Effects of syntactic features on sentence-picture matching in Broca's aphasics: a reply to Drai and Grodzinksy (2005)

We reanalyzed the data in , considering individual patients' responses to different sentence types to be non-independent events. The analyses revealed effects of two of the three factors identified by Drai and Grodzinsky--constituent movement and passive mood. The result is inconsistent with the trace deletion hypothesis; we conclude that features of syntactic structure other than constituent movement are relevant to understanding performance variation in patients with Broca's aphasia.     

Patterns of comprehension performance in agrammatic Broca's aphasia: a test of the Trace Deletion Hypothesis

We tested the core prediction of the Trace Deletion Hypothesis (TDH) of agrammatic Broca's aphasia, which contends that such patients' comprehension performance is normal for active reversible sentences but at chance level for passive reversible sentences. We analyzed the comprehension performance of 38 Italian Broca's aphasics with verified damage to Broca's area, who completed sentence-to-picture matching tasks using active and passive reversible sentences as stimuli. The results failed to confirm the predictions made by TDH: only a small minority (15%) performed at chance on passive sentences and better than chance on active sentences. Furthermore, the distribution of the 38 subjects' performance on passive sentences differed from that predicted by the TDH since many more subjects performed at better-than-chance levels than expected. We discuss the implication of these results for claims about the distribution of language processing mechanisms in the brain.     

Comprehension of passives in Broca's aphasia

Drai and Grodzinsky have statistically analyzed a large corpus of data on the comprehension of passives by patients with Broca's aphasia. The data come, according to Drai and Grodzinsky, from binary choice tasks. Among the languages that are analyzed are Dutch and German. Drai and Grodzinsky argue that Dutch and German speaking Broca patients should be relatively good (that is, perform above chance) on comprehension of passive sentences, since in Dutch and German passives the relative order of the object and lexical verb is the underlying order and hence no movement takes place. We will demonstrate that both their linguistic arguments and their selection of Dutch data are invalid.     

Dysprosody in Broca's aphasia: a case study

A detailed acoustic analysis of timing, intensity, and fundamental frequency (F0) at different levels of linguistic structure was conducted on the speech output of a Broca's aphasic who was a native speaker of Thai. Timing was measured with respect to syllables, phrases, and sentences in connected speech. Intensity variation at the sentence level was measured in connected speech. F0 variation associated with the five Thai tones was measured in both isolated words and connected speech. Results indicated that timing was differentially impaired depending upon complexity of articulatory gesture and size of the linguistic structure. Timing, as well as intensity, was aberrant at the sentence level. In contrast, F0 contours of the five tones were spared at all levels of linguistic structure. Findings are interpreted to support the view that dysprosody in Broca's aphasia is more applicable to speech timing than to F0.     

Semantic operations in aphasic comprehension: implications for the cortical organization of language.

We provide data on the neurological basis of two semantic operations at the sentence level: aspectual coercion and complement coercion. These operations are characterized by being purely semantic in nature; that is, they lack morphosyntactic reflections. Yet, the operations are mandatory (i.e., they are indispensable for the semantic well formedness of a sentence). Results indicate that, whereas Broca's patients have little or no trouble understanding sentences requiring these operations (performance was above chance for all conditions), Wernicke's patients performed at normal-like levels only for sentences that did not require these operations. These findings suggest that sentence-level semantic operations rely very specifically on the integrity of the cortical area associated with Wernicke's aphasia, but not on the region corresponding to Broca's aphasia. In the context of other findings from lesion and imaging studies, this evidence allows a view of the cortical distribution of language capacity that is drawn along a linguistic line, one which distinguishes syntactic from semantic operations.     

