Search

Pairing target stimuli (A and B) with different events (X and Y) instead of a common event enhances the distinctiveness of the target stimuli. In addition, after the distinctiveness training, pairing of X and Y will heighten the similarity between A and B.

The relationship between reward schedule and serial learning is examined in relation to experiments with rats. The issue of molar versus molecular approaches is also discussed.

Rats acquired a preference for an aqueous odor (almond) presented in simultaneous compound with sucrose. Separate presentations of saccharin reduced this preference in rats with ad-lib access to food during training or at test, but not in rats that were hungry during both training and test. In contrast, separate presentations of sucrose reduced the preference for the almond irrespective of deprivation state during training and test. We interpret the results to mean that a hungry rat forms odor-taste and odor-calorie associations, and its motivational state on test determines which of these associations controls the preference. In contrast, a rat that is not hungry during training only forms an odor-taste association, and its performance on test is independent of its level of hunger.

Four experiments assessed the effects of trial (T) and intertrial (I) durations on magazine approach behavior in rats. In Experiments 1 and 2, different groups of animals were conditioned with various combinations of I and T durations. The rate of acquisition, in terms of the number of trials required to reach various acquisition criteria, generally was faster in groups trained with large I:T ratios. There also were differences in rate of acquisition and terminal response rates between groups trained with identical I:T ratios but with different absolute I and T durations. Differences evident at the end of conditioning persisted during a common test with various combinations of I and T durations. Experiments 3 and 4 provided a more specific test of the predictions of 2 general classes of theories and found results that were consistent with those theories that characterize group differences as indicative of differences in learning, rather than in performance.

Rats are capable of utilizing intramaze cues to guide arm selection, suggesting the cues were perceived in a spatial representation. Rats were placed in a radial arm maze offering at least seven correct choices for three consecutive days, designed to display textured inserts as clues. Rotation and removal of the clues did not significantly lower correct choices, while transposition did.

Competition between contextual occasion setting and associative processes was indicated by experiments with rats using appetitive conditioning. The presence of a discrete predictive stimulus interfered with occasion setting by contextual stimuli. The rats were separated into groups that learned to predict nonreinforcement of an auditory target stimulus based on contextual stimuli (CTX), a visual stimulus that was compounded with the target (CI), or both (CTX+CI). A second auditory target was inconsistently reinforced. The visual stimulus controlled behavior in both the CI and CTX+CI groups, while the contextual stimuli controlled behavior of the CTX group.

Three experiments with a conditioned suppression paradigm examined rats' long-term memory for the duration of a conditioned stimulus. Forgetting of a previously acquired temporal discrimination increased with longer retention intervals, but was not greatly affected by training levels of either 45 or 90 trials. Retention intervals of 3, 7, 15, and 20 days were tested. The temporal discrimination gradient became progressively flattened, with little or no forgetting at 3 and 7 days, reaching a maximum after 20 days.

A retention period of 15 days between preexposure and conditioning caused attenuation of the effect of preexposure to the unconditioned stimulus (US) in flavor aversion learning of rats. No effect was observed, however, when the retention interval was interposed between conditioning and testing. In addition, conditioned aversion without US-preexposure was not affected by retention intervals. These results suggest that preexposure affects the formation of an association between the conditioned and unconditioned stimuli, rather than causing interference at the testing stage.

Evidence for both forward and backward excitatory serial conditioning is consistent with a behavior systems account of conditioning. In experiments with rats, the serial elements exhibited asymmetrical control, especially during compound conditioning. In addition, the specific responses seemed to be more strongly conditioned in forward conditioning than in backward conditioning. Some problems with the traditional associative account can be solved by using a behavior systems framework that incorporates both the adaptive-evolutionary and associative approaches.

Rats that observed demonstrators pushing joysticks in a particular direction were likelier to then push it in the same direction only if the demonstrator received food and a tone was sounded as a reward. A 5-sec delay between the act and the reward did not affect this imitation, nor did intermittent rather than continuous rewards. However, observers that saw demonstrators rewarded with tone or food alone were not likely to select the chosen direction for their own pushes.

Rats first given food pellets at regular intervals later conditioned poorly to a tone in conjunction unless food pellets were also given at wider intervals before and after the tone period. These results challenge the contingency school and theories of conditioning. However, the blocking ability of unconditioned stimuli (US) in a US-US association may explain the results. Modifications to the Rescorla-Wagner theory may also account for them.

