Brain 1996 Caplan 933 49

Brain (1996), 119, 933-949
Location of lesions in stroke patients with deficits in syntactic processing in sentence comprehension
David Caplan, Nancy Hildebrandt and Nikos Makris
Neuropsychology Laboratory, Massachusetts General Hospital, Boston, USA Correspondence to: David Caplan, MD, Neuropsychology Laboratory; Vincent Burnham 827, Fruit Street, Boston, MA 02114, USA
Summary
Sixty patients, 46 with left-hemisphere strokes and 14 with right-hemisphere strokes, and 21 normal control subjects were tested for the ability to use syntactic structures to determine the meaning of sentences. Patients enacted thematic roles (the agent, recipient and goal of an action) in 12 examples of each of 25 sentence types, which were designed to test a wide variety of syntactic operations. Both right- and left-hemisphere damaged patients performed worse than control subjects on syntactically complex sentences, and left-hemisphere patients performed worse than righthemisphere patients. Eighteen patients with left-hemisphere strokes underwent CT scanning to image the perisylvian association cortex. There was no difference between the performance of patients with anterior and posterior lesions, and no correlation between the degree of impairment and the size of lesions in different regions of the perisylvian cortex. These results are consistent with the view that syntactic processing involves an extensive neural system, whose most important region is the left perisylvian cortex. When these results are combined with those of other studies, the picture that emerges is one in which, within this cortical region, this system manifests features of both distributed and localized processing.
Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012
Keywords: localization of language functions; syntactic comprehension deficits; localization of syntactic processing; syntactic comprehension in stroke patients Abbreviations: CVA = cerebrovascular accident (stroke); rCBF = regional cerebral blood flow
Introduction
Sentences are the level of the language code at which the meanings of individual words are related to each other to express information about events and states in the world (Jackson, 1874). This information indicates who is doing what to whom (thematic roles), which adjectives are associated with which nouns (attribution of modification), what pronouns and other 'referentially dependent' items are related to (coreference), and other similar semantic information. The ability to express this propositional information contributes in an important way to the power that human language has as a vehicle for thought and communication. The propositional content of a sentence is determined by the syntactic structure of that sentence (Chomsky, 1986). Individual words are assigned to different syntactic categories (e.g. noun, verb, preposition). These categories are organized into hierarchical structures (e.g. noun phrase, verb phrase) in which particular phrases stand in specific relationships to one another (e.g. subject of the verb, object of a preposition). Oxford University Press 1996 Propositional meaning is determined by these relationships. For instance, in the sentence 'The dog that scratched the cat chased the bird', readers understand that 'the dog' is the agent of 'chased', despite a sequence of words'the cat chased the bird'that in other circumstances could be taken to express a proposition. The sentence is understood this way because of the position of the words 'the cat' and 'the dog' in the syntactic structure of the sentence: 'the dog' is the subject of 'chased' and 'the cat' is the object of 'scratched' and has no syntactic relationship to 'chased' (Fig. 1). Different aspects of syntactic structure determine different aspects of meaning. In the sentence discussed above, the syntactic relationships of subject and object determine the thematic roles played by noun phrases. In other sentences, different syntactic relationships determine other aspects of meaning, such as what a pronoun (e.g. 'him') or a reflexive (e.g. 'himself') refers to, which items are modified by an adjective, etc. (Chomsky, 1986). For instance, in the sentence
934
D. Caplan et al. distinctly human cognitive function, whose neural basis is therefore of considerable neurobiological significance (Pinker. 1994). However, the location of the syntactic processors that operate during sentence comprehension remains unclear. It has been suggested that the ability to process syntactic structure in sentence comprehension is carried out in a neural net based in the left-hemisphere, whose most active portion is Broca's area and adjacent parts of the frontal language zone(Mesulam, 1990; Damasio, 1992; Zurif et al., 1993). Evidence supporting this localization comes from the fact that a significant number of patients with Broca's aphasia have difficulty understanding sentences in which syntactic structure must be used to determine meaning (Caramazza and Zurif, 1976; Schwartz et al., 1980; Caplan and Futter, 1986). One study of regional cerebral blood flow (rCBF) using I5O-PET has shown a localized increase in rCBF in part of Broca's area (the pars opercularis) when subjects made acceptability judgements for syntactically more complex compared with syntactically less complex sentences (Stromswold et al., 1996). Studies of event-related potentials have also identified an early negative wave in the left frontal region associated with aspects of syntactic processing (Neville et al., 1991; Kluender and Kutas, 1993). However, this evidence does not clearly settle the issue of how the brain is organized for syntactic processing for several reasons. First, other data imply that Broca's area is not the only brain region in which syntactic processing occurs. Many patients with lesions outside Broca's area have been described with syntactic comprehension disorders (Seines et al., 1983; Caplan et al., 1985; Caplan and Hildebrandt, 1988; Tramo et al., 1988). Secondly, there is evidence that Broca's area is not needed for syntactic processing. Many agrammatic patients with Broca's aphasia demonstrate sensitivity to grammatical structure in grammaticality judgement and other tasks (Linebarger et al., 1983; Tyler, 1985), suggesting that they can assign the syntactic structure of a sentence even if they cannot use it to determine sentence meaning. Other agrammatic patients with Broca's aphasia have shown intact syntactic processing in sentence-picture matching and enactment tasks (Miceli et al., 1983; Caplan et al., 1985; Nespoulous et al., 1988; R. Berndt, C. Mitchum and A Haendiges, unpublished data). Thirdly, there are limitations to the database supporting localization of syntactic processing in Broca's area. Most of the investigators who have documented syntactic processing impairments in agrammatic patients have not reported specific aspects of lesions, and it is known that lesions in patients with Broca's aphasia often extend well beyond Broca's area (Mohr et al., 1978; Vanier and Caplan, 1990). The results of another, less well-controlled PET study implicated regions other than Broca's area in syntactic processing (Mazoyer et al., 1993). Results of eventrelated potential studies have also suggested that a more posterior wave (the P600 or SPS) is related to aspects of syntactic processing (Neville et al., 1991; Hagoort et al., 1993). Some studies that report this wave have found it to be maximal in amplitude over the right hemisphere (Osterhout
NP COMP
/ \
Det
II
N bird
The dog that 0 scratched the cat chased
Fig. 1 Diagram of the syntactic structure of the sentence The dog that scratched the cat chased the bird', illustrating the hierarchical organization of categories that determines the fact that 'the dog' is the agent of 'chased'.
'Mary's picture of her intrigued Susan', 'her' cannot refer to 'Mary', because of the syntactic relationship between 'her' and 'Mary', while in the sentence 'Mary's picture of herself intrigued Susan', 'herself can only refer to 'Mary', because of this syntactic relationship. The syntactic relationship between 'her' or 'herself and 'Mary' that determines whether they can be related is known as 'c-command' (Reinhart, 1983) and is different from that which relates 'the dog' and 'chased' in 'The dog that scratched the cat chased the bird'. Most researchers believe that determining the meaning of a sentence requires the assignment of a syntactic structure (parsing) and the use of that syntactic structure in conjunction with the meanings of the words in the sentence to determine the meaning of the sentence (sentence interpretation). Parsing and sentence interpretation are thought to involve a number of processes and operations that are specific to the construction of the particular syntactic relationships that determine different aspects of meaning (Frazier, 1987a, b, 1989, 1990; for an alternative view, see MacDonald, 1994). In addition, parsing and sentence interpretation are thought to require a processing resource system, whose size affects the efficiency and even the feasibility of assigning a syntactic structure and understanding a sentence (Just and Carpenter, 1992). As an illustration of this resource system, consider the sentence 'The man that the woman that the child hugged kissed laughed'. Most readers cannot assign the thematic roles in this sentence, though they can do so relatively easily in the two sentences that combine to form it'The man that the woman kissed laughed', and 'The woman that the child hugged kissed the man'. The trouble subjects have understanding the sentence 'The man that the woman that the child hugged kissed laughed' is thought to arise because they do not have sufficient working memory capacity to maintain the intermediate products of computation they generate in mind while processing the incoming words in this complex structure. Syntactic processing is an important candidate for a
Lesions in syntactic comprehension deficits and Holcomb, 1992, 1993), raising the question of whether event-related potentials have the necessary spatial resolution to be definitive in determining the neural sites of language processing. For these reasons, the functional neuroanatomy of syntactic processing and the role that Broca's area plays in this function remain unsettled areas. In this study, we report on 60 stroke patients in whom syntactic comprehension deficits were well-characterized behaviourally. In 18 patients, left-hemisphere lesions were visualized by CT scanning. The results provide data relevant to the functional neuroanatomy of syntactic processing during sentence comprehension.
935
were included as a benchmark for the patients' performance on the comprehension task. All patients or their spouses and the normal subjects gave their informed consent to participate in the study, which had the approval of the local ethical committee.
