WHO DIFFER IN READING AND SPELLING ABILITIES D avid Stead Mather
B achelor of S cien ce, McGill U n iversity, 1959
M aster of A rts, Sta te ‘U niversity o f New York in Platt.',burg, 1984 A C ' f ’ F P T i-' i A D isserta tio n Subm itted in P artial F u lfillm en t
”* •* 1 of the R equirem ents for the D eg ree of
F A CU L TY
01
j i R A O U A US'i UDlES
4
b
DOCTOR OF PHILOSOPHY Interdisciplinary I,, / / ’ / • Y ' y :> £ / t *A I Language Arts P sychologyWe a c c ep t this dissertation as conform ing to the required standard
Dr* P .O . Evans, Supervisor (D epartm ent o f C om m unication and S o c ia l Foundations)
Dr. M. Joschko, Supervisor (D epartm ent of Psychology)
wBr. J. L. Hilf, M ember (D epartm ent o f P sy ch o lo g ica l Foundations in Education)
Dr. R.E. Tinney, Member (DepartmerU/o'f Psychological Foundations In Education)
Dr. P. Satz, ExternaflSxai?iiner (University of California - Los A ngeles)
@ DAVID STEAD MATHER, 1992 U n iversity of V ictoria
All righ ts reserved. D isserta tio n may not be reproduced in w hole or in part, by photocoping or other m eans, w ithout the perm ission o f th e
M(( hue! Josrhko, Ph.D.
ABSTRACT
Few stu dios have in v e stig a te d the d iffe r e n c e s betw een poo? readers/poor sp e lle r s (SRD) and good readers who dem on strate u nexpected ly poor sp ellin g sk ills (SSD), Those, that have done so have been mainly concerned w ith search in g for p sy cltolin gu istic sim ila r itie s and d iffe r e n c e s . This dissertation project is b eliev ed to be the first com parison o i the tw o disorders on a task that did not involve the use o f alphanum eric stim u li.
From a review of n eu rop sych ological and language arts research into reading and spelling failure, it was hy p o th esized that both SSD and SRD m ight be d iffe r e n tia te d from good reader/superior sp ellers (GRS) by their resp onses to a line o r ie n ta tio n task w hich had b een proven to be a valid indicator o f right hem isphere fu n ctio n . A previous study had found that this task, presen ted
co n cu rren tly w ith right or le ft hand tapping, discrim inated b etw een good and noor N a tiv e Indian readers (S tellern . C ollins, C ossairt Sc G utierrez, 1986). T he
th e o r e tic a l underpinning o f the current stu dy su ggested that th e se resu lts m ay have b een more c lo s e ly rela ted to sp ellin g than reading ab ility.
E m pirical support for this h yp oth esis was sought by com paring the perform an ce o f SSD, SRD and GRS early a d o le sc e n ts on the concurrent
ta p p in g -lin e o rien ta tio n judgem ent task. The data supported this hyp oth esis in that the SSD and SRD groups d iffered from the GRS group in d em onstrating
-h o w ev er, all t-h ree groups perform ed sim ilarly wit-h resp ect to rate and a c c u r a cy in judging line o r ie n ta tio n . As these result's w ere ambiguous as to w hether the right hand tapping in te r fer e n c e exp erien ced by th e poor spellers was the result o f
d iffe r e n c e s in hem isphere processing o f sp a tia l stim uli, other possible ex p la n a tio n s are co n sid er e d in th e discussion,
Examiners:
Dr. t* .0 . E v ads', Supervisor (D epartm ent o f C om tnunicalibn antTSocia 1 Foundations)
Dr, M. JoschK m Tsunervisof'(D epartm ent o f Psychology)
Efe J , L, H ill, M em ber (Depart rifent o f P sy ch o lo g ica l Foundations in Education)
Dr, R .fe, T inney, M em beirT E fepartm ent^^Psycholdgical Foundations in E ducation)
Dr. P . S a tz , E xtern al E xannnerT U niversity o f C alifornia - Los A ngeles)
-C o n te n ts . , ...
i v
T ables ... vi
Figures
...
vj.i
D edication ... vig Chapter I: Introduction . . J. 1.1 Purpose of the Study ... 51.2 R esearch H y p o t h e s e s ... . 6
1.3 Significance of the Study ... 8
1.4 Lim itations o f th e Study . . . 8
1.5 D efinition of Terms . ... 9
Cliapter I t Literature R e v i e w ... 11
2.1 Current Theories o f Reading and Spelling Disorders. . . . 11
2.2 SSD and SRD Sim ilarities and D ifferen ces ... . 16
2.3 Reading-Spelling D issociation ... 18
2 .4 The Dual Tash Paradigm ... 23
2.5 Summary o f the Literature . . . 30
2.6 Purpose of the Present Study ... 31
Cliapter l i t METHODS, PROCEDURES AND FINDINGS ...33
3.1 Subjects ... 33
3.2 M aterials ... 34
3.3 Procedures ... 37
Cliapter IV: DISCUSSION ... 50
4.1 Conceptual and M ethodological Problems . . . 51
4.2 Individual D ifferen ces ... 53
4.3 L eft Hemisphere Visuospatial Dysfunction ... 54
Chapter V: Conclusions and Im plications . . . ... 89
5.1 Findings R ela ted to Existing R e s e a r c h ... 61
5.2 Conclusions and Implications from this Study ... 62
R eferen ce s ... 65
Appendix A: Individual Subject Demographics and T est S c o r e s 81
1. Meant; iirrd S t a n d a r d D e v i a t i o n s l o r D e m o g r a p h i e s . . . . J 5 2
.
M e a n s a n d S t a n d a r d D e v i a t i o n s f o r J LO T A c c u r a c y S c o r e s a n d T a p p 1 no a n d J LO T D e c r e m e n t S c o r e s ...3 9 3 . S u m m a r y o f A n a l y s i s o f V a r i a n c e o f B a s e l i n e T a p p i n g P e r f o r m a n c e l o r SKI), SSD, a n d GRS G r o u p s . ...4 1 4 . G r o u p Oompa “i c o n o f B a s e l i n e T a p p i n g R a t e M e a n s a n d S t a n d a r d D e v i a t i o n s . . . ... 42 rx. S u m m a r y o f A n a l y s i s o f V a r i a n c e o f t h e E f f e c t s o f L i n e O r i e n t a t i o n J u d g e m e n t o n S i g h t - a n d L e f t - H a n d T a p p i n g R a t e s f o r SRD, SSD a n d GRS G r o u p s . . . . ...4 3 6 . S u m m a r y o f A n a l y s i s o f V a r i a n c e o f R i g h t - a n d L e f t - H a n d T a p p i n g E f f e c t s o n JLOT L i n e J u d g e m e n t P r o c e s s i n g R a t e f o r SRD, SSD a n d GRS G r o u p s . . . 4 6 7 . S u m m a r y o f A n a l y s i s o f V a r i a n c e o f R i g h t - a n d L e f t - H a n d T a p p i n g E f f e c t s o n L i n e O r i e n t a t i o n J u d g e m e n t ( J L O T ) A c c u r a c y . . . 481. Judgem ent o f Line O rientation * ,16
Z . R ight-an d L eft-H an d Tapping In terferen ce During Judgem ent o f ...44
3. JLOT P erforam an ce Errors Without Concurrent Tapping and With 49
D e d ic a te d to Brian, Adam , Graham and M ichael, whose g ifte d n e ss m o tiv a te d this pro ject.