The tendency of rats to forget stimulus attributes can be reversed by pretest exposure to the original training context. A 90-sec exposure helped to restore their memory for at least 30 min, but a delay of more than 60 min led the rats to forget again. This means that forgetting stimulus attributes could be caused by a retrieval deficit, with time-dependent effectiveness. Forgetting stimulus attributes may therefore work the same way as the forgetting of target responses.

Two experiments with rats examined the effect of response competition on the trial-spacing (TS) effect. In an autoshaping experiment, use of a sugar solution but not food pellets as the unconditioned stimulus was associated with a strong TS effect. In a contextual conditioning experiment, feeder responses were associated with training with food pellets, whereas rearing responses were associated with presentation of the sugar solution. The results are best explained in terms of response competition, with competition occurring between lever-pressing and rearing but not feeder responses.

Two experiments were conducted with rats to investigate the effect of signals for reward omission on performance. Results demonstrate that association of a conditioned stimulus with omission of food results in control of specific responses. Withdrawal was generally higher in nonreinforced conditions compared to reinforced trials. Results are compared with previous experiments using opossums. The rats exhibited greater individual variability in response compared to the opossums. Effects of intertrial intervals and possible mechanisms underlying the withdrawal responses are also discussed.

The blocking effects of nonexplicit occasion setters were investigated in three experiments with rats. Blocking was not produced by nonexplicit occasion setters, although explicitly trained occasion setters did result in effective blocking. The nonexplicit procedure involves reinforcement of the conditioned stimulus (CS) when presented with the occasion setter, but the CS is never presented without the occasion setter. In the explicit procedure, the CS is reinforced in the presence of the occasion setter and appears without reinforcement when the occasion setter is absent. Implications for the associative account of occasion setting are discussed.

The associative status of an explicitly unpaired conditioned stimulus (CS) was investigated in three experiments with rats. With short intertrial intervals, the explicitly unpaired CS acted as a conditioned excitor. The explicitly unpaired CS also acted as a conditioned inhibitor following extinction of the CS or extinction of the contextual cues only. The results may be explained by associations between the CS and context, the context and the unconditioned stimulus (US), context-CS and CS-US or all of the above. Predictions of comparator theories of performance were contradicted by the experimental results.

In Experiment 1, rats experienced 2 stimuli (A and N) each preceded by the same event (food) or by different events (food preceded 1 but not the other). N was then paired with shock, and the generalization of conditioned suppression to A was assessed. Generalization was more marked when A and N had been experienced along with a common antecedent. In Experiment 2, 3 stimuli (A, B, and N) were presented in initial training. For 1 group, A and N were preceded by food and B was not; for a 2nd group A alone was preceded by food. In each group, suppression generalized more readily from N to the stimulus that had received the same initial training as had been given to N. Experiment 3 found that generalization was not enhanced between stimuli when 1 had preceded food in initial training and 1 had followed it. These results demonstrate that stimuli that have shared a common antecedent will come to be treated as equivalent.

In four experiments, rats learned initial response-outcome (R-O) associations in an instrumental training task. They then were given training of those same responses either with another outcome (Experiments 1 and 2) or with nonreward (Experiments 3 and 4). The presence of the original R-O associations was detected by measuring the depressive effect on responding of devaluing that O by pairing with LiCl. The results indicated that the original R-O associations remained intact through both potentially interfering treatments.

A rat's preference for food of a given flavor can be substantially enhanced by allowing it to interact with a conspecific demonstrator that has recently eaten food of that flavor. The heuristic value of treating such socially induced enhancement of flavor preference as an instance of Pavlovian conditioning was examined in three experiments. Conceiving of the smell of the food as a conditional stimulus and other cues emanating from the demonstrator rat as an unconditional stimulus, we determined whether each of three common Pavlovian phenomena - blocking, overshadowing, and latent inhibition - would occur. Using experimental parameters that readily produce socially induced enhancement of flavor preference, none of the three Pavlovian phenomena were found.

