Overshadowing in matching-to-sample: Reduction in sample-stimulus control by differential sample behaviors

Separate groups of pigeons were trained to high levels of accuracy on 0-delay matching-to-sample with sample-response requirements that were either differential or nondifferential with respect to the sample stimuli. Differential subjects produced the comparisons by completing a differential-reinforcement-of-low-rates-of-responding 3-sec (DRL 3″) requirement during one sample and a fixed-ratio (FR 10) requirement during the other. Nondifferential subjects produced the comparisons by completing the same schedule requirement (either DRL 3″ or FR 10) for both samples. Following acquisition to criterion, the DRL and/or FR sample-response requirements were replaced by a nondifferential single-peck (CRF) requirement in order to assess the degree to which the samples had acquired control over choice in each group. This change disrupted performance in all subjects, but the disruption was greater for the differential birds, which generally performed at lower levels of accuracy and required more sessions of retraining to reach criterion levels of accuracy than the nondifferential birds. Follow-up experiments revealed that comparison choices by the differential birds were primarily controlled by their DRL vs. FR sample-specific behaviors. The relatively poor performance of the differential group during testing with CRF requirements suggests that the cue arising from the birds’ differential sample behaviorshad also overshadowed the sample stimuli for conditional control over choice. The unique, and rather unusual, aspect of this overshadowing effect is that it occurred in spite of the fact that the overshadowed cue (that provided by the samples) was necessary for producing the cue that resulted in overshadowing (the differential sample behaviors). This finding has potentially important implications for the differential outcomes effect in conditional discrimination learning and for attentional processes in compound-cue situations in general.     

Retardation and facilitation of matching acquisition by differential outcomes

Pigeons were trained on many-to-one matching-to-sample with food and no-food outcomes that were either differential or nondifferential with respect to the sample stimuli. In the differential condition, outcomes were correlated with the correct comparison-alternatives-for half of the subjects, and were uncorrelated with those alternatives for the remaining subjects. Relative to non-differential training, matching acquisition was facilitated in the correlated condition but retarded in the uncorrelated condition. These results clearly demonstrate that differential outcomes do not affect conditional discrimination learning merely by enhancing the discriminability or distinctiveness of the samples with which they are associated. Rather, they apparently give rise to another discriminative cue (viz., an outcome expectancy), which can either enhance or interfere with performance, depending on its predictive validity.     

A differential-outcome effect in pigeons using spatial hedonically nondifferential outcomes

We examined the extent to which nonhedonically different differential outcomes involving feeder location control pigeons’ comparison choices in matching to sample. In Experiment 1, we showed that differential feeder location outcomes associated with each of two samples can facilitate delayed-matching accuracy. In Experiment 2, we found positive transfer following training on two matching tasks with differential feeder location outcomes when samples from one task were replaced by samples from the other task. In Experiment 3, we found that when differential-outcome expectations could no longer serve as the cues for comparison choice, sample stimuli continued to exert some control over choice of comparisons. The results indicate that differential outcomes (involving feeder location) that presumably do not differ in hedonic value are sufficient to control comparison choice. Thus, the differential hedonic value of the outcome elicited by the sample does not appear to be a requirement of the differential-outcome effect. Furthermore, these differential outcomes appear to augment matching accuracy, but they do not eliminate control by the samples.     

Some relationships between outcome expectancies and sample stimuli in pigeons’ delayed matching

Two sets of experiments examined how differential outcomes affect conditional stimulus control by the samples in delayed matching-to-sample. Pigeons were initially trained on symbolic delayed matching with reinforcing outcomes that were either differential or nondiffereatial with respect to the samples. In one set of experiments, the outcome manipulation involved different (p = 1.0 vs. 0.2) versus the same (p = 0.6) probabilities of food; in the other, food and no-food outcomes were used. Following initial acquisition and mixed-delay tests, the matching procedure in each study was discontinued while the samples were nondifferentially reinforced with the same probability of food, or with food and no food, respectively. When later retested on delayed matching with those nondifferential outcomes, birds initially trained with different reinforcement probabilities matched at the same levels of accuracy as those trained with the same probability. By contrast, birds initially trained with food versus no-food outcomes showed lower levels of matching accuracy than their nondifferential controls. Subsequent transfer tests showed that matching performances by the differential birds in both studies had been originally cued in part by differential outcome expectancies. Apparently, the expectancies based upon different probabilities of food provided a source of conditional stimulus control that did not compete with the samples. By contrast, the expectation of food versus no food reduced (overshadowed) sample-stimulus control.     