INTRODUCTION
S p ecific reading disability (SRD) and sp ecific spelling disability (SSD) are term s which have been adopted to describe disorders of symbol processing that have y et to be explained. Specific reading disability, also referred to as dyslexia, has received a, great deal o f research attention over the past 50 years, usually aim ed at uncovering an underlying gen etic or constitutional explanation (see Aaron, 1989, for a recent review). In contrast, interest in SSD is fairly recent and is focused more on drawing comparisons betw een the spelling patterns o f SRD and SSD in the hope of obtaining a b etter understanding of the underlying linguistic ch aracteristics of both disorders (Bruck 8c Waters, 1988; Frith, 1980, 1985; Schwartz, 1983). The type of spelling disability exhibited by good readers/poor spellers (SSD) has been found to differ from that displayed by poor readers/poor spellers (SRD) both in the nature o f the spelling disorder (Frith, 1980, 1985'j dorm,
1981) and o f the underlying cognitive d eficit (Naidoo, 197?.;, Nelson 8t Warrington, 1974). However, sim ilarities have been observed in that both may be associated w ith poor handwriting skill (Share, Silva 8t Adler, 1987) and both may respond to rem ediation which em phasizes the motor patterns used in writing (Bradley, 1981).
Frith (1980) found that the major d ifference betw een good early adolescent readers who d iffered significantly in spelling sk ill was in the number rather than th e type nf errors, Frith described this d ifferen ce as a breakdown at th e le v el of
graphem e (letter) selection . While both groups produced phonologic ally intact m isspellings, the unexpectedly poor spellers appeared to lack the m ental store of visual representations of words which the good spellers called upon to se le c t the correct le tte r when more than one could be used to represent a given sound (e.g . th eir/th ere). Sloboda (1980), from research with adults, arrived at a similar hypothesis: "... good spellers store information about which letters ... word
contains whilst less good spellers store mainly information about which phonemes a word contains" (p, 246). Jorm (1981), in comparing SSD and SRD spelling
pat terns, of 10-year-old children, also concluded that SSD was characterized by d iffic u lties at the grapheme selection sta g e but further observed that the
phonological spelling accuracy o f good reader/poor spellers exceeded that of poor readers. To make m atters somewhat more com plex, d ifferences in phonological accuracy have also been observed in subtypes of poor readers (Boder, 1973; Sweeney & Rourke, 1978).
In attem pting to account for the phonetically accurate but orthographically (visually) irregular spelling performance o f poor readers, Sweeney and Rourke (1978) hypothesized that the fundamental problem may be "a d eficit in the evocation of language-related visualizations" (p. 214). They also suggested that this may represent a le ft hemisphere d e fic it, recalling Luria's (1973) observations that lesions within the le ft parieto-tem pero-occipital region in adults in terfere w ith the production Of visual im ages which prompt.* or are prompted by,
appropriate verbal labels. Thus both the Frith (1980) and Sweeney and Rourke stu d ies suggested that phonetically accurate misspellings may1 represent a phonsm e-graphem e visualization d eficit common to both SSD and SRD subjet+s.
If Luria's observations apply, it may be that both groups d iffer from good
reader/spellers (GRi!) in le ft hemisphere processing of nonverbal visual stim uli. In other words, an absence of le ft hemisphere language - related visualization may be associated with the presence of atypical non-language-relatod visualization cap acity. Evaluating this hypothesis is the focus of this dissertation.
The formulation of this study was influenced by the theoretical stan ce of Frith (1988) who hypothesized that SSD might be the consequence of prematurely learned stra teg ies which had failed to merge with others learned in la ,r phases o f developm ent (see Kirk, 1983, for a sim ilar opinion), She also hypothesized that d yslexia may represent a developm ental arrest in the learning of alphabetic skills, possibly due to a disorder c-f phonological awareness, From this perspective, a language-related visualization d efic it may represent a failure to learn how to encode visual spelling patterns in th e le ft hem isphere. An important question is whether this is due to a general visual encoding d eficit or whether it is sp ecific to alphanumeric symbols. The latter possibility is suggested by dual-task research which compared dyslexics and good readers/good spellers (GRS) on dichaptic stim ulation tasks, requiring the identification o f both m eaningless shapes und le tte r s presented sim ultaneously to each hand w ith sight occluded (W iteison, 1976). W iteison found that, for the haptic identification of nonsense shapes, d yslexics behaved as if they had two right hemispheres, That is, while the two groups dem onstrated comparable le ft hand (right hemisphere) perform ance, d yslexic right hand (le ft hemisphere) performance was sign ificantly b etter than GRS from age 10 to age 14 (the upper test lim it). With le tte rs, a right hand (le ft hemisphere) advantage was obtained with GRS whereas the opposite pattern (le ft
hand-right hemisphere advantage) was found with dyslexics. This pattern was con sisten t from age 8 to the upper age test lim it of 14 years. These results suggest that dyslexics differ from GRS in the manner in which th ey encode alphabetic stim uli and that this is not due to a general le ft hemisphere d e fic it in visu alization .
Le Doux, Wilson, and Gazzaniga (1977) speculated that hemispheric
sp ecia liza tio n for nonverbal tasks occurs by default, in consequence of language developm ent in le ft hemisphere space. From this perspective, W itelson's (1976) findings m ay imply a "default mechanism" d eficit in dyslexia, one that may be related to d ifficien cies in the hemispheric encoding of alphabetic sym bols and which may even be responsible for the posited developm ental arrest in alphabetic learning skill (Frith, 1985).