Materials
Twelve sentences of each of twenty-five sentence types were presented (see Appendix). Subsets of these sentences have previously been used to test comprehension in stroke patients (Caplan et al., 1985; Caplan and Hildebrandt, 1988), patients with closed head injury (Butler-Hinz et al., 1990), and patients with dementia of the Alzheimer's type (Rochon et al., 1994). Six sentence types contained only 'full' noun phrases, which are noun phrases like 'the dog' or 'the cat' that refer directly to items in the real world, and six sentence types contained pronouns or reflexives ('himself or 'him'), which have to be related to another noun phrase in order to make reference to an item in the world. The remaining 13 sentence types contained what are known as 'empty noun phrases' (Chomsky, 1986)items such as the understood subject of 'to jump' in the sentence 'John promised Bill to jump' or the object of 'scratched' in the sentence 'The dog that the cat scratched chased the mouse'. To require that subjects structure these sentences syntactically and not simply rely on real-world knowledge to determine the correct meaning of these sentences, all sentences were constructed such that any noun could have accomplished or been the recipient of the action of any verb and could have been referred to by any pronoun, reflexive or empty noun phrase in the sentence. Thus, the sentences were structured so as to assess a subject's ability to process a wide range of syntactic structures in sentence comprehension.
Method
Patients
Forty-six patients (29 English subjects and 17 French) with left-hemisphere vascular lesions and 14 patients (nine English and five French) with right-hemisphere vascular lesions participated in this study. Patients were recruited from hospital and rehabilitation facilities in the Montreal area. The 29 English left-hemisphere patients consisted of 15 males and 14 females, aged 20-88 (mean 63) years. The 17 French left-hemisphere patients included 10 males and seven females aged 38-70 (mean 57) years. The level of education for both the English and French subjects ranged from grade school through college. Subjects were classified as being right-handed or having anomalous dominance (Geschwind and Galaburda, 1985) based on questions drawn from the Edinburgh Handedness Scale that were answered by the patient, the patient's spouse, or another close informant. All except three English subjects and one French subject were right-handed. CT scans were obtained for nine English and nine French patients. The nine English patients for whom CT scans were obtained included three male and six female patients, eight right-handed and one ambidextrous, with a mean age of 60 years. The nine French patients for whom CT scans were obtained included seven male and two female patients, all of them right-handed, with a mean age of 52 years. The nine English right-hemisphere patients consisted of four males and five females, aged 48-86 (mean 65) years. The five French right-hemisphere patients included two males and three females aged 27-79 (mean 54) years. Level of education for the both English and French subjects ranged from grade school through college. All except one English and two French subjects were right-handed.
Procedure
Sentences were divided into three batteries. The first battery contained active, passive, dative and relative-clause sentence types. The second battery contained sentence types with one proposition and a reflexive or a pronoun, and matched sentences with full noun phrases. The third battery contained sentence types with two propositions and either full, reflexive, pronoun or empty noun phrases. Nouns in the first battery were animal names, and nouns in the second and third batteries were either a definite concrete noun phrase ('the old man', 'the boy') or a relational noun phrase ('the father', 'his friend'). The three batteries were given in the same sequential order, with the first battery first and the third last, with a training period before each. Subjects were tested individually in testing rooms at the hospitals or rehabilitation facilities or in their homes. At the onset of each session, the experimenter indicated the names of the objects (animals or dolls) to each patient and then tested the patient's ability to identify these objects one at a
Normal subjects
Twenty-one normal subjects (11 English and 10 French) were also tested on a subset of 22 sentences types. The 11 English control subjects were aged 16-76 (mean 59) years. The 10 French control subjects were aged 52-73 (mean 64) years. Handedness was not recorded in the control subjects, who
936
D. Caplan et al.
AC-PC
Fig. 2 Lateral view of the left-cerebral hemisphere in the human showing the cortical regions of interest considered in this study, and cerebral sulci. The shadowed area with the asterisk in it is not included in region of interest Tl. F3t = inferior frontal gyrus/pars triangularis; F3o = inferior frontal gyrus/pars opercularis; SG = supramarginal gyrus; AG = angular gyrus; Tl = superior temporal gyrus; ce = central sulcus; prc = precentral sulcus; sf = superior frontal sulcus; if = inferior frontal sulcus; aar = anterior ascending ramus of the sylvian fissure; ahr = anterior horizontal ramus of the sylvian fissure; phr = posterior horizontal ramus of the sylvian fissure; par = posterior ascending ramus of the sylvian fissure; st = superior temporal sulcus; it = inferior temporal sulcus; poc = postcentral sulcus; ip = intraparietal sulcus; im = intermediate sulcus of Jensen; ag = angular sulcus; ao = anterior occipital sulcus; lo = lateral occipital sulcus. (Modified from Rademacher et al., 1992.) time and in series. Patients who could not reliably point to all objects in the set were excluded from further testing. Subjects were told that the purpose of the experiment was to test their abilities to understand 'who did what to whom' in the sentences. Subjects were instructed to indicate 'who did what to whom' by acting out the sentence using the items provided. The experimenter emphasized that subjects did not need to show details of the action of the verb, but had to clearly demonstrate which item was accomplishing the action and which receiving it. Practice sessions were given for each battery, during which some easy and some difficult sentence types were presented. During these practice sessions, the experimenter did not correct errors that a patient made, but did ask for repetitions and revisions of responses in which it was not clear which item initiated and which item received an action. Practice continued until the patient's actions could be clearly interpreted. The experimenter then read each experimental sentence with a normal, neutral intonational contour and recorded the subject's response (for details of the task, see Caplan et al., 1985; Caplan and Hildebrandt, 1988).
Neuroimaging
Eighteen patients with left-hemisphere strokes underwent CT scanning. Scans were performed from 7 days to 7 years after the onset of the stroke (one on day seven, two on day eight, and the remainder from 3 months to 7 years after the onset of the stroke). In 17 subjects, a special protocol was used to obtain CT images. Scans were supervised by a neuroradiologist. The subject was carefully positioned so that the imaging plane was parallel to the canthomeatal (CM) line, which runs almost parallel to the bicomissural (ACPC) line (Tokunaga et al., 1977; Fox et al., 1986). A Scout film was obtained with a radio-opaque marker on the skull perpendicular to the canthomeatal line at the point of the
external auditory meatus, which corresponds to the position of the posterior commissure. This marker was visible as a white dot on the left side of the head in all CT images, and served to help verify the angle of the scan (Vanier et al., 1985). The brain in its entire width was imaged in series of single slices of 5 mm width in 14 patients and of 10 mm width in four patients. These 17 scans were mapped onto the Talairach and Tournoux (1988) atlas, whose templates are parallel to the bicomissural line. The anatomical regions defined in the Talairach and Tournoux atlas correspond roughly to cytoarchitectonic fields (Sanides, 1964; Rademacher et al., 1993), and the atlas has been the basis for localization of changes in rCBF and regional cerebral blood volume in activation studies (Fox et al., 1985; Belliveau et al., 1991; Fox and Lancaster, 1993). The remaining scan, which was obtained at a different angle, was matched to the templates corresponding to its angle of imaging in the Damasio and Damasio (1989) atlas. The templates of the Damasio and Damasio atlas were normalized to the Talairach dimensions so that volumes of lesions and regions of interest would be comparable across the 18 scans. Film images were traced on transparencies and magnified to match templates in the appropriate atlas. One magnification factor for each brain was used, which was the ratio of the maximum longitudinal axis of a selected CT scan slice to the longitudinal axis of its corresponding template. On a slice per slice basis, guided by key landmarks (mainly the sylvian fissure and the hemispheric margins), the surface of the lesion was matched to the surface of the atlas template. This matching required small amounts of spatial stretching, rotation and translation. Volumetric analysis of each lesion was performed by measuring the surface occupied by the lesion in each normalized CT scan slice, multiplying it by the thickness of the corresponding template, and summing
Lesions in syntactic comprehension deficits Table 1 Mean percentage correct for each sentence type and each subject group
LCVA (n = 46) Baseline sentences without referentially dependent noun phrases Two-place active Three-place active Conjoined Active conjoined theme Three reflexive expressions Simple active reflexive expressions Sentences with overt referentially dependent noun phrases Reflexives, simple noun phrase Pronouns, simple noun phrases Simple active reflex Simple active reflex (friend of X subj) Simple active pronoun (friend of X subj) Simple active pronoun Sentences with empty referentially dependent noun phrases Two-place passive Truncated passive Two-place cleft object Three-place passive Three-place cleft object Subject-object relative Object-subject relative Object-object relative Subject-subject relative Passive conjoined agent Object control Subject control Noun phrase-raising/pass object control RCVA (n = 14)
937
CTRL (n = 23)
92 77 69 81 68 62 74 63 88 63 66 57 70 67 70 46 54 37 49 42 60 65 69 49 52
95 83 91 97 88 87 88 86 96 88 91 86 89 90 88 73 67 55 70 69 83 91 77 68 61
100 100 99 * 97 78 99 98 78 77 78 78 100 * 99 96 96 88 95 97 94 * 98 93 96

