Pure alexia letterbyletter reading n n Introduction This

Pure alexia (letter-by-letter reading) n n Introduction This form of acquired dyslexia has been recognized for more than a century, the French neurologist Dejerine having described it in 1892. It was called "pure" alexia because it is the only form of acquired dyslexia known in which writing and spelling could be unaffected; in all other forms of acquired dyslexia, some form of impairment of writing and spelling will be seen. The first case studied systematically from a cognitive-neuropsychological (i. e. model-based) point of view were those of Warrington and Shallice (1980), who referred to the disorder as "word-form dyslexia", and Patterson and Kay (1982), who referred to it as "letter-by-letter reading". There has been much work on this form of acquired dyslexia over the past decade.

Jules Dejerine (1849 -1917) Alessia pura senza agrafia: Ipotesi in termini di disconnessione: gli stimoli visivi sono percepiti normalmente dal lobo occipitale destro ma non possono essere trasferiti all’area del linguaggio a causa della lesione del corpo calloso

The symptoms of pure alexia. n n Word and nonword reading are very slow, and the patient normally reads by spelling out the word or nonword aloud (naming each letter in left-to-right sequence) - hence the term "letter-by-letter reading". If all the letters are named correctly, then the word or nonword is likely to be then read aloud correctly. But reading will be very slow and the time to read a word will vary strongly with the number of letters in the word. Here are typical data from a case of pure alexia (GT, March 1991): Notice particularly that the time taken between when a word is presented is a linear function of the number of letters in the word. Every extra letter adds an extra 4. 36 seconds to her reading time. That shows she is identifying the word by identifying each of its letters first, and she needs about 4. 36 second to identify each letter. Spelling and writing can be intact and usually is. So the patient can write a passage but then can't read it, especially because it is in handwriting, which is particularly hard for letter-by-letter readers to read, since the individual letters aren't clearly separated in handwriting. Single letter recognition can be intact, though sometimes it is impaired as well as word recognition. There is usually no aphasia i. e. understanding and producing spoken language is usually normal.

The typical lesion locations in pure alexia. n n Almost all reported cases of pure alexia have had two lesions, in the same two places. The first is left occipital; damage here is usually extensive enough that the patient has a right hemianopia. That lesion, however, would not make reading of words presented to the left visual hemifield abnormal. The second lesion usually also present is a lesion of the splenium, which is a posterior region of the corpus callosum. This must mean that, in people with just a right hemianopia, their normal reading of words presented to the right visual field is achieved by (a) initial visual processing by right visual cortex and then (b) transfer of the processed information via the splenium to reading centres in the left hemisphere. When the splenium is damaged, this right-to-left transfer can't occur in the normal way. Something must be transferred, however, or else reading by the left hemisphere couldn't happen. So presumably the slow and letter-by-letter reading that is achievable reflects the way transfer to the left hemisphere happens when the normal transfer pathway (the splenium) is not available.

Pure alexia n Pure alexia: a possible explanation is that what the brain damage has done here is make the letter identification system work very slowly and one letter at a time when it is sending information to the visual word recognition system and letterto-sound rules. Normally this system is fast and deals with all letters at once. This can't explain why some pure alexics show covert reading i. e. some knowledge of the meanings of words obtained well before they can read the word aloud. That might be something that the right hemisphere of the brain does; the diagram we are using here is only a diagram of the left hemisphere's reading system, and says nothing about what kinds of reading might be possible for the right hemisphere.

Covert (unconscious) reading in pure alexia. n One of the nine-letter words presented to GT (see the figure above) was the word professor. This took her more than thirty seconds to read; but about three seconds after its onset, she turned briefly and smiled at the person who was doing the testing - who was in fact a professor. This suggests, at an anecdotal level, that she had accessed the meaning of the word long before she could read it aloud (or even consciously recognize it). This is referred to as covert word recognition, and experimental studies, beginning with Shallice and Saffran (1986), have shown that it happens in many (but not all ) cases of pure alexia. Pure-alexic patients asked to do visual lexical decision, or semantic classification (is the name of an animal or not? ), with stimuli presented briefly (1 -2 seconds) will initially refuse, since they will (correctly) say that at such short exposures they cannot report anything from the display. When cajoled to "guess", however, many patients score very much above chance in such situations, even though they feel certain that they are always just guessing.