Dual-task tim e-sharing procedures have been used to infer cerebral la teralization of verbal and non-verbal a c tiv ities (H ellige & Longstreth, 1981; Kinsboume & Hiscock, 1983). Commonly these procedures involve tapping with one hand while performing verbal or spatial tasks such as reading aloud from Unfamiliar tex t or solving nonverbal block design problems (e.g ., H ellige 8c L ongstretch, 1981). Tapping in terferen ce is taken as indicating that th e
hem isphere on the opposite (contralateral) side o f the performing hand is involved in th e dual processing o f both tasks. Stellern, Collins, C ossairt, and G utierrez (1986) Used this paradigm to investigate the hemispheric sp ecialization o f N ative A m erican Indian children. They employed the Judgement of Line Orientation T est (JLOT) (Benton, Va?ney, 8c Hamsher, 1978) which has been dem onstrated in studies o f unilateral brain disease to be a strong measure of right hemisphere in teg rity (Benton, Hannay 8c Varnay, 1975; Poizner, Klima, Bellugi, 1987).
that N ative Indians process spatial tasks in the le ft hemisphere, in view of studies which have suggested that N ative Indian languages require greater righ ts
hemisphere involvem ent in language processing (cf. Rogers, Ten Houten, Kaplan, & Gardener, 1977). Instead, in a post hoc comparison, they found that reading ability predicted which hand experienced the greater tapping disruption dutdng concurrent JLOT perform ance. Over 95% of the good readers dem onstrated le ft hand (right hemisphere) interference, whereas 80% of the poor readers
dem onstrated right hand (left hemisphere) disruption.
In this dissertation, the concurrent tapping - line orientation judgement procedure is used to evaluate the lateralization of spatial processing in Caucasian early ad olescen t SSD, SRD and control subjects with good reading and superior spelling skills (GRS), As the Stellern e t al. (1986) study did not control for spelling ability, the present research should provide a clearer picture of the relationship betw een perform ance on this in terferen ce task and reading/spelling proficiency.
1.1 Purpose o f th e Study
The purpose of this study is to in v estig a te whether SSD and SRD right-handed early a d olescen ts dem onstrate abnormal right hand tapping disruption when perform ing a line-orientation task th at has been validated as a clinically useful indicator of right hemisphere impairment (Benton, Varney & Hamsher, 1978), In younger normal Caucasian reader/spellers (grades 2 to 5) this task has been found to disrupt le ft and right hand tapping performance equally, leading the
(Hiscock, Antoniuk, Prisciak, & von H essert, 1985). However, after the age o f 10 there is a transition from a phonological stage, wherein words are spelled largely by sounding out their phonemic elem ents, to a visual-orthographic sta g _•
characterized by greater use of visual word patterns (Nolen Sc M cCartin, 1984; Sm ith, 1992). In other words, at approximately 11 years of a ge, verbal lab els appear to becom e amalgamated with visual word im ages (Ehri, 1980), a developm ent that may depend on le ft hemisphere p arieto-tem p ero-occip itai function (Luria, 1913). Speculatively nonverbal spatial tasks, such as the JLOT, may be latevalized to the right hemisphere by default with th e result that early adolescent GRS may dem onstrate only le ft hand disruption w hile perform ing the JLOT. On the other hand if early adolescent SSD and SRD subjects d em on strate a right hand tapping decrem ent, such a finding would be com patible w ith Frith's (1985) hypothesis that spelling disability results from an arrest at a
pre-orthographic stage of alphabetic learning. That is, a failure to develop orthographic representation may leave functional space in th e le ft hem isphere for processing line orientation judgements.
1.2 R esearch Hypotheses
The sp ecific questions addressed in the study w ith 11 to 14 year old su bjects are as follows:
1. Right hand tapping disruption resulting from a concurrent line orien tation judgem ent task will be significantly greater for SSD and SRD as com pared to GRS.
This question is, exploratory in nature, designed to contribute to the
reader/spellers the line orientation judgement task could be exp ected to disrupt le ft hand tapping perform ance since this spatial task has been shown to be a robust indicator of right hemisphere integrity (Benton, Hannay & Varney^ 1975), Right hand tapping disruption in poor spellers would therefore indicate atypical le ft hem isphere spatial processing t' ac may be related to the d ifficu lty these subjects have with correctly ordering the spatial components (i.e*, letters) of words.
2. SSD and SRD right hand performance will not be Significantly different on thif' task.
This question addresses the issue of whether the right hand tapping disruption during a line orientation judgem ent task (cf. Stellern, Collins, C ossairt, &
G utierrez, 1986) is related to reading ability. The th eoretical underpinnings of this study (see Chapter E) suggest that any such in terferen ce is more likely to be related to spelling rather than reading dysfunction.
3. The right hand disruption differences w ill not be sign ifican tly related to either the ra te or accuracy of perform ance on the line orientation judgem ent task.
This question addresses the issue of whether poor spellers w ill perform the non-manual line orientaton judgement task at the expense o f disruption on the manual tapping task. H iscock e t al. (1985) found this to be the case w ith children who w ere without d e fic its in the areas o f reading or spelling.
1.3 Significance o f th e Study
This study grew out of the researcher's interest in whether th e cause of dyslexia is constitutional, environmental, or a com bination of both. MostJ
neuropsychological research has been focused on uncovering an underlying g en etic i ' T constitutional explanation of reading disorders (see Aaron, 1989, for a recent review ) in contrast to the psycholinguistic orientation o f the far le ss prevalent SSD research. This study is designed to explore whether both SSD and SRD share a com m onality Which is relevant to spelling ability. If th is hypothesis is validated, it may spark further neuropsychological research into the sim ilarities and
d ifferen ces betw een SSD and SRD. It may also lead to more research em phasis on environm ental-constitutional interactions since it is d ifficu lt to envision au innate neurological deficit; common to both SSD and SRD, that would com prom ise
spelling but not reading.
1.4 Lim itations of th e Study
The dual task interference task, used in this study to infer cerebral processing d ifferen ces, does not provide a direct measure o f the neuronal su bstrates involved in >;he spelling process. For exam ple, if this procedure is found to discrim inate betw een the hemispheric processing of good and poor spellers th ere is no Way of knowing whether both SSD and SRD are using the same intrahem ispheric stra teg y to solve the task. Right hemisphere lesions involving frontal, tem poral, parietal, and p arietal-occipital cortical areas have all been found to com prom ise judgem ent of line orientation (Benton, Hanny, & Varney, 1975). The most th a t may be
that both SSD and SRD dem onstrate atypical le ft hemisphere involvem ent in spatial tasks such as the JLOT.
A second lim itation is that there is lim ited control for in telligen ce due to d iffic u ltie s in obtaining school permission for administering IQ tests.
A third lim itation is that tim e constraints made it necessary to rely on reports and/or school psychologist's assessm ents in order to elim inate subjects whose academ ic d e fic it may have resulted from neurological, m edical, socioeconom ic, or em otional fa cto rs. This procedure may not have identified children whose
learning d iffic u lties may have resulted from preschool factors such as, for exam ple, inner ear infections or the sequellae of pre-natal or natal birth injury.
