Grammatical class in lexical production and morhpological processing: Evidence from a case of fluent aphasia

We present the case of a fluent aphasic patient who is impaired at producing nouns relative to verbs in picture naming, sentence completion, and sentence generation tasks, but is better at both producing and comprehending concrete nouns than abstract nouns. Moreover, he displays a selective difficulty in producing the plural forms of some nouns and pseudowords presented as nouns, but was able to produce the phonologically identical third-person singular forms of corresponding verb homonyms and of the same pseudowords presented as verbs. This pattern of performance casts doubt on the hypothesis that grammatical class effects are always epiphenomena of more general semantic impairments that affect the naming of actions or of concrete objects, and suggests that these effects may arise instead from damage to syntactic processes pertaining specifically to the grammatical properties of words. We also discuss the implications of such damage for models of morphological processing.     

The noun/verb dissociation in language production: varieties of causes

We report the performance of two patients who presented with complementary deficits in naming nouns relative to verbs: EA performed far worse with nouns than verbs, while MR performed worse with verbs than nouns. The two patients' grammatical category-specific deficits could not easily be explained in terms of damage to specific types of semantic knowledge prototypically associated with nouns (visual properties) and verbs (action features). One of the two patients, MR, also presented with a selective deficit in processing verbal as opposed to nominal morphology, in line with her impairment in naming verbs. The other patient, EA, showed no impairment in producing nominal and regular verbal morphology. The contrasting patterns of grammatical category-specific deficits in naming and morphological processing, along with other recently reported patterns, are interpreted as providing support for the claim that semantic and grammatical properties independently contribute to the organisation of lexical processes in the brain.     

THE NOUN/VERB DISSOCIATION IN LANGUAGE PRODUCTION: VARIETIES OF CAUSES

Abstract

We report the performance of two patients who presented with complementary deficits in naming nouns relative to verbs: EA performed far worse with nouns than verbs, while MR performed worse with verbs than nouns. The two patients' grammatical category-specific deficits could not easily be explained in terms of damage to specific types of semantic knowledge prototypically associated with nouns (visual properties) and verbs (action features). One of the two patients, MR, also presented with a selective deficit in processing verbal as opposed to nominal morphology, in line with her impairment in naming verbs. The other patient, EA, showed no impairment in producing nominal and regular verbal morphology. The contrasting patterns of grammatical category-specific deficits in naming and morphological processing, along with other recently reported patterns, are interpreted as providing support for the claim that semantic and grammatical properties independently contribute to the organisation of lexical processes in the brain.     

Differences between nouns and verbs after anterior temporal lobectomy

To test the claim that lesions of left anterior and middle temporal cortical structures specifically impair processing of nouns but not verbs, 56 left-hemisphere-language-dominant patients who had undergone anterior temporal lobectomy (ATL) completed tasks assessing confrontation naming of pictured objects and actions, generation of synonyms for nouns and verbs, and semantic lexical judgments about nouns and verbs. Compared with right ATL patients, left ATL patients were impaired across different tasks that assessed naming and comprehension of high-imageability as well as low-imageability nouns. These groups did not differ, however, in verb naming or comprehension on most tasks. Results are consistent with the hypothesized specialization of left temporal lobe structures for processing nouns and suggest that naming problems commonly seen after left ATL extend beyond difficulties with retrieving object names and may be related to subtle disturbances in comprehension of the meanings underlying nominal word forms.     

An influence of syntactic and semantic variables on word form retrieval

We report the case of DPI, an aphasic patient who shows a phonological impairment in production that spares certain syntactic and semantic categories. On a picture naming task, he produces mostly phono-logical paraphasias, and the probability of producing a correct response depends on the frequency and length of the target word. This deficit occurs in the presence of spared ability to find the grammatical gender of the items that he cannot name, intact conceptual knowledge, and very good reading and word repetition. Therefore, we conclude that DPI's deficit is restricted to the phonological retrieval of a correctly selected lexical entry. However, production errors are not uniform across semantic and syntactic domains. Numerals and names of days and months are totally spared compared to matched controls. In addition, abstract nouns and verbs are significantly less affected than concrete nouns, even when variables affecting phonological retrieval (frequency, length, syllabic structure) are controlled for. This suggests that a functional organisation in terms of semantic and syntactic variables exists at the level of phonological retrieval. We discuss these findings in light of current models of speech production.     