See Appendix for descriptions of sentence types. LCVA = left CVA patient; RCVA = right CVA patient; CTRL = control subject; * = not tested on this sentence type.
these values across the templates in which the lesion appeared. Five anatomical regions of interest that correspond to brain structures within the left-hemisphere language zone were defined on the Talairach and Tournoux atlas. The regions were defined following the criteria described by Rademacher et al. (1992), which rely primarily on the morphology of the cerebral sulci. Lesion size in each region of interest was calculated as described above, and expressed as a percentage of the volume of each region of interest. The five regions of interest in which we performed volumetric analyses are shown in Fig. 2, and were defined as follows. Region of interest 1. This is a region in the superior temporal gyms, which corresponds to Brodmann's cytoarchitectonic area 22. It is defined posteriorly by a coronal plane that passes through the dorsal end of the posterior ascending ramus of the sylvian fissure, ventrally by the superior temporal sulcus, anteriorly by a coronal plane passing through the posterior end of the temporal pole, and dorsally by the posterior horizontal ramus of the sylvian fissure. The cortex corresponding to Heschl's gyrus and the central portion of the planum temporale (areas 41 and 42) are excluded from this region of interest. Region of interest 2. This region corresponds to pars
triangularis and the portion of frontal operculum that underlies the pars triangularis. It is part of the inferior frontal gyrus and represents Brodmann's cytoarchitectonic area 45. It is defined posteriorly by the anterior ascending ramus of the sylvian fissure, ventrally by the anterior horizontal ramus of the sylvian fissure, anteriorly by the coronal plane that passes through the rostral end of the anterior horizontal ramus, and dorsally by the inferior frontal sulcus. Region of interest 3. This includes the pars opercularis and the part of frontal operculum that underlies the pars opercularis. It is is also part of the inferior frontal gyrus. It corresponds to Brodmann's cytoarchitectonic area 44. It is defined posteriorly by the precentral sulcus, ventrally by the posterior horizontal ramus of the sylvian fissure, anteriorly by the anterior ascending ramus of the sylvian fissure, and dorsally by the inferior frontal sulcus. Region of interest 4. This refers to the angular gyrus and corresponds to Brodmann's cytoarchitectonic area 39. A coronal plane that passes through the caudal end of the anterior occipital sulcus is its posterior border, and the anterior occipital sulcus together with the superior temporal sulcus form its ventral border. A coronal plane that passes through the inferior tip of the intermediate sulcus of Jensen
938
D. Caplan et al. A point to note is that the English patients with (L)CVAs performed better than their French counterparts, and the difference between them and the English (R)CVA patients did not reach statistical significance. However, in other studies using the same task, English and French aphasic patients have performed at the same levels (Caplan et al., 1985), and the English (L)CVA patients' level of performance was lower than that of the (R)CVA patients in the present study. Further analyses {see below) indicate that the English (L)CVA patients showed the same impairment in syntactic processing as the French (L)CVA patients. The somewhat better than expected overall performance of the English (L)CVA patients who were tested in this study is therefore probably an atypical feature of the (L)CVA patients in this sample. We caution against concluding from this feature of this group's performance that (L)CVA patients in general perform at the same level as (R)CVA patients on this task. There are many factors that enter into the performance of this task, and that could have contributed to lowered performance in the patient groups. To investigate whether specifically syntactic aspects of sentence processing were affected by lesions in either hemisphere, we compared performance on sentences that were more syntactically complex with that on matched sentence types that were syntactically less complex. Syntactic complexity was determined in the following manner. In most English sentences, the subject noun phrase precedes the verb, and the verb is followed by an object and then by one or more prepositional phrases. The noun phrases in subject and object position are usually assigned the thematic roles of agent (the perpetrator of an action) and theme (the person or item upon whom the action is enacted), respectively. The noun phrases in the prepositional phrases play other thematic roles, such as goal, beneficiary, etc., depending upon the preposition that is present. This order of thematic rolesagent-theme-goal (or other)is known as the canonical thematic role order for English. It has been shown that sentences with empty noun phrases are more difficult than sentences with either full noun phrases or pronouns or reflexives when the order of thematic roles in the sentence deviates from the canonical agent-theme-goal order (Caplan et al., 1985; Schwartz et al., 1987; Caplan and Hildebrandt, 1988). This complexity is attributable to processing the syntactic structure of these sentences, as opposed to their length or other factors. Thus, to test specifically syntactic aspects of sentence processing, seven sentence types with empty noun phrases in which thematic roles occurred in non-canonical order were matched for number of nouns, thematic roles and propositions with four sentences in which thematic roles appeared in a canonical order (see Appendix). For example, the sentence 'The monkey that the cow hit pushed the goat' was compared with the sentence 'The cow hit the monkey and pushed the goat'. Comparison of performance on these sentences with noncanonical thematic role order to performance on these
constitutes its anterior limit, and the intraparietal sulcus defines the dorsal border of the angular gyrus. Region of interest 5. This corresponds to the supramarginal gyrus and the parietal operculum, and represents Brodmann's cytoarchitectonic area 40. Its caudal border is a coronal plane passing through the inferior end of the intermediate sulcus of Jensen, its ventral limit is defined by the posterior horizontal and posterior ascending rami of the sylvian fissure, and the superior temporal sulcus. Anteriorly, the postcentral sulcus is its border, and dorsally it is delimited by the intraparietal sulcus. Because of concerns related to the accuracy of the parcellation of the Talairach and Tournoux atlas according to the Rademacher et al. (1992) criteria, a grosser parcellation system was also used. This parcellation grouped together regions anterior to the pre-central sulcus (regions of interest 2 and 3) into a single 'anterior' region of interest. This region corresponded to the traditional Broca's area (Brodmann's areas 44 and 45). A second region of interest was formed by combining regions of interest 1, 4 and 5 into a single 'posterior' region that included perisylvian association cortex posterior to the post-central gyrus. Finally, all five regions of interest were combined into a single region of interest that reflected the entire perisylvian association cortex.
Results Performance on sentence comprehension task

Performance of patients and control subjects on the sentence comprehension task is shown in Table I. The mean overall accuracy on the 25 sentences of the 21 normal subjects was compared with that of the group of 46 patients with cerebrovascular accident (CVA) to the left hemisphere (L) and the group of 14 patients with CVA to the right hemishere (R) in a 3X2 between-subjects ANOVA with subject type [normal controls, (L)CVA, (R)CVA] and language (English, French) as orthogonal factors. There was a main effect of group \F{2.11) = 20.56, P < 0.001]. There was no main effect of language. There was. however, an interaction between group and language type 1/(2,77) = 3.48, P = 0.036]. Simple effects showed that patients with both left and right CVAs were less accurate overall than normal controls. For the French patients, the (R)CVA group (overall accuracy = 83%) was significantly better than the (L)CVA group (overall accuracy = 48%). For English patients, (R)CVA patients (overall accuracy = 82%) did not differ from the (L)CVA group (overall accuracy = 73%). The English (L)CVA patients performed significantly better than the French (L)CVA patients. The difference between the (R)CVA patients and the English (L)CVA patients was not statistically significant. This pattern of results indicates that damage to both hemispheres affects sentence comprehension, with greater effects following left-hemisphere damage than right-hemisphere damage.
Lesions in syntactic comprehension deficits

1.0-1
939
0.9-
0.8C D
</>
CD CO CO
6 CJ
C CD
6-
0.7-
o
CD
Q.
CD
.a
0.5-
Controls (R)CVA (L)CVA
2-
0.4 Simple Complex
Sentence type Fig. 3 Performance of control subjects and patients with (L)CVA and (R)CVA on matched sets of syntactically simple and syntactically complex sentences.
-0.2
0.0
0.2
0.4
0.6
Syntactic complexity score Fig. 4 Graph of the number of subjects with left and right CVAs showing different magnitudes of syntactic complexity effects.
sentences with canonical thematic role order tests the integrity of syntactic processing. The results are shown in Fig. 3. The data were analysed in a 3X2X2 ANOVA with subject type [normal controls, (L)CVA, (R)CVA] and language (English, French) as between-subject factors and syntactic type (complex, simple) as a within-subject factor. There were main effects of group [F(2,77) = 29.9, P < 0.001] and syntactic type [F(l,77) = 66.9, P< 0.001]. There was a groupXlanguage type interaction [F(2,77) = 3.3, P = 0.04] and a groupXsyntactic type interaction [F(2,77) = 18.7, P < 0.001]. No other effects were significant. The effect of group and the interaction of group and language showed the same patterns as the results reported above for performance on all sentence types. The effect of syntactic type and the interaction of group and syntactic type are relevant to the question of whether patients showed deficits in syntactic processing. Simple effects showed that performance of all groups was better on the simple sentences than on the complex sentences. For simple sentences, control subjects performed better than both patient groups, and the two patient groups did not differ in their performance. For complex sentences, control subjects performed better than both patient groups, and patients with right-hemisphere lesions performed better than those with left-hemisphere lesions. These results indicate that both leftand right-hemisphere damaged patients have more difficulty comprehending sentences that require more complex syntactic operations to be understood than sentences that do not. This deficit is greater for patients with left-hemisphere lesions than for patients with right-hemisphere lesions, but it arises with damage to either hemisphere. The English (L)CVA patients showed the same disturbances of syntactic processing as were seen in the French (L)CVA patients (the three-way interaction of languageXsentence typeXhemisphere was not significant).
Effects of right-hemisphere lesions