A fourth lim itation is that SSD and GRS groups were not ea sily form ed. In each category, many o f the students selec ted by their teachers did not m eet the spelling discrepancy criteria. The results of this study are therefore of lim ited gen erality for good readers. That is, classroom teachers w ere, on average, only able to id en tify approximately fiv e percent o f their good readers as having a marked contrast betw een reading and spelling ability. However th is was not the case for the SRD group as, in all cases, poor reading was accom panied by poor spelling.
1.5 D efin ition o f Terms
The follow ing term s are defined for the purpose of this document •
D yslexia - A developm ental reading disorder originating from unknown c a u s e s .
L aterality - A term that is used to mean lateral sp ecialization o f the cerebral hem ispheres.
Orthography - The study o f spelling and th e conventions of correct spelling.
P arietal lobe - A terra used to define a region underlying the parietal skull bone of each cerebral hemisphere that is involved in spatial functions.
Phonological awareness - A term used to describe the ability to d e te c t th e order Of sounds within words.
Temporal lobe - A term used to define the lobe of each hem isphere that is situated below and in front o f the parietal lobe and is involved in phonological functions.
LITERATURE REVIEW
In this chapter literature is reviewed from a number of areas of research which have been identified as important to this study* These include studies (a) that h,\ve been concerned with examining the etiology of reading and spelling disorders, (b) which have provided a basis for understanding how spelling may be d issociated from reading perform ance, (c) which are relevant to the dual task paradigm used to infer hemispheric processing stra teg ies and (d) which have described subtype d ifferen ces of disabled readers and spellers.
2.1 Current Theories o f Reading and Spelling Disorders.
Most of the research into dyslexia (sp ecific reading disability) has adopted the prem ise that th is disorder is attributable to developm ental problems o f basic constitutional (neurological) origin (Aaron, 1989; VellutinO, 1979)* Geschwind and a sso cia tes (Geschwind 8c Behan, 1982; Geschwind 8c Galaburda, 1985a, 1985b; Galaburda, Sherman, Rosen, Aboitiz 8c Geschwind, 1985; Galaburda, Signoret, 8c R onthal, 1985), on the basis o f structural anom alies d etected in autopsiad d yslexic brains, have theorized that the majority o f ca ses of developm ental dyslexia spring
from prenatal m icroscopic cellular anom alies. Briefly, the theory sta te s that d yslexia is due to an ex cess o f testosteron e during prenatal developm ent which, in som e m ales, slow s the maturation of the left hem isphere. This putative! y
-a ctiv -a te s compens-atory functions in the right hemisphere le-ading to -a sh ift in hand preference (left) and the developm ent o f superior visuospatial a b ilities (e.g., Symmes & Rapoport, 1972)* The underdevelopment of the le ft hem isphere is hypothesized to foster later reading and writing disturbances. The theory also postulates th at testosterone im pacts on the thymus gland, leading to an increased risk for autoimmune disorder in later life , In sum, the theory predicts a triadic association involving handedness (left), cognition (reading/language ability) and im m unological status (autoimmune disorder! in m ales who are exposed prenataliy to ex cessiv e amounts o f testosteron e (see Geschwind & Galaburda, 1986, for a recent summary). ,
This "imbalance hypothesis" has a ttra cted considerable a tten tion . For exam ple Aaron (1989), in a pilot study, obtained WISC-R scores from two
10-year-old dyslexic m onozygotic twins. Both showed a large d iffere n ce b etw een their verbal and performance IQ scores (PIQ>VIQ by 40 and 50 points). Aaron concluded that:
The fa c t that both boys show a similar degree of discrepancy renders birth injury unlikely. The influence of prenatal
genetic-neurochem ical factors is a more likely explanation, (p. 129) However Satz and Soper (1986) have criticized the hypotheses advanced by Geschwind and Galaburda for significant design and m ethodological d eficien cie s. Further Hynd and Semrud-Clikeman (1989), in an exten sive review o f postm ortem and neuroimaging (CT/MRI) studies of dyslexics, concluded th at th e available evidence does not establish th e claim that the brains o f developm ental d y slexics are characterized by deviations in normal brain asymmetry:
Although there should be no doubt that the neUroimaging studies in general and the post m ortem /cytoarchitectonic studies sp ecifica lly suggest that sym m etry in the region of the planum tem porale and
p arieto-occip ital cortex is found with significantly greater
frequency than in th e normal population, there is neither su fficien t agreem ent among the neuroimaging studies nor independent
verification o f (the postm ortem /cytoarchiteetonic) findings to be more affirm ative with regard to this question, (p. >176-477) _ Hynd and Semrud-Clikeman concluded that m ethodological d eficien cies ch a racterize the literature relating dyslexia to brain morphology, particularly regarding th e diagnosis of dyslexia* appraisal of handedness and neurolinguistic d e fic its, and the lack of evidence that unusual brain asym m etries are unique to th e d yslexic syndrome.
Satz (1990) provided an important critique of the "imbalance" hypothesis, taking a d ifferent p ersp ective. From a review o f studies which have assessed sp eech and language recovery a fter early childhood brain lesions, ha concluded that pre- or postnatal brain damage would have to Occur bilaterally, in order to account for dyslexia. This is because le ft hemisphere lesions which occur before th e first birthday are likely to cause a sh ift of both speech and handedness to the right hemisphere without sign ificantly a ffectin g the subsequent developm ent of reading and writing skills. The im balance hypothesis, according to Satz, cannot explain why this d oes not occur in the ca se of dyslexia.
A d ifferent approach to Understanding dyslexia, and one that is closer to the position taken in th is study, has been advanced by Frith (1985) who em phasized the need for understanding the interaction o f constitutional and environm ental
factors:
If there were an innate d eficien cy in the brain structure underlying a sp ecific skill, one would then ex p ect a particular type o f reading fa ilu re.... A t the price o f fa ith in predestination this theory essen tially dispenses with developm ental factors. Thus, it ignores not only m aturational p rocesses and their com plex interaction w ith the environm ent, but also Social, cultural, and educational fa cto rs that influence the acquisition o f literacy, (p. 302)
Frith did notj however, abandon the notion that a brain d eficit is primary in th e etio lo g y o f dyslexia. What she argued against is a direct comparison b etw een reading d isab ilities resulting from brain injury (acquired dyslexias) and those which are developm ental in nature. For the la tter, she stressed the need to study the interaction o f developm ental factors w ith impairments in basic cognitive processes.
Frith's adoption o f this th eoretical position was influenced by her studies of children who w ere ex cellen t readers but poor spellers (Frith, 1980). She
hypothesized that the spelling disability might be the consequence o f prem aturely learned stra teg ie s which had failed to m erge with others learned in later phases of developm ent (Frith, 1985). Kirk (1988) was o f the sam e mind, arguing that
spelling disability may represent "too much, too soon" (p. 16), pointing to research by Bradley (1981) which indicated that reading and spelling may represent, very d ifferen t skills. While Frith did not sp ecifica lly suggest that d yslexia might be the result o f premature learning, she did form ulate a plausible theory (Frith, 1985) that it may represent a developm ental arrest in the learning of alphabetic sk ills, possibly due to a disorder of phonological awareness.
