brain onset of speech after left hemispherectomy in a nine ... · when examined at age 11:5 by a...

braini0107

Brain (1997),120,159–182

Onset of speech after left hemispherectomy in anine-year-old boyFaraneh Vargha-Khadem,1 Lucinda J. Carr,2,3 Elizabeth Isaacs,1 Edward Brett,2 Christopher Adams4

and Mortimer Mishkin5

1Cognitive Neuroscience Unit,2Neurosciences Unit, Correspondence to: Dr Faraneh Vargha-Khadem, CognitiveInstitute of Child Health and Great Ormond Street Hospital Neuroscience Unit, Wolfson Centre, Mecklenburgh Square,for Children, 3Human Neurophysiology Unit, Department London WC1N 2AP, UKof Physiology, University College London,4Departmentof Neurosurgery, Radcliffe Infirmary, Oxford, UK, and5Laboratory of Neuropsychology, National Institute ofMental Health, Bethesda, Maryland, USA

SummaryCase Alex, with Sturge–Weber Syndrome affecting the left comparable IQ (range, 40–68) but with early development

of speech and language, suggests that, surprisingly, Alex hashemisphere, failed to develop speech throughout earlyboyhood, and his comprehension of single words and simple suffered no permanent disadvantage from his protracted

period of mutism and severely limited comprehension.commands remained stagnant at an age equivalent of 3–4years. But then, following left hemidecortication at age 8.5 Although the findings in Alex, as in other left-

hemispherectomized patients, indicate definite limits to theyears and withdrawal of anticonvulsants when he was morethan 9 years old, Alex suddenly began to acquire speech and cognitive and linguistic capacity of the isolated right

hemisphere, Alex’s achievements appear to challenge thelanguage. He also showed an unusual degree of residualmotor capacity on his right side. Alex’s remarkable progress widely held view that early childhood is a particularly

critical period for acquisition of speech and language orin learning speech and language, and the development of hisother cognitive abilities, were measured periodically from any of their selective aspects, including phonology,

grammar, prosody and semantics. It is concluded thatthe age of 9 to 15 years. His most recent scores on tests ofreceptive and expressive language place him at an age clearly articulated, well structured, and appropriate

language can be acquired for the first time as late as age 9equivalent of 8–10 years. Comparison with the level offunction attained in these domains by nine other left years with the right hemisphere alone.hemispherectomized patients with early onset of disease and

Keywords: congenital aphasia; hemispherectomy; late speech onset; right hemisphere language; motor reorganization

Abbreviations: BPVS 5 British Picture Vocabulary Scale; CELF-R5 Clinical Evaluation of Language Fundamentals—Revised; 1DI5 the first dorsal interosseus muscle of the hand; TROG5 Test for Reception of Grammar; WISC-R5Wechsler Intelligence Scale for Children—Revised; WISC-III5 the newly standardized WISC; WORD5 Wechsler ObjectiveReading Dimensions

IntroductionFailure by a child to acquire speech is extremely rare. When severe mental retardation, or, at the other, from an ischaemic

or traumatic injury limited to the perisylvian areas andthe failure appears in the absence of deafness or extremesensory deprivation, it is likely to be due to bilateral thereby arresting the development of speech selectively

(Landauet al., 1960; Vargha-Khademet al., 1985). However,neuropathology encroaching on one or more of thethalamocortical systems subserving speech and language. speech may fail to develop in the absence of all of the

foregoing factors. For example, children with the Landau–This bilateral encroachment may result, at one extreme, fromthe type of widespread cerebral abnormality that causes Kleffner syndrome are aphasic (Landau and Kleffner, 1957),

© Oxford University Press 1997

160 F. Vargha-Khademet al.

and it has been hypothesized that this is due to bilateral Shortly after birth, a port-wine stain was noted over theleft side of his face in the distribution of the first division ofinterference with the functioning of the cortical speech and

language areas by the centrencephalic seizures associated the trigeminal nerve. Located principally on his forehead,the stain extended back over his skull for some distance, andwith this disorder (Paetauet al., 1991). In the present

report, we document another case of arrested speech also forward to his left upper eyelid and down over the leftside of his nose.development that was presumably caused by bilateral

functional interference with the cortical speech and language At age 6 days, Alex had an episode of jerking of the rightlimbs lasting ~3 min. This was soon followed by a secondareas, but with a different aetiology. The case is of a boy,

Alex, now aged 15 years, with the Sturge–Weber Syndrome attack of 1–2-min duration and then, 2 days later, by athird attack. Treatment with phenobarbitone was effective,affecting the left hemisphere, who remained without speech

and arrested at a mental age of ~3–4 years throughout his however, and when he was later discharged home he wasbreast-feeding well and was clearly alert and vigorous.early childhood. More remarkably, at the age of nearly

9.5 years, 10 months after left hemidecortication and the Because he had shown two of the classical signs of Sturge–Weber Syndrome, i.e. port-wine naevus and contralateralsubsequent withdrawal of anticonvulsant medication, Alex

suddenly began to develop speech. The emergence of seizures (Aicardi, 1992), he was admitted at 4 weeks of ageto Great Ormond Street Hospital for Children, London, forexpressive language was soon followed by improvement in

his receptive language and other verbal intellectual skills to further investigation. Physical examination, including a fullneurological evaluation, was normal. A CT scan showed aa current level of overall language ability equivalent to that

of an 8–10-year-old. A brief report of this case has already normal ventricular system, but there was some widening ofthe left sylvian fissure and focal widening of the sulci in theappeared (Vargha-Khadem and Isaacs, 1992).

Early childhood is a particularly critical period for left posterior parietal region, indicative of left hemiatrophy.Following a change in medication from phenobarbitone todeveloping speech and verbal comprehension, and, by

implication, for functional development of the underlying sodium valproate at the age of 6 weeks, Alex had anotherattack, twitching on the right side and losing consciousness.neural systems, such that if language competence is not

acquired or reacquired during the first few years of life, He then suffered from 10 to 20 seizures daily for 10 daysdespite a return to the original anticonvulsant and the additionspeech and language are likely to be grossly deficient

thereafter. This conclusion is based on several sources of of others. The seizures consisted predominantly of weaknessof the right limbs with jerking of the right leg and right arm,evidence, including (i) the dramatic cases of individuals

deprived of language experience during early childhood (e.g. and twitching of the right side of the face. From the time ofthat episode until his surgery more than 8 years later, despiteCurtiss, 1977); (ii) congenitally deaf children exposed to sign

language only later in life (Newport, 1984; Johnson and the administration of high doses and various combinationsof anticonvulsants, he suffered from similar bouts of seizuresNewport, 1989); (iii) older children and adults learning a

second language (Oyama, 1976; 1978; Johnson and Newport, several times a year, each bout lasting for 2–3 days beforedissipating.1989) and (iv) cases with left hemisphere damage sustained

after early childhood (Rasmussen and Milner, 1977; Patterson By the age of 6 months Alex was moving his right limbsless than his left, and right-sided hyperreflexia was evident.et al., 1989). In all such instances, even though some language

ability is acquired or reacquired, the final product tends to By 7 months he was showing clear signs of right hemiparesisand right hemianopia. As a baby, Alex did not babble orfall far short of normal, and is usually characterized by

markedly defective phonology and grammar. However, the gurgle very much and was reported to be a silent infant. Atthe age of 21 months, he was evaluated by his paediatricianexact age that defines the end of the critical period for

learning a language, and using it effectively, is still unknown. who noted that he showed age appropriate social interactionand play despite the absence of vocalization or attempts atTo our knowledge, no previously reported child has acquired

a first spoken language that is clearly articulated, well single words. Alex began to walk only at 30 months, and hewas not toilet-trained until 36 months. When evaluated atstructured and appropriate after the age of ~6 years (case of

Isabelle; Skuse, 1988). The case reported here extends that the age of 33 months by an educational psychologist, hisintellectual development was noted to be globally delayed,upper limit substantially. Alex’s achievement suggests that,

surprisingly, the complex neural system underpinning speech and, in particular, despite normal hearing, he had developedno speech, his utterances at this time being limited toand language may remain viable despite disuse throughout

the first 9 years of life, with this linguistic system having babbling.When Alex was between the ages of ~4 and 8 years,been limited, moreover, to an isolated right hemisphere.

several attempts were made to obtain formal indices of hisintellectual ability and language comprehension, but, althoughhe was sociable and even affectionate with adults, the severityMedical history

Alex was born in November 1980 following a normal of his behaviour disorder precluded reliable measurement.Having remained essentially speechless (see below), hepregnancy. He was delivered by forceps due to delay in the

second stage of labour, but he required no resuscitation. tended to rely on pointing, gestures and Makaton symbols (a

Onset of speech after left hemispherectomy 161

Fig. 2 PET scans showing diffuse hypometabolism of Alex’s lefthemisphere (displayed on right of each scan) compared withFig. 1 Magnetic resonance image showing Alex’s atrophic andnormal metabolism of the right hemisphere.diffusely calcified left hemisphere (displayed on right of figure)

together with his enlarged right hemisphere.abnormal blood vessels (Figs. 3 and 4). The hemisphere itselfwas firmer than normal due to the diffuse calcification that

system of verbally labelled pictographs), although his use ofhad been detected on the CT scan.signing and his motivation to communicate were limited. The surgical removal was performed by a modifiedMoreover, his concentration was poor, and it was difficult totechnique (Adams, 1983) designed to prevent the delayedgain his attention for more than a few minutes at a time. Atcomplications that are otherwise often associated withthe age of 6 years, Alex was placed in a school for childrenhemispherectomy. Specifically, after the removal, which waswith severe learning difficulties, where he remained until thecarried out without any particular difficulties, the foramen ofage of 13 years, when he was transferred to a school forMonro was plugged with muscle to prevent blood within thechildren with moderate learning disabilities. subdural cavity from irrigating the ventricular system, and

In June 1988, when Alex was aged 7:7 (years:months),the dural flap was sutured to both the tentorium and the duraadditional brain imaging was carried out. MRI and CT scanslining the anterior and middle cranial fossae to minimize therevealed an atrophic and diffusely calcified left hemisphere,volume and the surface area of the subdural cavity.

Alex made a very satisfactory postoperative recovery andtogether with an enlarged right hemisphere (Fig. 1); and awas able to be discharged home 13 days after surgery. SincePET scan showed diffuse hypometabolism on the left sidethere was no recurrence of seizures, anticonvulsant treatmentbut seemingly normal metabolism on the right (Fig. 2).was gradually reduced and then finally discontinued 9 monthsElectroencephalography performed at this time disclosed nopostoperatively, when he was aged 9:3. Alex remains wellseizure activity but did reveal that the amplitude andand seizure free to the present time (aged 14:11).frequency of background activity over the left hemisphere

were abnormally low. Based on these findings, surgicalremoval of the left hemisphere for relief from both the Resultschronic seizures and the anticonvulsant medication wasInformed parental consent was given for Alex’s investigation,strongly recommended. which had the approval of the local ethics committee.

One year later, when Alex was aged 8:6, a lefthemispherectomy was carried out. At operation the bone was

Neuropsychological evaluationfound to be filled with large venous channels, and the duraLanguage and cognition before and 6 monthswas much thicker than normal due to infiltration by numerous

small arterioles. Upon opening the dura a grossly abnormalafter surgeryhemisphere was revealed, dusky blue in appearance, the piaThe first successful evaluation of Alex’s cognitive abilities

was carried out when he was aged 8:2, 4 months before hisand arachnoid having also been infiltrated throughout by


Fig. 3 Lateral view of Alex’s surgically removed, atrophic left hemisphere.

Fig. 4 Horizontal section through Alex’s atrophic left hemisphere.

hemispherectomy, and a second was performed when he was comprehension of single words and simple commandsexceeded his expressive ability. Indeed, assessment of his9:0, 6 months after surgery, but before complete withdrawal

of the anticonvulsants. On both occasions, he was given expressive language before surgery was precluded, since hewas unable to score at all on formal tests of this function.the Griffiths (1970) Mental Development Scales, which

measure overall development in six areas of function. As His expressive vocabulary consisted of one correctlyarticulated word (‘mumma’), several sounds which he usedindicated in Table 1, Alex’s scores in these areas seldom

rose much above a mental age of 3 years. His only substantial consistently to refer to objects (e.g. ‘oof’ for ‘shoe’), andword approximations (e.g. ‘oo’, ‘ee’, ‘or’, ‘i’, for the numbersgain during the 10 months that separated the two testing

sessions was in Practical Reasoning, on which he showed an two to five). At this stage, Alex was able to repeat one digit.Receptive vocabulary was measured with the British Pictureincrease in mental age of 9 months, consonant with the

increase in his chronological age (Table 1). At this age level, Vocabulary Scales (BPVS) (Dunnet al., 1982), on whichAlex performed significantly better than his nearly nonexistentthe items comprising the Practical Reasoning subtest involve

simple concepts such as choosing the larger of two objects expressive vocabulary would suggest, obtaining a scoreequivalent to that of a child aged 4:1.or knowing the meaning of ‘money’.