The battle for Broca's region

The intense effort to characterize Broca's region has produced many views on its anatomy and function. Here, we present the leading approaches and consider ways to adjudicate among them empirically. Anatomically, we focus on the cytoarchitecture of Brodmann areas 44 and 45, which constitute Broca's region. Functionally, we discuss four views: action perception, working memory, syntactic complexity and syntactic movement. We compare these views, reflect on how they can be distinguished experimentally and review relevant aphasia and functional magnetic resonance imaging (fMRI) studies. Although no single hypothesis accounts for all of the data, the role of Broca's region in language comprehension is best explained by the syntactic movement account. This conclusion opens the door for an attempt to define general principles for the neural representation of linguistic knowledge.     

Proper and common nouns: form class judgments in Broca's aphasia

Studies of agrammatic Broca's aphasics' comprehension of sentences containing articles have demonstrated profound deficits. It has not been clear whether the impairments are due to an inability to isolate the article in the stream of speech, or to difficulty in the construction and/or interpretation of various syntactic, semantic, or pragmatic levels of representation. This paper reports three experiments on Broca's aphasics' ability to distinguish between common nouns (e.g., "a rose") and proper nouns (e.g., "Rose"). This grammatical form class decision is signaled by the presence or absence of an article, and is represented at the lexical node level of linguistic analysis. The three experiments demonstrated that Broca's aphasics point to pictures representing classes of objects when asked to point to "the X" and point to pictures representing unique individuals when asked to point to "X". Thus, they were shown to be able to use the presence or absence of an article to determine lexical category. Their performance was especially accurate in an oral language context which was highly redundant and in a written language context where patients themselves could control the rate of information flow. They were quantitatively impaired, relative to controls, in a third study, which made higher information processing demands. Moreover, in this third study nonsense syllables preceding the noun which are phonologically similar to a known article were much more likely to evoke the misclassification of its noun as common than were phonologically distinct nonsense syllables. These data indicate that Broca's aphasics indeed have some difficulty isolating the article in the stream of speech. Nevertheless, detailed analyses of aphasics' performance revealed their ability to distinguish between common nouns and proper nouns even under these demanding conditions. Taken together, the three studies show that insofar as agrammatic patients are able to keep track of the presence or absence of articles, they can make a grammatical decision at the lexical node level of linguistic analysis. We conclude, then, that agrammatic Broca's aphasics are particularly impaired in the use of articles to construct and/or interpret phrasal constituents.     

Agrammatic Broca's aphasia is not associated with a single pattern of comprehension performance

One influential hypothesis posits that the brain regions implicated in Broca's aphasia are responsible for specific syntactic operations that are necessary for the comprehension and production of sentences (Grodzinsky, 1986, 1990, in press). The empirical basis of this hypothesis is the claim that Broca's aphasics have no difficulty understanding sentences in the active voice (and other "canonical" sentence types, such as subject relatives and clefts with negative predicates), but perform at chance level with passive voice constructions (and other "noncanonical" sentences such as object-gap relatives and object clefts). In the face of well-established results indicating that Broca's aphasics can exhibit several different performance patterns on these sentence types, Grodzinsky, Pi?ango, Zurif, and Drai (1999) argued that these conflicting results do not challenge the theory when the data are analyzed appropriately. They carried out a creative statistical analysis of the comprehension performance of published cases of Broca's aphasia and concluded that all of these cases are in agreement with the predicted pattern: chance on passives and 100% correct on actives. Here we show that the statistical reasoning adopted by Grodzinsky et al. (1999) is flawed. We also show that the comprehension performance of a substantial number of the Broca's aphasics in their own sample does not conform to the pattern required. Rather, contrary to these authors' claim, Broca's aphasia is not associated with a consistent pattern of sentence comprehension performance, but allows for a number of distinct patterns in different patients.     

