Extinction of responding maintained by timeout from avoidance

The resistance to extinction of lever pressing maintained by timeout from avoidance was examined. Rats were trained under a concurrent schedule in which responses on one lever postponed shock on a free-operant avoidance (Sidman) schedule (response-shock interval = 30 s) and responses on another lever produced 2 min of signaled timeout from avoidance on a variable-ratio 15 schedule. Following extended training (106 to 363 2-hr sessions), two experiments were conducted. In Experiment 1 two different methods of extinction were compared. In one session, all shocks were omitted, and there was some weakening of avoidance but little change in timeout responding. In another session, responding on the timeout lever was ineffective, and under these conditions timeout responding showed rapid extinction. The within-session patterns produced by extinction manipulations were different than the effects of drugs such as morphine, which also reduces timeout responding. In Experiment 2 shock was omitted for many consecutive sessions. Response rates on the avoidance lever declined relatively rapidly, with noticeable reductions within 5 to 10 sessions. Extinction of the timeout lever response was much slower than extinction of avoidance in all 4 rats, and 2 rats continued responding at baseline levels for more than 20 extinction sessions. These results show that lever pressing maintained by negative reinforcement can be highly resistant to extinction. The persistence of responding on the timeout lever after avoidance extinction is not readily explained by current theories.     

The effects of cocaine on behavior maintained by timeout from avoidance.

Rats were trained on concurrent schedules under which responses on one lever postponed shock (avoidance) and responses on the other lever produced brief (2-min) periods of signaled timeout from avoidance. For 6 rats, timeout from avoidance was programmed on a variable-interval 45-s schedule that generally resulted in rates that were lower than those on the avoidance lever. For another 6 rats, timeout was arranged on a variable-ratio 15 schedule that produced higher baseline rates. Cocaine (3 to 40 mg/kg) produced large, dose-dependent increases in behavior maintained by timeout in both groups of rats. Avoidance responding was also generally increased by cocaine, but the increases were of lesser magnitude. Increases in response rates were seen across a broad range of doses on behavior maintained by either interval or ratio schedules, an outcome that was unexpected on the basis of most studies of cocaine on food-maintained behavior. These results were similar to those of previous studies of the effects of amphetamine on behavior maintained by timeout from avoidance and suggest that stimulant drugs affect behavior maintained under a shock-postponement schedule differently than they affect behavior maintained by timeout from avoidance.     

Increased reinforcement when timeout from avoidance includes access to a safe place

Three experiments investigated the reinforcing value of access to a safe place during timeout from an avoidance schedule. Rats were trained on conjoint schedules in which responding both postponed shock on a free-operant avoidance schedule and produced periods of timeout on fixed-ratio schedules. In some conditions, a shelf was inserted into the operant chamber during timeout, enabling subjects to get off the grid floor. The combination of timeout and shelf maintained substantially higher response rates than the baseline avoidance schedule with ratio requirements as high as 90 (Experiment I). Adding the shelf to timeouts in one component of multiple fixed-ratio schedules of timeout resulted in higher response rates in the component where the shelf was included (Experiment II). When timeouts with and without the shelf were arranged on concurrent schedules, the shelf-timeout combination was preferred, even when of shorter duration than timeout alone (Experiment III). In all three experiments, subjects climbed on the shelf, although all shocks were cancelled during timeout periods. The results could not be accounted for solely in terms of the reinforcing properties of changes in shock rates, but required an interpretation that ascribed conditioned reinforcing value to stimuli associated with such changes.     

Variable-interval schedules of timeout from avoidance

Rats were trained on concurrent schedules in which pressing one lever postponed shock and pressing the other occasionally produced a 2-min timeout during which the shock-postponement schedule was suspended and its correlated stimuli were removed. Throughout, the shock-postponement schedule maintained proficient levels of avoidance. Nevertheless, in Experiment 1 responding on the timeout lever was established rapidly, was maintained at stable levels on variable-interval schedules, was extinguished by withholding timeout, was reestablished when timeout was reintroduced, and was brought under discriminative control with a multiple variable-interval extinction schedule of timeout. These results are in contrast with Verhave's (1962) conclusion that timeout is an ineffective reinforcer when presented to rats on intermittent schedules. In Experiment 2 the consequence of responding on the timeout lever was altered so that the shock-postponement schedule remained in effect even though the stimulus conditions associated with timeout were produced for 2 min. Responding extinguished, indicating that suspension of the shock-postponement schedule, not stimulus change, was the source of reinforcement. By establishing the reinforcing efficacy of timeout with standard variable-interval schedules, these experiments illustrate a procedure for studying negative reinforcement in the same way as positive reinforcement.     