It was noted during the two evaluations that Alex’s Six months postoperatively his expressive language showed


Table 1 Mental age scores on the Griffiths Mental Development Scales before and after surgery

Alex’s chronological age (years:months)

8:2 9:0 12:9 14:2 14:11

Time pre/postoperatively (years:months) 20:4 10:6 14:3 15:8 16:6Griffiths mental-age scores

Locomotor 2:4 2:6 – – –Personal and social 3:1 3:2 5:0 – –Hearing and speech 1:9 2:0 6:10 7:5 7:8Eye/hand coordination 2:3 2:10 5:6 6:0 6:3Performance 3:11 3:6 4:2 5:5 5:3Practical reasoning 2:7 3:4 7:0 7:2 7:8

no gains, again precluding testing, while his receptive enunciated and intelligible sentences. He still had someproblems with certain sounds in words (saying, for example,vocabulary on the BPVS actually declined slightly to the‘crock’ for clock, ‘van’ for ran, and ‘vash’ for wash), resultinglevel of a child aged 3:9.in his speech resembling that of a ‘well spoken foreigner’.When examined at age 11:5 by a speech therapist, hisarticulation was again described as mostly intelligible butOnset of speech and subsequent languagestill containing some of the phonetic errors noted earlier.developmentAlso, his respiration was uneven, affecting his prosody andOne month after total withdrawal of anticonvulsantstress. Now, 3.5 years later, following intensive speechmedication, when Alex was aged 9:4, his parents reportedtherapy, his speech sounds are essentially normal andthat he suddenly started uttering syllables and single wordstherefore highly intelligible.(seeAppendix 1). Within several months of this dramatic

breakthrough, he had progressed to producing full sentences.Naming.Tested for the first time at age 10:7 on the WordWith the onset of speech, his behaviour became far more

Finding Vocabulary Scale (Renfrew, 1977), which measuresmanageable, and he began to take an active interest in

vocabulary knowledge and naming, Alex obtained an agethe world around him, asking questions and demanding

equivalent of 6:3, and then at age 11:8 he reached an ageexplanations. By the age of 10:0, he could converse withequivalent of 7:10 to 8:1. His lexical retrieval was alsocopious and appropriate speech, involving some fairly longevaluated at this time with the Oldfield and Wingfieldwords. At a neurological examination carried out at this time,(1965) Object Naming Test, which requires the rapid naminghis newly acquired speech allowed his personality to shineof 36 line drawings of objects. Alex obtained a score of 22through, and he was described as being pleasant and outgoing,out of 36, and his mean reaction time for accurate namingwith an independent attitude. was 2.9 s. His accuracy was comparable with that of children

Between the ages of 9:4 and 14:11, Alex’s linguistic andat the same verbal mental age as Alex at the time of testingintellectual abilities were examined in detail, allowing for (i.e. 6:2; see below). With the exception of one drawingquantitative assessment of the development of his emerging(anchor) that he failed to identify and two others (piano andspeech and language and other cognitive skills. Within eachstethoscope) that he misnamed (‘pinanyo’ and ‘se...scope’),domain reported below, those aspects in which Alex hasAlex’s errors were either semantic (e.g. he called a bagpipeshown the strongest gains are described first, followed by‘violin’, and a book ‘comic’) or perceptual (e.g. he named athose in which he continues to be deficient. horseshoe ‘toilet seat’, and a microscope ‘robot’). When

retested at the age of 14:2, he obtained a score of 26out of 36, and his mean reaction time had decreased to

Expressive ability 1.5 s. His accuracy was now comparable with that of normalchildren ranging in age from 8:0 to 8:11 years [mean score

Articulation.On the Edinburgh Articulation Test (Anthony 6 SD 5 26.3062.66 (F. Vargha-Khadem and N. Vargha-et al., 1971), a measure of speech sound development, Alex’sMajzub, unpublished)], and above his verbal mental age atspeech sounds at age 9:4 consisted mainly of consonant–this time (i.e. 7:5;seebelow). Error analysis indicated thatvowel pairs; and some of these were mispronounced, suchAlex failed to identify two drawings (metronome and anvil),as ‘da’ for ga, and ‘ra’ for la. They were thus too restrictedwith the remainder of his mistakes being once again eitherin range to yield a score, but he was estimated to have anperceptual ( e.g. ‘coffee maker’ for microscope, ‘telescope’age equivalent of 20 months. By age 10:7, however, Alexfor syringe), semantic (e.g. ‘nail’ for screw) or part nameshad reached ceiling on this test, attaining an age equivalent(e.g. ‘pipes’ for bagpipe, ‘horoscope’ for stethoscope).of .6:0. At about this time, Alex was seen by his educationalpsychologist, who noted that he was speaking in wellInformation and grammar.The development of these


Morphology.Not all of Alex’s syntactical abilities showedthe same degree of improvement. For example, at age11:5 he was given a test to assess his ability to producemorphological markers (Vargha-Khademet al., 1991). In thistest, adapted from Berko (1958), an incomplete sentence isread to the child, accompanied by a drawing illustrating thesubject of the sentence. The child’s task is to producethe appropriate morphological marker that completes thesentence. The test is designed to tap knowledge of Englishmorphology through the use of nonsense drawings andnonsense words (e.g. chot, chotter; ponner, ponnest; etc.),but it also includes a control condition with real words(e.g. jog, jogger; bigger, biggest; etc.) to assess productioncapability. The norms for children of the same mental ageas Alex (i.e. 6:2) at the time he was tested are 16.261.9 outof 20 for words and 11.763.9 out of 20 for nonwords(E. Isaacs, S. Taylor and F. Vargha-Khadem, unpublished).In contrast, Alex scored only 11 out of 20 on words (butseebelow, section on Overall expressive language ability) anddropped even further to two out of 20 on nonwords.

Phonological imitation.Alex was also evaluated on aFig. 5 The development of Alex’s expressive language ability asmeasured by the Renfrew Action Pictures Test. test of the ability to repeat nonwords immediately after their

presentation (Gathercole and Baddeley, 1989), providing ameasure of phonological imitation based on working memory,aspects of Alex’s expressive ability was evaluated with the

Renfrew (1971) Action Pictures Test, which assesses both which is commonly impaired in children with disorderedlanguage development (Tayloret al., 1989; Gathercole andGrammar, reflecting the complexity of syntax, and

Information, reflecting the richness of content and vocabulary. Baddeley, 1990). The test consists of 40 nonwords, 10 eachof one to four syllables in length (e.g. ‘prindle’, ‘trumpetine’).As already indicated, Alex could not be examined with this

test in the period before, or shortly after, surgery. Even when When tested at age 11:0, Alex made 16 errors, 15 of whichwere on the items of more than one syllable. This score ismedication was first discontinued, he was below the 2-year-

old level on both measures. Between the ages of 9:4 and close to the mean performance of 5-year-olds (28.764.8;Gathercole and Baddeley, 1989). The test was readministered10:0, however, he showed a dramatic increase on both, and

he then continued to improve until the age of 14:2 when he twice with the one-syllable items eliminated when Alex was12:7 and 14:2, but his performance was virtually unchangedreached ceiling performance, i.e. an age equivalent of 8:5

(Fig. 5). The level he has attained is exemplified by his (15 and 14 errors, respectively).An analysis of ‘sound clusters’ carried out on Alex’s tapedresponse at age 14:2 to Renfrew Card Number 6 (which

shows a girl lying beside broken eyeglasses at the bottom of repetition of nonwords at age 12:7 revealed nine out of 28cluster errors, six of which were on the phonemes b, p, anda set of stairs): ‘she’s fallen down the stairs and broke

her glasses’. t when each was followed by a liquid (e.g. bl, pl, br, pr). Ananalysis carried out when he was 14:2 yielded essentially thesame results. A novel pattern observed on this latest occasionLength of utterance.Another aspect of Alex’s dramatic

speech development is reflected in the length of his sentences. was regularization of nonwords into real words (e.g. ‘presto’in lieu of ‘blistow’; ‘plasterer’ in lieu of ‘brasterer’).On the Bus Story Test (Renfrew, 1969), where a short story

accompanied by appropriate pictures is read to the child for For comparison, we also administered a real-wordrepetition test (Gathercoleet al., 1991), consisting of 40immediate recall, Alex, then aged 10:11, produced a mean

sentence length (11.6 words) matching that of a 6- to 7-year- words, 10 each of 2–4 syllables in length (e.g. ‘thimble’,‘problematic’). At age 12:7, Alex made only five errors, allold (seeAppendix 2A). Alex’s improvement in producing

complex sentences was documented in situations other than on three- and four-syllable words, and then, at age 14:2, onlyone error, on a four-syllable item (‘computated’ in lieu offormal testing. Spontaneous utterances, such as ‘I wash my

face in the sink’, ‘I got one ‘crock’ (clock) upstairs in my ‘complicated’). Moreover, he rarely hesitated and made fewarticulation errors. It is clear that lexical knowledge facilitatesroom’, ‘I did saw one fish in Headon Park—very big one

indeed—this size it was’ (gesturing with hands apart), ‘that his repetition performance substantially.wind-up toy will slide because it is on a slippery surface’,demonstrate the complexity which his speech had attainedPhonological awareness.To explore further his phono-

logical awareness, which is a strong predictor of literacybetween the ages of 10:5 and 10:11.


Table 2 Clinical Evaluation of Language Fundamentals—Revised

Language evaluation item Alex’s CELF-R subtestscores at chronological age(years:months)

14:1114:2

Expressive languageWord structure 14 12Formulated sentences 3 3Recalling sentences 3 4Sum of these three subtest scores 20 19

Expressive score (standard score) 78 76

Receptive languageLinguistic concepts 3 3Sentence structure 6 9Oral directions 3 3Sum of these three subtest scores 12 15

Receptive score (standard score) 59 67

Total language score (standard score) 67 70

Fig. 6 The development of Alex’s receptive vocabulary asmeasured by the British Picture Vocabulary Scale (BPVS) and theacquisition in both normal (Bryant and Goswami, 1987) andTest for the Reception of Grammar (TROG). The TROG score athemiplegic children (Muter, 1994), we administered theage 9 years is extrapolated from the age equivalents given by

Phonological Awareness Protocol devised by Muter (1994).Bishop (1982).This protocol comprises tests of rhyme detection, rhymeproduction, syllable and phoneme identification and seg-mentation, phoneme deletion, sound blending and speech on Morphology). His other subtest scores, reflecting the

ability to formulate and reproduce complex sentencs, fellrate. Results indicated that Alex’s phonological awarenessskills at age 14:11 were at the level of a 5- to 6-year-old ~2 SDs below the mean.(seeAcademic attainments), his deficits being much morepronounced on the phoneme than on the syllable subtests.

Receptive abilityOverall expressive language ability.Finally, to obtain astandardized measure of overall expressive language ability,Vocabulary. As indicated earlier, Alex’s receptive

vocabulary on the BPVS had reached only the 4-year-oldwe administered the test of Clinical Evaluation of LanguageFundamentals-Revised (CELF-R; Semelet al., 1987) first level in the period before, and shortly after, surgery. After

speech onset, his verbal comprehension, like his expressivewhen Alex was 14:2 years and then 9 months later (14:11years). The CELF-R, designed to evaluate syntax, semantics ability, also began to show gains, but these gains were

more gradual relative to those in speech output (cf. Figsand memory, provides expressive, receptive, and total lan-guage scores (with a normal mean6 SD of 100615) and 5 and 6). The level he has now achieved, an age equivalent

of 8:1 at the chronological age of 14:11, requiresan age equivalent (range 5–15 years). The expressive languagescore is derived from the subtests of Word Structure (assessing comprehension of difficult items spanning a variety of

lexical types, including concrete nouns (e.g. ‘fern’ andknowledge of word structure rules in sentences), FormulatedSentences (evaluating formulation of simple, compound and ‘tusk’), abstract nouns (e.g. ‘isolation’), actions (e.g.

‘plastering’) and adjectives (e.g. ‘tubular’).complex sentences) and Recalling Sentences (measuringrecall and reproduction of surface structure in sentences asa function of syntactic complexity). The subtests designedSemantic/pragmatic comprehension.To assess Alex’s

sensitivity to, and understanding of, the semantic/pragmaticfor children aged 5–7 years were administered to Alex, assome of the subtests for older children either involve reading aspects of language, we adapted a test of reference pronouns

designed by F. Happe (personal communication). The testor place too great a demand on short-term memory. Asindicated in Table 2, Alex’s overall Expressive Language consists of 24 items, each containing two short sentences.