Predicate-argument structure as a link between linguistic and nonlinguistic representations

We present a study wherein a severe Broca's aphasic patient was trained to learn symbols representing both pure transitive and dative predicates--predicates differing in argument structure--in a visually based artificial language (c-ViC). We found a decrease in performance when two symbols, rather than one, were used to depict these "verbs." However, this decrease in performance was more pronounced for symbols representing pure transitive verbs--those that allow only one argument structure--than for symbols representing dative verbs--those that allow two different argument structures. Also, dative "verbs" yielded better performance when they were inserted in more complex, three-argument "sentences" than when they were inserted in two-argument "sentences." The opposite pattern was found for pure transitives. These results are discussed in terms of our claim that argument structure serves as a point of connection between linguistic information and non-linguistic visual information and in terms of the possibility that argument structure entries are shaped by the form in which visual information is parsed.     

Speech and language functions that require a functioning Broca's area

A number of previous studies have indicated that Broca's area has an important role in understanding and producing syntactically complex sentences and other language functions. If Broca's area is critical for these functions, then either infarction of Broca's area or temporary hypoperfusion within this region should cause impairment of these functions, at least while the neural tissue is dysfunctional. The opportunity to identify the language functions that depend on Broca's area in a particular individual was provided by a patient with hyperacute stroke who showed selective hypoperfusion, with minimal infarct, in Broca's area, and acutely impaired production of grammatical sentences, comprehension of semantically reversible (but not non-reversible) sentences, spelling, and motor planning of speech articulation. When blood flow was restored to Broca's area, as demonstrated by repeat perfusion weighted imaging, he showed immediate recovery of these language functions. The identification of language functions that were impaired when Broca's area was dysfunctional (due to low blood flow) and recovered when Broca's area was functional again, provides evidence for the critical role of Broca's area in these language functions, at least in this individual.     

Kicking over the traces: A note in response to Zurif and Piñango (1999)

Zurif and Pi?ango (1999) claimed that they excluded the four agrammatic patients reported by Druks and Marshall (1991) from their review article because two of the patients were nonnative speakers of Hebrew and because the Hebrew sentences we used in our investigations were ungrammatical. In Druks and Marshall (1991) we have shown that the presence or the absence of a trace in two types of Hebrew passives had no effect on the patients' performance. Two patients, without comprehension deficits, performed equally well on both types of passives and two patients, with comprehension deficits, were equally impaired on both types. We remind Zurif and Pi?ango of our previous response to the claims of ungrammaticality of our materials (Druks & Marshall, 1992) and argue that there were no justifiable reasons for excluding these cases from the review. We also comment on Zurif and Pi?ango's (1999) and Grodzinsky's (2000) new proposal that the association of agrammatic comprehension should be with Broca's aphasia and not with agrammatism.     

Object scrambling in Dutch Broca's aphasia

This study focuses on the production of sentences in which objects have been moved without a change in the order of the thematic roles. In Dutch, the basic word order is subject-adverb-object-verb. The object can be moved over the adverb; this is called object scrambling. The difference between the two word orders is pragmatic in nature: in the basic word order the focus is on the object, in the scrambled order, the focus is on the adverb. The aim of the present study is to evaluate if production of constructions with moved objects is impaired in Broca's aphasia and if so, whether that is for syntactic or pragmatic reasons. The results show that for individuals with Broca's aphasia, sentences with the scrambled word order are more difficult to produce than sentences with the basic word order, even if the scrambled order results in a pragmatically more acceptable sentence. This falsifies several theories of production in Broca's aphasia and shows an interesting parallel to the performance on comprehension tasks.     

Comprehension of contrastive stress by Broca's aphasics.