Larger reinforcement magnitude was found to increase efficiency of progressive-ratio lever-press responding as well as resistance to effort, but did not affect resistance to prefeeding in three experiments with rats. Extinction was more rapid in the group with larger magnitude of reinforcement. In addition, acquisition of the lever-press response was not affected by reinforcement magnitude. The reinforcement magnitudes tested were one versus three sucrose pellets.

Rats' vulnerability to later gastric ulceration from stress increased when a shock was followed by a rest period in the shock apparatus, but not when a waiting period in the apparatus preceded the same amount of shock. Results indicate that the effect on ulceration is the result of psychological processes rather than simply the amount of physical trauma. The results also provide evidence for backward conditioning, since aversion was associated with the context following the shock, but not with the context preceding it. It is apparent that the effect of time spent in the shock apparatus both before and after shock must be considered in any similar experiments.

Three experiments with rats examined the effect of a visual stimulus presented after an auditory cue in a task involving discrimination learning with delayed reinforcement. A 30-second visual stimulus interfered with discrimination learning, whereas a 10-second visual stimulus had little effect and a one-second visual stimulus enhanced learning. The retarding effect of the 30-second visual stimulus suggests that it acquired its own associative strength, whereas the one-second stimulus may have acted as a distractor to reduce generalization between the reinforced and non-reinforced auditory stimuli.

Cross-modal transfer of learning of grouping cues was demonstrated in two experiments with rats using series of reinforced (R) and nonreinforced (N) trials. Four sets of three-trial series were conducted in the pattern RNR,RNR,RNR,RNR with change in runway brightness, long intertrial interval or a tactile signal serving as grouping cues. Learning was better with grouping cues than with no grouping cues. Since grouping persisted across modalities and cues were not associated with specific reinforcement events, grouping cues cannot be interpreted as discriminative cues.

Three experiments involving conditioned lick suppression with water-deprived rats examined the effects of extinction of a comparator stimulus on Pavlovian response. The comparator stimulus was a background light present during conditioned stimulus (CS) training. CS response was facilitated when a short context exposure occurred between CS and unconditioned stimulus training, with a short post-training exposure similar to that occurring during training and with more extensive post-training exposure without similarity.

Differential outcome effect did not control performance in three experiments with rats using a discrimination task with visual stimuli and delayed reward. The results provide evidence that conditioned reinforcement rather than a differential outcome effect is responsible for performance enhancement when a signal is presented during reinforcement delay as compared to unsignalled delay. Presentation of a visual signal associated with the incorrect lever did not enhance discrimination, as would occur if differential outcome effect controlled performance. In fact, the stimulus actually impaired performance.

Competition occurred between inhibitory conditioning to a punctate stimulus and acquisition of negative contextual occasion setting in two experiments using a Pavlovian discrimination procedure with rats. Relative salience of the competing stimuli determined the outcome. The results contradict previous appetitive blocking studies which found no interaction between positive occasion-setters and excitatory stimuli and suggest that higher-order and simple associative memory systems are not independent.

Acquisition was found to be faster with fewer trials per session in three autoshaping experiments with rats. In the first experiment, acquisition was fastest with 40 1-trial sessions in comparison with 20 2-trial sessions and a control group. Additional experiments examined the effects of nonreinforced preexposure and of a trace procedure that interrupted the temporal contiguity between the conditioned and unconditioned stimuli. Implications of the trials per session effect for theoretical accounts of acquisition are discussed.

Target stimulus salience was found to affect operant feature positive discrimination learning in two experiments with rats. With a target of low salience, direct stimulus control was acquired by the feature. However, with a highly salient target, the feature became the occasion setter for responses controlled by the target. The differences were associated with target intensity during training, not during testing. Results are discussed in relation to several theories of modulatory versus direct stimulus control.

The role of incentive learning in instrumental performance following a shift in the degree of water deprivation was analyzed in three experiments. In Experiments 1A and 1B, rats trained to perform an instrumental action reinforced with either sucrose or maltodextrin solutions when in a high-deprivation state were subsequently shifted to a low-deprivation state and tested in extinction. This within-state shift in water deprivation reduced instrumental performance only when the animals had been exposed to the reinforcer in the low-deprivation state prior to instrumental training. In Experiment 2, a concurrent training procedure was used to assess whether the change in the value of the reinforcer brought about by preexposure was mediated by the contingency between the instrumental action and the reinforcer. Preexposure to the reinforcer under the low-deprivation state produced a selective reduction of the performance of the action upon which it was contingent during training when testing was conducted in extinction following a shift from the high- to the low-deprivation state. These experiments provide evidence that animals have to learn about the incentive value of a reinforcer in a particular motivational state through exposure to the reinforcer in that state.