The first sentence of each pair describes a situation using thescore was in the low normal range, but on one subtest (WordStructure) he scored in the high average range, indicating a names or titles of two persons, accompanied by a picture

illustrating the situation; the second sentence provides furthersubstantial improvement in performance on English morpho-logy over that recorded when he was 11:5 (seeabove, section information about the situation but uses pronouns instead of


names or titles (e.g. ‘The doctor cured the patient. He was the test’s ceiling, an age equivalent of.7:0, when he was10:7. Alex’s receptive verbal skills were also assessed atvery grateful’). After presentation of each item, a question

is asked which requires identification of the pronoun age 11:0 with the Token Test of Auditory LanguageComprehension (DeRenzi and Vignolo, 1962), which consistsreferent(s), e.g. ‘Who was very grateful?’. The 24 items,

consisting of three sets of eight items each, are presented in of commands of increasing complexity requiring the subjectto touch or move tokens that vary in shape, colour, and size.random order. In set A, the pronoun referent is determined

by the syntax and semantic content of the sentences (e.g. His performance on this test was uneven. On parts I–IV, thecommands are expressed in a syntactically simple form,‘Georgina went into the butcher’s shop. He said hello to her.

Who said hello to whom?’). In sets B and C, the pronoun initially with a single object (e.g. ‘touch the red circle’), butlater with a compound one (e.g. ‘touch the small yellowreferent is not fixed by the syntax but is determined

pragmatically (e.g. ‘The doctor cured the patient. He was circle and the large green square’). Alex’s score (37 out of40) on these four parts was age appropriate (Noll and Lass,very clever. Who was very clever?’). In set B, the pronoun

referent in the second sentence is the first of the two nouns 1979). Part V of the test introduces more complex syntacticalstructures, such as prepositions (e.g. ‘beside’, ‘between’),presented in the initial sentence; in set C, it is the second of

the two. At age 14:2 years, Alex correctly answered 15 adverbs (e.g. ‘quickly’, ‘slowly’) and clausal phrases (‘beforetouching the yellow circle, pick up the red square’). Alex’squestions (set A, 6; B, 5; C,4). This score is,1.5 SD below

the mean for normal children ranging in age from 10:2 to score (nine out of 22) on this part fell.3 SDs below themean for 8-year-olds.10:11 (n 5 19; E. Isaacs, S. Taylor and F. Vargha-Khadem,

unpublished).

Overall receptive language ability.To obtain aProsody. To determine Alex’s sensitivity to prosodic standardized measure of Alex’s receptive language abilitiesvariations in spoken sentences, we designed a test of receptiveacross different domains, we gave him the CELF-R (Semelprosody (E. Isaacs, S. Taylor and F. Vargha-Khadem,et al., 1987) at ages 14:2 and 14:11. Like the expressiveunpublished). In this test, the child listens through headphoneslanguage score, the receptive language score is based on threeto a tape containing four different sentences, each utteredsubtests: (i) linguistic concepts (assessing comprehension offour times. The sentences are affectively: neutral—‘He wasconcepts related to inclusion, exclusion, coordination, time,a tall man with brown hair’; happy—‘I’m really looking condition and quantity); (ii) sentence structure (like theforward to my summer holidays’; sad—‘I miss my best TROG, evaluating comprehension of grammatical rules atfriend Tim; he’s gone away’; angry—‘My brother is always the sentence level); (iii) oral directions (measuringtaking my toys without asking’. Each of these sentences iscomprehension, recall and execution of oral commands ofspoken in one of the four tones of voice (neutral, happy, sadincreasing length and complexity). Alex initially obtained aor angry), yielding four sentences in which tone and contentstandard receptive language score of 59 and then, 9 monthsare congruent and 12 in which they are incongruent. Afterlater, his score increased to 67, due to a gain on the Sentencelistening to each utterance, the child indicates the tone ofStructure subtest (Table 2). On the other two subtests, hisvoice by pointing to one of four line drawings of faces performance remained unchanged at 2 SDs below the mean;depicting the expressions corresponding to neutral, happy,it should be noted that these two measures closely resemblesad, and angry. Administered this test when he was 14:2, AlexPart V of the Token Test, on which Alex’s performance wascorrectly matched 10 tones of voice with their correspondingsimilarly restricted.facial expressions. This score is,1 SD below the mean of By summing over the standard scores obtained on thethe oldest normal children we tested (ages 10:2 to 10:11;n 5 Expressive and Receptive subtests, a standard total language19; E. Isaacs, S. Taylor and F. Vargha-Khadem, unpublished).score and an age equivalent can be calculated. As indicated

in Table 2, Alex’s age equivalent on this test rose from 8:1Grammar. Alex was given Bishop’s (1982) Test for (when he was aged 14:2) to 8:5 (when he was aged 14:11).Reception of Grammar (TROG), a test of sentence–pictureIt may be noted that on both occasions his expressivematching, on several different occasions. Tested for the firstlanguage abilities were superior to his receptive ones, with thetime at age 9:0, his score (three out of 20 blocks) placedmagnitude of the difference being halved on second testing.him at an age equivalent of 3:6 (Fig. 6). Retested 2 and 3years later, his age equivalent rose steadily to 8:0 (15 out of20 blocks), and it remained at the same level when he was

Cognitive development subsequent to speech14:2; but then in the space of 9 months (14:11) it improvedonsetto 10:0 years (17 out of 20 blocks).

Comprehension of complex sentences.The Reynell Intelligence. By age 11:0, Alex’s verbal skills haddeveloped sufficiently to allow the administration of theLanguage Scales (1977), which make use of miniature toys

and objects to assess verbal comprehension, yielded an age Wechsler Intelligence Scale for Children—Revised (WISC-R; Wechsler, 1976), and this test was administered onceequivalent of 3:11 when Alex was 9:4, but he then reached


Table 3 Alex’s intelligence quotients and IQ/Index Scores

IQ item Scores related to chronological age (years:months)

WISC-R UK WISC-III Improvement

11:0 11:9 14:2 14:11 (14:11–11:0)

Verbal IQ ,45 ,45 50 59 114Performance IQ ,45 ,45 50 53 18Full scale IQ ,40 ,40 47 52 112Verbal comprehension – – 54 63Perceptual organization – – 53 56Freedom from distractibility – – 52 52Processing speed – – 58 54

The corrected scores on the WISC-R (seetext) were below the floor for this test.

more when he was 11:9. At the ages of 14:2 and 14:11,Alex was retested for intelligence on Wechsler’s newlystandardized WISC-III (Wechsler, 1992). To allowcomparisons between the IQ scores derived from the old andnew standardizations, a correction (i.e. downscaling by afactor of 0.3 of an IQ point for each year sincestandardization) was applied to the WISC-R scores (Flynn,1984, 1987; Fuggleet al., 1992). Table 3 shows the verbalIQ, performance IQ, and full scale IQ that Alex obtained oneach of the four occasions.

The WISC-III also enables the calculation of standardizedindex scores for Verbal Comprehension, PerceptualOrganization, Freedom from Distractibility and ProcessingSpeed. The index scores obtained by Alex on each of thetwo testing occasions are also shown in Table 3. (Raw andscaled scores, as well as mental age equivalents for each ofthe subtests administered on the different occasions arepresented in Appendix 3.) As Table 3 and Appendix 3indicate, Alex has shown a consistent increase over the past4 years in all three IQ measures, particularly on selected

Fig. 7 The relationship between developments in mental age,subtests that contribute to the verbal scale. At the latestmeasured by the Griffiths Scales of Mental Development, andassessment (age 14:11), his full scale IQ was 52, while hislanguage abilities. Expressive language scores are averages of the

verbal IQ was 59. scores for Grammar and Information shown in Fig. 5; receptiveEvaluation of Alex’s cognitive development with the language scores are those for the BPVS shown in Fig. 6.

Griffiths Mental Development Scales (Griffiths, 1970), hisfifth assessment with this test, was carried out 6.5 years afterlagged behind his language abilities by ~18 months (seesurgery when he was aged 14:11 (Table 1). In contrast to theTable 4).very modest gain he had shown 6 months postoperatively, To determine whether Alex’s profile of having betterhis later improvements are impressive. If predictions werelanguage ability than overall intellectual function isbased on the overall increase in his mental age (i.e. 3 months)characteristic of left hemispherectomized patients ofthat had occurred during the 10-month interval separatingcomparable intelligence, but without mutism prior to theirthe first two assessments, an increase of 13.5 months wouldhemispherectomies, we tested four such patients on thehave been expected during the 45-month interval separatingWISC-III and the CELF-R. The only basis of selection wasthe second and third evaluations. In fact, Alex’s overalltheir ready availability for assessment with the relativelymental age across this interval increased by 33 months.new CELF-R test, these four cases being ones that wereFigure 7 illustrates the parallel between the gains in Alex’sconcurrently undergoing neuropsychological evaluation atmental age, as measured by the Griffiths Scales, and his gainsthe Great Ormond Street Hospital for Children. Like thein receptive and expressive language ability, as measured bypathology in Alex, the pathology in three of these patientsthe BPVS and the Action Pictures Test, respectively. Yet,was congenital in origin. Case 1 had congenital hemiplegia,

with onset of epilepsy at the age of 2 years following a veryAlex’s mental age since the age of 11:0 has consistently


Table 4 Alex’s IQ and language ability in terms of mental 1979). On each occasion, Alex attained a memory quotientage equivalents that was from 18 to 21 points higher than his full scale IQ

and from 13 to 17 points higher than his verbal IQ.Alex’s chronologicalOnce again, to determine whether this pattern isage (years:months)

characteristic of left hemispherectomized patients, we14:2 14:11 compared Alex’s scores with those of seven previously tested

cases in the same age band as Alex (seeTable 6). One patientMental age equivalents(No. 6) showed a pair of discrepancies (memory quotientVerbal IQ 6:8 7:4versus full scale IQ, and memory quotient versus verbal IQ)Performance IQ 6:5 6:6

Full scale IQ 6:6 6:11 similar to those of Alex, but none had spreads greater thanLanguage ability (CELF-R) 8:1 8:5 Alex’s, his latest memory quotient surpassing that of all

the others.Mental age equivalents for the above IQ scores are based onTable 7 presents Alex’s longitudinal profile on the Wechslerassigning subtest age equivalents of 6:0 when obtained scores

were,6:2 (seeAppendix 3). Memory Scale. His immediate and delayed recall of theTaylor Stories, Form I are reproduced in Appendix 2B toillustrate not only his memory performance but also hisprolonged seizure; an MRI carried out subsequently revealed

encephalemalacia of the entire left cerebral hemisphere. Case verbal ability. On some of the memory measures, e.g. third-trial learning of both related and unrelated Paired Associates,2 also had a congenital hemiplegia, with onset of seizures at

the age of 2 months, secondary to a parieto-occipito-temporal his age equivalent was close to his chronological age of 15(F. Vargha-Khadem, unpublished data).developmental anomaly. Case 3 presented with infantile

hemiplegia and seizures at age 3:4, following a haemorrhage On another measure of verbal-list learning (Children’sAuditory Verbal-Learning Test-2; Talley, 1993), where a listinto an extradural cyst, compressing the left hemisphere. The

remaining patient, Case 4, was diagnosed as having acquired of 16 words is presented five times for recall, Alex (thenaged 14:2 and 14:11) obtained an average standard score forRasmussen’s encephalitis (Rasmussen, 1978, 1983) at age

3:6. All four patients were testedù2 years after their level of learning of 84, which is at the lower limit of theaverage range. His average standard score for 20-min delayedhemispherectomy and had been weaned of their

anticonvulsants forù1 year. Table 5 shows the verbal IQ, recall of this list is 93, which is also in the average range.Not all of his Wechsler Memory subtest scores were atperformance IQ and full scale IQ scores, along with standard

scores on expressive, receptive, and total language indices this level, however. Consistent with his relatively poor visualperformance skills, his scores on immediate and delayedof the CELF-R for Alex and Cases 1–4. The mental age

equivalents are also shown for the tests of intelligence and Visual Reproduction were substantially lower than average(i.e. .2 SDs below those of 6–7-year-olds).total language. The comparison indicates that the language-

