Electrophysiological correlates of nouns and verbs: Grammatical class effects elicited by abstract nouns and verbs

There is no consensus as to whether the processing of single nouns and verbs requires access to grammatical class knowledge, and it is difficult to find evidence for differences in the brain between the representations of nouns and verbs as grammatical categories. In previous studies noun-verb differences were conflated with (i) differences between object and action knowledge; (ii) differences between motoric and sensory features of nouns and verbs; (iii) morphological markers of nouns and verbs. In the present study, we recorded ERPs while participants carried out a lexical decision (LD) task and a grammatical class judgement (GcJudge) task on abstract and concrete nouns and verbs. GcJudge task is expected to tap more directly than tasks previously used into information about the grammatical class of words. The materials for the experiments were selected to reduce or eliminate the confounding between nouns and objects, and verbs and actions. No differences in the LD task with respect to grammatical class in RTs or ERPs were found. In contrast, in the GcJudge task faster responses to abstract verbs compared to abstract nouns, and faster responses to concrete nouns compared to verbs were found. ERP responses complemented this result regarding the abstract items. More negative responses to abstract verbs than to abstract nouns in the left anterior regions, and more negative responses to abstract nouns than to abstract verbs were found in the posterior regions within a time frame from 350 to 650 ms. We suggest that the grammatical information attached to verbs and nouns is activated even when single words are processed. The detection of this effect may be dependent on the materials used (word lists that do not elicit motor and/or sensory activations in particular abstract nouns and verbs) and to some extent on the task (in this case grammatical class judgement, but other tasks may also be suitable)