The purpose of the study was to investigate the hypothesis that requiring a student to verbalize a newly discovered mathematical concept interferes with his ability to use that concept. Five semi-programmed lessons (dealing with function machines, exponents, marker games, geometry, and One Pile Nim) were prepared and taught to a random selection of 118 preservice elementary education majors. Results showed that the required verbalization group performed significantly better than the nonverbal group on the performance tests for the function machine lesson; the author concluded that this indicated that verbalization did not reduce the subject's ability to use a newly discovered concept. For the function machine lesson, the required verbalization group completed significantly fewer target tasks than the nonverbal group; the author concluded that required verbalization interferes with one's subsequent discovery power. The same relationships existed among means in the geometry lesson, although they were not significant. (Author/DT)