Several writers have recently implied that organisms may have been sensitive to an increase in the mere probability of reinforcement when no reinforcer is immediately contingent upon a response. I do not think that the possibility of a conditioned reinforcer has been satisfactorily eliminated as an explanation, but I will rest my case on the following experiment, which takes advantage of the fact that the role of a reinforcer is clearer in shaping behavior than in maintaining it. (p. 169)