IQ discrepancy was greater in Alex than in the others, his Similarly, his scores on both digit span (forward 4;backward 2) and block span (forward 4; backward 3) are upachievement in language being the highest of all.to 2 SDs below the norm for children of the same mentalage as his (Isaacs and Vargha-Khadem, 1989). As shown inLearning and memory.On the same four occasions that

the WISC IQ tests were administered to Alex, the Wechsler Table 6, a short memory span appears to be characteristic ofchildren with left hemispherectomy.Memory Scale, Forms I and II (Wechsler and Stone, 1945)

was also given (using age corrections and children’s storiesdesigned by L. Taylor as described in Kimura and McGlone,Visual perception and reproduction.Visual perception

Table 5 Alex compared with four other left-hemispherectomized cases on IQ and language function: scores (ageequivalents)

Subjects (and chronological ages in years:months)

Alex (14:11) Case 1 (7:1) Case 2 (18:7) Case 3 (15:11) Case 4 (7:0)

WISC-III or WPPSI-RVerbal IQ 59 (7:4) 52 (,6:2) 46 (,6:2) 58 (7:8) 63 (4:6)Performance IQ 53 (6:6) 49 (,6:2) 46 (,6:2) 48 (,6:2) 57 (4:5)Full scale IQ 52 (6:11) 47 (,6:2) 40 (,6:2) 49 (,6:2) 56 (4:5)

CELF-RExpressive language 76 (–) 50 (–) 62 (–) 70 (–) 67 (–)Receptive language 67 (–) 50 (–) 50 (–) 50 (–) 62 (–)Total language 70 (8:5) 50 (2:2) 53 (6:9) 59 (7:4) 63 (,5.0)Total full scale IQ 18 3 13 10 7Total verbal IQ 11 22 7 1 0


Table 6 Alex compared with seven other left-hemispherectomized cases on IQ and memory

Subjects (and their chronological ages in years:months)

Alex No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7(14:11) (10:5) (18:3) (18:2) (20:3) (15:11) (18:7) (10:2)

Wechsler IQVerbal IQ (VIQ) 59 57 69 69 64 58 46 48Performance IQ 53 48 70 63 77 48 46 50Full scale IQ (FSIQ) 52 48 68 64 68 49 40 46

Wechsler memory scaleMemory quotient (MQ) 73 53 61 66 57 63 59 54

Differences between MQ and IQsMQ–FSIQ 21 5 27 2 27 14 19 8MQ–VIQ 14 24 28 23 211 5 13 6

Digit spanForward 4 3 3 4 4 4 5 3Backward 2 2 3 3 2 4 2 2

Table 7 Wechsler Memory Scale (WMS Forms I and II)

WMS item Alex’s WMS scores related to his chronological age (years:months)

11:0 11:9 14:2 14:11

Memory quotient 58 60 67 73

Raw scoresPersonal and current information 1/6 1/6 3/6 3/6Orientation (temporal and spatial) 2/5 3/5 5/6 5/6Mental control 0 0 2 3Recall of prose passages

immediate 8.5/22 7.5/22 4.75/22 7.5/22delayed at 90 min 5.5/22 5/22 8.5/22 0.5/22*

Digit spanforwards 3 3 3 4backwards 0 0 3 2

Visual reproductionimmediate 1/14 3/14 3.5/14 4/14delayed recall at 40 min 1/14 0/14 3/14 1/14

Paired associate learningrelated pairs 5/6 6/6 6/6 6/6unrelated pairs 4/4 3/4 3/4 3/4

Delayed recall of paired associatesat 90 min: related pairs 4/6 5/6 6/6 5/6at 90 min: unrelated pairs 1/4 3/4 3/4 2/4

*By the end of this 150-min session, Alex had clearly become fatigued, and this is reflected in his uncharacteristically low score on 90-min delayed recall of the stories.

independent of motor output was investigated using the Finally, Alex was asked to draw four items (a box, bicycle,lion and elephant) from long-term memory. Figure 8 showsMotor Free Visual Perception Test (Colarusso and Hammill,

1972). This test evaluates five aspects of visual perception: his extremely impoverished representations of these items,confirming that, consistent with his low performance IQ andspatial relationships, visual discrimination, figure–ground,

visual closure and visual memory. When administered this restricted immediate and delayed memory for geometricdesigns (seeLearning and memory), his motor abilities laggedtest at the age of 14:2, Alex obtained an age equivalent of 7:5.

Visual reproduction was assessed through Alex’s copying behind his language abilities by ~3 years.of 24 geometric forms (Developmental Test of Visual–MotorIntegration, Beery, 1989) at ages 14:2 and 14:11. On bothAcademic attainments.After transferring in October 1993

to a school for children with moderate learning difficulties andoccasions, Alex obtained a standard score of,55 (less thanthe first percentile score), equivalent to age 5:6. having spent one academic term receiving formal tuition in


Fig. 8 Alex’s drawing of a box (top left); a bicycle (top right); a lion (bottom left) and an elephant(bottom right).

reading and writing, Alex was evaluated at age 14:2 in these degree of recovery of movement with the affected limb,which we describe next. Alex had an established righttwo literacy skills. He did demonstrate a sound knowledgehemiplegia before surgery, the arm being more severelyof letter names, recognizing all 26 letters of the alphabet. Inaffected than the leg. The hemiplegia was associated withprereading children, this ability is an important predictor ofright hemiatrophy, increased tone and hyperreflexia. Despiteliteracy acquisition (Muter, 1994). Yet, Alex still does notthis, he was able to hold a cup and catch a ball with bothpossess any phonetic decoding or spelling skills. Thus,hands, climb the ladder of a playground slide, kick a ballon the Wechsler Objective Reading Dimensions (WORD;with either leg, and even run.Wechsler, 1993), a test of reading accuracy for single words

Immediately following the hemispherectomy, Alex’sthat is independent of memory and reading comprehension,hemiplegia worsened, but over the next few monthshe obtained a standard score of only 46 (less than the firstmotor function on his right side gradually returned to itspercentile score; age equivalent,6:0 years); and at agepreoperative level. On neurological assessment at age 11:0,14:11, having completed four academic terms with instructionhe had a 3 cm shortening of the right arm and a 2 cmin this area, he showed no improvement (standard score ofshortening of the right leg. He was hypertonic and42). His extremely limited sight reading vocabulary precludedhyperreflexic on the right side, with an extensor plantarhis scoring any points on the reading comprehension subtestresponse and a typical hemiplegic gait, with his right legof WORD. Further, on the Nonword Reading Test (Snowlingcoming down in walking. In running, his right arm waset al., 1996), he failed to score at all (zero out of 20; ageslightly flexed, but he could run fairly fast, but awkwardly.equivalent,6 years). On the spelling subtest of WORD,He was unable to make independent finger movements withAlex correctly wrote to dictation only six letters of thehis right hand, but he could use this hand as a prop or toalphabet, and the words ‘no’, ‘pig’ and ‘to’ (standard score,grasp objects between thumb and fingers. For example, he51; less than the first percentile score; age equivalent 6:0).could hold a vest while taking it off or putting it on, thoughFinally, although Alex is able to read and write the numbershe could not manage buttons or socks or hold a utensil to1–8, he was unable to perform any of the arithmetic operationscut food. No associated or mirror movements were notedon the Basic Number Skills subtest of the British Abilitywhen he moved his limbs on either side.Scales (Elliott et al., 1983).

Quantitative measures of hand strength and dexterity wereobtained when Alex was aged 11:3. Grip strength was

Motor function before and after surgery assessed with a Kiddie Dynamometer (Lafayette InstrumentMotor ability with the paretic hand Company, 1989) for three trials with each hand. His meanAlthough Alex’s recovery is particularly impressive in the left-hand score (19.660.5 kg) was well within normal limits

for children of his age (Finlayson and Reitan, 1976); his meanarea of speech and language, he also shows an unusual


the transfer of 10 dowels as rapidly as possible from theback to the front of a pegboard. The time to complete thistask was averaged across five trials with each hand. Aswith tapping, his mean sorting time with the left hand(20.162.0 s) was somewhat longer than would be expectedfor an 11-year-old boy (10.8611.2 s; Annett, 1986); and hismean sorting time with the right (116.8638.3 s), althoughmarkedly impaired, once again indicated some residual motordexterity.

Alex’s residual motor function on the right side wasassessed by transcranial magnetic brain stimulation. Thiswas carried out with a Magstim 200 stimulator (MagstimCo, Dyfed, UK) equipped with a focal double-cone coil(150 mm). The coil design allows painless transcranialstimulation of selected areas of the motor cortex forassessment of functional corticospinal projections. In thepresent study, the stimulation was used to evoke responsesof the first dorsal interosseus (1DI) muscles of the hand, aswell as of the forearm extensor, biceps and deltoid muscles,each in turn.

Multiunit EMG responses were recorded with surfaceelectrodes (and neurolog modules) from the correspondingmuscles of the left and right upper limbs during weakbackground, volitional, contraction. In control subjects (n 510), focal magnetic stimulation of the hand area of thedominant motor cortex evoked responses in the contralateral1DI muscles only; that is, responses were not seen inthe ipsilateral 1DI muscles (Fig. 9A). In Alex, however,stimulation of the right motor cortex evoked EMG responsesfrom both contralateral and ipsilateral 1DI muscles (Fig. 9B).Compared with the contralateral responses, those evokedipsilaterally were smaller in amplitude, had a longer meanlatency (30.061.6 ms versus 21.562.9 ms;n 5 11), and,

Fig. 9 Effects of focal magnetic brain stimulation on EMG with the responses having been rectified and normalized foractivity recorded from preactivated left and right first dorsal

background EMG levels, were only 22% as large.interosseous (1DI) muscles. Responses to three trials areSimilar results were obtained for the other muscles of thesuperimposed. Stimulation is at time zero on the ms scale. (A) In

a healthy control subject, stimulation of the dominant left motor upper limb. In control subjects, responses were only seen incortex evokes responses in the contralateral right 1DI at the contralateral forearm extensor, biceps, and deltoidaround 22 ms. There is no ipsilateral response. (B) In Alex, muscles, whereas in Alex, both ipsilateral and contralateralstimulation of the isolated right hemisphere evokes responses in

responses were recorded in all these muscle pairs.both the contralateral and ipsilateral 1DI at around 21.5 msThe nature of the abnormal ipsilateral projections wasand 30 ms, respectively.

further explored by means of cross-correlation analysis. Fromthe multiunit EMG recorded as above, cross-correlogramsright-hand score (2.760.5 kg), although severely deficient,

indicated some residual function. were constructed between homologous left and right musclepairs by logging the times of occurrence of motor unit spikesFine motor control of the index finger was measured with

a manual tapper attached to an electronic counter. The number from each muscle [minimum of 3000 spikes;see Carret al. (1993) for details of the method]. When two spinalof taps in a 20-s period was averaged across two trials with

each hand. The mean tapping speed with his left index finger motoneuron pools are innervated by a branched final inputarising from a common source, the discharges are likely to(234 taps/min) was ~1 SD below the mean for 11-year-olds

(F. Vargha-Khadem and N. Vargha-Majzub, unpublished be closely synchronized, this being reflected as a short-duration central peak in the cross-correlogram (Sears andnormative data); his mean tapping speed on the right, where

he had to use his whole hand rather than the index finger Stagg, 1976). In the case of Alex, the cross-correlogramsfailed to show central peaks for any of the paired muscles,alone, was far slower (96 taps/min), but again it clearly

exhibited some preserved ability. suggesting that the spinal motoneurons supplying themembers of each pair were innervated from separateFinally, finger, hand and arm control were evaluated

with the Annett Pegboard Task (Annett, 1986) which requires cortical sources.


language functions showed remarkable gains. Indeed, hisPraxic functionsgains in the latter sphere are truly extraordinary, given thatAt age 12:7, Alex was administered tests of limb and oralhe had been essentially without speech until the age of nearlypraxis. The former test (R. Passingham and F. Vargha-9:5. Alex’s development in speech and language ability, asKhadem, unpublished) assesses the execution of meaningfulwell as the parallel improvement that took place in hisupper limb movements (e.g. ‘Drink from a glass’),other cognitive functions, will be considered below, after ameaningless movements (e.g. ‘Put your hand on your head’)discussion of his unchanged motor status.and the use of objects (e.g. ‘Show me what you do with

this’, the examiner pointing, say, to a pair of scissors). Thetest contains 15 items, five of each type. Two independentjudges rated Alex’s execution of the movements with his