N either Frith (1985) nor Kirk (1983) speculated on Why som e children m ay be at risk for reading or spelling disorders, or how phonological dysfunction may prevent normal reading progress. In contrast Mesker (Mesker, 1969; van Eyck,
1980), whose ideas have not received a tten tion by researchers into dyslexia, had som e interesting thoughts about the interaction betw een m aturational processes and schooling. Mesker observed that reading problems were associated with poorly established laterality as evidenced by a lack of clear-cu t dom inance in
handedness and eyedness (see Orton, 1937, for a similar viewpoint). From a neurological viewpoint he suggested that such children could be prematurely stim ulated to lateralization (e.g., with writing) and that this could lead to learning d iffic u ltie s la ter on in a child's developm ent. He therefore stressed the
im portance o f bimanual a ctiv ities (e.g. clay modelling) to fully develop sym m etric m otor skills prior to lateralization . Lamme (1982), a professor involved in early childhood curriculum design, was of a sim ilar opinion believing that bimanual coordination should be prerequisite to learning to w rite.
S atz and Sparrow (1970) also conceptualized the etiology of dyslexia within the co n tex t o f a developm ental rather than a disease model. They postulated that the d y slex ic child is delayed on a number of developm ental skills which are not obviously related to th e reading process. This position was first broadly suggested by Money (1966) who stated that:
...th e m ajority of reading disability cases w ill be classifiable not on the basis of brain pathology, but sim ply as representative of a lag in the functional developm ent of the brain and nervous system that subserves the learning of reading, (p. 35)
Satz and Sparrow (1970) extended this view by conceptualizing the behavioral signs within th e framework of le ft hem isphere integration. Satz and van Nostrand (1973) added further evidence in support of a possible lag in the maturation o f the le ft hem isphere while observing that the nature of the underlying mechanism s till rem ained obscure. In general they found support for the hypothesis that many o f the high risk children who enter pre-school, particularly boys, may no/e be
m aturationally or developm entally ready to cope with the early form al demands ot reading.
Others who have also favoured a maturations! lag explanation of dyslexia include Bender (1963), Cohn (1961), de Hirsch, Jansky and Langford (1966).
C rltch ley (1964) observed that: _
Most neurologists... would be reluctant to visu alize in developm ental dyslexia any focal brain lesion, dysplastic, traum atic, or
oth erw ise.... To do so would be to ignore the important factors o f im m aturity as applied to chronological a ge, cortical developm ent and process of learning, (p. 21)
The foregoing review has suggested that som e reading and spelling disorders may represent a developm ental arrest in the early sta g es of alphabet learning th a t cannot, with any certainty, be attributed to brain dam age. In contrast, th ere is som e convergence of opinion (Frith, 1985; Kirk, 1983; van Eyck, 1980) that certain children may not be biologically ready to learn to read a t the normal chronological age and may consequently adopt dysfunctional learning stra teg ies.
2.2
SSD and SRD Similarities and Differences
Much less is known about the neuropsychological ch aracteristics of SSD as compared to SRD. While there is som e evidence that SSD may be associated w ith poor handwriting skill (Share e t a l., 1987) and d iffic u lties in remembering physical d eta il of the printed page (Ormrod, 1986), there has been little effo r t spent in researching the visual or motor characteristics o f this disorder. Instead th e focus has been on psycholinguistic spelling comparisons b etw een SSD, SRD and GRS Children a t d ifferent age lev els (e.g., Bruck & Waters, 1988; Schlapp & Underwood, 1988; Schwartz, 1983; Smith, 1992; Waters, Bruck & Seidenborg, 1985). In general th ese studies have found that spelling d ifferen ces betw een these three groups em erge by grade 6 (early adolescence). A t this tim e poor spellers, regardless of
reading ability, appear to have problems converting phonemes into graphemes that are positionally (orthographicaliy) accurate. Two studies have found d ifferences b etw een SSD and SRD in phonetic spelling accuracy in this age range (Frith, ,1,980; Jorm, 1981), while another did not (Bruck & Waters, 1988). However comparing and interpreting these studies is problem atic since the relative emphasis on phonic s vs whole word instructional approaches was not reported.
Stronger evidence that th ere are variations in phonetic spelling accuracy, at lea st within SRD, is provided by an earlier neuropsychological investigation
(Sw eeney & Rourke, 1978). While this study also found that differences in spelling ab ility becom e more evident during early adolescence, it further observed th e em ergen ce o f d ifferen ces in sound-spelling accuracy which discrim inated betw een tw o major SRD subtypes that have been identified by d ifferent investigators in the field . In general th ese tw o subtypes are distinguished by a preference for auditory left-h em isp h ere stra teg ies or visual right-hemisphere strategies in reading tasks. They include th e audiles and v isiles of Money (1962) and Wepman (1968); the auditory and visual d yslexics o f Johnson and Myklebust (1967); the
dysphonetic-dyseidetic d yslexics o f Boder (1971, 1973); the L- c.nd P - types o f Bakker (1984); the dysphonemic sequencing and audiophonetic disorders o f Denckla (1979); and th e language and visuo-spatial disorder subtypes which have been id en tified by several investigators (e.g ., Lyon, 1983; Rourke, 1985; Satx 8t Morris, 1981). Mixed subtypes have also been identified that dem onstrated combinations o f both visual and auditory disorders (Denckla, 19?9; Lyon, 1983} M attis, French, 8c Rapin, 1975). It is possible that th ese two major subtypes of reading disability rep esen t d ifferen ces in learning sty le that are not sp ecific to dyslexia, as a
sim ilar contrast has been found with SSD (Bruck 8c Waters, 1988) and GRS (Baron, Treim en, W ill, 8c Kellman, 1980) subjects. Thus the influence o f this variable on th e tasks used in the present research may be consistent across the th ree groups to be evaluated.
In summary, studies which have examined the d ifference in spelling perform ance betw een early adolescents who d iffer in reading and/or spelling ab ility agree that poor spellers share a common d ifficu lty in producing
orthographically correct spellings. There is less agreem ent cn whether th ey d iffer w ith respect to the phonological accuracy of m isspellings. U ntil instructional variables are controlled in studies of this nature, it may be unwise to sp ecu la te on constitutional d ifferences in the phonological processing of print.