The presence of a syntactic complexity effect is expected following damage to the perisylvian region of the left hemisphere. However, it is somewhat surprising that an effect of syntactic complexity would arise following righthemisphere stroke. Its presence would be readily explained, however, if it were due to the performance of one or two patients, who might be right-hemisphere dominant for language. To determine whether this was the case, we calculated a syntactic complexity score for each patient, consisting of the average of his or her performance (expressed as percent correct) on the seven syntactically complex sentences subtracted from his or her performance on the matched syntactically simple baseline sentences. For the (R)CVA patients, these scores ranged from close to zero (-0.044), indicating the absence of a syntactic complexity effect, to 0.514, indicating a considerable effect. Thirteen of the 14 scores of the (R)CVA patients were positive, indicating that 13 of the 14 patients contributed to the syntactic complexity effect. As shown in Fig. 4. only one righthemisphere patient had a complexity score that was substantially above the mean score of the left-hemisphere group and that indicated a major syntactic complexity effect. With the exception of this one patient, the scores of the righthemisphere patients approximate a normal distribution. This pattern suggests that the syntactic complexity effect in the right-hemisphere patient group is not due to the performance of one or two patients, but is due to small complexity effects appearing in most patients.
Effects of left-hemisphere lesions

We analysed the performance of patients with left-hemisphere lesions to gain clues as to the determinants of their sentence
940
D. Caplan et al these 18 (L)CVA patients was compared with that of the remaining (L)CVA patients in a 2X2 [group (patients with and without scans) Xsentence type (syntactically complex versus baseline)] ANOVA. There were main effects of group [F( 1,44) = 5.9, /><0.02] and sentence type [F(l,44) = 101.5, P < 0.001), and a significant interaction between the two factors [F(l,44) = 6.4, P < 0.02]. Analysis of simple effects showed that subjects with and without scans performed at the same level on the baseline sentences, and that subjects with scans performed worse than those without scans on the complex sentences. The syntactic complexity index described above was also computed for each patient, and the magnitude of this index compared in the patients with and without scans using Student's t test. This analysis confirmed the results of the ANOVA in showing that the syntactic complexity index was greater in the patients with scans than the patients without (t = 2.52, P < 0.02). These results indicate that the sample of patients who were scanned had more difficulty with syntactic processing than the remaining patients. Given that the English (L)CVA population tested here performed somewhat better than other (L)CVA groups (see above), the fact that the 18 patients whose scans were examined were among the more affected patients provides reassurance that they are typical of aphasic (L)CVA patients. In addition, the fact that they had clear problems in syntactic processing makes the patients who were scanned good subjects for a study of the lesion sites associated with disorders of this function. The cortical extent of the lesion sites in the 18 (L)CVA patients who had undergone CT scans is shown in Fig. 5. Table 3 lists the percent of each region of interest occupied by the lesion in each patient, along with each patient's performance on the sentence comprehension task. To investigate the relationship between the site and size of these lesions and performance on the sentence comprehension task, t tests, correlational analyses and regression analyses were performed. First, we tested the hypothesis that patients with damage to different regions of the perisylvian association cortex differ in syntactic processing in sentence comprehension by using t tests. It was impossible to compare patients with purely anterior lesions with those with purely posterior lesions, because only one patient had a purely anterior lesion. We therefore compared the performance of six patients whose visible lesions were confined to the 'posterior' region of interest with that of the remaining 12 patients, whose lesions involved the 'anterior' region of interest (Broca's area). Independent / tests showed that neither overall accuracy on the 25 sentence types nor the syntactic complexity score differed between the two groups. Although the factor analysis suggested that a single factor accounted for most of the variance in patients' performances, previous research has shown that individual patients can have selective impairments affecting particular syntactic operations (Caplan and Hildebrandt, 1988). Moreover, as noted in the Introduction, psycholinguistic models postulate different
Table 2 Factor loadings and variance accounted for by each factor

Sentence type Factor 1 Factor 2 0.31956 -0.28987 -0.36908 -0.27535 0.01036 0.06886 -0.3242 0.23936 -0.03096 -0.27093 -0.16539 -0.07572 -0.33878 0.23839 -0.09189 -0.01246 0.19594 0.2747 0.07758 -0.19258 0.35182 0.52692 -0.05151 0.11918 0.36464 6.2 Factor 3 0.14589 0.01989 0.05229 0.20701 0.16013 -0.18102 0.04805 0.0569 -0.28533 -0.2347 0.02504 -0.15093 -0.28031 -0.37192 0.12815 -0.42088 0.18158 -0.1178 0.11746 0.2477 0.17586 -0.03802 0.31278 0.50633 -0.2248 5.0 Factor 4 -0.06984 -0.10883 0.15141 -0.04501 -0.03516 0.0709 -0.24518 -0.31975 -0.21175 0.13009 0.23109 0.3087 0.07827 -0.03572 -0.04886 -0.04885 0.14721 0.35834 -0.44259 -0.26493 -0.18863 0.11673 0.2676 0.28947 -0.03516 4.2
0.85955 ST19 ST4 0.85916 0.85897 ST6 ST14 0.85516 ST22 0.85381 ST5 0.85007 ST2 0.84246 ST15 0.8404 ST8 0.83743 ST11 0.83155 ST7 0.83128 ST10 0.82429 ST9 0.81878 ST13 0.79744 ST3 0.79737 ST12 0.79539 ST24 0.79485 ST23 0.78422 ST18 0.75663 ST21 0.74745 ST16 0.72315 ST20 0.70003 ST25 0.6973 ST17 0.67032 ST1 0.54815 Variance (%) 63.1
comprehension impairments. A factor analysis with varimax rotation was carried out on the 25 sentence types in order to explore the nature of the relationship between performance on these sentence types. Four factors were extracted. Loadings of sentence types on factors and proportionate variance contribution are shown in Table 2, which shows that the first factor accounted for almost two-thirds of the variance. With a cut of 0.5 for inclusion of a variable in interpretation of a factor, all variables loaded on the first factor, with only one sentence type (simple active reflexive) loading on the second factor, and one sentence type (subject control) loading on the third. This analysis, which replicates the results of Caplan et al. (1985) and extends those results to a much larger set of sentences, shows that patient performance is largely affected by a single factor. This factor has been thought of as the availability of the processing resource discussed in the introduction to this paper that constrains overall sentence processing ability (Caplan et al., 1985). This analysis thus suggests that a useful approach to the study of the localization of syntactic processing would be to see whether overall performance and the syntactic complexity score described above, which would be principally determined by the availability of this resource, differ as a function of lesion site and/or correlate with lesion size in a particular site. As indicated above, CT scans were obtained on 18 (L)CVA patients. To determine whether these patients were similar to other aphasic patients with (L)CVAs, the performance of

NC EG
AD ^~T~^ SD_-T-^ CD,
941
Fig. 5 Diagrams of the left lateral hemisphere of the human brain depicting the approximate extent of the lesion in each of the 18 subjects with (L)CVAs whose CT images were analysed. Using a stylized hemisphere as a template, each case was reconstructed from axial views of the CT scans. Shaded areas include both cortical and white matter lesions. Anglophone patients are presented in the left panel and francophone patients in the right.
Table 3 Individual lesion volumes as a percentage of normalized volume of regions of interest in perisylvian cortex, and performance on sentence comprehension task
Patient Normalized Percentage of normalized ROI occupied by lesion ROI2 (F3t) ROI3 (F3o) ROM (AG) ROI5 (SG) All ROIs (perisylvian) Performance Total percent correct 64 81 38 62 95 75 65 76 60 19 35 64 37 28 28 36 58 56 Syntactic complexity index 0.29 0.24 0.36 0.18 0.17 0.31 0.25 0.13 0.14 0.13 0.44 0.57 0.61 0.30 0.37 0.51 0.26 0.43 total lesion 3 volume (cm ) ROI1 (STG) English W.B. N.C. E.G. A.L. L.M. E.M. D.S. G.T. L.Z. French A.D. S.D. CD. H.D. R.M. F.R. F.S. G.S. J.V.
148 79 27 101 3 10 21 141 5 130 23 32 115 18 103 7 23 94
88 78 26 25 19 0 0 85 7 74 0 29 16 8 49 0 0 0
82 0 0 18 0 21 0 100 0 100 4 3 100 0 61 18 0 0
66 0 0 56 0 21 0 100 0 89 23 18 100 0 93 61 12 17
33 95 24 24 13 0 0 0 0 0 0 17 41 0 0 0 0 1
40 43 21 74 0 0 <1 25 1 71 57 16 75 8 4 0 4 11
56 51 35 48 6 4 <1 48 2 64 27 18 61 5 29 8 3 6
ROI = region of interest; STG = superior temporal gyms; F3t = pars triangularis; F3o = inferior frontal gyrus/pars opercularis; AG = angular gyrus; SG = supramarginal gyrus.
types of parsing operations that construct different aspects of syntactic form (Frazier, 1990). Therefore, using independent t tests, we analysed the performance of patients with and without lesions in the anterior region of interest on 19 separate measures that correspond to particular syntactic operations. Each of these measures consisted of the difference
between a baseline sentence and a matched sentence, that required the same processing as the baseline sentence, plus an additional syntactic operation. For instance, one measure of the ability to construct and interpret the passive form consisted of performance on active sentences (baseline) minus passive sentences that each had two noun phrases. The 19
942
D. Caplan et al.
Table 4 Measures of specific syntactic operations