Development of excitatory backward associations during the establishment of forward associations in a delayed conditional discrimination by pigeons

The development of excitatory backward associations in pigeons was demonstrated in three experiments involving conditional discriminations with differential outcomes. In Phase 1 of all three experiments, correct comparison choices following one sample were followed by food, whereas correct comparison choices following the other sample were followed by presentation of an empty feeder. In Phase 2, the food and no-food events that served as outcomes in Phase 1 replaced the samples. When the associations tested in Phase 2 were consistent with the comparison-outcome associations developed in Phase 1, transfer performance was significantly better than when the Phase 2 associations were inconsistent with the Phase 1 associations. In Experiment 1, an identity matching-to-sample task was used with red and green samples and red and green comparisons. In Experiment 2, a symbolic matching task was used with shape samples and hue comparisons, and it was shown that the backward associations formed were between the trial outcome (food or no food) and the correct comparison. In Experiment 3, it was determined that the transfer effects observed in these experiments did not depend on either the similarity of behavior directed toward the samples in the training and test phases, or the similarity of food and no-foodexpectancies generated by the samples in Phase 1 to food and no-foodevents presented as samples in Phase 2.     

The role of test stimuli in matching to compound samples by pigeons

Two pigeons matched to sample in a three-key operant conditioning chamber. In Experiment I, two different kinds of samples were presented on the center key.Element samples were members of one of two sample sets — colors (a red or blue disk) or lines (a vertical or horizontal orientation of a set of white lines). These samples were followed by their respective sample sets on the side keys as comparison stimuli.Compound samples consisted of a set of lines superimposed on a colored disk. Following these samples, either sample set could appear as comparison stimuli. Matching to compound samples was less accurate than matching to element samples. One interpretation is that sharing of attention among elements of a compound sample weakened stimulus control by each element. A different interpretation is that an element sample controlled matching better because it was physically identical to a comparison stimulus whereas a compound sample was not. Experiments II–IV evaluated this “generalization decrement” alternative by testing element- vs. compound sample control with both element and compound comparison stimuli. Irrelevant elements were added to form compound comparison stimuli, some of which were physically identical to a preceding compound sample, but never identical to an element sample. In all experiments, the addition of irrelevant elements of comparison stimuli reduced sample control. However, the generalization decrement hypothesis failed to predict how differences in performance maintained by element and compound samples were affected by different tests of sample control. Matching accuracy appeared to be independently determined by the number of elements in a sample and whether irrelevant elements were present during tests of sample control.     

Representation strength in pigeon short-term memory: Effect of delay training

An attempt was madeto manipulate the strength of internal stimulus representations by exposing pigeons to brief delays between sample offset and comparison onset in a delayed conditional discrimination. In Experiment 1, pigeons were first trained on delayed conditional discrimination with either short (0.5-sec) delays or no delays. When delays were increased by 2.0 sec, birds trained with a delay performed at a higher level than did birds trained with no delays. In Experiment 2, subjects were first trained on a delayed simple discrimination. Following a circle stimulus, responses to a white key were reinforced; however, following a dot stimulus, responses to the white key were not reinforced. The pigeons were then trained on a delayed conditional discrimination involving hue samples and line-orientation comparisons with differential outcomes. Choice of vertical following red yielded food; choice of horizontal following green yielded no food. Mixed delays were then introduced to birds in Group Delay, whereas birds in the control group received overtraining. When tested on a delayed simple discrimination with hue stimuli (red and green initial stimuli followed by white response stimulus), pigeons in Group Delay tended to perform at a higher level than did birds in the control group (i.e., although the birds in both groups responded more following red than following green, birds in Group Delay did this to a greater extent than did birds in the control group). Thus, experience with delays appears to strengthen stimulus representations established during training.     