Reductions in shock frequency and response effort as factors in reinforcement by timeout from avoidance

Rats' presses on one lever canceled shocks programmed after variable cycles, while presses on a second lever occasionally produced a 2-min timout during which the shock-delection schedule was suspended and its correlated stimuli removed. These concurrent schedules of avoidance and timeout were embedded in a multiple schedule whose components differed, within and across conditions, in terms of the programmed shock rate associated with the shock-deletion schedule. Analyses based on the generalized matching law suggest that the reduction in the response requirement correlated with termination of the avoidance schedule was a more important factor in the reinforcing effectiveness of timeout than was shock-frequency reduction, at least in 2 of 3 rats. After training in each condition, responding on the timeout lever was extinguished by withholding timeouts in both components over seven sessions. Resistance to extinction varied directly with the rates of both shock-frequency reduction and avoidance-response reduction experienced during training. Although reduction in response effort appeared to dominate shock-frequency reduction in the maintenance of responding, neither factor had a clear advantage in predicting the course of extinction.     

Variable-interval schedules of timeout from avoidance: effects of chlordiazepoxide, CGS 8216, morphine, and naltrexone.

Rats were trained on concurrent schedules in which pressing one lever postponed shock and pressing the other occasionally (variable-interval schedule) produced a 2-min timeout during which the shock-postponement schedule was suspended and its correlated stimuli were removed. These procedures provided a baseline for studying the effects of drugs on behavior maintained by different sources of negative reinforcement (shock avoidance and timeout from avoidance). Experiment 1 studied a benzodiazepine agonist, chlordiazepoxide, and antagonist, CGS 8216. Chlordiazepoxide (2.5-30 mg/kg) had little effect on avoidance responding except at higher doses, when it reduced responding. By comparison, responding on the timeout lever was increased in 5 of 6 rats. These effects were reversed by CGS 8216 (2.5-5 mg/kg) in the 2 rats tested, but CGS 8216 had no effect by itself. Experiment 2 studied an opiate agonist, morphine, and antagonist, naltrexone, with 3 rats. Morphine's (2.5-20 mg/kg) effects were opposite those of chlordiazepoxide: At doses that either increased or had no effect on avoidance responding, morphine depressed timeout responding. Naltrexone (5 mg/kg) reversed these actions but had no effect by itself.     

Variable-ratio schedules of timeout from avoidance: effects of d-amphetamine and morphine.

Rats were trained on concurrent schedules in which pressing one lever postponed shock and pressing the other lever produced periods of signaled timeout from avoidance on variable-ratio schedules. These procedures generated high rates of timeout-reinforced responding and provided a baseline for studying the effects of drugs on behavior maintained by different types of negative reinforcement (shock postponement vs. timeout). Morphine (2.5 to 10.0 mg/kg) reduced behavior maintained by timeout at doses that increased or had no effect on avoidance responding. In contrast, d-amphetamine (0.125 to 2.0 mg/kg) produced large increases in timeout responding at doses that had minimal effect on avoidance in rats trained on variable-interval and variable-ratio schedules. Thus, the event-dependent effects of morphine, observed in previous studies in which timeout responding was maintained at low rates by interval schedules, were replicated with high timeout rates maintained by variable-ratio schedules. The effects of d-amphetamine could also be described as "event dependent" because timeout responding was stimulated more than avoidance regardless of the maintenance schedule or baseline rate.     

