THE EFFECTS OF THE CODABILITY AND DISCRIMINABILITY OF THE REFERENTS ON THE COLLABORATIVE REFERRING PROCEDURE

Clark & Wilkes‐Gibbs (1986) proposed a model of verbal cooperation accounting for how the participants in a conversation coordinate in the making of references. The model is conceived in terms of how a process of mutual acceptance gets initiated, carried through, and completed in a task where people converse about arranging complex figures into one particular order. One aim of the present study was to test the replicability of Clark & Wilkes‐Gibbs' findings with French native speakers. It also aimed at specifying how the relative complexity of the referents as perceived by the interlocutors affects their referring procedure. Two aspects of this complexity were examined within a repeated referential task: the codability of the referents (assessed by the ease with which a figure can be interpreted and linguistically expressed) and their discriminability (assessed by the ease with which a figure can be differentiated from others). First, the present observations gained with French‐speaking interlocutors replicate Clark & Wilkes‐Gibbs' findings to a remarkable degree. Next they show that the more codable and discriminable the figures, the fewer words and turns the participants take in the task: more collaborative effort is required in referring to items that are harder to name and harder to tell apart. Finally they also show that the more codable and discriminable the figures, the more likely they are to be described with holistic than with segmental definite descriptions. 1991 The British Psychological Society