Interaction of sample dimension and sample-comparison mapping on pigeons’ performance of delayed conditional discriminations

Coding strategies developed in the acquisition of delayed conditional discriminations can be assessed by independently manipulating sample and comparison memory load. Two stimulus dimensions that can affect memory load were examined: Number of stimuli in the sample and comparison sets (two vs. four) was manipulated between groups in a 2×2 design, and discriminability of sample and comparison stimuli (hues vs. lines) was manipulated between counterbalancing subgroups and within subjects. The results indicated large effects of sample discriminability but not of comparison discriminability, evidence for retrospective coding. There was also a significant effect of number of stimuli in the comparison set (although only with hard-to-discriminate samples) but not of number of stimuli in the sample set, evidence for prospective coding. These findings suggest evidence for retrospective coding with easy-to-discriminate samples, independently of number of stimuli in the comparison set, and evidence for prospective coding with hard-to-discriminate samples.     

Pigeons transfer between conditional discriminations with differential outcomes in the absence of differential-sample-responding cues

Pigeons were trained to perform two independent zero-delay conditional discriminations involving the same differential outcomes (i.e., food vs. a feeder light) and were then tested by replacing the samples from one of the tasks with those from the other. Differential responding to the two samples was required in Phase 1, in Phase 2, in neither phase, or in both phases. Half the pigeons in each group were then tested with associations that were either consistent with the presumed outcome expectancies established during Phases 1 and 2 (positive transfer condition) or inconsistent with those expectancies (negative transfer condition). The magnitude of the transfer effect was largest in the group that could use differential sample responding as a cue to mediate transfer, but significant transfer effects were also found in the groups that could use only outcome expectancies. Thus, differential sample responding contributes to, but does not account for, the differential-outcomes effect.     

Delayed matching in pigeons with food and no-food samples: Further examination of backward associations

When pigeons are trained on a delayed conditional discrimination with presence versus absence samples and tested with delays, a bias to choose the comparison associated with the absence sample is observed with increasing delay. Additionally, when the samples consist of food versus no food, this trial-type performance difference is reversed on short-delay trials: a bias to choose the comparison associated with the presence sample develops with delay testing. This reversal in comparison bias at short delays has been attributed to a preference produced by backward associations between the hedonic samples and the nonhedonic choice stimuli. In the present experiment, we tested an alternative hypothesis, that the short-delay comparison bias is produced by proactive interference—in particular, from reinforcement obtained on the previous trial—by including a group trained with reinforcement on only half of the trials with a correct response. According to the proactive interference account, this group should have shown a smaller short-delay comparison bias than would the typical 100% reinforcement group. Instead, consistent with a backward-association interpretation, the magnitude of the short-delay comparison bias shown by the 50% group was significantly greater than that shown by the 100% group.     

Differential-outcomes effect using hedonically nondifferential outcomes with delayed matching to sample by pigeons

When differential outcomes follow correct responses to each of two comparison stimuli in matching to sample, relative to the appropriate control condition, higher matching accuracy is typically found, especially when there is a delay between the sample and the comparison stimuli. In two experiments, we examined whether this differential-outcomes effect depends on using outcomes that differ in hedonic value (e.g., food vs. water). In Experiment 1, we found facilitated retention when a blue houselight followed correct responses to one comparison stimulus and a white houselight followed correct responses to the other, prior to nondifferential presentations of food. In Experiment 2, we found facilitated retention again when a blue houselight followed correct responses to one comparison stimulus and a tone followed correct responses to the other, prior to nondifferential presentations of food. The results of both experiments indicate that the differential-outcomes effect does not depend on a difference in hedonic value of the differential outcomes, and they suggest that outcome anticipations consisting of relatively arbitrary but differential stimulus representations can serve as cues for comparison choice.     

Transfer of performance to new comparison choices following differential outcome matching-to-sample

Four experiments examined transfer of differential outcome performances to new choice responses in pigeons. Experiments 1A and 1B showed that new responses trained off a matching-to-sample baseline readily substituted for the choice alternatives in differential outcome matching, provided that they shared the same outcome associations as the alternatives they replaced. Experiment 2 showed that comparison responses trained on baseline, but in a task in which their different outcomes occurred equally often following each sample (viz., one-to-many matching), substituted for the choices in a standard, differential outcome task. Experiment 3 showed, somewhat surprisingly, that the choices in the latter task were likewise effective substitutes in one-to-many matching. These results pose separate challenges for standard two-process theory and for the bidirectional account of differential outcome performance, and they suggest other cues that pigeons may use to predict outcomes.     