AN INFLUENCE OF SYNTACTIC AND SEMANTIC VARIABLES ON WORD FORM RETRIEVAL

We report the case of DPI, an aphasic patient who shows a phonological impairment in production that spares certain syntactic and semantic categories. On a picture naming task, he produces mostly phono-logical paraphasias, and the probability of producing a correct response depends on the frequency and length of the target word. This deficit occurs in the presence of spared ability to find the grammatical gender of the items that he cannot name, intact conceptual knowledge, and very good reading and word repetition. Therefore, we conclude that DPI's deficit is restricted to the phonological retrieval of a correctly selected lexical entry. However, production errors are not uniform across semantic and syntactic domains. Numerals and names of days and months are totally spared compared to matched controls. In addition, abstract nouns and verbs are significantly less affected than concrete nouns, even when variables affecting phonological retrieval (frequency, length, syllabic structure) are controlled for. This suggests that a functional organisation in terms of semantic and syntactic variables exists at the level of phonological retrieval. We discuss these findings in light of current models of speech production.     

An exaggerated effect for proper nouns in a case of superior written over spoken word production

We describe a brain-damaged subject, RR, who manifests superior written over spoken naming of concrete entities from a wide range of conceptual domains. His spoken naming difficulties are due primarily to an impairment of lexical-phonological processing, which implies that his successful written naming does not depend on prior access to the sound structures of words. His performance therefore provides further support for the "orthographic autonomy hypothesis," which maintains that written word production is not obligatorily mediated by phonological knowledge. The case of RR is especially interesting, however, because for him the dissociation between impaired spoken naming and relatively preserved written naming is significantly greater for two categories of unique concrete entities that are lexicalised as proper nouns-specifically, famous faces and famous landmarks-than for five categories of nonunique (i.e., basic level) concrete entities that are lexicalised as common nouns-specifically, animals, fruits/vegetables, tools/utensils, musical instruments, and vehicles. Furthermore, RR's predominant error types in the oral modality are different for the two types of stimuli: omissions for unique entities vs. semantic errors for nonunique entities. We consider two alternative explanations for RR's extreme difficulty in producing the spoken forms of proper nouns: (1) a disconnection between the meanings of proper nouns and the corresponding word nodes in the phonological output lexicon; or (2) damage to the word nodes themselves. We argue that RR's combined behavioural and lesion data do not clearly adjudicate between the two explanations, but that they favour the first explanation over the second.     

Grammatical category dissociation in multilingual aphasia

Word retrieval deficits for specific grammatical categories, such as verbs versus nouns, occur as a consequence of brain damage. Such deficits are informative about the nature of lexical organization in the human brain. This study examined retrieval of grammatical categories across three languages in a trilingual person with aphasia who spoke Arabic, French, and English. In order to delineate the nature of word production difficulty, comprehension was tested, and a variety of concomitant lexical–semantic variables were analysed. The patient demonstrated a consistent noun–verb dissociation in picture naming and narrative speech, with severely impaired production of verbs across all three languages. The cross-linguistically similar noun–verb dissociation, coupled with little evidence of semantic impairment, suggests that (a) the patient has a true “nonsemantic” grammatical category specific deficit, and (b) lexical organization in multilingual speakers shares grammatical class information between languages. The findings of this study contribute to our understanding of the architecture of lexical organization in bilinguals.     

Grammatical class and context effects in a case of pure anomia: Implications for models of language production

Numerous variables have been used in previous attempts to account for the performance of aphasic patients on naming tasks. These include lexical/semantic factors such as frequency, and contextual factors such as whether the target is produced as a single word or in a sentence. This paper reports the case of a pure anomic patient (HY) who was strongly affected by both lexical and contextual factors in his naming. Beyond a strong frequency effect, a marked grammatical class difference was noted. Verb production was far superior to noun production in single word naming. Production of nouns was highly facilitated by provision of a semantically relevant sentence frame. These two findings were related to current models of single word and sentence production; a functional locus for HY's naming deficit is proposed.     

Functional neuroimaging of grammatical class: Ambiguous and unambiguous nouns and verbs

Neuropsychological and neuroimaging studies of grammatical-class differences suggest that nouns and verbs may be associated with different brain regions. The current functional magnetic resonance imaging (fMRI) study focused on auditory grammaticality judgements of two-word stimuli consisting of noun/verb ambiguous and unambiguous nouns and verbs preceded by either “the” or “to”. The fMRI results showed effects of class ambiguity in the left inferior frontal gyrus, possibly due to greater selection demands, and effects of grammaticality (“yes” vs. “no” response) in left superior temporal gyrus, consistent with greater activation for trials with greater conflict. In addition, posterior left temporal lobe regions showed increased activity on unambiguous trials for verbs than for nouns, consistent with prior imaging studies. Thus, in a task specifically focused on the grammatical (rather than semantic) aspects of words and that used morphologically simple nouns and verbs controlled for imageability, verbs preferentially activated posterior temporal but not frontal lobe regions. However, for ambiguous trials in this same region, nouns showed greater activation than verbs, suggesting that these effects can be modulated by class ambiguity.     

