Object and action naming in aphasia: A double dissociation?

Initially it was argued that the reason for verb deficits in aphasia and for the delay in acquisition is that verbs are more difficult to process than nouns (e.g., Saffran, Schwartz, & Marin, 1980). The complexity of verbs, however, cannot explain the phenomenon that some anomic patients present with superior verb production in the face of impaired noun production (e.g., Zingeser & Berndt, 1988). This finding led to the double dissociation view of noun-verb differences. In the present study we re-examine this view by reviewing past lesion studies reporting noun-verb dissociations in picture naming. We reviewed 37 studies that reported data of 269 patients published between 1984 and 2005 that report object and action naming performance in aphasic patients. We compiled information about the clinical diagnosis of each patient, and the direction and magnitude of difference between naming the two picture types. Of the 269 aphasic patients, 39 did not show a noun-verb difference, 28 presented with relative noun deficits and 201 patients with relative verb deficits. While the majority of patients with selective noun deficits were classified as anomics or Wernicke’s aphasics, patients with verb deficits were assigned different diagnostic categories, including anomia and Wernicke’s aphasia. Since there are theoretical arguments and empirical evidence (Bogka et al., 2003; Szekely et al., 2005; Druks et al., submitted) that verb processing might be more demanding than noun processing, in the review we distinguished between cases that presented with large and cases with smaller differences in the production of nouns and verbs. We argued that small differences may be due to verbs being more difficult and not due to true noun-verb differences. In our further analysis, therefore, we focussed only on cases that reported at least 30% difference between naming the two types of pictures. Using this criterion, 25 studies that described 42 patients remained. Among the 42 patients, 11 presented with relative noun deficits and 31 with relative verb deficits. The noun deficit group consisted of anomic and fluent patients only, while the verb deficit group consisted of both non-fluent (22, Broca’s aphasics and agrammatic patients) and fluent (8, anomic and Wernicke’s aphasic patients) and one patient with a mixed diagnosis. All patients with noun deficits, for whom information about lesion site was available, had lesions involving the left temporal lobe. For three of those patients the lesion was confined to the temporal lobe while for the remaining patients additional damage to the left frontal, parietal, and/or occipital lobes was also reported. Lesion sites for the verb deficit group were wide ranging: frontal, temporal, occipital, and/or parietal lobes and the basal ganglia. While the temporal lobe was always involved in noun deficits, frontal lobe lesions were not involved in all cases of verb impairment. The distribution of lesion sites and either noun or verb deficits showed that damage limited to the anterior part of the brain (frontal, parietal lobes or basal ganglia) always led to relative verb deficits, and damage limited to posterior regions - the temporal and occipital lobes always led to relative noun deficits. However, lesions involving both anterior and posterior structures led to indeterminate consequences: lesions described as entirely similar resulted sometimes in noun deficits and sometimes in verb deficits (e.g., De Renzi & Di Pellegrino, 1995; Breedin & Martin, 1996). The literature review showed that there are many more patients with verb deficits than noun deficits. It also showed that while there is a reliable association between noun deficits and anomia, verb deficits are not invariably associated with nonfluent aphasia. Furthermore, temporal lobe lesions are always present in noun impaired patients, but frontal lobe lesions are not always present in verb impaired patients. Finally, lesions underlying verb deficits tend to be more variable and widespread than lesions underlying noun deficits. These findings question the double dissociation view of noun-verb differences. First, the scope of the double dissociation is smaller than it had been implied in the literature because many cases of verb deficits can be accounted for by action naming being generally more difficult than object naming. Second, the more complex relationship between verb deficits and diagnostic category/lesion site than noun deficits and diagnostic category/lesion site suggest that noun deficits are not the mirror image of verb deficits. Third, the more widespread and variable lesions present in patients with verb deficits suggest that verb deficits may be due to a variety of reasons, while noun deficits, being associated with more compact lesions, are more specific. It is possible, therefore, that the underlying reason for noun and verb deficits are different in kind. If so, it is technically incorrect and theoretically misleading to conceptualize the relationship between the two forms of deficit as double dissociation.