Prospective representation: The effects of varied mapping of sample stimuli to comparison stimuli and differential trial outcomes on pigeons’ working memory

The mapping of sample stimuli onto comparison stimuli and the nature of trial outcomes were factorially manipulated in a delayed matching-to-sample procedure. In the many-to-one condition (MTO), responding to a particular comparison was correct following several sample stimuli, whereas in the one-to-many condition (OTM), responding to several comparison stimuli was correct following a particular sample. Probabilities of reinforcement for correct responding to comparison stimuli were either differential (DO) or nondifferential (NDO). Four groups of pigeons were trained under four combinations of mapping and outcome conditions, MTO-DO, MTO-NDO, OTM-DO, and OTM-NDO. Testing at delay intervals of 1, 2, 4, and 8 sec revealed significant effects due to both the mapping variable and the differential outcomes variable. It was argued that the poorer performance obtained in the OTM condition was due to the differential prospective coding requirements placed on reference and working memory by this mapping. In the OTM conditions, a greater number of response codes had to be retrieved from reference memory and multiple response codes may have overburdened working memory, which has a limited capacity.     

Symbolic representation in animals: Emergent stimulus relations in conditional discrimination learning

When two initially unrelated stimuli are associated with the same event (e.g., many-to-one conditional discriminations), an emergent relation can develop between those stimuli, allowing them to be interchangeable in other contexts. The source of the emergent relation appears to be a common representation of the two stimuli, and the nature of the representation may be retrospective, involving one of the initial stimuli (probably the stimulus that first enters into association). These emergent relations have been implicated in the establishment of stimulus equivalent relations. However, the failure to find strong evidence for other emergent relations thought to be involved in stimulus equivalence (Sidman, 1990; e.g., bidirectional associations, transitive relations, and functional equivalence associated with one-to-many conditional discriminations) suggests that the mechanisms underlying the emergent relations found using transfer designs following many-to-one training are different from those presumed to underlie stimulus equivalence. On the other hand, evidence for the common representation of stimuli associated with the same arbitrary event extends the nature of those representations beyond those well-established effects found when the common events are biologically important (e.g., the differential outcomes effect).     

Directed forgetting in monkeys

Based on several recent demonstrations of a directed forgetting effect in pigeons, three experiments were carried out in an attempt to demonstrate directed forgetting in three squirrel monkeys. During initial training with a delayed matching-to-sample procedure, retention tests were always given for sample stimuli followed by remember cues (R-cues) and were always omitted for sample stimuli followed by forget cues (F-cues). Retention of F-cued items was tested on probe trials after initial training. The first two experiments examined the effects of R- and F-cues on memory for slide-projected pictures, with different pictures used on each trial of a session. In Experiment 1, a complex design was used in which one or two sample pictures were presented on each trial; when two pictures were presented, both could be R-cued or F-cued, or one could be R-cued and the other F-cued. A simpler design was used in Experiment 2, with only single pictures presented as sample stimuli and half the trials within a session R-cued and the other half F-cued. In both of these experiments, no differential retention of R- and F-cued stimuli was found, even at a retention interval as long as 16 sec. In Experiment 3, a series of studies was performed to test for directed forgetting when only two sample stimuli were used repeatedly throughout training and testing. With two pictures as sample stimuli, clear evidence of directed forgetting was found in Experiment 3b. It is suggested that the directed forgetting effect may arise only when a small set of sample stimuli is used.     