Four experiments with rats demonstrated occasion setting with nonexplicit feature training. Nonexplicit training involves reinforcing the conditioned stimulus in the feature's presence, but not explicitly nonreinforcing in its absence. The training used two brief conditioned stimuli (CSs), each paired with a different extended feature stimulus. The CSs subsequently elicited more response when paired with the feature they were trained with than when paired with the alternate feature. Further experiments ruled out unconditioned responding and enhanced associability as explanations of the effect.

Three experiments tested water-deprived rats using a conditioned lick suppression paradigm to evaluate the edges of the contingency space proposed by R.A. Rescorla. Rescorla defined the contingency space based on the probabilities of an unconditioned stimulus in the presence and absence of the conditioned stimulus. However, the change in probability was unable to account for the experimental results without additional post hoc assumptions. Other contingency metrics were examined, of which the phi correlation coefficient accounted for the most variance, 76.4%.

Variables affecting anticipatory contrast in rats under free-operant multiple schedules were tested in four experiments. The number of cycles presented per session was the crucial variable, with a single cycle producing the greatest anticipatory contrast effect. Previous studies obtained anticipatory contrast effects with consummatory response procedures but not with free-operant response procedures. The current results suggest that the different results in the previous experiments were due to the number of exposures rather than to the difference between consummatory and free-operant response procedures.

Runway experiments were used to study hierarchical organization of series information in rats. Brightness or tactile grouping cues were used to organize each of two trial series into two lists. The rats' ability to organize series memory into smaller representational units was demonstrated. The results suggest that the memory for the second list was represented as a single unit, or list chunk, which could be anticipated by items presented from the first list. Grouping cues were found to be functionally independent of the items organized.

The effect of a reinforcement signal on rats under a synthetic variable interval schedule for food reinforcement was investigated. The signal was found to enhance acquisition of the required response sequence in spite of a lower response rate. The result suggests that the signal enables acquisition of a more accurate response.

The influence on discrimination learning in rats of the interval between the reinforced stimulus and the reinforcer was investigated in two experiments. Shock was used as the reinforcer. Stronger discriminations were formed when the shock came immediately after the stimulus in comparison with a trace interval of 30 seconds. Contiguous reinforcement was also more effective in the context of concurrent training on a visual and an auditory discrimination.

Two experiments were conducted to estimate the retention of response-prevention effects using massed vs distributed treatments in a model of animal avoidance-learning. In Exp. I, 120 rats were trained to avoid shock in a one-way platform avoidance apparatus. Groups received response-prevention treatment or nontreatment in a 36-min. massed session or in several sessions distributed over a four-day period. In Exp. II, 160 rats were given two trials of escape training in a one-way shuttle box. Groups received response-prevention treatment or nontreatment in a 24-min. session of massed or distributed treatments delivered in one day. Subjects in both studies were tested using a passive-avoidance paradigm immediately following treatment, 24 hours later, and 30 days later. Analysis showed that response-prevention treatments were effective in reducing avoidance behavior and there were no significant differences in retention of avoidance associated with massed vs distributed response-prevention treatments. Implications for animals and humans are discussed, and researchers are encouraged to change from a criterion training procedure to an escape procedure since the latter is a closer analogue to the human condition.

Five rats were reinforced under variable-interval schedules with different average interreinforcement intervals (30 sec, 1 min, 2 min, and 4 min). Each animal was run only two sessions of each schedule. The interresponse times (IRTs) were recorded and analyzed. The autocorrelation function of the IRT series and of the IRT time series (number of responses per time interval) were calculated, and absence of periodicity in the subject's behavior was demonstrated. Frequency distribution of IRTs showed in all cases a similar shape and could be fitted to a gamma probability density function in 60% of cases with a signification level of .01 (Kolmogorov-Smirnoff test). The frequency distributions of the IRT time series were distributed as a Poisson process with a .05 significance level. These results suggest that during variable-interval schedules the responses of the animal can be modeled as a random process characterized by a gamma distribution, as a first approximation.