Motor ability with the paretic armunaffected (left) arm and hand, scoring each item on a four-Despite his right-sided hemiplegia, more pronounced in thepoint scale ranging from 0 (no response) to 3 (correct). Botharm than in the leg, Alex showed some residual fingerjudges rated Alex’s performance as 43 out of 45, indicatingmovements in his right hand before surgery, so that he couldthat he has little or no difficulty with limb praxis.grasp a metal bar to exert power (using a dynamometer), andThe test of oromotor praxis (Vargha-Khademet al., 1995)even grasp small objects (such as pegs) between the thumbconsists of seven sections (animal/machine noises, five items;and fingers (seeMotor ability with the paretic hand). Sincemeangingless noises, e.g. ‘click your tongue’, five items;no permanent change in his motor status was noted after thesinging, e.g. ‘hum a tune’, four items; nonvocal movements,hemispherectomy, it may be concluded that, though modest,e.g. ‘bite your lip’, 15 items; eye movements, e.g. ‘closethe residual motor function of his hemiplegic hand andyour left eye’, four items; sequences of movements, e.g.fingers is largely a reflection of motor reorganization brought‘stick out your tongue, lick your upper lip, and smack yourabout by the early prenatal onset of his left hemispherelips’, two items). The judges were somewhat discordant inpathology.their assessment, one rating Alex’s performance as 59 out of

Postoperatively, magnetic stimulation of Alex’s remaining108 and the other as 65 out of 108. It is clear, however, thathemisphere revealed the presence of abnormal ipsilateralAlex experiences far more difficulty executing instructedcorticospinal projections to the muscles of his hemiplegicmovements with his oral and facial musculature than witharm and hand. An earlier neurophysiological study (Carrthe musculature of his left upper limb.et al., 1993) of hemiplegic, but nonhemispherectomized,subjects provides evidence that such reorganization of motorpathways frequently follows congenital cortical damage. Thefindings in that study yielded evidence of abnormal ipsilateralDiscussionprojections from the undamaged motor cortex to theIt has long been recognized that extensive injury to the leftmotoneurons of the hemiplegic arm and hand in 63% (20cerebral hemisphere in an infant or a young child is compatibleout of 32) of the subjects examined. These ipsilateral pathwayswith the development of speech and language provided thatwere always associated with some residual motor functionthe right hemisphere is relatively intact. Indeed, if a left-of the hemiplegic upper limb. In 18 of the 20 individualssided lesion is sustainedin utero, there is an even greaterwith ipsilateral projections, the hemiplegia resulted fromopportunity for the escape not only of speech and languageprenatal or perinatal damage. Cross-correlation analysisfunctions but also of distal movements of the right handapplied to the multiunit EMG recorded from Alex’s paired(Dijkerman et al., 1993). The vascular malformation inupper limb muscles indicated that the corresponding spinalSturge–Weber disease, which is manifested by a port-winemotoneuron pools are innervated by different presynapticstain over the face and an associated venous angioma of theinputs. Taken together with the results of magneticleptomeninges, is related to a defect in the development ofstimulation, this suggests that separate projections fromthe cerebral blood supply during the early stages of gestationthe right hemisphere supply Alex’s paired ipsilateral and(Menkes, 1980). Thus, in Alex’s case the origins of leftcontralateral muscles.hemisphere pathology can presumably be dated to the first

Evidence of motor reorganization based on the results oftrimester ofin utero life, and, if predictions were to be basedmagnetic brain stimulation in hemispherectomized patientssolely on the timing of the pathology, impressive preservationwho had sustained congenital injuries was obtained previouslyof both limb movement and language functions would beby Beneckeet al. (1991) and Cohenet al. (1991). In theexpected. In fact, however, his preoperative seizure disorder,remaining hemisphere of Cohenet al.’s (1991) patient, theantiseizure medication, or both, severely limited the degreecortical representations of the ipsilateral and contralateralto which his right hemisphere could subserve speech andupper limb muscles differed topographically, the ipsilaterallanguage, such that only in the limb-movement domainones being located more rostrally and laterally than thedid he show any preservation of function during earlycontralateral. Beneckeet al. (1991) did not search for suchdevelopment. The later development of these functions,topographic differences in their patient, but they found thathowever, was dramatically different, his limb-movement

ability remained relatively unchanged, but his speech and the degree of normality of the ipsilateral muscle action


potentials did correlate roughly with the degree of residual hemispherectomy performed 4 months after the onset of theepilepsia partialis continua, the child’s seizures were arrested,motor capacity in the individuals’ paretic limbs.

The developmental origin of the pathways mediating the and she then showed a rapid resolution of the syndrome andits associated features. Not only did her speech return butipsilateral response is still unclear. Animal studies have

shown that novel functional ipsilateral corticospinal pathways her language ability improved to the point where she wasproducing complete phrases and began to read letters andmay develop following early unilateral hemispherectomy

(Hicks and D’Amato, 1970; Barth and Stanfield, 1990), and simple words.The child reported by Fusco and Vigevano (1991) sharesthe findings by Cohenet al. (1991) suggest that the same

may have occurred in their patient. It is also possible, at least three important features with Alex. First, like Alex,she showed no evidence of bilateral structural lesions orhowever, that pre-existing ipsilateral corticospinal pathways,

projecting either directly or via the brainstem, are stabilized pathology in the frontal opercular regions, damage that istypical of the Foix–Chavany–Marie Syndrome (Graff-or enhanced by early damage (Beneckeet al., 1991). Indeed,

both types of change could coexist (seeCarr et al., 1993). Radfordet al., 1986; Kuznieckyet al., 1989; Tatumet al.,1989). Secondly, both children showed evidence of oromotorHowever they arose, the abnormal ipsilateral projections

found in these two patients and in Alex presumably account apraxia, which in the case of Alex was present from an earlyage and still continues for certain movements of the oromotorfor their preserved ipsilateral motor abilities, since such

projections appear to be absent in patients who fail to show system, such as blowing, sucking through a straw, etc. (seePraxic functions) despite the current absence of any signs ofsuch functional preservation (Homberget al., 1991). Yet the

amount of sparing of Alex’s motor ability with his paretic a motor aphasia. And third, both children were withoutspeech during the period of presumed functional interferencelimb is severely limited compared with that seen in his

speech, reflecting major differences in the degree to which with the opercular or perislyvian areas on the right, withboth either regaining or acquiring this function only afterthe isolated right hemisphere can assume these two types of

motor function. their left hemispheres were removed. An important differencebetween the two is that, prior to her seizure onset, Fuscoand Vigevano’s 3-year-old child had already acquired somespeech, which was arrested for a period of only 4 months,Speech and oromotor praxis

The modest degree of cerebral reorganization that Alex whereas Alex spoke for the first time only after the age of9 years.showed before his surgery and maintained afterwards may not

have been limited to the limb movement sphere, despite the Other instances of reversible interference with speechand language functions were described recently by Morrellabsence of overt evidence of speech and language until after

the complete withdrawal of anticonvulsant medication. Theet al. (1995) in a series of children with Landau–KleffnerSyndrome (Landau and Kleffner, 1957). This syndrome israpid onset of his speech subsequent to withdrawal of

anticonvulsants suggests that at least some degree of characterized by the appearance, during the period of languagedevelopment, of severe, bilateral epileptiform activity over thecerebral reorganization for language processes had already

occurred, but that its expression was prevented by func- posterior temporal regions, which leads almost invariably toauditory agnosia followed by aphasia, often to the point oftional interference, presumably with the perisylvian areas on

the right, caused by the epileptogenic activity and/or the mutism. Morrellet al. (1995) discovered that the bilateralelectrographic abnormality in many of their cases had aanticonvulsant treatment. This interpretation is consistent

with the findings from the MRI scan, which gave no evidence unilateral origin; treatment of those children with a newsurgical technique, namely multiple subpial transectionthat calcifications typical of Sturge–Weber disease had

affected his right hemisphere, and from the PET scan, which (Morrellet al., 1989) throughout the area from which theabnormality originated, resulted in bilateral abolition of thehad revealed normal right-sided metabolism.

Other instances of functional interference with the epileptiform activity followed by a gradual recovery of speechfunctions to normal or near-normal levels. Each of thedevelopment of speech and language in young children have

recently been reported. For example, Fusco and Vigevano successfully treated children, unlike Fusco and Vigevano’scase (1991), had remained aphasic for periods ranging from(1991) describe the case of a 3-year-old child with left

hemimegalencephaly and delayed intellectual and language 1.5 to 4.5 years, although, like that case, they too hadacquired some speech prior to their illness. Thus, among thefunction, who developed progressive epilepsia partialis

continua on the left with a spread to the contralateral right cases with reversible aphasia of which we are aware, Alexstands alone in having acquired speech for the first time infrontal and central regions. As a result of the evolution of

her seizures, this child developed bilateral frontal opercular late childhood.syndrome (Foixet al., 1926), with its associated featuresof facio-pharyngo-glosso-masticatory diplegia resulting insevere oromotor apraxia and mutism. In this case, however,Expressive language

The vast majority of normal children progress through distinctunlike all other such cases that have been reported, theopercular syndrome was reversible; following left stages of speech development (Brown, 1973; Bateset al.,


1992), in which they first produce words (~11–13 months), than that of the four other left hemispherectomized patients(seeTable 5).then word combinations (~20–24 months) and, finally,

grammatical phrases (~28 months). The same pattern ischaracteristic of children whose first language is sign language(Newport and Meier, 1985; Meier, 1991) as well as of Receptive language

In the domain of receptive language in normal children,children who are language delayed (Gibson and Ingram,1983). In contrast to this protracted 1.5-year period with its studies have focused on comprehension of both single words

(Huttenlocher, 1974; Goldin-Meadowet al., 1976; Snyderrelatively well defined stages of speech acquisition, Alexprogressed within a few months from articulating consonantset al., 1981) and syntax (de Villiers and de Villiers, 1973;

Bates et al., 1992). Without exception, and regardless offor the first time to uttering single, intelligible words andthen to producing full-length sentences. This initial burst in method of testing, comprehension of single words precedes

and outpaces production by a large temporal gap. This robustspeech output has the appearance, at least, of a release frominterference permitting very rapid learning. Between the ages finding holds not only for normal children but also for

language-delayed subjects (Gibson and Ingram, 1983) andof 9:4 and 14:11, Alex’s expressive language abilities, asreflected in his production of well formulated sentences even for apes trained to communicate with symbols or

gestures (Savage-Rumbaughet al., 1986).conveying knowledge of both semantics and syntax, rosefrom an age equivalent of,2 years to one of.8 years. Normal children begin to show evidence of single-word

comprehension by ~9 months of age, and between the agesBetween the ages of 9:4 and 10:11, his mean length ofutterance increased dramatically from 0 to 11.6 words, not of 1:1 and 2:0 they amass a receptive vocabulary of 50–70

words, whilst producing only 10–20 words (Snyderet al.,only reflecting his increased ability to produce complexsentences but also presaging his ability to comprehend 1981; Goldin-Meadowet al., 1976). This gap between

comprehension and production often becomes magnifiedsentences containing virtually all grammatical forms (seebelow, Receptive language). It is noteworthy as well that his in language-delayed subjects. For example, the language-

delayed child reported by Gibson and Ingram (1983) had ascores on English morphology (Word Structure: ExpressiveLanguage Subtest of the CELF-R) and phonological imitation receptive vocabulary of 183 words, whereas his productive

vocabulary was limited to eight words. In general, nounof multisyllabic real words rose to the normal range for hischronological age. comprehension precedes verb comprehension, although

young children understand many verbs they do not produceYet not all domains of Alex’s expressive language abilitiesshowed the same impressive gains. His performance on (Savage-Rumbaughet al., 1993).