Pigeons’ short-term memory for temporal and visual stimuli in delayed matching-to-sample

In the present experiment, we compared directly pigeons’ short-term memory of temporal and visual stimuli in a delayed matching-to-sample task. The sample stimuli consisted of red and green lights presented for 5 and 30 sec, followed by a retention interval and blue and yellow comparisons. For subjects in the visual group, duration was irrelevant and the color of the sample was the conditional cue. For animals in the temporal group, color was irrelevant and duration of the sample was the conditional stimulus. The results showed that acquisition of the matching task was faster and accuracy was higher in the visual than in the temporal group. More importantly, memory of either sample generally declined at a similar rate when the duration of the retention interval was increased and when the intertrial interval was reduced. Taken together, the results indicate that with 1–8-sec retention intervals, short-term memory for temporal stimuli is similar to that found with color-visual samples. The findings are discussed in terms of retrospective and prospective processing.     

Coding of feature and no-feature events by pigeons performing a delayed conditional discrimination

Pigeons’ performance of a delayed conditional discrimination with presence versus absence of conditional (sample) stimuli was examined in two experiments. The pigeons showed steeper retention functions with feature (i.e., presence) samples (either food or yellow) than with no-feature (i.e., absence) samples (either no food or no yellow). These results suggest that pigeons code features and respond only by default to test stimuli (comparisons) associated with no features. In contrast, the overall superiority of performance on no-feature-sample trials compared with feature-sample trials in both the food/no-food- and yellow/no-yellow-sample tasks was reversed at a 0-sec delay in the food/no-food-sample group, but not in the yellow/non-yellow-sample group. This difference in results with hedonic versus nonhedonic samples suggests that the crossover in delay performance on food/no-food-sample trials is produced by the formation of backward associations between the food-associated comparison stimulus and the food sample.     

Common coding in pigeons: Partial versus total reversals of one-to-many conditional discriminations

Common coding in pigeons was examined using a delayed conditional discrimination in which each sample stimulus was associated with two different comparison stimuli (one-to-many mapping). In Experiment 1, pigeons matched circle and dot samples to red and green hues and vertical and horizontal line orientations. In Experiment 2, the samples were red and green and the comparisons were vertical and horizontal spatial positions (up vs. down and left vs. right). Following acquisition to high levels of accuracy in each experiment, the associations between the samples and either both sets or only one set of comparisons were reversed. Pigeons learned the total reversals faster than the partial reversals. These results suggest that when different comparisons are associated with a common sample, they may become functionally equivalent.     

Differential effects of omitting comparison stimuli on symbolic and identity matching to sample: Evidence for altered instructional processes in pigeon short-term memory

Two experiments were performed to determine the effects of omitting the comparison stimuli in a matching-to-sample task. In Experiment 1, birds were trained initially on both symbolic and identity matching to sample. Comparison stimuli were then omitted following the presentation of a particular sample stimulus, and this decreased the number of sample (observing) responses. The reintroduction of the comparison stimuli on subsequent probe trials revealed that the accuracy of symbolic matching was reduced to chance levels, while identity matching accuracy was significantly below chance. In Experiment 2, a similar procedure was employed; however, observing responses to the comparison-omitted samples were maintained by direct reinforcement (fixed ratio 20). Matching accuracy during probe trials was again at chance levels for symbolic matching but, contrary to Experiment 1, was significantly above chance for identity matching. The differential effects of omitting comparison stimuli on symbolic and identity matching trials in these two experiments were interpreted within a framework which assumes that instructional processes are altered by comparison-omission procedures.     

Emergent relations in pigeons following training with temporal samples

In two experiments, we investigated emergent conditional relations in pigeons using a symbolic matching-to-sample task with temporal stimuli as the samples and hues as the comparisons. Both experiments comprised three phases. In Phase I, pigeons learned to choose a red keylight (R) but not a green keylight (G) after a 1-s signal. They also learned to choose G but not R after a 4-s signal. In Phase II, correct responding consisted of choosing a blue keylight (B) after a 4-s signal and a yellow keylight (Y) after a 16-s signal. Comparisons G and B were both related to the same 4-s sample, whereas comparisons R and Y had no common sample. In Phase III, R and G were presented as samples, and B and Y were presented as the comparisons. The choice of B was correct following G, and the choice of Y was correct following R. If a relation between comparisons that shared a common sample were to emerge, then responding to B given G would be more likely than responding to Y given R. The results were generally consistent with this prediction, suggesting, for the first time in pigeons, the emergence of novel relations that involve temporal stimuli as nodal samples.