Baseline 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Two-place active (ST1) Two-place active (ST1) Three-place active (ST5) Active conjoined theme (ST13) Object control (ST16) Object control (ST16) Matched sentence P2 (ST2) PI (ST3) P3 (ST6) Passive conjoined agent (ST14) Subject control (ST17) Noun phrase-raising (English ST25) Passive object control (French ST25) Two-place cleft object (ST4) Three-place cleft object (ST7) Object-subject (STIO) Object-object (ST11) Subject-object (ST9) Subject-object (ST9) Subject-subject (ST12) Reflexives, simple (ST18) Reflexive, simple (ST20) Reflexive, complex NP (ST21) Pronouns, simple (ST19) Pronoun, simple (ST24) Pronoun, complex noun phrase (ST23) Operation Passive Passive Passive Passive Antecedent of pronoun Antecedent of noun phrase-trace Object relativization Object relativization Object relativization Object relativization Object relativization Centre embedding Centre embedding Antecedent of reflexive Antecedent of reflexive Antecedent of reflexive Antecedent of pronoun Antecedent of pronoun Antecedent of pronoun
Two-place active (ST1) Three-place active (ST5) Conjoined (ST8) Object-subject (ST10) Subject-subject (ST 12) Object-object (ST11) Object-subject (ST 10) Three reflexive expressions (ST15) Two-place active (ST1) Simple active relexive-expression, complex noun phrase (ST22) 17 Three reflexive expressions (ST15) 18 Two-place active (ST1) 19 Simple active relexive-expression, complex noun phrase (ST22)
measures are listed in Table 4. None of these comparisons was significant. We also explored the effect of lesion size within a region on performance through correlational analyses. We separately correlated (i) overall accuracy on the entire set of 25 sentence types, (ii) the overall syntactic complexity score described above, and (iii) the 19 separate measures that correspond to particular syntactic operations, with (a) normalized lesion volume in the language zone, (b) normalized lesion volume in each of the five regions of interest, and (c) normalized lesion volume in the anterior and posterior regions of interest. None of these 168 correlations were significant. To look for any non-linear relationships that might have obscured a correlation, we created separate plots for each of these 168 correlations. We found no evidence of non-linear relations in any of these plots. Finally, correlational analyses were performed to identify whether lesion size in a particular region of interest affected performance, once the effect of overall lesion size had been taken into account. The percentages of each region of interest occupied by a lesion were correlated against the residuals of two regression analysesone in which the normalized lesion volume in the language zone was regressed against the overall accuracy scores and one in which this value was regressed against the overall syntactic complexity scores. (NB In this analysis, the five regions of interest were reduced to three by combining the two frontal and the two parietal regions of interest, resulting in regions of interest that represent the frontal, parietal and temporal portions of the language zone. This reduction was undertaken to reduce the ratio of independent variables to cases, thereby allowing for the analysis of these data by regression analyses.) There were
no significant correlations of lesion extent in any of the regions of interest with the residuals of either of these regressions. This indicates that, when the effect of total perisylvian lesion volume is removed, there is still no particular area within this region in which lesion size correlates with the degree of syntactic processing impairment. These analyses indicate that there was no difference between the patients with and without anterior perisylvian lesions with respect to their overall level of performance, the magnitude of a syntactic processing deficit and the presence of impairments of specific parsing operations. They also indicate that there was no relationship between lesion size in either the anterior or posterior language area and the overall magnitude of a sentence comprehension deficit, the overall magnitude of a syntactic processing deficit, or the magnitude of deficits in specific syntactic operations. This suggests that lesions that are confined to the posterior perisylvian cortex have essentially the same effect on syntactic processing as lesions that affect both the posterior and anterior perisylvian regions. However, these analyses could be misleading because the patients were studied and scanned at different times in relation to their lesions. We undertook four analyses to determine whether this was likely to be the case. First, we determined that the mean interval from stroke to testing was the same in the patients with and without anterior lesions. For the 12 patients with anterior lesions, time since stroke ranged from 4 to 84 months with a mean of 25 months (SD = 25); for the six patients without anterior lesions, time since stroke ranged from 8 to 65 months with a mean of 28 months (SD = 8). Secondly, we correlated the duration of illness, measured in months between stroke and time of testing, with the overall syntactic complexity score. Because
Lesions in syntactic comprehension deficits Table S Performance (number correct of 12 trials) of five patients on different sentence types
Sentence types Patients E.M. Sentences with full noun phrase Two-place active Three-place active Conjoined Active conjoined theme Three reflexive-expressions Simple active reflexive-expression Sentences with pronouns or reflexives Reflexives, simple noun phrase Pronouns, simple noun phrase Simple active reflexive Simple active reflexive ('friend of X' subj) Simple active pronoun ('friend of X' subj) Simple active pronoun Sentences with empty noun phrases Two-place passive Truncated passive Two-place cleft object Three-place passive Three-place cleft object Subject-object relative Object-subject relative Object-object relative Subject-subject relative Passive conjoined agent Object control Subject control Noun phrase-raising/passivized object control Total correct See Appendix for descriptions of sentence types. 12 11 9 10 8 10 F.S. 12 5 7 9 4 1 G.S. 10 12 7 11 2 5 L.M. 12 12 12 12 12 12 L.Z. 10 7 3 9 11 12
943
8 9 12 11 11 12 12 12 9 9 2 5 7 7 9 10 11 6 5 227
7 5 12 0 6 9 2 8 3 0 1 0 4 0 4 1 5 2 2 109
8 2 10 5 6 8 10 11 11 6 7 6 9 5 6 7 9 0 0 173
12 12 12 12 12 12 12 12 12 12 12 7 10 9 12 12 12 12 6 284
8 10 12 9 6 7 9 7 8 7 8 2 2 3 3 12 9 5 2 181
the values for illness duration were not normally distributed, we also correlated the log of illness duration with the overall syntactic complexity score. Neither correlation was significant. Thirdly, we repeated all correlational analyses in a subset of 10 patients, who were tested between 3 and 24 months after stroke. These patients were also selected so as to exclude the patient whose scan was not taken along the canthomeatal plane, and to exclude the three patients whose scans were obtained within 2 weeks of stroke. None of the correlations were significant. Fourthly, though there were insufficient numbers of patients in this smaller selected set to compare those with lesions that only affected Broca's area with those with lesions that spared this region, we were able to select five patients with roughly equal-size small lesions, three of which primarily affected Broca's area, and two of which spared it. We analysed their patterns of performance, which are shown in Table 5. Three patients (E.M, F.S. and G.S.) had lesions that affected Broca's area. Overall performance did not correspond to the percentage of Broca's area that was affected in these patients. G.S., a 48-year-old French patient, had a lesion that occupied
<10% of Broca's area and extended through the pre- and post-central gyri to minimally affect the supramarginal gyrus. Her overall comprehension score was 58%. E.M., a 57-yearold English patient, had a lesion that was almost entirely restricted to Broca's area, of which it occupied ~20%. His overall comprehension score was 75%. F.S., a 60-year-old French patient, had a lesion that was also largely restricted to Broca's area and minimally affected the adjacent areas of the second frontal and pre-central gyri; the lesion occupied a little less than half of Broca's area. His overall comprehension score was 36%. The two patients with lesions that spared Broca's area also showed no correspondence between lesion size and degree of impairment. L.Z., a 62-year-old English patient, had a lesion primarily located in the superior temporal gyrus, of which it occupied 7%. Her overall comprehension score was 60%. L.M., a 21-year-old English patient, had a lesion that occupied 19% of the superior temporal gyrus and 13% of the angular gyrus. Her overall comprehension score was 95%. Similarly, a comparison of the performance of the three subjects with primarily Broca's area lesions and the two with lesions that spared this region showed no specific type of
944
D. Caplan et al. syntactic complexityindependent of sentence lengthon the performance of the right-hemisphere patients, provides support for the view that the right hemisphere plays some role related to assigning sentence structure and/or using it to determine sentence meaning. It is not yet clear what the role of the right hemisphere in syntactic processing is. The magnitude of syntactic processing impairments was roughly normally distributed in both the right- and left-hemisphere lesioned patients, with a greater degree of impairment in the left-hemisphere group. This suggests that there might be a reduction in resources available for syntactic processing that varies in its extent in both populations. The greater effect of left-hemisphere lesions could reflect a specialization of this resource capacity for syntactic processing within the left-perisylvian cortex; the role of the right hemisphere might be to provide a less specialized working memory capacity that makes a lesser contribution to syntactic processing (Caplan and Hildebrandt, 1988; Waters et al., 1995). More specific protocols will have to be used to determine exactly what aspects of the total sentence comprehension process are accomplished by the right hemisphere. In addition, to determine the specificity of any deficit in sentence comprehension for lesions in the territory of the middle cerebral artery of the right hemisphere, it will be necessary to study patients with frontal lesions and other lesions outside the perisylvian cortex and to compare their performances with those of patients with right-hemisphere perisylvian lesions. The left-hemisphere damaged patients performed more poorly and showed greater effects of syntactic processing than those with right-hemisphere lesions. It is possible that the poorer performance of the left-hemisphere patients and the presence of greater syntactic complexity effects in this population compared with the right-hemisphere patients is due to larger lesions in the left- than in the right-hemisphere groups. However, though this possibility cannot be ruled out without additional data, it seems unlikely that the lefthemisphere lesions were on average twice to three times as large as those in the right hemisphere. The most likely explanation for the poorer performance and greater syntactic complexity effects in the left-hemisphere patients is that their lesions affected neural structures that are more crucially involved in syntactic processing. The data from the 18 patients in whom CT scanning was available indicate that deficits in syntactic processing follow lesions in all parts of the perisylvian association cortex of the left hemisphere. Before considering the implications of this pattern for the functional neuroanatomy of language, we must ask whether the observed pattern of deficit-lesion relationships might reflect deficiencies in the methodology of this study. The most obvious limitation of this study is that only 18 subjects were scanned. However, this limitation should be seen within the context of previous research on this topic: the present study is the largest study of patients in whom radiological data have been obtained and who have been
sentence that was affected by lesions in either location. All five of the patients had problems comprehending at least some relative clauses. L.M., with the second largest lesion in the group of five patients, but the best overall performance, only had problems with sentences of this type and one other sentence type with an empty noun phrase. L.Z. (with a primarily temporal lesion) had difficulty with a set of sentences with pronouns and reflexives and full noun phrases, and with all the sentences with empty noun phrases except one type of passive sentence. The patients with primarily anterior lesions had a variety of impairments other than those affecting relative clauses. E.M. showed impairments in sentences in which the subject of the main clause was related to the subject of an embedded infinitive (subject control and noun phrase-raising sentences), in sentences with three full noun phrases, and in some sentences with reflexives or pronouns. F.S. showed impairments in all but the simplest sentence types. G.S. showed impairments in longer sentences with full noun phrases (conjoined sentences and sentences with three full noun phrases), all sentences with reflexives or pronouns except the most simple type, and several passive sentence types. Most of these patterns are interpretable in psycholinguistic terms, but present no particular pattern across the patients with lesions in particular locations. Overall, this more detailed analysis of single cases with small lesions of roughly comparable size, who were tested at about the same time after their strokes, illustrates that the degree of variability found in quantitative and qualitative aspects of patients' performances are not easily related to lesion location or the size of lesions in the anterior or posterior portion of the perisylvian association cortex.