In three experiments negative contrast in consummatory responding was measured, as well as the preference for the flavor associated with the lesser reinforcer. In all experiments one flavor of saccharin was given on days that chocolate milk occurred and another flavor was given on days with no chocolate milk. For some groups the saccharin preceded chocolate milk, for others the order of chocolate milk and saccharin was random, and for others the saccharin followed the chocolate milk. In all experiments rats suppressed consumption of the lesser reinforcer during training (negative contrast) but preferred the flavor of saccharin that had accompanied chocholate milk to the flavor of saccharin that occurred alone. This suggests that contrast is not a reflection of reduced reward value of te lesser reinforcer. Instead some negative reaction to receiving the lesser reinforcer during training must suppress consumption., Three experiments were conducted to measure negative contrast effects in the flavor preferences of rats. One flavor of saccharin was given with chocolate milk and a second saccharin flavor was given alone. The saccharin preceded, followed or was randomly varied with the chocolate milk. Although current saccharin consumption was suppressed when chocolate milk was also received, association with chocolate milk enhanced the saccharine's reward value whether it preceded or followed the milk. The reduced reward value of the lesser reinforcer did not cause negative contrast.

Water-deprived rats served in a conditioned lick suppression paradigm designed to probe the associative structure underlying a partially reinforced Pavlovian inhibitor (i.e., A + /AX [+ or -]) when Stimulus A (a flashing light) was either of high or low salience. Following inhibition training, subjects received either extinction of A, extinction of the training context alone, or remained in their home cages. The inhibitory potential of Stimulus X (white noise) was the assessed in a neutral context using both a summation test with an excitor other than Stimulus A (Experiment 1) and a retardation test (Experiment 2). When A was of high salience, the inhibitory potential of Stimulus X was directly related to the excitatory value of A and not the excitatory value of the training context. But when A was of low salience, the inhibitory potential of X was directly related to the excitatory value of the training context rather than the excitatory value of A. Although the high salience A, relative to the low salience A, overshadowed the context's role as a modulator of responding to Stimulus X, it did not appreciably overshadow context's potential to directly elicit conditioned responding; thus, a dissociation was seen in the ability of a punctate CS to overshadow a context as a function of the behavioral index of overshadowing. These results are discussed in the framework of the comparator hypothesis., Two experiments were conducted using conditioned lick suppression in water-deprived rats to investigate the effect of excitor salience on Pavlovian conditioned inhibition training. The first experiment used a summation test for inhibition; the second, a retardation test. Bright or dim lights served as high- or low-salience punctate excitors. The high-salience excitor overshadowed the training context, but the low-salience excitor did not. The results are interpreted in terms of the comparator hypothesis.

Three studies evaluated the role of feedback stimuli in a instrumental aversively motivated task. Rats were first given classical fear-conditioning trials (white noise-shock) in the start compartment of a two-compartment apparatus. They were then trained, in the absence of shock, to jump a hurdle from the start compartment to a 'safe' compartment. For some groups, the fear-arousing situational cues and white noise CS were present in both compartments; for others, these fear-arousing cues were absent in the safe box. Also varied was the presence or absence of an external feedback stimulus (darkness or a flashing light) after the instrumental response. The results indicated that feedback stimuli can improve performance but only when sufficient fear is present following the response. In addition, the data revealed that darkness can be a more effective feedback stimulus than a flashing light but only when fear of visual situational cues is present. The facilitation of instrumental performance attributable to the presentation of a feedback stimulus was found to occur as early as the second trial. The findings generally support a reinforcement, fear-reduction/relaxation, interpretation of the role of feedback stimuli., Three experiments using acquired-drive design with rats were conducted to evaluate the effectiveness of feedback stimuli in aversively motivated instrumental learning. Feedback stimuli improved performance by contributing to reinforcement through reducing fear. Darkness was found to be a more effective feedback stimulus than a flashing light in conditions involving visual fear-arousing cues. The reinforcing effect of feedback stimuli occurred after only a few trials. This result is inconsistent with conditioned inhibition but supports a conditioned-relaxation interpretation.

