Correspondence was developed between children's non‐verbal and verbal behavior such that their non‐verbal behavior could be altered simply by reinforcing related verbal behavior. Two groups of six children each were given food snack at the end of the day: for reporting use of a specific preschool material during free play (procedure A); and then only for reports of use which corresponded to actual use of that material earlier that day (procedure B). Initially, procedure A alone had little or no effect on the children's use of materials. Procedure B resulted in all of the children in one group actually using a specific material, and after repeating procedures A and B with this group across a series of different materials, procedure A alone was sufficient to significantly increase use of a specific material. Correspondence between verbal and non‐verbal behavior was produced such that, in this group of 4‐yr‐old disadvantaged Negro children, “saying” controlled “doing” 22 or more hours later. In the second group, procedure B initially did not increase the use of a specific material; rather, the children's reports decreased so as to correspond to the intermittent use of the material. It appeared from subsequent procedures with this group that maintenance of a high level of reporting was crucial to the saying‐then‐doing correspondence seen in the first group. 1968 Society for the Experimental Analysis of Behavior
