In Experiment I, a 3-min tone that preceded a free pellet of food suppressed variable-interval performances maintained by the same type of pellets, but failed to elicit conditioned changes in the heart rates and blood pressures of two rhesus monkeys. Initially severe, the prereward suppression became temporally discriminated to progressively later portions of the tone, and was maintained at an attenuated level for over four months. The suppression was apparently not caused by interfering autonomic respondents, nor was it superstitiously conditioned, since 21 of the initial 25 tone-food pairings took place outside of baseline sessions. In Experiment II, a 1-min light, paired with four free pellets of food, suppressed the variable-interval responding of a second pair of similarly trained monkeys. An interresponse-time analysis showed that in one subject, mild prereward suppression of responding developed through two stages. On early trials, response rate slowed by 10% throughout the prefood interval. On later trials, the animal suppressed by pausing for a like portion of the interval, most often near the end, but otherwise responded normally during the prefood signal.Both prereward and preaversive stimuli have now been shown to share the property of suppressing an animal's response rate when superimposed upon certain appetitive schedules. Emotional interpretations of the disruptions occasioned by preshock warning signals have been widely held ever since the expression "conditioned anxiety" was first used by Estes and Skinner (1941) However, what appears common to a number of emotional explanations of operant-respondent interactions is the proposition that suppression of operant behavior is dependent upon some internally conditioned events. For example, in the literature that speaks of preaversive suppression as fear conditioning (Kamin, 1965;McAllister and McAllister, 1971;Miller, 1951), changes in operant responding are never treated as the emotion itself, but rather as overt skeletal by-products of some normally unobserved conditioned physiological response, the latter (coupled with cognitive inferences) defining the real emotion. So described, an emotion is an intervening variable, linking certain antecedent conditions to a behavioral change. In turn, the change, suppression, is viewed as a quantitative index of the inferred emotional state, an eye on hidden events. These internal physiological responses are held to have been incompatible with, hence to have interfered with, the subject's skeletal activities during the preshock stimulus, resulting in a decrease in operant work.A similarly reasoned emotional interference hypothesis has been proposed by Azrin and Hake (1969) to account for the suppressant properties of prereward stimuli. Noting dur-93 1973, 20,[93][94][95][96][97][98][99][100][101][102][103][104] NUMBER I (JULY)