Four experiments used hungry rats leverpressing for food reinforcement. The effects of different stimuli serving to signal a delay of reinforcement on responding maintained by a variable interval (VI) 60-s schedule with a 5-s delay of reward were studied. In the initial two experiments, groups of rats responded faster with a discrete visual stimulus (Experiment 2) or diffuse auditory stimulus (Experiment 2) that either signalled the whole delay period or signalled the first part of the delay period relative to groups with an unsignalled delay. If only the second half of the delay period was signalled with the visual stimulus, then performance relative to the unsignalled condition was attenuated (Experiment 1). In contrast, performance with the diffuse auditory signal in the second half of the delay period was similar to that with no stimulus (Experiment 2). The results obtained with a stimulus filling the second half of the delay period were confirmed in Experiments 3 and 4. These data suggest that the effect of signalling reinforcement delays will depend upon the degree of interference produced by sign-tracking behaviors, and not on the blocking of response-reinforcer associations by intervening stimuli., Four experiments using food reinforcement of rats were conducted to investigate effects of signalled delay of reinforcement on attenuation of instrumental response. The different results obtained for visual and auditory stimuli suggest that the effect of signals depends on interference from sign-tracking behavior rather than on overshadowing of response-reinforcer associations.

In these experiments, we examined movement organization during spatial learning in modular mazes made of transparent Plexiglas tunnels. The orientation of adjacent tunnels provided the main direction cues, while the access to visual environmental cues was controlled by the position of a large position. The first series of experiments was conducted in a modified eight-arm radial maze with two parallel and six radial arms. An important neglect of one of the parallel tunnels was completely prevented when these arms were separated by a partition. Removing this screen during training induced a slow development of the neglect. In a second series of experiments, spontaneous alternation behavior (SAB) was tested in asymmetrical three-arm mazes in which the two goal arms were always 45[degrees] apart. Various configurations were designed in which the angle between the start arm and the nearest goal arm varied between 0 and 135[degrees]. In these mazes, SAB was reduced by an important neglect of the nearest goal arm, which was an inverse function of the angle between the start arm and the pair goal arms, i.e., maximum for 0[degree]and minimum for 135[degrees]. This bias was significantly decreased when a curtain separated the start arm from the nearest arm, restoring significant SAB. These results confirm and extend previous findings that the observed neglect is a direct consequence of movement rules which favor large angle transitions during an exploratory phase., Two experiments were conducted with rats using the radial arm maze task to study the effect of visual cues on movement organization. In both an eight-arm maze and spontaneous alternation tests, the rats developed strong tendencies to neglect parallel arms. Visual cues had little effect on this pattern, confirming the hypothesis that it is caused by movement rules favoring large-angle transitions during exploration of a new environment.

Two experiments used lick suppression by water-deprived rats to investigate the associative structure underlying the enhancement of negative summation often seen following operational extinction (i.e., X-) of a Pavlovian conditioned inhibitor (A+/AX-). Our expectation was that one consequence of inhibitory training with Stimulus X was X's accruing excitatory strength through mediation by Stimulus A, and that operational extinction of A attenuated this strength. Experiment 1 determined that, following A+/AX- training, operational extinction of X enhanced negative summation even when the same excitor was used both on the summation test and in inhibitory training (i.e., Stimulus). This suggests that the enhanced negative summation arises from extinction of a direct X[right arrow]US excitatory association, which is mediated by A during acquisition but not during subsequent retention. In Experiment 2, the AX compound was extinguished following A +/AX - training. This treatment was intended to maintain the X[right arrow]A association while extinguishing the X[right arrow]US association. Stimulus X was then tested in summation with an excitor other than A. Unlike extinction of X alone, extinction of AX caused no increase in the inhibitory potential of X; however, it produced a decrease in the excitatory value of A. Additionally, Experiment 2 found that the effectiveness of further inhibition training was similar to operational extinction of X in increasing X's inhibitory response potential. These results a consistent with the view that X, as a result of the AX- trials of inhibitory training, comes to possess a direct X[right arrow]US associative link, and that the negative summation potential of an inhibitor is a function of the discrepancy in excitatory associative strengths betweene putative inhibitor and the conditioned excitor used in inhibition training., Two experiments were conducted using lick suppression by water-deprived rats to examine mechanisms causing enhanced negative summation after operational extinction of a Pavlovian inhibitor. Results support the comparator hypothesis, which denies the existence of inhibitory associations, over the model proposed by R.A. Rescorla, in which negative summation is caused by loss of excitatory strength of the inhibitor.