia
• Peripheral dysgraphias
• Grapheme level impairment
• Allograph level impairment
• Motor pattern impairment

8
Phonological dysgraphia

• Shallice (1981) reported patient PR.
• PRs ability to spell low frequency familiar
  words was at an above-average level - for example
  he could spell the word "coniferous".
• Very poor at spelling nonwords.
• 2/10 four letter nonwords and 0/10 six-letter
  words.
• Errors to nonwords tended to be similar to real
  words 'lexicalisations'
• "na" -gt gn (from "gnat") "sult" -gt ault (from
  "assault")
• Had trouble generating letter sounds (/b/ for B).

9
Loss of nonlexical route

• Spelling in the absence of phonological mediation
  (i.e., poor phoneme awareness).
• Vocabulary or lexically based spelling.
• Spelling can thus proceed via an autonomous
  module called the graphemic output lexicon and
  does not require phonological mediation.
• Other patients (Baxter Warrington, 1986 Nolan
  and Caramazza, 1982 Roeltgen et al., 1986).

10
Surface dysgraphia

• Beauvois and Derousne (1981) patient RG.
• Spelling unambiguous words GOOD (90).
• Spelling irregular words POOR (40).
• Spelling nonwords GOOD (100).
• RG showed a regularity effect on spelling.
• e.g., yacht --gt yot.
• Could not spell homophones correctly either
• e.g., steak --gt stake.

11
Loss of lexical routes

• Spelling by sound.
• There is a modular phoneme to grapheme system for
  spelling words (nonwords OK).
• Over-reliance on nonlexical route due to
  impairment of lexical routes.
• Difficulties with exception words ("island") and
  ambiguous words ("team") because they require
  semantic knowledge (Graham, 2000).
• Other patients (Hatfield Patterson, 1983
  Parkin, 1993b Weekes and Coltheart, 1996).

12
Nonsemantic spelling?

• Bramwells case of word meaning deafness.
• This patient could spell to dictation words she
  could not understand from spoken word input.
• These words included irregularly spelled words
  such as the word Edinburgh.
• Patterson (1986) reported a similar patient GK.
• Some patients with dementia show the same pattern
  i.e. an absence of knowledge about the meaning of
  a word from any modality but have intact spelling
  of irregular words (Graham, 2000).

13
Loss of semantic and nonlexical routes

• Patterson (1986) proposed the direct route from
  the auditory input lexicon to the graphemic
  output lexicon (via the speech output lexicon)
  thus bypassing the semantic system.
• This could account for cases of people who can
  spell irregular words without comprehension.
• Recent evidence suggests that the semantic system
  supports the spelling of irregular words - the
  semantic glue hypothesis (Graham, 2000).

14
Deep dysgraphia

• Bub Kertesz (1982) reported patient JC.
• Made semantic errors when spelling words.
• e.g., time -gt clock
• e.g., sky -gt sun
• JC had better spelling of concrete than abstract
  nouns and this is called an imageability effect.
• JC also displayed impaired nonword spelling.
• 5/20 four letter nonwords, 0/17 eight letter
  nonwords.

15
Loss of nonlexical and nonsemantic routes

• Suggests that the route from semantics to the
  graphemic output lexicon was inoperative.
• JC did not make semantic errors when reading
  aloud so this is evidence in favor of separate
  reading and spelling systems in the brain.
• JC recovered completely (after 6mths) thus ruling
  out a developmental explanation of the dysgraphia
  and also showing the plasticity of spelling
  skills.

16
One orthographic lexicon or two?

• This dissociation between reading and writing
  underpins contemporary models of written language
  comprehension and production.
• Unitary model of writing/reading (Behrmann Bub,
  1992 Allport Funnell, 1988).
• Dual lexicon model of writing/reading (Ellis
  Young, 1996 Weekes and Coltheart, 1996).
• Weekes (1996) reported patient NW who was surface
  dysgraphic and surface dyslexic.
• Rehabilitation of spelling skills did not affect
  his reading performance and vice versa.