The syntax of single words: Evidence from a patient with a selective function word reading deficit

We describe the reading performance of a patient who has selective deficits for reading nonwords, function words, and morphologically complex words in isolation. His reading of highly abstract nouns and verbs, however, is relatively well preserved. He can recognise and comprehend the meaning of written function words, of derivational morphology, and of most inflectional morphology. We suggest that his deficit in reading grammatical morphemes is unrelated to his problems in reading nonwords and cannot be explained by their low semanticity and imageability. The patient's speech is ungrammatical but is not devoid of grammatical morphemes and his reading of functional elements improves when these are presented within the context of sentences. We argue that syntactic information relevant to individual lexical items including information about how the word may potentially be used within a phrase must be accessed during single word reading tasks (e.g., Levelt, 1989). This is particularly difficult for function words due to their linguistic specification, which is different from that of lexical categories (Chomsky, 1995). Both linguistic theory and Garrett's (e.g., 1982) model of sentence processing account for the patient's improved reading of function words in the context of sentences.     

The syntax of single words: Evidence from a patient with a selective function word reading deficit

We describe the reading performance of a patient who has selective deficits for reading nonwords, function words, and morphologically complex words in isolation. His reading of highly abstract nouns and verbs, however, is relatively well preserved. He can recognise and comprehend the meaning of written function words, of derivational morphology, and of most inflectional morphology. We suggest that his deficit in reading grammatical morphemes is unrelated to his problems in reading nonwords and cannot be explained by their low semanticity and imageability. The patient's speech is ungrammatical but is not devoid of grammatical morphemes and his reading of functional elements improves when these are presented within the context of sentences. We argue that syntactic information relevant to individual lexical items including information about how the word may potentially be used within a phrase must be accessed during single word reading tasks (e.g., Levelt, 1989). This is particularly difficult for function words due to their linguistic specification, which is different from that of lexical categories (Chomsky, 1995). Both linguistic theory and Garrett's (e.g., 1982) model of sentence processing account for the patient's improved reading of function words in the context of sentences.     

Action performance and action-word understanding: Evidence of double dissociations in left-damaged patients

It has been proposed that language and action representational systems overlap when the tasks used to assess them involve the same stimuli and require abilities acquired at similarly early developmental stage. We matched variables at task and stimulus level to test this hypothesis in a group of 12 left-damaged patients (and 17 controls). At the patients' group level, we replicated previously reported correlations between linguistic and nonlinguistic tasks. When performances were analysed individually, however, double dissociations were observed between the ability to imitate pantomimes and the ability to produce and comprehend the corresponding action verbs, as well as between the ability to use tools and the ability to comprehend the corresponding tool nouns. These findings suggest that processing action words is independent of the ability to produce the associated object-directed actions. Double dissociations were also found between the ability to comprehend action verbs and the ability to comprehend tool nouns. Moreover, action and tool naming showed differential effects of age of acquisition, suggesting that the two word categories meet the lexical organization by word class (nouns and verbs), even when related to identical action concept. Dissociations at behavioural level are supported by anatomical dissociations shown in the analysis of patients' lesions.     

Differential integration efforts of mandatory and optional sentence constituents

We investigated the online sensitivity of the semantic integration system to the different roles played by sentence constituents that are necessary (verbs and nouns) or optional (adjectives) for argument completion. We compared the effect of semantic incongruities introduced in both types of words on the N400 ERP component. Participants read sentences for meaning, half of which were rendered anomalous by an incongruent verb, noun, or an early/late adjective. Incongruent adjectives led to smaller N400 effects than did incongruent nouns and verbs, and the congruity effect for sentence-final adjectives was not significant. All incongruities are therefore not created equal: Incongruent optional sentence constituents create less of an integrative burden than incongruent mandatory sentence constituents, suggesting that online sentence integration processes are sensitive to the distinct roles played by different words in shaping sentence meaning.     

Nouns, verbs, objects, actions, and the animate/inanimate effect

We report an aphasic patient, Z.B.L., who showed a significant advantage for verbs compared to nouns in picture-naming tests. Within the object class, he performed better on animate things than on nonliving things in picture naming as well as in an "attribute judgement task". This pattern of performance is contrary to the central prediction of a recent proposal (Bird, Howard, & Franklin, 2000), which attributes noun-verb dissociation in aphasic patients to deficits in processing certain kinds of semantic features. This model proposes that conceptual representations of verbs have a lower proportion of sensory features than do representations of nouns; the same is proposed for inanimate versus animate items within the noun category. Noun deficits are assumed to arise due to impairment for the processing of sensory features. The model predicts that if a patient is more impaired for nouns than for verbs, he will also display more difficulty with animate than with inanimate objects. Contrary to predications derived from this theory, Z.B.L. performed better with animate than inanimate nouns.     