2.3 Reading-Spelling D issociation
One reason for the sparsity of neuropsychological research into SSD is the fa c t that it is not an academ ically threatening disorder. Two to fifte e n percent of the population of normal language users do not becom e proficient spellers but n everth eless exhibit average or above-average ab ilities in other areas o f language function (Frith, 1980; Nolen, 1980). Frith, however, has focused a tten tion on spelling disability by describing a group o f tw elve-year-old good readers who possessed unusually poor spelling skill. Her research suggested that "reading by e y e and spelling by ear" (p. 512) characterized th e print-processing behavior of th ese early adolescents. While able to distinguish betw een homophones (e.g. th eir/th ere) in reading, these were frequently m isspelled in w ritten work. Other research suggested that SSD adolescents may achieve word recognition in reading
by p„. rtial cu es (Frith, 1978); Ormrod, 1985) and may d iffer from GRS in having short-term memory problems with printed m aterial (Ormrod, 1986).
There is general agreem ent that word-recognition can be accom plished either through d irect phonological encoding of whole words or by indirect
phoneme-grapheme conversion of le tte r sequences (see Barron, 1986, for a review ). The indirect route uncovers word meaning by segm enting words into graphemes (lette r s or le tte r clusters such as oa, th, gh) which correspond to phonemes. T hese are then assem bled into strings which approximate spoken words. In contrast th e direct route uncovers meaning phcnologically, without first having to id en tify the individual parts o f words. It seem s a reasonable conjecture that SSD may represent a developm ental dissociation betw een th ese tw o routes for accessin g word meaning. D irect phonological a cce ss, without the ability to decode print graphem ically, could explain a partial-cue style o f reading as w ell as memory d e fic its for print d etail. The discrim ination o f homophones in reading but not spelling m ight be aided by contextual rather than spelling Cues. This
hypothesis m ay have relevan ce to the following studies which w ere designed to exam ine the cortica l localization o f functions involved in the sequential
processing o f language.
Kimura (1982), from an analysis o f neurological injuries restricted to either the le ft or right hem isphere, found both oral and manual sequencing control to be under le ft hem isphere control. Her research indicated that th ese two functions were more separable in the frontal than in the parietal region, suggesting that the la tter may play a general programming role that is then enacted through the frontal region. Thus injury to the parietal area a ffe c te d the ability to program
n on -rep etitive oral or manual m ovem ents (e.g., syllables or pantomimes).
Importantly, when the temporal system was in tact it was able to bypass d e fe c tiv e fron tal or parietal oral motor system s. That is, the left, temporal region-appeared to be not only necessary, but also su fficien t, for the reproduction of auditorily presented fam iliar speech m aterial. Other research, using elec tric a l stim ulation mapping techniques (Ojemann, 1981), has dem onstrated that oral m ovem ents are co rtica lly linked with the ability to discrim inate the parts (phonemes) of words, a com m onality that was found only in the le ft hemisphere motor (pre frontal) and sensory (parietal and temporal) lobes.
Kimura's (1982) and Ojemann's (1981) findings, taken together, indicate th at the le ft hemisphere parietal and tem poral system s can independently function in tasks which require awareness o f the phonological constituents o f words. This research may provide som e insight into the etiology o f both SSD and SRD. D yslexia is associated with a d elay in th e developm ent of phonological skills (B eech & Harding, 1984; Olson, Davidson, K liegl 8c D avies, 1984; Siegal 8t Linder,
1984) whereas such has not been documented in the case of SSD. Other research has shown th at phonological awareness is a strong predictor o f reading su ccess (see Wagner 8c Torgesen, 1987, for a review ). It is tem pting to suggest that th e d ifferen ce betw een SSD and SRD is that the form er is able to bypass dysfunctional p arietal lobe phoneme processing by utilizing a tem poral lobe facu lty for receiving word meaning by direct phonological conversion.
The possibility that the tem poral and parietal lobes may play distinct roles in print encoding receiv es support from Luria's (1970) observations o f the
hem isphere local brain wounds or tumors. Patients with le ft temporal lobe dam age had d ifficu lty finding the correct letter for the sound they wanted to w rite (e.g ., writing "pull" instead o f "bull" or "tome" instead of "dome"). J n contrast, patien ts with lesions to th e visual and spatial zones (occipital and parietal lobes), w hile p erfectly able to analyze speech sounds, showed marked d ifficu lty in recognizing and forming written letters. These subjects found it d ifficu lt "to visu alize the required structure among the parts of the le tte r and to put th e parts togeth er to form the whole" (p. 71). In this light it is suggestive that both SSD and SRD have b een found to be associated with handwriting d e fic its in a large sca le longitudi/ ial study (Share, Silva & Adler, 1987).
More recen t research, review ed by Farah, Hammond, Levine and Calvanio (1988), has supported a distinction betw een visual representation of appearance (forv^s, patterns, objects) in the temporal lobe and visual representation o f spatial location in th e parietal lob e. These two visual system s have been respectively term ed th e "what" system and the "where" system (Ungerleider & Mishkin, 1982). C onvergent evid en ce from several disciplines indicates that the latter (parietal) sy stem m ay be dysfunctional in d yslexia (Conners, 1990; Livingstone, Rosen, D rislane, & Galaburda, 1991; Mason, 1980).
Mason (1980) com pared skilled and less skilled high school reader's ab ility to id en tify a b riefly exposed le tte r embedded in a row o f dollar signs. Both groups perform ed id en tically when cued in advance as to the location o f the le tte r , how ever, only th e skilled reader's perform ance suffered when advance location inform ation was withheld. Mason concluded that "...the role o f perception in reading has been underestim ated because emphasis has been on item perception,
and th e perception of spatial location has been largely ignored" (p. 89). Sim ilarly, Conners (1990) from a comprehensive review of neurophysiological research on dyslexia, concluded that "abnormalities in Central parietal m echanism s for visual se le c tiv e attention cause both th e symptoms of dyslexia as w ell as the
accom panying abnormal eye movements" (p. 181). Finally Livingstone e t al. (1991) presented physiological and anatom ical evidence from studies of d yslexics
consistent w ith the hypothesis that the disorder is associated w ith abnorm alities in the visual system responsible for spatial lo ca lisa tio n . Without sim ilar
investigations into the nature o f SSD, one can only sp ecu late as to whether comparable "where" d e fic its would be Uncovered.
It is also impossible to ascertain from this evidence whether th e posited sp atial localization d e fic it in dyslexia is lim ited to the processing o f alphanumeric stim uli. Le Doux et a l. (1977) hypothesized that the manipulospatial superiority of the right hemisphere may occur by d efau lt, in consequence of language
developm ent in left hemisphere space. Further, Ehri (1985) has presented evidence which suggested that th e learning of spelling ex er ts a considerable influence on children's speech. She observed that it is much easier to d etach language from its com m unicative function, treat it as an object, and study if form when one has fixed visual-spatial sym bols to see and manipulate. If SS; md SRD involve an arrest in the developm ent of alphabetic spelling sk ills, it is
possible that both disorders may be associated with anomalous le ft hem isphere la teralization for nonverbal visual-spatial functions. In other words if, for
w hatever reason, the facu lty for visual-orthographic spelling has failed to develop in th e le ft hemisphere, the a ffe c te d space may be functional for other nonverbal tasks.