Avoidance of timeout from response-independent food: effects of delivery rate and quality

In three experiments, a rat's lever presses could postpone timeouts from food pellets delivered on response-independent schedules. In Experiment 1, the pellets were delivered at variable-time (VT) rates ranging from VT 0.5 to VT 8 min. Experiment 2 replicated the VT 1 min and VT 8 min conditions of Experiment 1 with new subjects. Finally, subjects in Experiment 3 could postpone timeouts from delivery of pellets that differed in quality rather than quantity (unsweetened versus sweetened pellets). In general, response rates and success in avoiding increased as a function of the rate and quality of the pellets. Also, performance efficiency increased as the experiments progressed, that is, the avoidance response occurred later and later in the response-timeout interval. The results support the conclusion that timeout from reinforcement has functional properties similar to those of more commonly studied aversive stimuli (e.g., shock).     

Warmup in free-operant avoidance as a function of the response-shock = shock-shock interval

Warmup effects, the repeated within-session transitions from ineffective to effective avoidance, were examined with rats on free-operant shock-delay procedures. The shock-shock and response-shock intervals were kept equal as they were varied. As measured by both response rates and shock rates, the magnitude of within-session change in performance was inversely related to the size of the manipulated intervals. The duration of warmup tended to decrease as the intervals were increased. This finding, that increased shock frequencies do not shorten the warmup, appears to be inconsistent with all interpretations of the warmup that have been offered to date. Late-session performances replicated general features of prior experiments, but differed with respect to details of secondary conclusions in previous reports. These differences may stem from the selection of especially proficient avoiders for previous experiments.     

Choice of timeout during response-independent food schedules

Rats' lever pressing terminated visual or auditory stimuli associated with fixed-time or variable-time schedules of food delivery and produced a timeout period during which food delivery could not occur. Lever pressing during a timeout period reinstated the food-associated stimuli and again permitted food delivery according to the fixed-time or variable-time schedules. The mean interfood interval ranged from 1 minute to 16 minutes (variable-time schedules) or 32 minutes (fixed-time schedules); the timer controlling schedule intervals did not stop during timeout periods. The percentage of session time spent in timeout increased when the mean interfood intervals were lengthened and decreased when the mean interfood intervals were shortened. Timeouts were initiated most frequently about half way between successive food deliveries (fixed-time schedules) or after 15 seconds or more had lapsed since the last food delivery (variable-time schedules). Elimination of food delivery increased the percentage of session time spent in timeout, and elimination of the timeout contingency decreased lever press rates. When timeout was produced only when the lever was held in the depressed position, little time was spent in timeout. The main determinants of timeout initiation and termination appeared to be the rate of food delivery, freedom of movement during timeout, and the stimulus change associated with initiation and termination of timeout.     

Warmup in avoidance as a function of time since prior training

On avoidance procedures, rats and pigeons typically show warmup effects, characterized by improving performance within sessions and loss of the improvement (“warmup decrement”) between sessions. Between-session losses were examined by varying the time between periods of avoidance training. In one experiment, rats lived fulltime in conditioning chambers while intermission intervals were varied. In a second experiment, the animals lived in home cages between sessions; timeout intervals were introduced at midession, producing recurrence of warmup in the second half-session. In both experiments, the warmup decrements increased substantially as the timeout or intersession intervals were increased from zero to 30 minutes. With intervals of 60 or 120 minutes, the decrements approached or exceeded those obtained with intervals of a day or more. When avoidance was interposed between appetitive sessions, the appetitive responding was disrupted, but this seemed unrelated to the warmup or to the proficiency of avoidance. The warmup in avoidance shares characteristics with transient punishment effects, with the Kamin effect, and with habituation phenomena, but it is premature to assume that they reflect common processes.     

Behavioral variability as avoidance behavior

This study aimed to investigate whether variable patterns of responses can be acquired and maintained by negative reinforcement under an avoidance contingency. Six male Wistar rats were exposed to sessions in which behavioral variability was reinforced according to a Lag contingency: Sequences of three responses on two levers had to differ from one, two or three previous sequences for shocks to be avoided (Lag 1, Lag 2 and Lag 3, respectively). Performance under the Lag conditions was compared with performance on a Yoke condition in which the animals received the same reinforcement frequency and distribution as in the Lag condition but behavioral variability was not required. The results showed that most of the subjects varied their sequences under the Lag contingencies, avoiding shocks with relatively high probability (≥ 0.7). Under the Yoke procedure, responding continued to occur with high probability, but the behavioral variability decreased. These results suggest that behavioral variability can be negatively reinforced.     