Research on sentence comprehension is less systematicexpressive language tasks involving manipulation ofnonwords (see sections on Morphology, Phonological than that on word comprehension, but it has been shown that

English-speaking children between the ages of 2 and 3 yearsimitation, and Phonological awareness ) remained severelyrestricted at a 5:0–5:6 age level (i.e. usuallyù3 years below respond appropriately to sentences in the active voice and

are sensitive to word-order cues (de Villiers and de Villiers,his performance on tests involving real words). Furthermore,on at least one nonword test (Phonological imitation), his 1973; Bateset al., 1984). The passive construction is more

difficult to master and develops gradually between the agesperformance showed virtually no improvement during aperiod of .3 years. Alex’s restricted performance on tasks of 4 and 7 years (de Villiers and de Villiers, 1973; Roberts,

1983). De Villiers and de Villiers (1973) reported that normalrequiring nonword manipulation suggests a deficiency inverbal representations at the level of syllables and, even children with a mean length of utterance of 1–1.5 words

could not use word-order cues in sentence comprehension.more, of phonemes. This deficiency in analysing wordsaccording to their phonetic features coupled with his limited Only when the mean length of utterance increased to 1.6–3

words could children comprehend word-order distinctions inmemory span for syllables, akin to his limited memory spanfor digits and blocks, could account for his difficulty on the active voice, and only with a mean length of utterance

of ù3 words could they comprehend reversible passives.nonword tests. This type of difficulty, as well as the tendencyto make semantic and perceptual errors in naming (seesection These and similar findings by Chapman and Miller (1975)

have led to the conclusion that, unlike the case with singleon Naming), appears to be pathonomonic of right-hemispherespeech (Zaidel, 1980; Gazzaniga, 1983; Pattersonet al., words, grammatical production (specifically word-order

constraints)precedessentence comprehension.1989; Patterson and Vargha-Khadem, 1991; Vargha-Khadem,et al., 1991). Measures of expressive language ability Because of the relationship between the mean length of

utterance and sentence comprehension, it is difficult to inferinvolving meaningful words are not sensitive to thisunderlying deficit and can result, as they do in the case of what level of receptive language ability Alex had acquired

prior to the onset of his speech. Although both his receptiveAlex, in a seemingly normal, albeit delayed, ability to generatethe different aspects of speech. In fact, it is noteworthy vocabulary and verbal comprehension at age 9:4 were at the

3- to 4-year-old level (seeVocabulary and comprehension ofthat despite the history of mutism until the age of 9:4(years:months), Alex’s expressive language score on the complex sentences), his mutism may well have left him

insensitive to word-order constraints, and indeed hisCELF-R, which fell in the low normal range, was higher


comprehension seems to have been limited to short sentences 7), his latest memory quotient of 73 falling in the low normaluttered in the active voice. Now, however, his receptive andrange and so surpassing his verbal IQ and full scale IQ byexpressive vocabulary (as measured by the BPVS and the15–20 points. Although this lead of memory quotient overVocabulary subtest of the WISC-III, respectively) have bothIQ is not atypical of left hemispherectomized patients (seerisen to the same range (8:1 versus 7:6 years, respectively)Table 6), Alex’s lead was the greatest among the cases thatand show the typical lead of word comprehension over wordwere studied, largely because his memory quotient was theproduction. This, coupled with the dramatic explosion earlierhighest. Furthermore, on some measures within the domainin his mean length of utterance, have provided him with theof learning and memory, e.g. learning and delayed recall ofsentence comprehension of an 8- to 9-year-old, consistentPaired Associates (seeTable 7) and level of learning on thewith the evidence that the isolated right hemisphere canChildren’s Auditory Verbal-Learning Test-2 (see Learningmediate receptive language (encompassing semantics, syntaxand memory), Alex obtained scores that were within theand prosody) at a fairly high level (Zaidel, 1980). normal range for his chronological age. The isolated right

Among the different aspects of receptive language, onlyhemisphere thus clearly has the capacity to subserve, alongthe processing of complex grammatical forms stands outwith expressive and receptive language abilities, verbalas particularly problematic for Alex. Evaluated with three learning and memory functions that are typically mediateddifferent tests (The Token Test of Auditory Languageby the left temporal lobe (Milner, 1975; Incisa della RocchettaComprehension: Section V, and the CELF-R receptiveet al., 1995; Gadianet al., 1996).language subtests of Linguistic Concepts and Oral The same cannot be said of immediate memory or ofDirections), Alex’s performance on this aspect was severelyimmediate and delayed reproduction of geometric designs.deficient. Each of these tests requires carrying out sequencesDespite his relatively high memory quotient, Alex’s memoryof actions to constantly changing instructions pertaining to aspan for both digits and spatial locations is quite limited,small set of visual stimuli differing in shape, colour, size, as is his immediate memory for word lists (see Learninglength, etc. Although these tests assess comprehension of theand memory). The evidence suggests that his isolated rightmore complex aspects of syntax (Lesser, 1976; Noll andhemisphere cannot effectively mediate the process ofLass, 1979; Lezak, 1983), it is also likely that performanceworking memory. Similarly, Alex’s immediate and delayedis limited by immediate memory, which normally counteractsreproduction of geometrical designs is restricted, and althoughthe interference among conflicting instructions across trials.there has been a gradual improvement over the years inSince Alex’s performance was age-appropriate on theimmediate reproduction, delayed recall is still very muchSentence Structure subtest of the CELF-R, which alsoimpaired and shows no significant gains. This nonverbalinvolves increasing demands on syntactical processing, hismemory deficit, like the relatively small increase in hisimpairment in carrying out the competing instructions could(nonverbal) performance IQ (see Tables 1 and 3), isbe due as much to his limited memory span as to any difficultypresumably a reflection of the mechanism of ‘crowding’he might have in understanding complex grammatical forms.(Teuber, 1975;seeIntelligence section).This issue requires further investigation, not only in Alexbut also in other patients whose speech and language aremediated by the right hemisphere. Intelligence

Comparison of Alex’s overall receptive and expressiveThe seizure disorder and/or its treatment with medicationlanguage profiles indicates that the former has lagged behindappear to have suppressed the early development not onlythe latter, although the difference is now considerably reducedof Alex’s speech and language but also of his intellectual(seeTable 2). Despite this lag, the comparison with otherfunctions generally. Despite somewhat divergent startingleft hemispherectomized patients of comparable intelligencepoints at ~6 months after operation, both his language abilityindicates that Alex achieved the highest level of overalland his mental age rose in parallel, although from the age ofreceptive language (seeTable 5), just as he did in expressive 12:9 onwards his mental age has lagged behind both hislanguage. His long history of mutism thus seems to havereceptive and expressive language functions by ~1.5 yearshad no deleterious effect on final language outcome. This(seeFig. 7, and Tables 4 and 5). The pattern is reminiscentconclusion is consistent with the finding by Morrellet al. of patients with Williams syndrome (Williamset al., 1961),(1995) that children with the acquired syndrome of Landau–who typically show preserved language abilities, particularlyKleffner, whose speech and language functions are restoredfor complex grammar, in the presence of marked intellectualfollowing surgical treatment, achieve approximately normal

retardation and deficits in visuospatial functions (Bellugilevels of language function despite having lost auditory

et al., 1992). Interestingly, children with Williams syndromeverbal communication skills for several years during early

embellish their speech with stereotyped, attention-gettingchildhood (seeSpeech and oromotor praxis).

phrases and questions (Bellugiet al., 1994) much like thosecontained in Alex’s speech, as illustrated in Appendix 2.

Such discrepancies between language ability and IQ raiseLearning and memorythe question of how phonology and grammar can be mediatedAs with language, Alex’s long-term memory showed steady

improvement between the ages of 11:0 and 14:11 (seeTable effectively when cognitive ability is so restricted. Any


proposal at this point is necessarily speculative, but one language functions are assumed by the right hemisphere dueto early damage to the left, the cost of this reorganization issuggestion comes from the burgeoning evidence in both

animals and humans that many forms of noncognitive skill borne by nonverbal functions, and not the reverse, presumablybecause of the higher priority of auro-vocal communicationlearning depend on interactions between the cerebral cortex

and the extrapyramidal system, consisting mainly of the basal skills in the everyday life of the developing child (Vargha-Khademet al., 1992).ganglia and cerebellum (Petri and Mishkin, 1994). Perhaps

real-word phonology and certain aspects of grammar are Another area of cognitive function in which Alex hasshown no improvement is literacy and numeracy skills.likewise noncognitive skills supported by this same

behavioural learning system (e.g. Lieberman, 1992) and so Although he has received systematic instruction in reading,writing and arithmetic since the age of 13 years, hiscan be acquired and maintained somewhat independently of

the cerebral systems that serve cognition. Applied to Alex, performance in these areas has remained at the basal level(,6 years). As indicated earlier, a reliable predictor ofthe proposal could help explain his dramatic progress in

speech and language despite the limitiations imposed by a literacy acquisition in both normal and hemiplegic childrenis performance on tests of phonological awareness (Bryantlow IQ.

Alex’s intellectual development after the onset of speech and Goswami, 1987; Muter, 1994). Because Alex’s ageequivalent on such tests is still 5–6 years, he will probablyand language has nonetheless been substantial (seeTable 3).

Furthermore, as in the case of his linguistic and verbal continue to have difficulty in learning to read and write byphonetic decoding. However, a programme based on whole-memory abilities, his subsequent intellectual development

relative to that of other left-hemispherectomized children word recognition might well enable him to acquire a sightvocabulary and eventually amass a reading and writingdoes not seem to have been adversely affected by his

prolonged period of mutism (seeTable 5). lexicon equal to that of a 7- to 8-year-old.There is a dearth of information on developmentalStudies of intellectual outcome after hemispherectomy in

patients who had sustained congenital or perinatal injury dyscalculia, particularly regarding the association of thisdisorder with deficits in other academic attainments. It isindicate that as many as 40–70% of the cases show an

increase in IQ ofù10 points (Verityet al., 1982; Lindsay often the case, however, that children with low intellectualabilities who have severe problems with the acquisition ofet al., 1987; Beardsworth and Adams, 1988; Vargha-Khadem

and Polkey, 1992). A smaller proportion, namely 10–25%, reading and writing also have pronounced deficiencies informing number concepts. Given Alex’s profile to date, itare reported to show a 10-point or greater decline. In general,

the latter patients are more likely to have suffered from an seems unlikely that he will be able to make substantial gainsin this area.early seizure disorder. Alex of course also suffered from

early onset of intractable seizures, yet his increase in full It is noteworthy that Alex’s full scale IQ accurately predictshis academic attainments but not his linguistic or verbalscale IQ of 12 points clearly places him in the former group.

Because the development of intellectual ability does not memory abilities. The latter abilities are more consistent withhis social than with his intellectual skills. Such a pattern isplateau until ~17–18 years of age (Wechsler, 1992), it is

likely that Alex will continue to make at least modest gains compatible with the notion that language development derivesits impetus from the child’s need for communication andin IQ during the next 2–3 years. Already, however, the gains

he has made are striking, considering that, as late as age 9 social interaction (Locke, 1993), a need that Alex’s linguisticachievements have amply fulfilled.years, he still had a mental age of ~3 years.

Alex’s increase in mental age is not equal across all aspectsof intellectual function. On every type of mental assessment—the Griffiths Mental Developmental Scales, the Wechsler

Implications for the critical period hypothesisIntelligence Scale and the Wechsler Memory Scale—therewas a clear discrepancy between his scores on the subtestsand right hemisphere speech

Alex’s case is special in two respects. First, despite his veryof verbal and nonverbal skills. In each case, his verbalperformance approximated that of a normal 7- to 8.5-year- late start, his linguistic skills have developed thus far to the

level of an 8- to 10-year-old, his progress being more or lessold, ~1.5 years below the levels he has attained in receptiveand expressive language. By comparison, his nonverbal equivalent across all aspects of both expressive and receptive

language, including semantics, prosody, grammar andabilities have remained even more limited. The one exceptionis his visuo-perceptual ability (as assessed by the Motor Free phonology. This outcome is markedly different from that in

the two language-deprived individuals, Genie (Curtiss, 1977)Visual Perception Test and the Picture Completion subtest ofthe WISC-III), where he obtained scores with age equivalents and Chelsea (Curtiss, 1989), who first began to speak at the

ages of 13.5 and 32 years, respectively. Although both ofof 7:5 and 8:9. On all other nonverbal measures, hisperformance has risen only to the level of a 5- to 6-year- these young women eventually acquired sizeable vocabularies

and some knowledge of word-order, their phonology andold. This discrepancy in the development of verbal versusnonverbal abilities is consistent with the ‘crowding’ syntax have remained highly abnormal, resulting in speech

that is poorly enunciated and grammatically bizarre. Inhypothesis (Teuber, 1975), which proposes that when


contrast, Alex speaks not only clearly and fluently but also a low IQ and poor working memory. Hinging on the answeris whether the right cerebral hemisphere has exactly the samein grammatically correct sentences. Previously, on the basis of

evidence gained from studies of congenitally deaf individuals potential as the left for the development of speech andlanguage functionsper se.(Johnson and Newport, 1989), hemispherectomy patients