Comprehension of stress as a determiner of reference for pronouns was compared in eight patients with Broca's aphasia (BA) and five age-matched control subjects. The subjects were asked to listen to sentences in which the stressed or unstressed condition of the pronoun was a critical criterion for the establishment of reference. For each sentence, subjects were shown three pictures and asked to point to the correct referent of the pronoun. While the controls were nearly perfect in both the stressed and unstressed conditions, BA patients were significantly worse than normals, showing chance performance in both cases. However, a significant disparity was found in the BA patients' selection of the object NP as the referent under stressed and unstressed conditions, indicating that BA subjects are, indeed, sensitive to the stress patterns of pronouns. It was thus hypothesized that the BA patients' chance performance was the result of an inability to implement their knowledge of stress during the processing of sentences involving discourse-related linguistic operations, such as the establishment of pronoun reference (Grodzinsky, Wexler, Chien, Marakovitz, & Solomon, 1993). To test this hypothesis, a second experiment was conducted in which discourse-related operations were eliminated. In this second experiment, comprehension of stress by the same two groups was compared in tasks involving purely morphosyntactic processes. The contrastive stress patterns of otherwise homophonous compound nouns and adjectival phrases (e.g., BLACKboard, black BOARD), rather than those of pronouns, were examined. In this grammatically "simpler" experiment (i.e., without discourse-related operations), BA subjects scored significantly above chance in their comprehension of sentences involving compound nouns; unexpectedly, however, these same subjects did not show significantly above-chance performance in their comprehension of sentences containing adjectival phrases. Nevertheless, the results obtained in these two experiments seem to support the view that aphasic patients may have a lack of processing capacity, resulting in more errors during the processing of discourse-related linguistic constructions.     

To end ambiguous sentences

A study of the time required to complete ambiguous sentences suggested that: even though Ss are unaware of the ambiguity while completing sentences, they take more time to complete ambiguous sentences than unambiguous ones: the degree of difficulty in completing ambiguous sentences is related to the linguistic level at which the ambiguity occurs: sentences containing two ambiguities are more difficult to complete than those containing only one, and when these two ambiguities occur at different linguistic levels, these sentences are harder to complete than when both occur within the same linguistic level: ambiguity may affect the grammaticality and relevance of completions; and may cause stuttering and laughter, even without awareness of the ambiguity. An attempt to fit these results to several theories of the processing of ambiguous sentences led us to the conclusion that ambiguity interferes with our understanding of a single meaning of a sentence, and that the degree of interference varies with the linguistic level at which the ambiguity occurs.     

To end ambiguous sentences

A study of the time required to complete ambiguous sentences suggested that: even though Ss are unaware of the ambiguity while completing sentences, they take more time to complete ambiguous sentences than unambiguous ones: the degree of difficulty in completing ambiguous sentences is related to the linguistic level at which the ambiguity occurs: sentences containing two ambiguities are more difficult to complete than those containing only one, and when these two ambiguities occur at different linguistic levels, these sentences are harder to complete than when both occur within the same linguistic level: ambiguity may affect the grammaticality and relevance of completions; and may cause stuttering and laughter, even without awareness of the ambiguity. An attempt to fit these results to several theories of the processing of ambiguous sentences led us to the conclusion that ambiguity interferes with our understanding of a single meaning of a sentence, and that the degree of interference varies with the linguistic level at which the ambiguity occurs.     

Kicking over the Traces: A Note in Response to Zurif and Piñango (1999)

Zurif and Piñango (1999) claimed that they excluded the four agrammatic patients reported by Druks and Marshall (1991) from their review article because two of the patients were nonnative speakers of Hebrew and because the Hebrew sentences we used in our investigations were ungrammatical. In Druks and Marshall (1991) we have shown that the presence or the absence of a trace in two types of Hebrew passives had no effect on the patients' performance. Two patients, without comprehension deficits, performed equally well on both types of passives and two patients, with comprehension deficits, were equally impaired on both types. We remind Zurif and Piñango of our previous response to the claims of ungrammaticality of our materials (Druks & Marshall, 1992) and argue that there were no justifiable reasons for excluding these cases from the review. We also comment on Zurif and Piñango's (1999) and Grodzinsky's (2000) new proposal that the association of agrammatic comprehension should be with Broca's aphasia and not with agrammatism.     

Production of non-canonical sentences in agrammatic aphasia: limits in representation or rule application?