Ratio responding as a function of concurrent avoidance schedules, yoked shocks, and ratio value

Fixed-ratio food-reinforced responding in rats was studied alone and with concurrent shock avoidance or with concurrent response-independent shocks matched to those that occurred in the avoidance condition. Under each condition, fixed-ratio size was increased over successive daily sessions. Fixed-ratio response rate generally passed through a maximum as a function of fixed-ratio size. Decreased fixed-ratio responding at values beyond the maximum occurred when (1) the time to complete a fixed ratio approximated the response-shock interval of the avoidance schedule, (2) the shock rate increased, and/or (3) the ratio requirements were so high that ratio strain occurred. Avoidance rates decreased slightly as fixed-ratio size increased.     

Rapid acquisition of discrete-trial lever-press avoidance: effects of signal-shock interval

Acquisition of discrete-trial lever-press avoidance learning was studied in three experiments. Experiment I compared a new training procedure, which produces rates of lever-press avoidance learning comparable to those obtained in shuttle boxes, with a “conventional”, less efficient training procedure. A factorial design was used to compare continuous versus intermittent shock and a long-variable versus a short-fixed signal-shock interval. Learning was best in the groups trained with the long and variable interval and poorest in those trained with the short and fixed interval. Type of shock had no effect. Experiment II separated the effects of duration from those of variability of the signal-shock interval. Fixed and variable intervals of 10 and 60 sec were tested and duration was the only significant factor. Experiment III addressed the effect of the differential opportunity to avoid provided by long signal-shock intervals by varying this interval from 10 to 60 sec in 10-sec steps. Only the 10-sec group showed slow acquisition relative to the others. Analysis of avoidance response latencies showed that the distributions for all groups were positively skewed and that skewness increased with increasing duration of the signal-shock interval. At intervals longer than 20 sec, the animals made progressively less use of their increased opportunity to respond. The data do not support the opportunity-to-respond interpretation of the effects of duration of signal-shock interval and suggest that some type of inhibitory process may block lever-press avoidance learning at intervals as short as 10 sec. The significance of these findings for species-specific defense reaction and preparedness theories was emphasized.     

Matching in concurrent variable-interval avoidance schedules

After pretraining with multiple variable-interval avoidance schedules, two rats were exposed to a series of concurrent variable-interval avoidance schedules. Responses on two levers cancelled delivery of electric shocks arranged according to two independent variable-interval schedules. The ratio of responses and time spent on the two levers approximately matched the ratio of shocks avoided on each. Matching to the number of shocks received was not obtained. Concurrent variable-interval avoidance can therefore be added to the group of positive and negative reinforcement schedules that can be expressed in the quantitative framework of the matching law.     

Acquisition and maintenance of postshock response pattern in nondiscriminated avoidance with rats

Five experimentally naive albino rats were placed under a nondiscriminated lever-press avoidance schedule in which the delay to the next shock for responses after a shock was longer than the delay for responses after a response. Four rats acquired the postshock response pattern and maintained it for a prolonged period. The results revealed that postshock responding was under operant control and was not purely shock-elicited. It was suggested that the two kinds of response-shock interval, i.e. the shock-response-shock interval and the response-response-shock interval, could and should be independently controlled in nondiscriminated avoidance schedules.     

Effects of sleep deprivation on free-operant avoidance

Two studies examined effects of sleep deprivation on free-operant avoidance by rats. In Experiment 1, a 5-s shock-shock (SS) interval and 20-s response-shock (RS) interval produced baseline performances, which were reestablished after each experimental manipulation. Once baselines were established, animals were exposed to 24, 48, or 96 hr of sleep deprivation and equivalent periods of home cage and food restriction as a control condition. Compared to baseline, sleep deprivation increased response rates by increasing the proportion of brief interresponse times (IRTs); response rates changed little in the control conditions. Percentage of shocks avoided did not systematically change across conditions. In Experiment 2, the RS interval was manipulated (10, 20, and 40 s), while the SS interval (5 s) and level of sleep deprivation (48 hr) were held constant. Across RS intervals, sleep deprivation increased response rates via a shift toward brief IRTs. In addition, sleep deprivation increased the percentage of shocks avoided as an inverse function of RS intervals.     