The theoretical interpretation of pure alexia. n n n Consider first deep dyslexia. If it is correct that reading in deep dyslexia is carried out by a right-hemisphere system, then that shows that at least for some kinds of words (particularly concrete nouns) word reading can be quite fast and even reasonably accurate when being done by the right hemisphere. Since pure alexics have intact right hemispheres, why then can't they read at least this kind of word reasonably promptly? But they can't: the slow letter-by-letter reading strategy is till evident even when the stimulus is a common concrete noun. A plausible account is as follows. Pure alexics have two reading options available to them: * They could use their intact right hemisphere reading system. That would make nonword reading impossible, the reading of function words and abstract words very difficult, and morphological and visual errors would occur (since these are all aspects of reading via the right hemisphere system). But at least when a word could be read it could be read reasonably promptly. * Or they could use their intact left hemisphere reading system. That would allow reasonably accurate reading of all types of word and nonword, but it would require first of all transfer of information from the right hemisphere to the left via a nonsplenial pathway, and that transfer will make reading slow and letter-by-letter. It is the second option they choose. That option is not available when presentation duration is short; here any reading they can do is via the first option. That is how they achieve above-average performance on lexical decision and semantic categorization tasks with brief presentation. This account predicts that, when presentation is short, reading by pure alexics should show similarities to deep dyslexic reading, and recent work by Saffran, Coslett and colleagues has provided evidence that this is so.

The theoretical interpretation of pure alexia. n n This all looks very promising, but what still needs to be explained is this: the pure alexic appears completely unaware of the availability of the first of these reading options, believing that he or she is simply guessing when presentation is brief. Why this unawareness? On this approach, it is not appropriate to try to interpret pure alexia in terms of some pattern of damage to a model of normal reading. The normal reading system in the left hemisphere is not damaged at all. The problem is not in that system, but in the input to it, which is abnormal because of the splenial damage. Here we need to know what's coming across the nonsplenial pathways. For example, are letters identified in the right hemisphere and then their identities submitted, slowly and serially, to the left hemisphere reading system via nonsplenial parts of the corpus callosum? Or is what's sent over raw visual information (e. g. bundles of visual features corresponding to letters), and these bundles are identified as specific letters by the letter identification level of the left-hemisphere reading system? We won't understand pure alexia fully until we know the answer to this question. But whatever the answer is, it does seem that pure alexia should not be viewed as arising from damage to the normal reading system itself.

Prognosis and treatment of pure alexia. n n Pure alexia sometimes resolves rapidly in the weeks or months after brain injury, becoming much milder if not completely resolved. This is not so for all cases, however. Where reading remains impaired, several rehabilitation techniques have been applied. Moyer (1979) and Moody (1988) used a technique involving repeated reading-aloud practice with familiar text passages and then transfer to new text passages, and this did improve reading to some extent. Coslett, Saffran and their colleagues have begun to explore a new approach to rehabilitation of pure alexia, based on the evidence given above that in, pure alexia, there is a right-hemisphere reading option of which the patients are unaware. If pure -alexic patients can learn to adopt a reading mode that makes some use of that option, their reading might become faster and less laborious.

References n n n Behrmann, M. and Mc. Leod, J. (1995) The rehabilitation of letter-by-letter reading. Neuropsychological Rehabilitation, 5, 1/2, 149 -180. Bowers, J. S. , Arguin, M. , & Bub, D. N. (1996). Fast and specific access to orthographic knowledge in a case of letter-byletter surface alexia. Cognitive Neuropsychology, 13, 525 -568. Bowers, J. S. , Bub, D. N. , & Arguin, M. (1996). A characterization of the word superiority effect in a case of letter-by-letter surface alexia. Cognitive Neuropsychology, 13, 415 -442. Coslett, H. B. & Saffran, E. M. (1989). Evidence for preserved reading in pure alexia. Brain, 112, 327 -329. Coslett, H. B. & Saffran, E. M. (1994). Mechanisms of implicit reading in pure alexia. In Farah, M. J. & Ratcliff, G. (Eds. ). The Neuropsychology of high-level vision. Hillsdale, N. J. : Erlbaum. Coslett, H. B. , Saffran, E. M. , Greenbaum, S. & Schwartz, H. (1993). Reading in pure alexia: The effect of strategy. Brain, 116, 21 -27. n Moody, S. (1988) The Moyer reading technique re-evaluated. Cortex, 24, 473 -476. n Moyer, S. B. (1979) Rehabilitation of alexia: A case study. Cortex, 15, 139 -144. n There will also be a Special Issue of the journal Cognitive Neuropsychology devoted to pure alexia, to appear early in 1998.