The study reported here compares two linguistically informed hypotheses on agrammatic sentence production, the TPH [Friedmann, N., & Grodzinsky, Y. (1997). Tense and agreement in agrammatic production: Pruning the syntactic tree. Brain and Language, 56, 397-425.] and the DOP [Bastiaanse, R., & van Zonneveld, R. (2005). Sentence production with verbs of alternating transitivity in agrammatic Broca's aphasia. Journal of Neurolinguistics, 18, 59-66]. To explain impaired production of non-canonical sentences in agrammatism, the TPH basically relies on deleted or pruned clause structure positions in the left periphery, whereas the DOP appeals to limitations in the application of movement rules. Certain non-canonical sentences such as object-questions and object-relative clauses require the availability of nodes in the left periphery as well as movement to these nodes. In languages with relatively fixed word order such as English, the relevant test cases generally involve a coincidence of left periphery and movement, such that the predictions of the TPH and the DOP are identical although for different reasons. In languages with relatively free word order such as German, on the other hand, it is possible to devise specific tests of the different predictions due to the availability of scrambling. Scrambled object sentences, for example, do not involve the left periphery but do require application of movement in a domain below the left periphery. A study was conducted with German agrammatic subjects which elicited canonical sentences without object movement and non-canonical scrambled sentences with object movement. The results show that agrammatic speakers have a particular problem with the production of scrambled sentences. Further evidence reported in the study from spontaneous speech, elicitation of object relatives, questions and passives and with different agrammatic subjects confirms that non-canonical sentences are generally harder to produce for agrammatics. These findings provide evidence in favor of the DOP and it will be argued that a cross-modal explanation of agrammatic deficits is possible if two factors--movement and canonicity--are taken into consideration.     

Word order and finiteness in Dutch and English Broca's and Wernicke's aphasia

The effect of two linguistic factors in Broca's and Wernicke's aphasia was examined using Dutch and English subjects. Three tasks were used to test (1). the comprehension and (2). the construction of sentences, where verbs (in Dutch) and verb arguments (in Dutch and English) are in canonical versus non-canonical position; (3). the production of finite versus infinitive verbs. Proportions of errors as well as types of errors made by each aphasic group are similar on the sentence comprehension and sentence anagram tasks. On the verb production task the performance pattern is, again, the same, but the error types are different. The discussion focuses on how the similarities and differences across languages and across aphasia types may be interpreted with respect to the underlying deficit in Broca's and Wernicke's aphasia.     

Speech prosody in Broca's aphasia

Because speech prosody is thought to be impaired in Broca's aphasia, we conducted three experiments using Broca's aphasics and nonneurological control subjects in order to determine to what extent this statement is accurate. Using three acoustic measures of speech prosody—sentence-final fundamental frequency (F₀) fall, F₀ declination, and sentence-final lengthening—we found that some aspects of prosody were spared while others were abnormal. All Broca's aphasics, regardless of degree of impairment, exhibited sentence-final F₀ fall. F₀ declination was present in simple sentences but was absent over longer domains; moreover, the amount of declination was correlated with the degree of linguistic impairment. Sentence-final lengthening was clearly absent in Broca's aphasia; in fact, sentence-terminal words were actually shorter than their sentence-initial and medial counterparts.     

Dissociation of algorithmic and heuristic processes in language comprehension: Evidence from aphasia

Three groups of aphasic patients, Broca's, Conduction, and Wernicke's, and a nonaphasic patients control group were tested for comprehension of object-relative center-embedded sentences. The sentences were of three types: sentences in which semantic constraints between words allowed the subjects to assign a correct semantic reading of the sentence without decoding the syntax, sentences in which semantic constraints were relaxed and for which a correct reading was only possible with knowledge of syntactic relationships among words, and sentences which described highly improbable events. The subjects' task was to choose which of two pictures captured the meaning expressed in the sentence. Broca's and Conduction aphasics performed near perfectly on sentences where they could use semantic information. Their performance dropped to chance when they had to use syntactic information. These results support a neuropsychological dissociation of heuristic and algorithmic processes based primarily, though not exclusively, on semantic and syntactic information, respectively.