Discussion
The results of this study provide information about the neural structures that are involved in sentence comprehension. They show that sentence comprehension is affected by lesions in both the left and the right hemisphere, more so by the former. This finding is consistent with other reports in the literature (De Renzi and Fagiolini, 1978). The more specific results of this study pertain to syntactic processing in sentence comprehension. They add to the evidence that lesions throughout the left perisylvian association cortex are associated with disorders affecting this process. They also raise the question of a possible contribution of the right hemisphere to this aspect of sentence processing. We shall discuss the results for the right and the left hemisphere separately. The performance of the right-hemisphere population on syntactically complex sentences was significantly lower than that of the normal control subjects. The analysis of the performance of individual patients makes it unlikely that the poorer performance of right-hemisphere patients than control subjects was due to a few patients in this group who were right-hemisphere dominant for this aspect of language processing. The finding that there were significant effects of
Lesions in syntactic comprehension deficits tested for syntactic processing in sentence comprehension. In addition, the comprehension data available on each patient consists of performance on 12 examples of each of 25 sentence types, while most previous studies provide results for at most three or four sentence types. Thus, though the limited number of patients requires that caution be exercised in accepting these results, the data constitute the largest dataset presently available on this topic and can provide a tentative basis for theory construction. A second concern is that differences in the time from stroke to testing could have affected the results. Several analyses speak against this possibility. The magnitude of the functional impairment was not correlated with time since stroke. Analyses of a subset of subjects whose strokes were between 3 months and 2 years of testing were identical to the analyses of the larger group. Though neither of these findings rules out the possibility that patients with longer periods of recovery could have lesser deficits, or that there could be a non-linear relationship between time since lesion and degree of recovery, they combine to make these possibilities less likely to have obscured the relationship of lesion location and size to impairments on this task. Moreover, it should be born in mind that cortical re-organization poststroke cannot explain the presence of deficits following lesions to brain regions not premorbidly involved in the exercise of the function. Therefore, the finding that patients with lesions in many parts of the perisylvian cortex have syntactic processing deficits has implications for the functional neuroanatomy of this aspect of language processing. A third concern is that in this study we tested sentence comprehension through the use of a single task, object manipulation and that the results thus reflect the demands of this task as well as those of sentence comprehension. To address the concern that the results of an object-manipulation task might not generalize to other tasks, in a separate study, we correlated the performance of 17 aphasic patients on 10 sentence types on object manipulation and sentence-picture matching tasks (Caplan et al., 1995). The Spearman rank order coefficient (p) for performance on the sentence types across the two tasks was 0.66 (P = 0.04). Thus, at least for the overall measure of sentence processing, patients' performance on the object manipulation task correlates with a very different task, and can be taken as an externally valid measure of their sentence processing capacities. A fourth issue is that this study tested sentence comprehension in an 'off-line' taskone that reflects the end-point of the comprehension process, rather than examine 'on-line' syntactic processingi.e. the time course of constructing syntactic representations. It has been claimed that on-line measures reveal deficits in aspects of syntactic processing in Broca's aphasia, and this has been taken to implicate Broca's area as the locus of certain syntactic operations in sentence comprehension (Zurif et al., 1993). If this claim is correct, we must identify the source of the qualitatively similar offline impairments in patients with and without Broca's area
945
lesions in this study. One possibility is that certain on-line operations are affected by lesions in Broca's area, while others that interfere with the same final product of the comprehension process are affected by lesions elsewhere. More detailed on-line studies of syntactic processing in aphasia and of the consequences of different disturbances of on-line processing for final comprehension are needed before these possibilities can be settled. Fifthly, in the present study we have not attempted to define the white matter tracts that are lesioned in these patients. Damage to these tracts can cause de-efferentation and de-afferentation of cortical areas, with functional consequences that may be similar to those caused by lesions to the regions themselves (Klippel, 1908; Geschwind, 1965; Kosslyn et al., 1993). The possible importance of white matter lesions is highlighted by case D.S., who had a very small cortical lesion, but who only responded correctly to 65% of the sentences and who had a syntactic complexity index of 0.25. It is possible that analyses that took white matter tracts into account might reveal a higher degree of localization of syntactic processing. A final concern is that CT scanning primarily identifies areas of necrosis, and is not very sensitive to the presence of hypoperfusion or hypometabolism in cerebral tissue. Several studies in which investigators used [l8F]fluorodeoxyglucose PET, have demonstrated larger areas of hypoperfusion than those shown to be necrotic by CT scanning in aphasic patients (Metter et al., 1983, 1984, 1986, 1987, 1989, 1990). It has been suggested that cortical areas that are hypometabolic may not sustain normal cognitive functions (Kosslyn et al., 1993). Therefore, it is possible that a greater degree of localization might be observed if measurements of metabolism, blood flow and/or oxygen extraction were used to assess CNS damage. One can hope that future work will bring together different measures of CNS function with extensive and detailed cognitive analyses of patients' deficits, to address these issues. Until such time, we are limited to the currently available data. Accepting these data, provisionally, we can ask what their implications are for the functional neuroanatomy of the left perisylvian association cortex for syntactic processing. The fact that lesions in all parts of the perisylvian cortex affected syntactic processing is consistent with one of two models that have been proposed for the functional neuroanatomy of cognitive processes. One is a model according to which there are significant individual differences in the localization of syntactic processing across the population (Caplan, 1987a, b). The second is a neural net model, in which representations are distributed throughout a region of the brain (McClelland and Rumelhart, 1986; McClelland and Kawamoto, 1986; McClelland et al., 1989). Evidence against individual differences in localization of syntactic processing comes from the recent PET study, referred to above (Stromswold et al., 1996), that showed an increase in rCBF during syntactic processing in the pars opercularis of Broca's area in all subjects studied. However,
946
D. Caplan et al.
only strongly right-handed college-educated males, between the ages of 20 and 30 years, with no left-handed family members were studied in that experiment. If there are individual differences in the localization of parts of the neural system that is responsible for syntactic processing, these differences may be related to sex, handedness, familial handedness, age, educational level or other factors. The present study did not include enough patients to determine whether correlations between the degree of impairment in syntactic processing and the size of lesions in particular locations would be greater if the correlations were confined to subjects of a certain age, sex, educational level or handedness profile. Larger studies, both involving deficitlesion correlational analyses and functional activation in neurologically normal subjects, with more subjects in each of these groups are needed to explore this issue. The distributed neural net model maintains that the neural system that is responsible for syntactic processing includes a cortical region that extends along the sylvian fissure. This model would predict impairments in syntactic processing after lesions throughout this region, and thus is compatible with the results of the present study. Distributed models could possibly also be compatible with the evidence for localization found in the PET study of Stromswold et al. (1996). It has been shown that neural net models can achieve some degree of internal structure; i.e. neural nets that are trained to accomplish a function frequently develop in such a way that a particular stimulus maximally activates a particular subset of the units in the net (Plaut and Shallice, 1993). This could correspond to a distributed system in which there is a local increase in activity, observable as an increase in rCBF, when a particular stimulus is processed. There is one aspect of the data that poses a challenge to the distributed neural net model; namely, the finding that there was no correlation between total lesion size and severity of impairment. Most neural net models obey the principle of mass action (Lashley, 1950), such that the larger the loss of computational elements, the greater the overall decrement in performance (McClelland and Rumelhart, 1986; Patterson et al., 1989). Possible areas of research thus include an effort to see if a neural net model that develops an internal structure can be lesioned in such a way that there are similar effects of lesions throughout the net but no effect of overall lesion volume on performance. In summary, the data presented here are consistent with the conclusion that several regions of the left perisylvian cortex form critical parts of a neural system responsible for syntactic processing. Other data suggest some degree of localization of this function within the pars opercularis. The complete picture is consistent with the models proposed by Mesulam (1990) and Damasio (1992). which involve both distributed and localized features. Many aspects of these models remain to be developed to account for the entire pattern of results seen in both deficit-lesion correlational studies and functional neuroimaging studies with normal subjects.
Acknowledgements
We wish to thank Pierre Delplas, David Kennedy, Randall Benson, Jeremy Schmahmann and David Gow for assistance obtaining, registering and interpreting the CT scans and performing the regression analyses. The work reported here was partially supported by a grant from the National Institute of Deafness and other Communication Disorders (DC00942).
References Belliveau JW, Kennedy DN Jr, McKinstry RC, Buchbinder BR, Weisskoff RM, Cohen MS, et al. Functional mapping of the human visual cortex by magnetic resonance imaging. Science 1991; 254: 716-19. Butler-Hinz S, Caplan D, Waters G. Characteristics of syntactic comprehension deficits following closed head injury versus left cerebrovascular accident. J Speech Hear Res 1990; 33: 269-80. Caplan D. Discrimination of normal and aphasic subjects on a test of syntactic comprehension. Neuropsychologia 1987a; 25: 173-84. Caplan D. Neurolinguistics and linguistic aphasiology. Cambridge (UK): Cambridge University Press, 1987b. Caplan D, Futter C. Assignment of thematic roles to nouns in sentence comprehension by an agrammatic patient. Brain Lang 1986: 27: 117-134. Caplan D, Hildebrandt N. Disorders of syntactic comprehension. Cambridge (MA): MIT Press, 1988. Caplan D, Baker C, Dehaut F. Syntactic determinants of sentence comprehension in aphasia. Cognition 1985; 21: 117-75. Caplan D, Waters GS, Hildebrandt N. Sentence comprehension in aphasic patients in sentence-picture matching tests. Brain Lang 1995; 51: 147-9. Caramazza A, Zurif EB. Dissociation of algorithmic and heuristic processes in language comprehension: evidence from aphasia. Brain Lang 1976; 3: 572-82. Chomsky N. Knowledge of language: its nature, origin, and use. New York: Praegen 1986. Damasio AR. Aphasia [see comments]. [Review]. N Eng J Med 1992; 326: 531-9. Damasio H. Damasio AR. Lesion analysis in neuropsychology. New York: Oxford University Press, 1989. De Renzi E, Fagiolini P. Normative data and screening power of a shortened version of the Token Test. Cortex 1978; 14: 41-9. Fox PT, Lancaster JL. Proceedings: BrainMap: Workshop I. San Antonio (TX): Research Imaging Center/UTHSCSA, 1993. Fox PT, Perlmutter JS, Raichle ME. A stereotactic method of anatomical localization for positron emission tomography. J Comput Assist Tomogr 1985; 9: 141-53. Fox PT, Mintun MA, Raichle ME, Miezin FM, Allman JM, Van Essen DC. Mapping human visual cortex with positron emission tomography. Nature 1986; 323: 806-19. Frazier L. Sentence processing: a tutorial review. In: Coltheart M,