Visual dyslexia n n n n Two cases of acquired visual dyslexia were described by Marshall and Newcombe (1973), but there has been very little subsequent research on this disorder. The reading errors made by such patients are largely or exclusively visual - that is, the error response is a word which shares many letters with the stimulus, such as: * arrangement -> "argument" * calm -> "claim". My impressions are * that in visual dyslexia, nonword reading also leads to these kinds of errors i. e. a nonword such as belm might be misread as a visually similar word such as "beam" or "bell". * that the visual errors do not arise at the letter level, because when the patient is asked to spell the letters aloud this may be done correctly, even though a visual error occurs when the patient is trying to read the word itself aloud. So if you ask the patient to spell the word aloud first and then read it, you often see this kind of responding: calm --> "C". . . "A". . . "L". . . "M". . . "claim" These, however, are just impressions based on brief contact with such patients, and need to be documented by systematic research.

Reference n Marshall, J. C. and Newcombe, F. (1973) Patterns of paralexia: A psycholinguistic approach. Journal of Psycholinguistic Research, 2, 175 -199.

Neglect dyslexia n n n Introduction. Extensive damage to the right hemisphere often leads to the condition known as unilateral neglect. The patient will fail to pay attention to the left side of space, so may engage in conversation only when the speaker is in the right side of space, may eat only the food on the right side of the plate, may dress only the right side of his or her body, and may deny ownership of his or her left arm and leg. Reading may also be affected in the same way, i. e. the right side of the page will be neglected, and in single word reading the right side of the word may be neglected. However, neglect dyslexia itself is a separate disorder, because recent research (e. g. Patterson and Wilson, 1990; Haywood, 1996) has shown that a number of the patients with this form of acquired dyslexia do not show unilateral neglect outside the domain of reading. In these patients, processing in the left half of space is impaired only when what is being processed is a letter-string.

The symptoms of neglect dyslexia n n In neglect dyslexia, reading errors arise because letters in a particular position in a word or nonword are neglected. Almost all cases are cases of left neglect dyslexia i. e. letters at the left of the stimulus are omitted or misidentified. Hence in left neglect dyslexia, one sees such errors as river read as "liver", cabin read as "robin", or liquid read as "squid". However, cases of right neglect dyslexia, where errors arise with the rightmost letters of the letter-string, have also been reported.

The theoretical interpretation of neglect dyslexia n n Caramazza. and Hillis (1990) have proposed a model of early visual processing in reading and used that model to offer an interpretation of various forms of neglect dyslexia. In their model, letter strings go through three stages of processing prior to word recognition, as follows: Retinocentric level: here each letter is represented as a bundle of visual features associated with a specific location on the retina. When neglect dyslexia is occurring at this level, it is because stimuli to the left of the fovea (the fixation point) are being neglected. Therefore a patient with neglect dyslexia at this processing level will show no neglect dyslexia if all the letters of a word are presented in the right visual field (i. e. if the patient fixates to the left of a word that is being read). Stimulus-relative level: here the feature bundles of the retinocentric level have been integrated so each bundle is seen as a particular whole object; and a string of, say, 5 letters is seen as a string of 5 objects. Neglect at this level is relative to such a string: the leftmost object in such a string will be neglected regardless of where that string is on the retina. So if a patient's neglect is at this level, the neglect will occur even for words presented entirely in the right visual field. Graphemic level: -->

Graphemic level: The input from the stimulus-relative level is used to identify each object at that level as a letter. At the graphemic level letters are represented abstractly, not in terms of their specific cases, so the uppercase, lowercase and handwritten forms of a particular letter are represented identically. Moreover, if the letter string is presented in a distorted way such as: n or even mirror-imaged such as:

n these distortions will be eliminated and the string will be represented as [c] [a] [p] at the graphemic level. Information about where exactly the letters were in space or whether they were mirror-imaged or not is discarded. If it is at this level that the left neglect dyslexia is present, this means that with mirror-imaged presentation, the [c] of will be the one that is neglected, even though it is the rightmost letter on the page - because it is the leftmost letter in the graphemic representation. Caramazza and Hillis (1990) described a patient who showed exactly this pattern of performance.

