This research describes a method that initially establishes a mand repertoire as the first component in producing echoic and tact repertoires in young children with limited verbal repertoires. The subjects were 3 nonverbal children with autism, aged 2.5 to 3.5 years. The results indicated that establishing a mand repertoire as the starting point for echoic training produced the acquisition of an initial echoic repertoire in all 3 children within the first 10 sessions. Two of the children also began tacting during the same period. A mand repertoire was acquired by all subjects by the sixth session. Negative vocal behavior was rapidly reduced without the use of aversive contingencies by shaping inapproapriate vocalizations into acceptable vocal mands. Our discussion analyzes those contingencies that contribute to the effectiveness of the manding procedure and compares this procedure with the conventional vocal imitation model. We also discuss variables that may contribute to the long-term success of language training programs and describe strategies for preventing language delay in typical infants and children.
