The literature contains indications that the time intervals between responses, i.e., the interresponse times (IRTs), made by rats depend on which IRTs have been followed more often by food reinforcements. In the present study evidence on the occurrence and nature of this dependence was obtained by comparing IRT distributions with reinforced-IRT distributions, and by altering the allocation of reinforcements to IRTs (reinforcement of an IRT refers to reinforcement of the terminal response of that IRT).The first indication that the IRT distribution is affected by the relative reinforcement of IRTs of different lengths came from the observation of Skinner (4) that fixed interval rein-1 This article is based on a thesis submitted in partial fulfillment of the requirements for the Ph.D. degree at Harvard University, 1955.

Animals learn to avoid with the Sidman procedure even though the avoidance response is not followed by the termination of any warning stimulus in the environment. What reinforces this response? The accepted explanation has been that the avoidance response is reinforced when it terminates other behavior that has become aversive by pairing with shock. However, the reinforcement may also be derived from the temporal discriminations that develop with Sidman avoidance. These and other temporal discriminations show that the animal has available some events that vary with the postresponse time. The shock will closely follow the temporal stimuli at long postresponse times and would be expected to make them aversive. The stimuli at short postresponse times would have a relatively low aversiveness due to their more remote relation to shock. Since the avoidance response changes a long postresponse time to a short one, that response would be followed by a decrease in aversiveness which would reinforce it. When sharp temporal discriminations are present, reinforcement from the decrease in aversiveness of temporal stimuli probably plays a dominant role in maintaining the avoidance response. This formulation fits the available data and has adequate answers for the objections that have been raised to earlier conceptions of the role temporal discriminations might play in Sidman avoidance. Although under some conditions the reinforcement in Sidman avoidance seems to be primarily due to the decrease in aversiveness of temporal stimuli, under other conditions there probably is reinforcement from the termination of conditioned aversive responses.

Pigeons were given repeated two-day conditionings alternating with eight-day extinctions using a trial procedure. One group had different key colors during each of the first five conditioning-extinction pairs; another group had the same key color throughout. Total extinction responses of both groups were quite constant over successive extinctions. This finding differs from the rapid declines found in most previous studies with bar-press and key-peck responses. The difference probably was due to our longer extinctions, because responses early in each extinction did decrease. However, that decrease was neutralized by increases in responses late in each extinction. The two opposite changes indicate the influence of two different factors during repeated extinctions, with neither factor having much stimulus specificity. The reduction of early responses may result from feeding changes confounded with extinction. The increase in later extinction responses may result from a decrease in the effect of unreinforced responses after their repeated occurrence. Comparisons among the above studies indicate that the following influences on repeated extinctions are likely, though in view of extensive confouinding of variables these conclusions are tentative until carefully controlled comparisons are made. A decrease in successive extinctions seems to be favored by (1) use of bar-press or key-peck responses rather than alleys or jumping stands, (2) measurement of number of responses in extinction rather than latencies or running times, (3) massing of extinction responses by free-operant techniques or short ITIs as compared with several minute or 24-hr ITIs, (4) increase in extinction trials per extinction, (5) not giving reinforcements following the extinction and within the same daily session, (6) increasing the difference between the ITI during extinction and the ITI between the last unreinforced trial and the first reinforced trial, and (7) continuous rather than intermittent reinforcement.Most of the above studies involved rather brief extinction periods, often only part of a session, and few data are available concerning repetition of long extinction periods. Jenkins (1961) gave pigeons two 10-day extinctions and found the usual decline, but the duration of his second conditioning was less than one 181 1976, 26,[181][182][183][184][185][186][187][188][189][190] NUMBER 2 (SEPTEMBER)

Rats' responding was stabilized for over 35 days on 4-min variable-interval reinforcement. Reinforcements per hour for 4-sec wide classes of interresponse times were then separately controlled by adjusting those for each class to the variable-interval values that had just prevailed. This produced little or no change in interresponse times, indicating that the new procedure was substantially equivalent to a variable-interval schedule. The variableinterval schedule produced a high and stable conditional probability of interresponse times in the 0-to 4-sec class, associated with a peak in reinforcements per hour for this class.Reducing the reinforcements per hour for this class while raising that for another class (by 3.3 reinforcements per hour) significantly reduced the conditional probability of 0-to 4-sec interresponse times. Restoring the 3.3 reinforcements per hour to the 0-to 4-sec class significantly elevated the conditional probability of interresponse times in this class. Hence, it is concluded that the distribution of interresponse times produced by a subject during some variable-interval schedules is determined partly by the relative reinforcement of different interresponse times that the variable-interval schedule provided.

A differential-reinforcement-of-other-behavior (DRO) schedule with trials and delayed reinforcement was investigated. Periodically a wheel was briefly available to rats, followed six seconds later by brief availability of a bar. Variable-ratio food reinforcement of wheel turns was adjusted to give 95% turns. After variable-ratio-five reinforcement of bar presses produced 100% pressing, then separate ratio schedules were used for presses following turns (turn presses) and presses following nonturns (nonturn presses). Increasing nonturn-press reinforcements decreased turns, even though total reinforcements increased. Reversal by decreasing nonturn-press reinforcements raised turns, though with hysteresis. Thus food reinforcement increased nonturns even though delayed six to ten seconds after nonturns, a delay that greatly reduces response reinforcement. Those and other results indicate that the turn decrease was not due to reinforcement of competing responses. Evidence against other alternatives, and the reduction of responding by increased reinforcement, indicate that the term inhibition is appropriate for the phenomenon reinforced. Response-specific inhibition appears appropriate for this particular kind, since its effects are more specific to particular responses than Pavlovian conditioned-inhibition. Response-specific inhibition seems best considered a behavioral output comparable to responses (e.g., both reinforcible) but with important properties different from responses (e.g., different reinforcement-delay gradients).