editor. Attention and performance XII: The psychology of reading. Hove: Lawrence Erlbaum, 1987a: 559-86. Frazier L. Theories of sentence processing. In: Garfield JL, editor. Modularity in knowledge pepresentation and natural-language understanding. Cambridge (MA): MIT Press, 1987b: 291-307. Frazier L. Against lexical generation of syntax. In: Marslen-Wilson W, editor. Lexical representation and process. Cambridge (MA): MIT Press, 1989: 505-28. Frazier L. Exploring the architecture of the language-processing system. In: Altmann GTM , editor. Cognitive models of speech processing: psycholinguistic and computational perspectives. Cambridge (MA): MIT Press, 1990: 409-33. Geschwind N. Disconnexion syndromes in animals and man. Brain 1965; 88: 237-94, 585-644. Geschwind N, Galaburda AM. Cerebral lateralization: biological mechanisms, associations, and pathology: IIII. A hypothesis and a program for research. [Review]. Arch Neurol 1985; 42: 428-59, 521-52, 634-54. Hagoort P, Brown C, Groothusen J. The syntactic positive shift (SPS) as an ERP measure of syntactic processing. Lang Cognit Process 1993; 8: 439-83. Jackson HH. On the nature of the duality of the brain. Med Press 1874; 1: 19-41. Just MA, Carpenter PA. A capacity theory of comprehension: individual differences in working memory. Psychol Rev 1992; 99: 122-49. Klippel MM. Discussion surl'aphasie. Rev Neurol 1908; 16:611-36. Kluender R, Kutas M. Subjacency as a processing phenomenon. Language and Cognitive Processes 1993; 8: 573-633. Kosslyn S, Daly P, McPeek R, Alpert N, Kennedy D, Caviness V. Using locations to store shape: an indirect effect of a lesion. Cereb Cortex 1993; 3: 567-82. Lashley KS. In search of the engram. Sympos Soc Exp Biol 1950; 4: 454-82. Linebarger MC, Schwartz MF, Saffran EM. Sensitivity to grammatical structure in so-called agrammatic aphasics. Cognition 1983; 13: 361-92. MacDonald MC. Probabilistic constraints and syntactic ambiguity resolution. Lang Cognit Process 1994; 9: 157-201. Mazoyer BM, Tzourio N, Frak V, Syrota A, Murayama N, Levrier O, et al. The conical representation of speech. J Cognit Neurosci 1993; 5: 467-79. McClelland JL, Kawamoto AH. Mechanisms of sentence processing: assigning roles to constituents of sentences. In: McClelland JL, Rumelhart DE, editors. Parallel distributed processing, Vol. 2. Cambridge (MA): MIT Press, 1986. McClelland JL, Rumelhart DE, editors. Parallel distributed processing, Vol. 2. Cambridge (MA): MIT Press, 1986. McClelland JL, St John M, Taraban R. Sentence comprehension: a parallel distributed processing approach. Lang Cognit Process 1989; 4: 287-335.
947
Mesulam M-M. Large-scale neurocognitive networks and distributed processing for attention, language, and memory. [Review], Ann Neurol 1990; 28: 597-613. Metter E, Riege W, Hanson W, Jackson C, Kempler D, VanLancker D. Comparison of metabolic rates, language and memory, and subcortical aphasias. Brain Lang 1983; 19: 33^17. Metter EJ, Riege WH, Hanson WR, Camras LR, Phelps ME, Kuhl DE. Correlations of glucose metabolism and structural damage to language function in aphasia. Brain Lang 1984; 21: 187-207. Metter EJ, Kempler D, Hanson R, Jackson C, Mazziotta JC, Phelps ME. Cerebral glucose metabolism: differences in Wernicke's, Broca's and conduction aphasias. Clin Aphasiol 1986; 16: 97-104. Metter EJ, Kempler D, Jackson CA, Hanson WR, Riege WH, Camras LR, et al. Cerebellar glucose metabolism and chronic aphasia. Neurology 1987; 37: 1599-606. Metter EJ, Kempler D, Jackson C, Hanson W, Mazziotta JC, Phelps ME. Cerebral glucose metabolism in Wernicke's, Broca's and conduction aphasia. Arch Neurol 1989; 46: 27-34. Metter EJ, Hanson WR, Jackson CA, Kempler D, van Lancker D, Mazziotta JC, et al. Temporoparietal cortex in aphasia. Arch Neurol 1990; 47: 1235-8. Miceli G, Mazzucchi A, Menn L, Goodglass H. Contrasting cases of Italian agrammatic aphasia without comprehension disorder. Brain Lang 1983; 19: 65-97. Mohr JP, Pessin MS, Finkelstein S, Funkenstein HH, Duncan GW, Davis KR. Broca aphasia: pathologic and clinical. Neurology 1978; 28: 311-24. Nespoulous J-L, Dordain M, Perron C, Ska B, Bub D, Caplan D, et al. Agrammatism in sentence production without comprehension deficits: reduced availability of syntactic structures and/or of grammatical morphemes? A case study. Brain Lang 1988; 33: 273-95. Neville H, Nicol JL, Barss A, Forster Kl, Garrett MF. Syntactically based sentence processing classes: evidence from event-related brain potentials. J Cognit Neurosci 1991; 3: 151-65. Osterhout L, Holcomb PJ. Event-related brain potentials elicited by syntactic anomaly. J Mem Lang 1992; 31: 785-806. Osterhout L, Holcomb PE. Event-related potentials and syntactic anomaly: evidence of anomaly detection during the perception of continuous speech. Lang Cognit Process 1993; 8: 413-37. Patterson K, Seidenberg MS, McClelland JL. Connections and disconnections: acquired dyslexia in a computational model of reading processes. In: Morris RGM, editor. Parallel distributed processing: implications for psychology and neurobiology. Oxford: Clarendon Press, 1989: 131-81. Pinker S. The language instinct: the new science of language and mind. London: Allen Lane, 1994. Plaut DC, Shallice T. Deep dyslexia: a case study of connectionist neuropsychology. Cognit Neuropsychol 1993; 10: 377-500. Rademacher J, Galaburda AM, Kennedy DN, Filipek PA, Caviness VS Jr. Human cerebral cortex: localization, parcellation, and morphometry with magnetic resonance imaging. J Cognit Neurosci 1992: 4: 352-74.
948
D. Caplan et al.
Tokunaga A, Takase M, Otani K. The glabella-inion line as a baseline for CT scanning of the brain. Neuroradiology 1977; 14: 67-71. Tramo MJ, Baynes K, Volpe BT. Impaired syntactic comprehension and production in Broca's aphasia: CT lesion localization and recovery patterns. Neurology 1988; 38: 95-8. Tyler LK. Real-time comprehension processes in agrammatism: a case study. Brain Lang 1985; 26: 259-75. Vanier M, Caplan D, CT-scan correlates of agrammatism. In: Menn L, Obler LK, editors. Agrammatic aphasia. Amsterdam: Benjamins, 1990: 97-114. Vanier M, Lecours AR, Ethier R, Habib M, Poncet M, Milette PC, et al. Proportional localization system for anatomical interpretation of cerebral computed tomograms. J Comput Assist Tomogr 1985; 9: 715-24. Waters GS, Caplan D, Rochon E. Processing capacity and sentence comprehension in patients with Alzheimer's disease. Cognit Neuropsychol 1995; 12: 1-30. Zurif E, Swinney D, Prather P, Solomon J, Bushell C. An on-line analysis of syntactic processing in Broca's and Wernicke's aphasia. Brain Lang 1993; 45: 448-64.
Rademacher J, Caviness VS Jr. Steinmetz H, Galaburda AM. Topographical variation of the human primary cortices: implications for neuroimaging, brain mapping, and neurobiology. Cereb Cortex 1993; 3: 313-29. Reinhart T. Anaphora and semantic interpretation. London: Croom Helm, 1983. Rochon E, Waters GS, Caplan D. Sentence comprehension in patients with Alzheimer's disease. Brain Lang 1994; 46: 329^9. Sanides F. The cyto-myeloarchitecture of the human frontal lobe and its relation to phylogenetic differentiation of the cerebral cortex. J Himforsch 1964; 6: 269-82. Schwartz M, Saffran E, Marin O. The word order problem in agrammatism I: Comprehension. Brain Lang 1980; 10: 249-62. Schwartz MF. Linebarger MC, Saffran EM, Pate DS. Syntactic transparency and sentence interpretation in aphasia. Lang Cognit Process 1987: 2: 85-113. Seines OA, Knopman DS, Niccum N, Rubens AB, Larson D. Computed tomographic scan correlates of auditory comprehension deficits in aphasia: a prospective recovery study. Ann Neurol 1983; 13: 558-66. Stromswold K, Caplan D, Alpert N, Rauch S. Localization of syntactic comprehension by positron emission tomography. Brain Lang 1996. In press. Talairach J, Tournoux P, Co-planar stereotaxic atlas of the human brain. Stuttgart: Thieme, 1988.
Received July 12, 1995. Revised November 13, 1995. Accepted December 29, 1996
949
Appendix: Sentence types