In summary, a case can be made for the functional independence of the tem poral and parietal lobes in processing d ifferent asp ects of phonological and visual representation. While there is mounting evidence of le ft hemisphere
parietal lobe anom alies in SRD, research into SSD is too lim ited to do anything but sp ecu late on whether th e tw o disorders are similar in this respect. However, studies which have in vestigated the e ffe c ts o f localized brain injuries on language perform ance provide som e rationale for hypothesizing that both SSD and SRD may have a dysfunctional parietal "where” system for spelling and that d ifferen ces in the tem poral "what" system function may be relevant to the contrast in reading ab ility. That is, SRD and SSD may be similar with respect to parietal system d e fic its for sp atially locatin g the le tte r components of words but d ifferent in that reading disability may additionally involve temporal lobe problems with the phonological id en tifica tio n of print meaning. As the line orientation judgem ent task to be Used in the present study can be regarded as a "where" rather than a "what" task (M cGee, 1979), it may provide a means for inferring whether the p u tative parietal lobe d e fic it in SSD and SRD is restricted to verbal stim uli.
2.4
The Dual Task Paradigm
The dual task paradigm has becom e a frequently used behavioral method to infer hem ispheric language lateralization. Its basic assumption is that an
asym m etrical decrem ent in tapping perform ance of the hands produced by various concurrent tasks versus tapping alone indicates that the manual task and the co gn itive task are com peting for resources from the sam e hemisphere (Kinsbourne & H iscock, 1983). B ecause the finger tapping movement of the hand is primarily
controlled by the cerebral hemisphere on the opposite side o f the body (Brinkman 8c Kuypers, 1972), greater right hand in terference produced by the concurrent task purportedly indicates more le ft hemisphere involvem ent for the cogn itive a ctiv ity (H ellige 8t K ee, 1990; Kinsbourne 8c H iscock, 1983).
This paradigm was first introduced by Kinsbourne and Cook (1971) who reported that in right handers, vocalizing disrupted the concurrent balancing of a dow el rod with the right hand but not the le ft. This asym m etry of in terferen ce was attributed to th e lateralization o f speech to the le ft hem isphere. Numerous subsequent studies have reported similar asym m etrical e f f e c t s for normal
right-handed monolingual speakers (see Kinsbourne 8c Hiscock, 1983, for a com prehensive review ). More recent language research w ith this procedure has exam ined the influence of handedness (Orsini, Satz, Soper 8c Light, 1985; Simon 8c Susstnan, 1987), stuttering (Webster, 1989), bilingualism (Furtado 8c W ebster, 1991), preprogramming of tapping responses (Inhoff, 1990), rate and variability of tapping (HiscoCk, Cheesm an, Inch, Chipuer, 8c G raff, 1989) and controlled vs autom atic verbal production (Wiegersma 8c Wijnmaalen, 1991) oil th e concurrent task. Of relevance to the present investigation, Hiscock e t al. found tapping rate, rather than variability, to be the more sen sitive measure o f la tera lity . Further, depending on whether the tapping task was performed under speed or con sisten cy instructions, the rate of tapping either decreased or increased as reading was conjoined w ith tapping. The right hand rate was observed to change more than the le ft hand rate irrespective o f the direction o f change. This la tter finding led H iscock et a l. to sugp 'st that these particular tasks, when conjoined, a ffe c t each other via cross-talk rather than as a result o f com petition for som e general resource such as attention.
Various models have been put forth to account for concurrent-task
perform ance (Green & Vaid, 1986; Heuer & Wing, 1984; Hiscock, 1986). In general th e th eoretical assumptions underlying this procedure are closely tied tCu
m ethodological issues (Hisock, 1986). These concerns are examined in d eta il by Green and Vaid (1986) and include variables extrinsic to the paradigm (i.e., individual d ifferen ce variables) and those intrinsic to it (i.e. methods for d ifferen tia l allocation o f attention, counterbalancing procedures, and m easurem ent o f concurrent performance).
A ttem pts to dem onstrate greater le ft - than right-hand interference using nonverbal tasks have produced equivocal findings. One possible reason is that it is d iffic u lt to find in terferen ce tasks that are su fficien tly lateralized to the right hem isphere. For exam ple, som e studies have used concurrent musical tasks o f various typ es and have reported results ranging from greater right- then left-hand disruption to equal disruption o f both hands (Johnson & Kozma, 1977; Hicks, 1975; Lomas & Kimura, 1976). Unfortunately, the use of musical tasks is problem atic because la tera l sp ecialization for various asp ects of music processing is still poorly understood (e .g ., G ates & Bradshaw, 1977).
R esults have been som ewhat more promising for concurrent tasks that require visuospatial processing. It has been found that such a ctiv ities either disrupt concurrent le f t - and right-hand performance equally (Bowers, Heilman, Satz, & Altm an, 1978; McFarland & Ashton, 1978a) or disrupt the le ft hand more than the right (H ellege & Longstreth, 1981; McFarland & Ashton, 1978b, 1978c). Those studies which dem onstrated sym m etrical interference involved resp ectively fa cia l recognition and concealed figure d etection. The ones which found greater
left-h an d interference may have tapped additional cognitive resources as
resp ectiv ely th ey required solving a block design problem, and performing running m em ory identification o f shapes and fa ces, Further, in contrast to Bowers e t a l., the more d ifficu lt McFarland and Ashton fa cia l recognition design required recognition and tapping to be performed simultaneously whereas Bowers et al. discontinued tapping during the facial identification phase o f the experim ent.
Studies w ith children have follow ed two paths. Much o f the research has addressed the question of developm ental change in cerebral dom inance for sp eech . This line of investigation has yielded quite consistent results. Verbal in terferen ce asym m etries resembling those occurring in adults are found in young children (Kinsbourne & McMurray, 1975) and remain invariant in magnitude across the age range o f 3 to 12 years (Hiscock & Kinsbourne, 1980; White 8c Kinsbourne, 1980). Visuospatial lateralization of line orientation was exam ined in one study (H iscock, e t a l., 1985) and generated bilaterally equal in terference w ith children in grades 2 to 5. The second line o f research has been concerned with learning disabled subjects and the findings are more diverse. For exam ple, Dalby and Gibson (1981) found that recitation produced bilaterally equivalent in terferen ce in "dysphonetic" and "dyseidetic" (subtypes o f reading disability) boys while producing th e ex p ected right hand asym m etric in terference in average readers. A non-verbal task
(Ravens Progressive M atrices) yielded left-greater-th an -righ t in terferen ce for average readers and d yseid etic boys, but not for the dysphonetic group. However other research (Sperry, 1974) has dem onstrated that this task does not reliably distinguish betw een le ft and right hemisphere processes.