(Pattersonet al., 1989), second-language learners (Johnsonand Newport, 1989), and language-deprived children,including Isabelle (Skuse, 1988), a case could be made thatAcknowledgementscompetence in a language can develop fully only if learningWe wish to thank the following individuals for their generousis begun no later than ~6 years of age. The outcome in Alex,help with this investigation: Alex and his family for theirhowever, indicates that the upper limit for the start of effectiveunfailing patience and cooperation during the many hours oflearning can be raised byù3 years, to ~9 years. It is of testing; Dianne Gumley, Susan Oxbury and the late Elizabethinterest that while this age places the upper limit considerablyBeardsworth, who collected the data presented in Table 1;higher than the above estimate, it is still well within the critical Sue Carne for providing the longitudinal data presented inperiod postulated earlier by Lenneberg (1967). According toFigs 5 and 6; Katie Price for performing the ‘Sound Cluster’Lenneberg’s proposal, which was based on differences inanalysis; Val Muter for administering and analysing thelanguage learning under conditions of mental retardation inPhonological Awareness Test Battery; Kate Watkins andchildren versus adults, and on differences in recovery fromDeborah Christie for preparing Figs 5, 6 and 7; Johnlesion-induced aphasia in children versus adults, languageOxbury for his reports on Alex’s neurological status; Harrycan still develop fully provided that it is acquired or reacquiredChugani for the MRI and PET scans; John Stephens and thebefore the onset of puberty. Such a limit suggests, in turn,Human Neurophysiology Laboratory at University Collegethat the hormonal changes of puberty bring an end to theLondon, for their help in collecting and analysing thekind or the degree of cerebral plasticity that is needed forneurophysiological data; and the Bobath Centre for thethe organization or reorganization of a relatively normalTreatment of Children with Cerebral Palsy for the use ofspeech and language system. If this view is correct, thentheir magnetic stimulator. This work was supported in partAlex’s achievement was possible because the cerebral systemby MRC Grant No. 8901132 to F.V.-K., an MRC Trainingmediating his speech and language was still within theFellowship to L.J.C., and the Sir Clavering Fison Visitingprepubertal stage of plasticity when he began to learn them.Professor Award to M.M.This set of proposals does not preclude the possibility thatthe earlier the learning, the better the outcome (seebelow),with the degree of plasticity during the critical period having

Referencesthe form perhaps of a downward gradient to puberty (JohnsonAdams CBT. Hemispherectomy—a modification. J Neurol

and Newport, 1989) rather than a flat profile followed by aNeurosurg Psychiatry 1983; 46: 617–9.

sudden drop.Aicardi J. Diseases of the nervous system in childhood. London:The second way in which Alex’s case is special is that,MacKeith Press, 1992.unlike the congenitally deaf, the language deprived, or

second-language learners, he has achieved linguisticAnnett M. The Annett Peg Moving Test of Hand Skill. Hull:competence with an isolated right hemisphere. Despite thisUniversity of Hull, 1986.remarkable achievement, Alex clearly continues to have

Anthony A, Bogle D, Ingram TTS, McIsaac MW. Edinburghparticular difficulty with at least two aspects of speechArticulation Test. Edinburgh: Churchill Livingstone, 1971.and language. First, on the receptive side, he has poor

comprehension of complex verbal material; secondly, on theBarth TM, Stanfield BB. The recovery of forelimb-placing behaviorin rats with neonatal unilateral cortical damage involves theexpressive side, he is poor at manipulating phonemes andremaining hemisphere. J Neurosci 1990; 10: 3449–59.syllables, reflected as impairment in segmenting words,

repeating nonwords and in producing morphological markersBates E, MacWhinney B, Caselli C, Devescovi A, Natale F,with nonwords. We proposed above that these deficienciesVenza V. A cross-linguistic study of the development of sentenceare due not to his prolonged period of mutism but rather tointerpretation strategies. Child Dev 1984; 55: 341–54.his low IQ, including his limited memory span. If this

Bates E, Thal D, Janowsky JS. Early language development and itsproposal is correct, then inasmuch as low IQ and shortneural correlates. In: Boller F, Grafman J, editors. Handbook of

memory span appear to be common consequences of leftneuropsychology, Vol. 7. Amsterdam: Elsevier, 1992: 69–110.hemispherectomy in childhood, some of the derivative effects

Beardsworth ED, Adams CBT. Modified hemispherectomy forin all such cases will necessarily be poor comprehensionepilepsy: early results in 10 cases [see comment]. Br J Neurosurgof complex verbal material and difficulty in manipulating1988; 2: 73–84. Comment in Br J Neurosurg 1989; 3: 137.phonemes and syllables. An important question not answered

by our results is whether the linguistic costs of early leftBeery KE, Buktenica NA. Developmental Test of Visual-Motorhemisphere dysfunction in Alex, as well perhaps as in otherIntegration (VMI), 3rd Revision. Cleveland: Modern Curriculum

Press, 1989.left hemispherectomized children, can be ascribed entirely to


Bellugi U, Bihrle A, Neville H, Doherty S, Jernigan T. Language, motor and tactile-perceptual functions in normal children. PerceptMot Skills 1976; 42: 475–81.cognition, and brain organization in a neurodevelopmental disorder.

In: Gunnar MR, Nelson CA, editors. Developmental behaviouralFlynn JR. The mean IQ of Americans: massive gains 1932 to 1978.

neuroscience. The Minnesota Symposia on Child Psychology, Vol.Psychol Bull 1984; 95: 29–51.

24. Hillsdale (NJ): Lawrence Erlbaum, 1992: 201–32.Flynn JR. Massive IQ gains in 14 nations: what IQ tests really

Bellugi U, Wang PP, Jernigan TL. Williams syndrome: an unusualmeasure. Psychol Bull 1987; 101: 171–91.

neuropsychological profile. In: Broman S, Grafman J, editors.Atypical cognitive deficits in developmental disorders: implications Foix C, Chavany JA, Marie J. Diplegic facio-linguo-masticatricefor brain function. Hillsdale (NJ): Lawrence Erlbaum, 1994: 23–56.d’origine cortico sous-corticale sans paralysie des membres. Rev

Neurol 1926; 33: 214–9.Benecke R, Meyer B-U, Freund H-J. Reorganisation of descendingmotor pathways in patients after hemispherectomy and severeFuggle PW, Tokar S, Grant DB, Smith I. Rising IQ scores in Britishhemispheric lesions demonstrated by magnetic brain stimulation.children: recent evidence. J Child Psychol Psychiatry 1992; 33:Exp Brain Res 1991; 83: 419–26. 1241–7.

Berko J. The child’s learning of English morphology. Word 1958; Fusco L, Vigevano F. Reversible operculum syndrome caused14: 150–77. by progressive epilepsia partialis continua in a child with left

hemimegalencephaly. J Neurol Neurosurg Psychiatry 1991; 54:Bishop DVM. TROG: Test for Reception of Grammar. Abingdon

556–8.(Oxon): Medical Research Council, 1982.

Gadian DG, Isaacs EB, Cross JH, Connelly A, Jackson GD, KingBrown RW. A first language: the early stages. Cambridge (MA):

MD, et al. Lateralization of brain function in childhood revealedHarvard University Press, 1973.

by magnetic resonance spectroscopy. Neurology 1996; 46: 974–7.Bryant P, Goswami U. Phonological awareness and learning to read.

Gathercole SE, Baddeley AD. Evaluation of the role of phonologicalIn: Beech JR, Colley AM, editors. Cognitive approaches to reading.

STM in the development of vocabulary in children: a longitudinalChichester: John Wiley, 1987: 213–43.

study. J Mem Lang 1989; 28: 200–13.Carr LJ, Harrison LM, Evans AL, Stephens JA. Patterns of central

Gathercole SE, Baddeley AD. The role of phonological memory inmotor reorganization in hemiplegic cerebral palsy. Brain 1993; 116:

vocabulary acquisition: a study of young children learning new1223–47.

names. Br J Psychol 1990; 81: 439–54.Chapman RS, Miller JF. Word order in early two and three word

Gathercole SE, Willis C, Emslie H, Baddeley AD. The influencesutterances: does production precede comprehension? J Speech Hear

of number of syllables and wordlikeness on children’s repetition ofRes 1975; 18: 355–71.

nonwords. Appl Psycholing 1991; 12: 349–67.Cohen LG, Zeffiro T, Bookheimer S, Wassermann EM, Fuhr P,

Gazzaniga MS. Right hemisphere language following brainMatsumoto C, et al. Reorganization in motor pathways following a

bisection: a 20-year perspective. [Review]. Am Psychol 1983; 38:large congenital hemispheric lesion in man: different ipsilateral

525–37.motor representation areas for ipsi- and contralateral muscles. JPhysiol (Lond) 1991; 438: 33P. Gibson D, Ingram D. The onset of comprehension and production

in a language delayed child. Appl Psycholing 1983; 4: 359–75.Colarusso RP, Hammill DD. Motor-Free Visual Perception Test(MVPT). Novato (CA): Academic Therapy Publications, 1972. Goldin-Meadow S, Seligman MEP, Gelman R. Language in the

two-year-old. Cognition 1976; 4: 189–202.Curtiss S. Genie: a psycholinguistic study of a modern-day ‘WildChild’. New York: Academic Press, 1977. Graff-Radford NR, Bosch EP, Stears JC, Tranel D. Developmental

Foix-Chavany-Marie syndrome in identical twins. Ann Neurol 1986;Curtiss S. The independence and task specificity of language. In:

20: 632–5.Bornstein A, Bruner J, editors. Interaction in human development.Hillsdale (NJ): Lawrence Erlbaum, 1989: 105–37. Griffiths R. The abilities of young children. London: Child

Development Research Centre, 1970.De Renzi E, Vignolo LA. The Token Test: a sensitive test to detectreceptive disturbances in aphasics. Brain 1962; 85: 665–78. Hicks SP, D’Amato CJ. Motor-sensory and visual behavior after

hemispherectomy in newborn and mature rats. Exp Neurol 1970;de Villiers JG, de Villiers PA. Development of the use of word

29: 416–38.order in comprehension. J Psycholing Res 1973; 2: 331–41.

Homberg V, Stephen KM, Netz J. Transcranial stimulation of motorDijkerman HC, Jones S, Polkey C, Vargha-Khadem F, Weiskrantz

cortex in upper motor neurone syndrome: its relation to the motorL. Sensorimotor functioning after a hemispherectomy, a case study

deficit. Electroencephalogr Clin Neurophysiol 1991; 81: 377–88.[abstract]. Soc Neurosci Abstr 1993; 19: 616.

Huttenlocher J. The origins of language comprehension. In: SolsoDunn LM, Dunn LM, Whetton C, Pintilie D. British Picture

RL, editor. Theories in cognitive psychology. Potomac (Mont):Vocabulary Scale Windsor (Berkshire): NFER-Nelson, 1982.

Lawrence Erlbaum, 1974: 331–68.Elliott CD, Murray DJ, Pearson LS. The British Ability Scales.

Incisa della Rocchetta A, Gadian DG, Connelly A, Polkey CE,Revised ed. Windsor (Berkshire): NFER-Nelson, 1983.

Jackson GD, Watkins KE, et al. Verbal memory impairment afterright temporal surgery: role of contralateral damage as revealed byFinlayson MAJ, Reitan RM. Handedness in relation to measures of


1H magnetic resonance spectroscopy and T2 relaxometry. Neurology American Sign Language. Papers Reports Child Lang Dev 1984;23: 1–22.1995; 45: 797–802.

Isaacs EB, Vargha-Khadem F. Differential course of developmentNewport EL, Meier RP. The acquisition of American Sign Language.of spatial and verbal memory span: a normative study. Br J DevIn: Slobin DI, editor. The crosslinguistic study of languagePsychol 1989; 7: 377–80. acquisition, Vol.1: the data. Hillside (NJ): Lawrence Erlbaum, 1985:

881–938.Johnson JS, Newport EL. Critical period effects in second languagelearning: the influence of maturational state on the acquisition ofNoll JD, Lass NJ. Use of the Token Test with children: twoEnglish as a second language. Cognit Psychol 1989; 21: 60–99. contrasting socioeconomic groups. In: Boller F, Dennis M, editors.

Auditory comprehension: clinical and experimental studies with theKimura D, McGlone J. Children’s stories for testing LTM. In:

Token Test. New York: Academic Press, 1979: 125–32.Kimura D, McGlone J, editors. Neuropsychological Test Manual.London: DK Consultants, 1979. Oldfield RC, Wingfield A. Response latencies in naming objects.

Q J Exp Psychol 1965; 17: 273–81.Kuzniecky R, Andermann F, Tampieri D, Melanson D, Olivier A,Leppik I. Bilateral central macrogyria: epilepsy, pseudobulbar palsy,Oyama S. A sensitive period for the acquisition of a non-nativeand mental retardation – a recognizable neuronal migration disorderphonological system. J Psycholing Res 1976; 5: 261–83.[see comments]. Ann Neurol 1989; 25: 547–54. Comment in Ann

Oyama S. The sensitive period and comprehension of speech.Neurol 1990; 27: 693–4, comment in: Ann Neurol 1990; 28: 840–2.Working Papers Bilingualism 1978; 16: 1–17.