17
Grapheme level impairment

• Miceli, et al (1985) reported patient FV.
• No aphasia but frequent spelling errors such as
  additions, deletions, transpositions and letter
  substitutions in writing to dictation (e.g b for
  p).
• Affected oral spelling aloud and handwriting.
• Affected both word and nonword spelling.
• Copying was intact.
• Normal letter formation (intact motor skills).
• Problem at the grapheme level where abstract
  letter identity forms are stored and ordered.

18
Allograph level impairment

• Goodman Caramazza (1986).
• MW.
• Letter substitutions in handwriting but not
  spelling aloud (so graphemic level is OK)
• starve -gt starze
• pierce -gt tierce
• Letters were well formed - (graphic motor OK)
• The assignment of shape of written letter form
  from a graphemic unit is impaired - this is
  called the allograph level of representation.

19
Other patients

• Patient JEC (Weekes, 1994).
• He made frequent case errors (e -gt E).
• Wrote upper case letters well but some lower case
  letters were written poorly (e.g. b p q d).
• The spatial similarity of the letters was the
  best predictor of substitutions, omissions etc.
• JEC also displayed some surface dysgraphia
  leading some to argue that peripheral writing
  errors result from loss of semantic memory
  (Patterson, Hodges, 1997 Graham, 2000).

20
Types of error

• letter substitution errors apple --gt aBBel
• letter omission errors swing --gt swin
• letter addition errors across -gt acccross
• case substitution errors Queen --gt quEEn
• Letter substitution errors were predicted by the
  spatial similarity of the lower case target to
  other lower case letters e.g., b, p, d, q.
• crab --gt craB ladder --gt laDDer
  apple --gt aBBel

21
Graphic motor pattern impairments

• Selecting letters for written output
• Baxter Warrington (1986) patient I.D.T.
• Problem in retrieving graphic motor patterns.
• Could spell aloud (so grapheme level OK).
• Could copy words and letters well and describe
  the shape of letters (so allograph level OK).
• Couldn't write at all - not even three letter
  words.

22
Other patients

• Executing (spatial dysgraphia)
• Ellis, Young Flude (1987) patient VB.
• Know the sequence of movements but problem in
  executing this sequence.
• Patient tends to write down one side of page.
• Patient omits or duplicates letters and strokes.
• Problem attending to visual and kinaesthetic
  feedback during writing?

23
Summary

• Studies of dysgraphic patients reveal that
  spelling and writing depend on multiple routes.
  and the functional architecture of the normal
  spelling/writing system is highly modular.
• The evidence from surface dyslexic and dysgraphic
  patients suggests that there is more than one
  orthographic lexicon used for reading and writing
  although this hypothesis awaits converging data
  from brain imaging.

24
Readings

• Parkin Chapter 8.
• Ellis Young (Chapter 78).
• Graham, N. (2000). Dysgraphia in dementia.
  Neurocase, 6, 365-376.
• Graham, N. Patterson, K Hodges, J (2000). The
  impact of semantic memory loss on spelling
  evidence from semantic dementia.
  Neuropsychologia,38, 143-163.
• Weekes, B (1994). A cognitive neuropsychological
  analysis of allograph errors from a patient with
  acquired dysgraphia. Aphasiology, 8(5), 409-425.
• Weekes, B.S. Coltheart, M. (1996). Surface
  dyslexia and surface dysgraphia Treatment
  studies and their theoretical implications.
  Cognitive Neuropsychology, 13(2), 277-315.

25
Weekes and Coltheart (1996)
26

• One example of this is the fact that the
  dissociation between letter naming and word
  recognition provides strong evidence for the
  modularity of letter specific and word level
  representations (Berhmann and McLeod, 1995).
• Similarly, the dissociation between impaired
  reading and preserved writing (the latter being a
  more difficult task) is among the strongest
  pieces of evidence in favor of the view that
  reading and writing are modular skills.
• The phenomenon of alexia without agraphia is also
  germane to the issue of whether there is one
  lexicon (Behrmann and Bub, 1992) or two lexica
  (Weekes and Coltheart, 1996) for reading and
  writing.