Stellern e t al. (1986) appears to be the only study which has exam ined the role of tapping in terferen ce in older children using a test (line orientation judgement) which has b een found, at le a st with a iUlts, to be a robust indicator of right hem isphere processing (Benton e t al., 1978; Poizner, et al., 1987; Warrington & Rabin, 1970) The te s t has also been found to be a correlate of right hemisphere cogn itive s ty le in children aged 7 to 12 years (Stellarn, Marlow, 8c Jacobs, 1983).
Importantly, a sim ilar relation betw een cognitive sty le and spatial orientation judgem ent was found with undergraduate college students performing the dual
task paradigm (Bowers & LaBarba, 1988). This study used a spatial orientation (SO) subtest from th e Guildford-Zimmerman Aptitude Survey (Guilford 8c
Zimmerman, 1953) to se lec t right-handed subjects who scored in the upper (high SO) and low er (low SO) quartiles. Both groups were then required to concurrently tap w ith eith er their le ft or right hands while solving this su btest. Successful perform ance required maintaining a boat on course by com pensating its angle rela tiv e to horizon changes presented in a sequence of pictures. The construct valid ity o f th is su btest was dem onstrated in a factor analysis reported by McGee (1979). Bowers and LaBarba observed significantly greater right hand disruption w ith high SO subjects and sign ificantly greater le ft hand disruption with low SO subjects. T hese results suggest le ft hemisphere involvement in high SO subjects and right hem isphere involvem ent in low SO subjects in processing the
boat-horizon task. U nfortunately reading and spelling lev els w ere not measured. This inform ation would have been o f particular relevance to the present study in view o f reports that dyslexic performance on a spatial orientation map-walking task changes abruptly after age 10 from poor to superior perform ance (Denckla,
1985) and that d yslexic fam ily members frequently dem onstrate superior
visuospatial a b ilities without concom itant reading d eficits (Denckla, 1979} Gordon, 1980, 1989; Symmes & Rapoport, 1972). Conceivably, both SSD and SRD may be associated w ith high le f t hemisphere spatial ability as a "by-product" o f their failure to develop visual word pattern (orthographic) knowledge.
The dual task stud ies which have been review ed thus far have all been concerned w ith obtaining inferential evidence for hemispheric lateralization o f cogn itive ab ilities. For about the same number of years this paradigm has also been used to explore th e characteristics o f verbal and visual alphanumeric short term memory a b ilities (see Baddeley, 1992, for a recent review ). This research provided evidence for a m ulti-com ponent short term memory model controlled by a lim ited cap acity atten tion al system term ed the "central executive" and
supported by a t least tw o a ctiv e "slave" system s: an articulatory or phonological "loop" and a visuospatial temporary store or "sketchpad." The articulatory loop was assumed to be responsible for maintaining and manipulating speech-based inform ation whereas th e sketchpad was assumed to perform similar functions for visuospatial inform ation involving letters or numbers, For exam ple, visuospatial tasks such as tracking a spot o f light disrupted perform ance in a verbal m em ory task requiring spatial im agery whereas there was no e f f e c t in an analagous non-spatial ro te verbal memory condition (Baddeley, Grant, Wight, & Thomson,
1975).
The research of Baddeley and colleagues (Baddeley, 1992) is particularly relevant to th e present Study because of a recent rapprochement betw een
processing. L ogie (1986), on the basis of spatial im agery interference evidences, and Farah et a l. (1988), from neuropsychological and psychophysiological data, independently concluded that this memory system may have separate visual and spatial com ponents. As noted earlier, Farah et ah found evidence in support o f the possibility that th ese components may resp ectively involve the le ft
hemisphere tem poral and parietal lobes. While Baddeley (1992) stressed that the two com ponents are d ifficu lt to tea se apart in normal subjects, the evidence which has been presented in this review suggests that this may not be th e ca se with SSD. That is, th e SSD short term memory d eficit for printed m aterial (Ormrod, 1986) may represent an impairment of the spatial (parietal) com ponent of visuospatial memory, a ffectin g "where" rather than "what" alphanumeric decoding. Thus, while a ccess to to meaning o f print may be unaffected, there is d ifficu lty in converting phonemes into graphemes that are positionally (spatially) acceptab le (Bruck & W aters, 1988).
A recent conceptual concern with, d ual-task/tim e sharing studies is that the asym m etric in terferen ces observed may be a sta tistica l a rtifa ct of baseline d ifferen ces and not hem ispheric lateralization (Sussman, 1989; Willis & Goodwin,
1987). As expressed by Willis and Goodman:
A potential problem w ith analyzing raw scores generated through th ese paradigms is that they may be insensitive to lateralized in terferen ce e f f e c t s . This is because research participants are usually right handed and interhand comparisons of initial (i.e., baseline) tapping rates typ ically favor the right. Given this initial descrepancy b etw een th e hands, d ifferen tial in terferen ce e ffe c ts associated with the concurrent perform ance Of an unrelated task may be due tp in itial d ifferen ces in tapping speed rather than
lateralization e f f e c t s . In this respect, interference might be greater for right- than left-handed tapping because, due to th e higher range o f in itial values for the right hand, there is a aigher possible range for reduction, (p. 719)
H ow ever several studies have demonstrated that baseline asym m etries (i.e., manual dominance) are neither a necessary or su fficien t condition for producing asym m etric dual-task interference (Cherry & K ee, 1991; Hiscock e t al., L989; Kee & Cherry, 1990).
In summary, the dual task paradigm has engendered extensive research in the domains o f both neurospsychology and cognitive psychology. In both areas there has been much more emphasis on its use for investigating verbal rather than spatial inform ation processing. However, a recent convergence o f evidence betw een the two disciplines in the interpretation of visuospatial processing system s, prom ises to heighten research a ctiv ity using dual-task m ethodology. While its use in investigating visuospatial processing d ifferences in learning disorders has produced equivocal results, a post hoc analysis of a N ative Indian study (Stellern et a l., 1986) points to a need for further research w ith these populations.
2.5
Summary of the Literature
This review began with a look at current theories regarding th e etio lo g y o f sp e c ific spelling (SSD) and sp ec ific reading (SRD) disorders. From this
exam ination it becam e evident that a great deal o f uncertainty remains as to the rela tiv e im portance o f constitutional and environm ental factors in the
developm ent of th ese d isabilities. As Ehri observed: "We cannot le g itim a tely conclude that the source of dyslexia lies in individual learner d eficien cie s until We have p erfected the w ay we teach children to read and spell" (1989, p. 364).