References n n n Caramazza, A. and Hillis, A. (1990). Levels of representation, coordinate frames and unilateral neglect. Cognitive Neuropsychology, 7, 391 -445. Haywood, M. (1996) Neglect dyslexia without visuospatial neglect. Unpublished Ph. D thesis, Macquarie University. Patterson, K. , and Wilson, B. A (1990). A rose is a rose or a nose: A deficit in initial letter recognition. Cognitive Neuropsychology, 7, 447 -477.

Agrafia: modelli neurologici ðAgrafia afasica ð Agrafia aprassica ð Agrafia pura ð Agrafia spaziale ð Agrafia motoria

Acquired dysgraphia n n n Writing and spelling can be impaired in various ways by brain damage. Many of these parallel the various forms of acquired dyslexia: * Phonological dysgraphia: the patient can spell words but not pronounceable nonwords such as VIB or SLINT * Surface dysgraphia: the patient can spell nonwords such as VIB and SLINT, and also words whose spellings obey English spelling rules such as RUB or SPOT, but words that don't obey those rules are spelled as if the did e. g. "yacht" is spelled YOT or "blood" is spelled BLUD * Deep dysgraphia: semantic errors are made when spelling to dictation, so the patient might spell "parrot" as CANARY or "time" as CLOCK. There also peripheral dysgraphias, where the problem is at the extreme output level: the correct spelling is known, but letters are poorly or wrongly formed. There's a good discussion of this in the Ellis & Young text, Human Cognitive Neuropsychology.

Agrafia: modelli cognitivi n n Disgrafia superficiale (agrafia lessicale, disgrafia ortografica, spelling fonologico) Agrafia fonologica Agrafia profonda Deficit del buffer grafemico

Disgrafia superficiale n n n Deficit derivante da un disturbo della procedura lessicale Assente in lingue ad ortografia regolare(? ) Errori fonologicamente plausibili n n n Regolarizzazioni: inglese answer ----> anser; francese: monsieur ----> messieu

n Luzzatti et al. (1998) hanno trovato che n il 46% di disgrafici italiani mostravano un disturbo misto (sbagliavano sia parole irregolari che non-parole) n il 34% un danno prevalente alla routine lessicale (disgrafia superficiale) n il 13% un danno prevalente alla routine fonologica) sub-lessicale (disgrafia

n Esempi di ambiguità ortografica in italiano: n n n scie vs sce (scienza e non scienza) cie e gie vs ce e ge (cielo e non celo; igiene e non igene quo vs cuo (liquore, quota) (in nord Italia): balia vs baglia; geranio vs geragno (in quasi tutta Italia): libro e non libbro, febbre e non febre, biblico e non bibblico, pubblico e non publico

Agrafia fonologica è è Buona capacità nello scrivere parole (regolari e irregolari); cattiva capacità nelle scrivere non-parole (94 vs 18 nel paziente originario di Shallice 1981) I pazienti sono di norma in grado di ripetere le nonparole, escludendo un’incapacità a percepire o ripetere una stringa di fonemi.

Agrafia profonda Il paziente non è in grado di scrivere non parole che è in grado di ripetere èEffetto di immaginabilità e classe grammaticale nella scrittura di parole èParole concrete meglio delle astratte; meglio dei funtori èPresenza di errori semantici (moon invece di star; cake invece di bun, Newcombe e Marshall, 1980). è

Deficit del buffer grafemico n n Deficit presente nello spelling scritto od orale; sia per parole che non-parole Disturbo dipende dalla lunghezza dello stimolo Errori influenzano l’ordine sequenziale delle lettere (e. g. transposizione). Errori nelle geminate: avisso per avviso; leccula per cellula; cortecciia per corteccia.

Relazione tra dislessia e disgrafia acquisita n Sono descritti vari pazienti che presentano forme diverse di dislessia e disgrafia: n R. G. (Beavuois e Dérousné, 1979; 1981) era dislessico fonologico ma disgrafico superficiale n Un quadro simile è stato riportato in un paziente italiano (D. R. ). n DR leggere meglio le parole (81%) che le non-parole (43%); scriveva meglio le parole regolari (28%) che quelle ad ortografia ambigua (10%). ------->

Relazione tra dislessia e disgrafia acquisita n n Viceversa, non vi era differenza nello scrivere parole e non-parole Questi pazienti suggeriscono la presenza di due lessici indipendenti per la lettura e la scrittura in età adulta.