The role of grammatical gender during word recognition

This article concerns the role of grammatical gender during noun recognition. Tt starts with a description of French grammatical gender. This description shows that there is no rule of general application that accounts for the noun repartition in masculine and feminine categories. However, the nominal categorization system reveals formal regularities related to the distribution of noun endings in the two categories of gender. For example, the ending "-at" mostly appears in masculine nouns. Thus, the probability for a noun ending in "-at" to be masculine is higher than its probability to be feminine. Several studies have shown that French speakers have an implicit knowledge of these regularities and exploit them in order to categorize a noun or a pseudo-word. The absence of categorization rubs suggests that grammatical gender information is associated to each noun in the mental lexicon and the presence of exploited regularities suggests that gender information might be retrieved without access to the noun representation. There is no representational model that postulates a double way to retrieve gender information. However, it is precisely the hypothesis of a double way that allows for an influence of grammatical gender during the recognition of isolated nouns. Recent data are compatible with this hypothesis. Finally the article provides a review of experimental data concerning the role of grammatical gender on word recognition by investigating the influence of gender information provided by the context in which a noun appears. Research reveals an influence of grammatical context, although the intervention of this influence remains controversial.