Why the leash constrains the dog: the impact of semantic associations on sentence production

The production of language is one of the most complex and amazing skills in humans. Increasing evidence demonstrated that associative relations (e.g., car - garage) play an important role during concept formation but during speech production the effects and processing of associations are highly debated. Hence, the present study investigated the impact of associations and different SOAs on the production of sentences (Experiment 1) and on naming objects (Experiment 2). In an adapted version of the picture-word interference task, participants were asked to name two pictures using a standardized sentence (e.g., "The car is to the left of the trousers"). Thereby, a simultaneous (SOA is 0 ms) or slightly preceding (SOA is -150 ms) auditory or visual distractor had to be ignored. Distractors were related to the first noun (for example: "The car is to the left to the trousers", distractor: "garage") or to the second noun (distractor: "belt") or unrelated to both nouns (distractor: "bottle") of the sentence. At simultaneous presentation, visual and auditory distractors related to the first noun of the sentence prolonged naming responses (i.e., interference). For slightly preceding distractors, only visual presentation induced interference for the first noun of the sentence. During no condition, longer naming responses were found for the second noun of the sentence. These effects suggest that associatively related concepts are active during speech production and can be competitors, i.e., they lead to semantic interference. In Experiment 2, subjects had to name an object (e.g., car) while ignoring a visually presented distractor (e.g., motor). The stimulus set was the same as in Experiment 1. The results showed a facilitation effect if the distractor and the target were associatively related. Overall, the current results provide new insight in the models of speech production: while during single word production, associations facilitate naming, they interfere during sentence production. Hence, associations have an important influence on producing speech but the impact is varied by the context, i.e., single word or sentential.     

Action performance and action-word understanding: evidence of double dissociations in left-damaged patients

It has been proposed that language and action representational systems overlap when the tasks used to assess them involve the same stimuli and require abilities acquired at similarly early developmental stage. We matched variables at task and stimulus level to test this hypothesis in a group of 12 left-damaged patients (and 17 controls). At the patients' group level, we replicated previously reported correlations between linguistic and nonlinguistic tasks. When performances were analysed individually, however, double dissociations were observed between the ability to imitate pantomimes and the ability to produce and comprehend the corresponding action verbs, as well as between the ability to use tools and the ability to comprehend the corresponding tool nouns. These findings suggest that processing action words is independent of the ability to produce the associated object-directed actions. Double dissociations were also found between the ability to comprehend action verbs and the ability to comprehend tool nouns. Moreover, action and tool naming showed differential effects of age of acquisition, suggesting that the two word categories meet the lexical organization by word class (nouns and verbs), even when related to identical action concept. Dissociations at behavioural level are supported by anatomical dissociations shown in the analysis of patients' lesions.     

Treatment of thematic mapping in sentence comprehension: implications for normal processing

Abstract

The inability of some aphasic patients to interpret semantically reversible sentences has been hypothesised to arise from failure to link the grammatical roles of nouns (e.g. subject, object) to their corresponding thematic roles (e.g. agent, patient). Several previous attempts to improve patients' thematic mapping abilities have demonstrated a range of treatment effects of considerable relevance to the development of cognitive models of sentence processing. This study reports a new treatment approach to thematic mapping impairment; it succeeded in improving auditory sentence comprehension in a chronic aphasic patient with a long-standing comprehension deficit. Generalisation of improvement to auditory comprehension of sentences with untreated verbs, to comprehension of written sentences, and to tests using pictorial and videotaped materials not used in treatment, place constraints on the range of possible interpretations of the functional locus of treatment effects. Two areas that did not show significant improvement following treatment included auditory comprehension of sentences lengthened with modifiers, and spoken production of active and passive sentences that express correct thematic roles. These null effects are also interpreted as providing information relevant to models of sentence processing, including the role of working memory in sentence comprehension and the nature of thematic mapping procedures in comprehension and production.     

Cerebellar involvement in verb generation: an fMRI study

A possible role of the human cerebellum in the generation of verbs corresponding to presented nouns has been suggested. Previous functional brain imaging studies have compared generation of verbs with the reading of nouns as a measure of verb generation. In the present fMRI study involving healthy human subjects, the effects of speech articulation and motor imagery associated with verb production were investigated in greater detail. Generation of verbs to visually presented nouns was compared to a condition in which subjects read those same verbs that had been individually generated by each subject. Activation in lobule HVI/Crus I of the right cerebellar hemisphere was found as a measure of verb generation. In contrast, reading of verbs as a measure of speech articulation evoked cerebellar activations in both left and right paravermal lobule VI. These results suggest an involvement of the right lateral cerebellar hemisphere in linguistic functions during verb generation. Alternatively, effects of inner speech could also possibly explain the results.