Baseline (simple) sentences: no referentially dependent noun phrases (1)* Two-place active The frog hit the monkey. (5)* Three-place active The rabbit passed the cow to the goat. (8)* Conjoined The monkey scratched the rabbit and patted the elephant. (13) Active conjoined theme The frog patted the monkey and the elephant. (15) Three reflexive-expressions The old man knew that his friend scratched the boy. (22) Simple active reflexive-expression ('friend of X' subject) The father of the boy pointed to the old man. Overt referential dependencies ('himself, 'him') (18) Reflexives, simple noun phrase The old man said that the father hit himself. (19) Pronouns, simple noun phrase The old man believed that the father tickled him. (20) Simple active reflexive The old man kicked himself. (21) Simple active reflexive ('friend of X' subject) The father of the boy scratched himself. (23) Simple active pronoun ('friend of X' subject) The father of the boy kicked him. (24) Simple active pronoun The old man tickled him. Empty referential dependencies (2)f Two-place passive The monkey was hit by the frog. (3) Truncated passive The rabbit was patted . (4)^ Two-place cleft object It was the cow that the rabbit kissed . (6^ Three-place passive The elephant was given to the monkey by the frog. (7)1 Three-place cleft object It was the goat that the rabbit gave to the cow. (9)f Subject-object relative The monkey that the rabbit grabbed shook the goat. (10) Object-subject relative The goat hit the rabbit that grabbed the cow. (11) Object-object relative The monkey tickled the frog that the goat shook . (12) Subject-subject relative The frog that held the cow caught the elephant. (14) Passive conjoined agent The elephant was hit by the monkey and the frog. (16)* Object control The old man told the father to pray. (17) Subject control to sleep. The old man swore to the father _ (25)f Noun phrase-raising (English) The old man seems to the father to be bending over. Passivized object control (French) Le vieillard a t incit par le p re manger. *Baseline sentence type used in syntactic complexity analysis; Sentence types with empty referential dependencies and noncanonical word order.

Brain 1996 Caplan 933 49

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Brain 1996 Caplan 933 49

Hochgeladen von

Copyright:

Verfügbare Formate

Brain (1996), 119, 933-949

The dog that 0 scratched the cat chased

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

100 100 99 * 97 78 99 98 78 77 78 78 100 * 99 96 96 88 95 97 94 * 98 93 96

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Results Performance on sentence comprehension task

Lesions in syntactic comprehension deficits

Controls (R)CVA (L)CVA

0.4 Simple Complex

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Effects of right-hemisphere lesions

Effects of left-hemisphere lesions

Table 2 Factor loadings and variance accounted for by each factor

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Lesions in syntactic comprehension deficits

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

148 79 27 101 3 10 21 141 5 130 23 32 115 18 103 7 23 94

82 0 0 18 0 21 0 100 0 100 4 3 100 0 61 18 0 0

Table 4 Measures of specific syntactic operations

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Lesions in syntactic comprehension deficits

Lesions in syntactic comprehension deficits

Appendix: Sentence types

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Downloaded from http://brain.oxfordjournals.org/ by guest on May 2, 2012

Das könnte Ihnen auch gefallen