An experimental analysis of ongoing verbal behavior: reinforcement, verbal operants, and superstitious behavior.
Shaping guides live adult conversation toward any topic you pick, and random feedback spawns superstitious speech.
01Research in Context
What this study did
The experimenter talked with two adults. Every time the adult said something close to a hidden topic, the experimenter nodded and said "mm-hmm." That was the only reinforcement.
The topic changed each phase: first sports, then jobs, then back to sports. The last phase gave random beeps instead of social praise. The design was ABAB so each person served as their own control.
What they found
When praise followed sports words, the adults quickly filled the air with sports talk. When praise followed job words, the chat shifted to careers.
Random beeps created weird, superstitious themes. One adult started listing cities; another rattled off numbers. The moment real praise returned, the old topic snapped back.
How this fits with other research
Mace et al. (1990) did the same thing with preschoolers. Kids who could describe their own button pressing soon pressed at the shaped rate. Together the studies show shaping works from age 3 to adult.
Rosenfeld et al. (1970) hid the contingency inside a tape recorder. Adults thought they were running the study, yet their own request forms still shifted. Critchfield (1996) makes the effect visible in real conversation, not hidden tasks.
Davison et al. (1989) saw rats chew tokens after algorithmic shaping. Random beeps here created the same kind of odd, off-target talk. Both papers warn: non-contingent feedback breeds superstitious side habits.
Why it matters
Your social attention is a reinforcer. If you want a client to stay on topic during conversation, reinforce comments that fit the theme. If you toss out praise randomly, you may accidentally strengthen odd tangents. Track your own nodding and "uh-huh"—it shapes talk faster than you think.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick a social topic, reinforce on-topic comments with smiles and nods, and watch the chat stay anchored.
02At a glance
03Original abstract
Four adult humans were asked to asked to "find" and talk about a particular topic to a person in an adjoining room, and were instructed that they would hear a short beep (the only form of reply from the other person) when they were talking about the topic, or were "close" to the topic. In Session 1, the experimenter in the adjoining room presented the beeps in the manner of shaping, or the differential reinforcement of successive approximations, "toward" the designated topic. In Session 2, the same conditions were in effect but the experimenter was unable to hear the subject and the beeps were presented noncontingently in a way that roughly matched the frequency and distribution of presentations in Session 1. In Session 3, shaping conditions were again in effect but with a different topic than that designated for Session 1. Audio recordings were transcribed in a way that was designed to show the progress of shaping over time. These and additional forms of supporting data and accompanying rationale are presented and discussed in detail. Issues raised by the methodology and results of the experiment include the nature of the verbal operant, superstitious verbal behavior, and a variety of methodological issues relevant to the experimental analysis of ongoing or continuous verbal behavior.
The Analysis of verbal behavior, 1996 · doi:10.1007/BF03392908