A review of empirical studies of verbal behavior.

Oah et al. (1989) read every verbal-behavior experiment they could find. They grouped the studies by Skinner’s six verbal operants: mand, tact, echoic, intraverbal, textual, and transcription. They asked, 'Does the data really show these are separate skills?'

The paper is a narrative review, not a meta-analysis. The authors describe trends rather than crunch numbers.

Across dozens of studies, each operant behaved like its own skill. Teaching mands did not automatically teach tacts, and vice versa. The same pattern held for speaker and listener roles.

Because skills did not spill over, the team recommended training each operant and each repertoire separately.

Fox et al. (2001) later put the idea into action. They used differential reinforcement of appropriate intraverbals to cut attention-maintained swearing in an adult with ID. The study is a direct example of 'train speaker repertoires independently.'

Goldstein (2002) and Cui et al. (2023) both cite the 1989 paper when they review autism language treatments. They show the verbal-operant model is still the backbone of modern interventions such as FCT and milieu teaching.

King et al. (2020) warn that most behavior-analytic reviews, including this one, never report clear search methods. So treat the conclusions as expert opinion, not gospel.

If you write VB-MAPP programs, this review gives you license to keep goals narrow. Train 'mands for missing items' before you worry about 'tacts of the same items.' Check generalization after each operant, not at the end. The paper also reminds you to assess speaker and listener skills separately—an echoic fill-in does not guarantee a listener response.