Discriminated timeout avoidance in pigeons: the roles of added stimuli

Two experiments examined pigeons' postponement of a signaled extinction period, or timeout (TO), from an ongoing schedule of response-dependent food delivery. A concurrent-operant procedure was used in which responses on one (food) key produced food according to a variable-interval schedule and responses on a second (postponement) key delayed the next scheduled TO according to a response-TO (R-TO) interval. A series of response-independent stimulus changes on the food key temporally partitioned the R-TO into three equal segments (S1, S2, and S3). Postponement responses, in addition to postponing TO, also reinstated S1, the stimulus correlated with the greatest temporal distance from TO. In Experiment 1, the R-TO interval was manipulated systematically across blocks of sessions (conditions) at a given ratio of R-TO:TO duration. This R-TO:TO ratio was manipulated across blocks of conditions (phases). Postponement response rates varied inversely with R-TO interval in each phase. Changes in the R-TO:TO ratio did not produce consistent differences except at the 1:10 ratio for some pigeons, where it disrupted postponement responding in some conditions. Most of the postponement responses occurred in the presence of S2 and S3, the stimuli most proximal to TO, whereas most of the food-key responses occurred in S1. In Experiment 2, the R-TO contingencies were systematically manipulated in the presence of the time-correlated stimuli. In one set of conditions, the R-TO contingencies were made either ineffective or less effective in the presence of one or more stimuli. Postponement responses typically shifted to stimuli in the presence of which responses were relatively more effective. Postponement responses decreased markedly when the added stimuli were removed, and then recovered when the stimuli were reinstated. Results from both experiments indicate that the added stimuli in a discriminated TO-avoidance procedure serve predominately discriminative functions, delineating periods during which behavior is maximally effective. The results parallel those obtained in shock-avoidance procedures, providing further evidence that TO functions as an aversive stimulus.     

Response-shock delay as a reinforcer in avoidance behavior

After rats received preliminary training to avoid shock on a discrete-trial retractable-bar avoidance procedure, the procedure was changed such that responses retracted the lever but did not affect the rate of shock. Responses only delayed the onset of shock. About half of the animals under these procedures responded consistently on almost 100% of the discrete-trial cycles over days. When short latencies maximized the response-shock delay, animals tended to make short-latency responses. When long latencies maximized the response-shock delay, animals tended to make long-latency responses. When all response latencies produced the same response-shock delay, animals made differing average-latency responses. And, when responses did not delay shock, most of the animals primarily engaged in shock-elicited responding while the other animals engaged in preshock responding.     

Lever attacking and pressing as a function of conditioning and extinguishing a lever-press avoidance response in rats

Six experimental rats were conditioned to press one of two available levers to avoid shock. The levers registered bites as well as presses. For four of these rats, shock was contingent on lever bites when a specified time period had elapsed after the previous shock. An extinction period, in which only periodic noncontingent shocks were presented, followed avoidance training. Six yoked-control rats received the same sequence of shocks as did the corresponding experimental rats in both the conditioning and extinction phases. All six experimental rats repeatedly bit the avoidance lever. Four bit it more than the nonavoidance lever during conditioning, and five bit it more during extinction. Five of the six experimental rats consistently bit the levers many more times during each session than did their respective control rats, suggesting that avoidance conditioning facilitated lever biting. Rates of lever biting and pressing by all of the experimental rats and by some of the control rats were highest immediately following shock throughout both phases. During later portions of the intervals following shock, characteristic effects of conditioning and extinction were observed. This finding suggests that extinction of avoidance behavior by unavoidable shock presentations can be demonstrated more readily when shock-elicited responding is extricated from the data.