Lafayette Instrument Company. Instructions for the 78010 HandPaetau R, Kajola M, Korkman M, Hamalainen M, Granstrom ML,Dynamometer. Lafayette: Indiana, 1989.Hari R. Landau-Kleffner syndrome: epileptic activity in the auditory

Landau WM, Kleffner FR. Syndrome of acquired aphasia with cortex. Neuroreport 1991; 2: 201–4.convulsive disorder in children. Neurology 1957; 7: 523–30.

Patterson K, Vargha-Khadem F. Neuropsychological observationsLandau WM, Goldstein R, Kleffner FR. Congenital aphasia: aon the affinity between reading and phonological abilities. Mindclinicopathologic study. Neurology 1960; 10: 915–21. Lang 1991; 6: 140–5.

Lenneberg EH. Biological foundations of language. New York: JohnPatterson K, Vargha-Khadem F, Polkey CE. Reading with oneWiley, 1967. hemisphere. Brain 1989; 112: 39–63.

Lesser R. Verbal and non-verbal memory components in the TokenPetri H, Mishkin M. Behaviorism, cognitivism and theTest. Neuropsychologia 1976; 14: 79–85. neuropsychology of memory. Am Sci 1994; 82: 30–7.

Lezak MD. Neuropsychological assessment. 2nd Ed. New York:Rasmussen T. Further observations on the syndrome of chronicOxford University Press, 1983. encephalitis and epilepsy. Appl Neurophysiol 1978; 41: 1–12.

Lieberman P. Could an autonomous syntax module have evolved?Rasmussen T. Hemispherectomy for seizures revisited. Can J Neurol[Review]. Brain Lang 1992; 43: 768–74. Sci 1983; 10: 71–8.

Lindsay J, Ounstead C, Richards P. Hemispherectomy for childhoodRasmussen T, Milner B. The role of early left-brain injury inepilepsy: a 36 year study. Dev Med Child Neurol 1987; 29: 592–600.determining lateralization of cerebral speech functions. Ann NY

Acad Sci 1977; 299: 355–69.Locke JL. The child’s path to spoken language. Cambridge (MA):Harvard University Press, 1993. Renfrew CE. The Bus Story. Old Headington (Oxford): CE

Renfrew, 1969.Meier RP. Language acquisition by deaf children. Am Sci 1991;79: 60–70. Renfrew CE. Action Picture Test. 2nd Ed. Old Headington (Oxford):

CE Renfrew, 1971.Menkes JH. Textbook of child neurology. 2nd Ed. Philadelphia:Lea and Febiger, 1980. Renfrew CE. The Word Finding Vocabulary Test. Old Headington

(Oxford): CE Renfrew, 1977.Milner B. Psychological aspects of focal epilepsy and itsneurosurgical management. Adv Neurol 1975; 8: 299–321.

Reynell JK. Reynell Developmental Language Scales. Windsor(Berkshire): NFER-Nelson, 1977.Morrell F, Whisler WW, Bleck TP. Multiple subpial transection: a

new approach to the surgical treatment of focal epilepsy [seeRoberts K. Comprehension and production of word order in Stage

comment]. J Neurosurg 1989; 70: 231–39. Comment in: J NeurosurgI. Child Dev 1983; 54: 443–9.

1989; 71:629–30.Savage-Rumbaugh S, McDonald K, Sevcik RA, Hopkins WD,

Morrell F, Whisler WW, Smith MC, Hoeppner TJ, deToledo-MorrellRubert E. Spontaneous symbol acquisition and communicative use

L, Pierre-Louis SJC, et al. Landau-Kleffner syndrome: treatmentby pygmy chimpanzees (Pan paniscus). J Exp Psychol Gen 1986;

with subpial intracortical transection. Brain 1995; 118: 1529–46.115: 211–35.

Muter V. Phonology and learning to read in normal and hemiplegicSavage-Rumbaugh ES, Murphy J, Sevcik RA, Brakke KE, Williamschildren [PhD Thesis]. University of London, 1994.SL, Rumbaugh DM. Language comprehension in ape and man.Monogr Soc Res Child Dev 1993; 58: 3–4.Newport EL. Constraints on learning: studies in the acquisition of


Sears TA, Stagg D. Short-term synchronization of intercostal Vargha-Khadem F, Isaacs EB, Papaleloudi H, Polkey CE, WilsonJ. Development of language in six hemispherectomized patients.motoneurone activity. J Physiol (Lond) 1976; 263: 357–81.Brain 1991; 114: 473–95.

Semel E, Wiig EH, Secord W. CELF-R: Clinical Evaluation ofVargha-Khadem F, Isaacs EB, Van der Wert S, Robb S, Wilson J.Language Functions – Revised. London: PsychologicalDevelopment of intelligence and memory in children withCorporation, 1987.hemiplegic cerebral palsy: the deleterious consequences of early

Skuse DH. Extreme deprivation in early childhood. In: Bishop seizures. Brain 1992; 115: 315–29.D, Mogford K, editors. Language development in exceptional

Vargha-Khadem F, Watkins K, Alcock K, Fletcher P, Passinghamcircumstances. Edinburgh: Churchill Livingstone, 1988: 29–46.R. Praxic and nonverbal cognitive deficits in a large family with a

Snowling M, Stothard S, McLean J. Graded Nonword Reading Test.genetically transmitted speech and language disorder. Proc NatlBury St Edmunds (Suffolk): Thames Valley Test Company, 1996. Acad Sci USA 1995; 92: 930–3.

Verity CM, Strauss EH, Moyes PD, Wada JA, Dunn HG, LapointeSnyder L, Bates E, Bretherton I. Content and context in early lexicalJS. Long-term follow-up after cerebral hemispherectomy:development. J Child Lang 1981; 8: 565–82.neurophysiologic, radiologic, and psychological findings. Neurology

Talley JL. Children’s Auditory Verbal Learning Test – 2. Odessa 1982; 32: 629–39.(FL): Psychological Assessment Resources, 1993.

Wechsler D. Wechsler Intelligence Scale for Children—Revised.Tatum WO, Coker SB, Ghobrial M, Abd-Allah S. The open Windsor (Berkshire): NFER-Nelson, 1976.opercular sign: diagnosis and significance [see comment]. Ann

Wechsler D. Wechsler Intelligence Scale for Children. 3rd edition.Neurol 1989; 25: 196–9. Comment in: Ann Neurol 1990; 27: 213–4.

Sidcup (Kent): Psychological Corporation, 1992.Taylor HG, Lean D, Schwartz S. Pseudo-word repetition ability inWechsler D, Stone CP. Wechsler Memory Scale. San Antonio (TX):learning-disabled children. Appl Psycholing 1989; 10: 203–19. Psychological Corporation, 1945.

Teuber H-L. Recovery of function after brain injury in man. In: Wechsler D. Wechsler Objective Reading Dimensions (WORD).Porter R, Fitzsimmons DW, editors. Outcome of severe damage toSidcup (Kent): Psychological Corporation, 1993.the central nervous system. CIBA Foundation Symposium 34.

Williams JCP, Barratt-Boyes BG, Lowe JB. Supravalvular aorticAmsterdam: Elsevier, 1975: 159–90.stenosis. Circulation 1961; 24: 1311–8.

Vargha-Khadem F, Isaacs E. Speech onset in boy aged 10 followingZaidel E. The split and half brains as models of congenital

left hemispherectomy and withdrawal of anticonvulsants [abstract].language disability. In: Ludlow CL, Doran-Quine ME, editors. The

Soc Neurosci Abs 1992; 18: 1214.neurological bases of language disorders in children: methods anddirections for research. NINCDS Monograph No. 22. BethesdaVargha-Khadem F, Polkey CE. A review of cognitive outcome after(MD): U.S. Department of Health, Education, and Welfare, Publichemidecortication in humans. In: Rose FD, Johnson DA, editors.Health Service, National Institute of Neurological andRecovery from brain damage: reflections and directions. New York:Communicative Disorders and Stroke, 1980: 55–86.Plenum Press, 1992: 137–51.

Vargha-Khadem F, Watters GV, O’Gorman AM. Development ofspeech and language following bilateral frontal lesions. Brain LangReceived December 7, 1995, Revised April 25, 1996.

Accepted August 23, 19961985; 25: 167–83.


Appendix 1Words acquired by Alex in the first 3 months following onset of speech, i.e. between the ages of 9:4and 9:7 (from his mother’s diary)head toes cold there moon himhand toothpaste sun not book mousearm(s) water it no bus miceear(s) hair is yes dog morenose eye on do boat manmouth hide in two mat toy(s)tooth hear at off ball meatteeth hurt out light dice nicefoot hot here light off milk

Appendix 2(a) BUS STORY Immediate recallOnce upon a time there was a very naughty bus. While his Once on up a time... there was a nice bus and he decided todriver was trying to mend him, he decided to run away. run away. He made funny faces at each other. He wentHe ran along the road beside a train. They made funny faces through the country knocking down the people and theat each other and raced each other. But the bus had to go on policeman blewed his whistle so loud. He couldn’t stop andalone, because the train went into a tunnel. He hurried into he rode through the countryside and he jumped over thethe city where he met a policeman who blew his whistle and fence and rushed through the fields. And the cow said ‘moo’shouted, ‘Stop, bus’. and he said ‘good gracious, goodness me’ and he wentBut he paid no attention and ran on into the country. He roaring down the hill. He fell into the water with a bigsaid, ‘I’m tired of going on the road’. So he jumped over a splash. The end.fence. He met a cow who said ‘Moo, I can’t believe my eyes’. Age equivalent for recall: 7 years.The bus raced down the hill. As soon as he saw there waswater at the bottom, he tried to stop. But he didn’t knowhow to put on his brakes. So he fell in the pond with a splashand stuck in the mud. When his driver found where he was,he telephoned for a crane to pull him out and put him backon the road again.

(b) TAYLOR STORY I Immediate recallA hen had six little yellow chickens. One morning she took There was a hen, chicks and she took some... them for athem for a walk to look for something to eat. They found walk. A dog came to play with them. The hen didn’t likesome seeds and sand. A dog came to play with them. The the dog and she flew at him and made him run away. A duckhen did not like the dog so she flew at him and made him and a bus... about a hen.run away. Age equivalent for recall:,6 years

Delayed recallThe chickens... the mother hen took them for a walk to findsomething to eat. They found some seeds but a dog cameand do you know what she did? She flew at him andfrightened him away.Age equivalent for recall:,6 years.

TAYLOR STORY II Immediate recallThree boys built a house in the woods. They put a table and Three boys left the door opened. When they came back theytwo old chairs in it. There was a basket full of apples under found two little pigs eating the apples. One afternoon theythe table. One afternoon they went away and left the door went for a walk. When they came back they saw three pigsopen. When they come back they found two little pigs eating eating the apple.the apples. Age equivalent for recall:,6 years.

Delayed recallOne afternoon three boys was going for a walk... for thewoods. Suddenly they came back... they found three littlepigs eating the apples.Age equivalent for recall:,6 years.


Appendix 3 IQ subtests: raw score, scaled score (age equivalent)

Alex’s chronological age (years:months)

(11:0) (11:9) (14:2) (14:11)

Verbal testsInformation 4, 1 (,6:2) 5, 1 (6:2) 9, 1 (7:6) 12, 4 (9:2)Similarities 6, 3 (6:2) 4, 2 (,6:2) 9, 2 (6:10) 13, 6 (9:6)Arithmetic 0, 1 (,6:2) 5, 2 (6:2) 9, 1 (,6:2) 10, 1 (,6:2)Vocabulary 13, 1 (6:2) 14, 1 (6:2) 19, 2 (7:6) 19, 2(7:6)Comprehension 7, 3 (6:2) 7, 3 (6:2) 8, 1 (,6:2) 9, 1 (,6:2)Digit span 2, 1 (,6:2) 2, 1 (,6:2) 7, 2 (,6:2) 8, 2 (,6:2)

Performance testsPicture completion 6, 1 (,6:2) 12, 5 (6:6) 16, 4 (8:2) 17, 5 (8:6)Picture arrangement 2, 1 (,6:2) –, – (–) 9, 2 (6:2) 9, 2(6:2)Block design 0, 1 (,6:2) 0, 1 (,6:2) 6, 1 (,6:2) 10, 1 (,6:2)Object assembly –, – (–) –, – (–) 9, 1 (,6:2) 16, 2 (6:2)Coding 11, 1 (,6:2) –, – (–) 17, 1 (,6:2) 17, 1 (,6:2)Mazes –, – (–) 6, 1 (,6:2) –, – (–) –, – (–)Symbol search –, – (–) –, – (–) 15, 3 (6:2) 14, 2 (,6:2)

brain onset of speech after left hemispherectomy in a nine ... · when examined at age 11:5 by a...

Documents