A study was conducted in light of a neo-Piagetian theory of cognitive development to interpret the effect of priming executive schemes for describing relevant features in a referential communication task. Forty-eight 4-year-old children were divided into primed and nonprimed groups and were asked to describe in isolation a black and white photograph of a dog, or a person's facial expression well enough so that the adult listener could pick out the referent from among an array of nonreferent photographs of the same stimulus set. If the speaker's spontaneous initial encoding was not adequate for distinguishing the referent from the nonreferent array, two forms of feedback were provided by the listener: (1) non-specific verbal feedback of the form ("I still don't know which one it is. Is there anything else you can tell me about it?") and, if an adequate description was still not given; (2) specific visual contrasts (i.e., nonreferents) that fit the description already provided by the subject but were inconsistent with respect to those relevant features not yet mentioned were provided. The cognitive style field dependence-independence and mental processing capacity were assessed pre-experimentally, as well as the priming effect; vocabulary repertoire was assessed post-experimentally. Differences in communication performance between the two groups were not significant except in complex interactions with sex and stimulus order. Explanation for the relative lack of treatment effect were discussed with respect to a production deficiency and a mediation deficiency. (Author/MP)