Many human activities are related to verbal behavior. Verbal behavior is a type of behavior that is reinforced through the mediation of a listener specially prepared by a verbal community. Critics consider this mediation as insufficient to explain at least two linguistic phenomena: learning subtle verbal responses, such as grammatical constructions, and linguistic productivity. An important feature of verbal behavior is that speaker and listener can be the same person. A person can stimulate herself and be reinforced automatically. The term parity is used to describe a particular type of response in which a speaker, who is also listener, perceives if his verbal responses diverge or converge in relation to verbal models. It is assumed that the convergence is automatically reinforcing. Among the verbal models are vocals descriptions in the passive voice. At least two studies have investigated the role of modeling, explicit reinforcement and automatically reinforcement through parity in the control of vocal verbal responses. Preschool children participated in the studies. During the experimental phases, children heard the experimenter describe drawings in the passive voice and then were asked to describe similar drawings. Descriptions in the active voice or undefined voice were explicitly reinforced with stickers and praise; descriptions in the passive voice were followed with minimal vocalizations ("hã", "hum"). In both experiments, participants began to use the passive voice after these responses have been modeled, and emitted new sentences in the passive voice. A detailed analysis of reported studies allows the identification of limitations, including: lack of control over maturation and reactivity to the experimental conditions; lack of pre-experimental measures on the existence or extent of the passive voice repertoire, speaker-listener repertoire and tact repertoire of the stimuli presented; lack of determination on the reinforcing value of explicit consequences; lack of research on the type of conditioning that gives rise to the reinforcing effect of achieving parity with verbal vocal models. The present research aims to investigate the role of modeling, explicit reinforcement and automatic reinforcement through parity in the selection and maintenance of verbal responses emitted in the passive voice. Preschool children will participate. The procedure will consist on four stages. In the first two stages relevant pre-experimental repertoires will be accessed. On the third stage a reversal design with a conditions sequence of A-B-C-B1-C1-D-A1 will be used. A and A1 are control conditions. On conditions B and B1 the experimenter will provide models on the passive voice, through the description of the drawings, and he will explicit reinforce descriptions emitted by the participant in the active voice or undefined voice, description in the passive voice will be followed by minimal vocalizations. On conditions C and C1 the reinforcement criteria will be the same as the previous conditions, but there will be no modeling by the experimenter. On condition D new drawings will be used, the reinforcement criteria will remains the same and there will be no modeling by the experimenter. Participants that emit less than 30% of descriptions on the passive voice, during the previous stage, will participate on the fourth stage. The same experimental design from stage three will be used, but the reinforcement criteria will be reversed. The research results may help to establish functional relationships and increase understanding about learning subtle verbal responses, as passive voice, and linguistic productivity. As a replication, the study verifies the reliability and extends previous researches.
News published in Agência FAPESP Newsletter about the scholarship: