Categorization of natural movements by pigeons: visual concept discrimination and biological motion

In three experiments, pigeons were exposed to a discriminated autoshaping procedure in which categories of moving stimuli, presented on videotape, were differentially associated with reinforcement. All stimuli depicted pigeons making defined responses. In Experiment 1, one category consisted of several different scenes of pecking and the other consisted of scenes of walking, flying, head movements, or standing still. Four of the 4 birds for which pecking scenes were positive stimuli discriminated successfully, whereas only 1 of the 4 for which pecking was the negative category did so. In the pecking-positive group, there were differences between the pecking rates in the presence of the four negative actions, and these differences were consistent across subjects. In Experiment 2, only the categories of walking and pecking were used; some but not all birds learned this discrimination, whichever category was positive, and these birds showed some transfer to new stimuli in which the same movements were represented only by a small number of point lights (Johansson's “biological motion” displays). In Experiment 3, discriminations between pecking and walking movement categories using point-light displays were trained. Four of the 8 birds discriminated successfully, but transfer to fully detailed displays could not be demonstrated. Pseudoconcept control groups, in which scenes from the same categories of motion were used in both the positive and negative stimulus sets, were used in Experiments 1 and 3. None of the 8 pigeons trained under these conditions showed discriminative responding. The results suggest that pigeons can respond differentially to moving stimuli on the basis of movement cues alone.     

High-order concept formation in the pigeon

After 30 days of operant training, with pecking responses to aerial photographs containing man-made objects reinforced with food, and no food reinforcement for pecking on photographs not containing man-made objects, a discrimination to the two classes of photographs was obtained. The discriminative response generalized to photographs with which the pigeons had no previous experience. This study demonstrates that pigeons are capable of forming relatively high-order concepts. Some possible stimulus properties controlling the discrimination are discussed.     

Pigeons concurrently categorize photographs at both basic and superordinate levels

We studied categorization in pigeons, using carefully controlled photographs. Within daily sessions, 4 pigeons had to classify each of 32 photographs into either its proper basic-level category (cars, chairs, flowers, or people; four-key forced choice procedure) or its proper superordinate-level category (natural or artificial; two-key forced choice procedure). The pigeons successfully classified the same stimuli at both levels. Overall, the pigeons learned the basic discrimination more quickly than the superordinate discrimination, but this difference was reliable only for artificial stimuli (cars and chairs), not for natural stimuli (flowers and people). The pigeons also exhibited reliable discrimination transfer to novel photographs, attesting to the open-endedness of these basic and superordinate categories.     

Context specificity of operant discriminative performance in pigeons: II. Necessary and sufficient conditions

Six experiments were performed to explore the necessary and sufficient conditions for producing context specificity of discriminative operant performance in pigeons. In Experiment 1, pigeons learned a successive discrimination (red S+/blue S−) in two chambers that had a particular odor present and between which they were frequently switched. The birds subsequently learned the reversal (blue S+/ red S−) in one of these chambers with a different odor present. When switched to the alternative chamber, although the odor and the reinforcement contingency were still appropriate to the reversal, performance appropriate to the original discrimination recurred in subjects for which the houselights were on during training and testing but not for those for which the houselights were off. This indicated the importance of visual contextual cues in producing context specificity. Experiment 2 showed that the frequent switching between boxes in initial training was of no consequence, presumably because the apparatus cues were highly salient to the subjects. Experiment 3 showed significantly less context specificity when odor cues were omitted. Experiment 4 showed that simply using a different reinforced stimulus in each phase of training was ineffective in producing context specificity. Experiment 5 showed that the generalization test procedure used in Experiment 4 was sensitive to context specificity when discrimination-reversal training was used with different odors in the two training phases. Experiment 6 replicated the results of Experiment 4, but then showed that when different odors accompanied the two training phases, context specificity was obtained with the single-stimulus paradigm. Thus in both single-stimulus and discrimination-reversal paradigms, redundant odor cues potentiated learning about apparatus cues.     

Categorical Discrimination of Human Facial Expressions by Pigeons: A Test of the Linear Feature Model

Pigeons were trained to discriminate human facial expressions, happiness and anger, in a go/ no-go discrimination procedure. Five pigeons learned to discriminate photographs of the happy and angry faces of 25 different people and showed high levels of transfer to novel faces expressing the training emotions. The pigeons directed their pecks predominantly to the mouth, eyes, or the area between these features. The pigeons were then tested with familiar stimuli in which the upper and lower parts of the face were manipulated separately by substitution or removal of facial features ('eyes-and-eyebrows' and 'mouth'). It was shown that the salience of particular features differed considerably among the birds, but that a linear feature model adequately accounted for discriminative performance of the birds with these stimuli. Furthermore, the discrimination was maintained when these features were inverted. Thus, the so-called Thatcher illusion did not occur. It is suggested that the discrimination is based not on a feature configuration or perceptual gestalt but on an additive integration of individual features.     

Discrimination of artificial polymorphous categories by rhesus monkeys (Macaca mulatta)

Monkeys (Macaca mulatta) were trained to discriminate between sets of artificial stimuli such as those used by Jitsumori (1993) for pigeons and humans. The stimuli were arrays of symbols differing along three two-valued (positive or negative) dimensions. The discrimination required was between polymorphous categories in which a positive stimulus was defined by possession of any 2 out of 3 positive features. Of the 5 monkeys, 3 learned the discrimination much faster than did pigeons, but transfer to novel stimuli was less impressive than had been shown in pigeons. The 3 monkeys showed high levels of transfer to the stimuli that contained either all 3 positive or all 3 negative features, but 2 of the 3 monkeys failed to show transfer to stimuli that had 1 of the 3 features replaced with a novel one. Analysis of the monkeys' performance raised doubts on the additive integration of features but supported learning of feature combinations as a basis for the discrimination of polymorphous categories by this species.     

Long-term memory of pigeons for stimulus-outcome associations involving multiple stimuli, each seen in a single brief trial

Experiment 1 explored performance of pigeons in two versions of a shortterm recognition memory procedure. In one version responding to entirely novel slides was rewarded, and responding to familiar slides (slides seen once, for 10 sec) was not rewarded; in the other version, responding to familiar slides was rewarded. Performance was initially below chance in both versions of the procedure. This result indicated that in this procedure associations were formed between the slides and the outcome (reward or non-reward) that followed their presentation. The result also suggested that the true capacity of pigeon recognition memory cannot be assessed using these procedures, as performance is inevitably disrupted by the bird's associative memory. The tendency of pigeons to form one-trial associations was exploited in Experiment 2. Phase 1 consisted of 16 two-session cycles: in Session 1 of each cycle, birds were shown 20 novel slides and were rewarded for responding to 10 of those slides; in Session 2, the same slides were shown again, with the same reinforcement contingencies. The birds showed significant overnight retention of the one-trial associations formed in Session 1 of each cycle. Phase 2 showed significant retention over periods of more than 20 days of associations involving 320 slides seen twice only. Phase 3 re-exposed for nine daily sessions one of the sets of 20 slides used in Phases 1 and 2; a high level of discrimination emerged rapidly and 4 (of 8) birds showed, by the end of training, no overlap in response rates to positive and negative slides. Comparative implications of the results are discussed.     

Within-session reversal learning in rhesus macaques (<Emphasis Type="Italic">Macaca mulatta</Emphasis>)

In a midsession reversal (MSR) task, animals are typically presented with a simple, simultaneous discrimination (S1+, S2?) where contingencies are reversed (S1?, S2+) half-way through each session. This paradigm creates multiple, relevant cues that can aid in maximizing overall reinforcement. Recent research has shown that pigeons show systematic anticipatory and perseverative errors across the session, which increase as a function of proximity to the reversal trial. This behavior has been theorized to indicate primary control by temporal cues across the session, instead of the cues provided by recent reinforcement history that appear to control behavior shown by humans. Rats, however, appear to be guided by recent reinforcement history when tested in an operant context, thereby demonstrating behavior that parallels that seen in humans, but they appear to be guided by temporal cues when tested in an open-field apparatus, showing behavior more akin to that seen in pigeons. We tested rhesus macaques (Macaca mulatta) on the MSR with a computerized simultaneous visual discrimination to assess whether they would show errors indicative of control by time or by recent reinforcement history. When a single reversal point occurred midsession, rhesus macaques showed no anticipation of the reversal and a similar level of perseveration to rats tested in an operant setting. Nearly identical results also were observed when the monkeys were trained with a single, variable reversal point or with multiple, variable reversal points within a session. These results indicate that temporal cues are not guiding response flexibility in rhesus macaque visual discrimination.     

Discrimination of five-dimensional stimuli by pigeons: Limitations of feature analysis

Pigeons were trained to discriminate between stimuli constructed using five orthogonal two-valued features. The stimuli consisted of stylized monochrome drawings of seeds. Two different training procedures (conditional and simultaneous discrimination) were used. In the first two experiments, the discrimination required was between polymorphous categories, in which a positive stimulus was defined as one in which three or more of the five features took their positive values. Discrimination in both experiments was imperfect; the pigeons' behaviour only came under the control of a subset of the available features (one to three in Experiment 1, three or four in Experiment 2). In Experiment 3, single features had to be discriminated, while the remaining features varied. It was found that all five features of the “seed” stimuli could be discriminated, but one of them was exceptionally difficult. The results show that pigeons do not reliably use all the features available to them when making category discriminations. This casts doubts on feature analysis as a basis for the excellent performance pigeons show when required to discriminate between categories of natural objects.     

Labeling and family resemblance in the discrimination of polymorphous categories by pigeons

Two experiments examined whether pigeons discriminate polymorphous categories on the basis of a single highly predictive feature or overall similarity. In the first experiment, pigeons were trained to discriminate between categories of photographs of complex real objects. Within these pictures, single features had been manipulated to produce a highly salient texture cue. Either the picture or the texture provided a reliable cue for discrimination during training, but in probe tests, the picture and texture cues were put into conflict. Some pigeons showed a significant tendency to discriminate on the basis of the picture cue (overall similarity or family resemblance), whereas others appeared to rely on the manipulated texture cue. The second experiment used artificial polymorphous categories in which one dimension of the stimulus provided a completely reliable cue to category membership, whereas three other dimensions provided cues that were individually unreliable but collectively provided a completely reliable basis for discrimination. Most pigeons came under the control of the reliable cue rather than the unreliable cues. A minority, however, came under the control of single dimensions from the unreliable set. We conclude that cue salience can be more important than cue reliability in determining what features will control behavior when multiple cues are available.     

Contextual conditional discriminations

Three experiments used a discriminated operant procedure to study conditional discrimination learning in rats. The first experiment showed that rats were capable of learning a biconditional discrimination in which two contexts served as conditional cues signalling the reinforcement contingencies associated with two discriminative stimuli. The discrimination was learned equally well when one discriminative stimulus signalled food, the other its absence, and when one stimulus signalled food, the other extinction plus mild footshock.

In Experiment 2 it was shown that prior training on such a conditional discrimination enhanced the subsequent context specificity of simple conditioning relative to control groups of animals for whom the prior training had not been conditional. Experiment 3 showed that a reversal of the significance of one pair of discriminative stimuli produced no spontaneous reversal in performance to a second, target, pair.

The pattern of results is best accounted for by an analysis of contextual conditional discrimination learning in terms of stimulus configurations and offers no support for the notion that rats may learn a general conditional rule or set.     

Factors affecting conditional discrimination learning by pigeons

In Experiment 1 (within subjects) and Experiment 2 (between subjects) it was shown that the sequential training of pigeons on a color discrimination and then on its reversal, each in a different floor-tilt/texture context, failed to produce conditional control of discriminative performance by those contexts. Daily alternation between the two problems (with correlated contexts) was successful, however. In each of these experiments conditional control was better reflected in generalization test performance in extinction than during sessions of training with reinforcement.     

Behavioural contrast and autoshaping

This experiment was designed to investigate the importance of autoshaping to a signal for reinforcement in the production of behavioural contrast. Two groups of pigeons were given discrimination training on a mult VI-EXT schedule: the stimuli present in the two schedule components shared common attributes, but were distinguished by the presence or absence of a visual feature. For one group (the feature positive group) the feature signalled the availability of reinforcement. For the other group (the feature negative group) the feature signalled nonrein-forcement, and for this group there was no stimulus element which unambiguously signalled reinforcement. The feature positive group showed a higher response rate during the VI component of the mult VI-EXT schedule than the feature negative group. This finding was interpreted as support for the autoshaping explanation of behavioural contrast. The results differed from those of Jenkins and Sainsbury (1969, 1970) in that both the feature positive and the feature negative groups showed discrimination learning.     

Discrimination of polymorphous stimulus sets by pigeons

Pigeons were trained to perform a visual discrimination between stimulus sets in which the presence of any two of three positive features made a stimulus positive, while any two of three negative features made it negative (there were thus three different positive and three different negative stimuli). After training, the birds were exposed to test stimuli that contained either all three positive or all three negative features. In Experiment I three pigeons were successfully trained by a successive method, and subsequently responded to the test stimuli as though they were positive or negative respectively. In Experiment II four pigeons were trained by a simultaneous method. Three learned the discrimination and generalized appropriately to the test stimuli, but they showed no preference between positive test and positive training stimuli, nor any consistent difference in speed of response to them; and similar results were found for negative stimuli. It is argued from this that the pigeons learned to respond to the stimuli as patterns (configurations of features) rather than to the constituent features, but that they generalized to the test stimuli by using the common features. The experiments show that pigeons could in principle learn to discriminate natural polymorphous classes (such as “pigeon” or “person”) without using any single feature, but neither the present experiments nor earlier ones demonstrating discriminations of such natural classes establish that pigeons make use of polymorphous concepts in the same way as people.     

Choice and response contingencies

Two experiments investigated the extent to which response contingencies influence the choice between two schedules of reinforcement by exposing pigeons to a concurrent-chains procedure in which reinforcers in one terminal link were response-independent, and in the other terminal link, response-dependent. In Experiment 1, the pigeons were indifferent between an aperiodic, response-independent schedule and an aperiodic, response-dependent schedule that required a minimum rate of responding. This finding limits the generality of a required-rate contingency as a determinant of choice, which contingency had been previously demonstrated in a context of periodic reinforcement to evoke preference for an alternate schedule. In Experiment 2, the pigeons preferred a periodic, response-independent schedule to a periodic, response-dependent schedule that shared a feature with a required-rate schedule: there was a requirement to respond early in the interreinforcement interval, when responding produced reinforcement only later. The results of the two experiments suggest the following general interpretation: pigeons prefer a second schedule to the extent that the response contingencies of the first schedule must be satisfied during discriminable periods of nonreinforcement.     

Spatial and non-spatial visual discrimination learning in zebrafish (Danio rerio)

Zebrafish (Danio rerio) provide an excellent model for assessment of molecular processes of neurodevelopment. To determine the functional importance of molecular events during neurodevelopment, we have developed methods for assessing learning in zebrafish in a three-chambered fish tank. In the first study, simple escape response was assessed. Zebrafish tested with a moving net learned to escape to another chamber more rapidly over the six sessions of training than the fish with the still net which did not learn. Upon reversal of the contingencies, the fish switched to the inactive net rapidly learned to suppress the escape response and fish formerly in the inactive net condition learned to avoid the moving net. In the second study, spatial discrimination learning was assessed. Zebrafish were trained on a right-left position discrimination to avoid the active net. Zebrafish showed significant improvement in escape responses over six sessions of training with three trials per session. In the third study, red-blue non-spatial discrimination learning was assessed. There was a significant improvement over the first six training sessions. With the reversal of contingencies, there was a significant decline of performance. With continued training, the fish again significantly improved avoidance. These studies found an effective motivational stimulus and procedure for studying escape behavior in zebrafish; a procedure whereby zebrafish would learn both spatial and non-spatial discrimination. These methods are being developed to help determine the functional importance of molecular events during zebrafish neurodevelopment. Accepted after revision: 20 August 2001 Electronic Publication     

Choice as a function of local versus molar reinforcement contingencies.

Rats were trained on a discrete-trial probability learning task. In Experiment 1, the molar reinforcement probabilities for the two response alternatives were equal, and the local contingencies of reinforcement differentially reinforced a win-stay, lose-shift response pattern. The win-stay portion was learned substantially more easily and appeared from the outset of training, suggesting that its occurrence did not depend upon discrimination of the local contingencies but rather only upon simple strengthening effects of individual reinforcements. Control by both types of local contingencies decreased with increases in the intertrial interval, although some control remained with intertrial intervals as long as 30 s. In Experiment 2, the local contingencies always favored win-shift and lose-shift response patterns but were asymmetrical for the two responses, causing the molar reinforcement rates for the two responses to differ. Some learning of the alternation pattern occurred with short intertrial intervals, although win-stay behavior occurred for some subjects. The local reinforcement contingencies were discriminated poorly with longer intertrial intervals. In the absence of control by the local contingencies, choice proportion was determined by the molar contingencies, as indicated by high exponent values for the generalized matching law with long intertrial intervals, and lower values with short intertrial intervals. The results show that when molar contingencies of reinforcement and local contingencies are in opposition, both may have independent roles. Control by molar contingencies cannot generally be explained by local contingencies.     

Procedural antecedents of behavioral contrast: a re-examination of errorless learning

Behavioral contrast reliably occurred in pigeons following errorless discrimination training, contrary to Terrace's (1963) observations. In the main experiment, a 60-sec green keylight, associated with a variable-interval 30-sec schedule of reinforcement alternated with a 60-sec period of extinction when the key was dark. Such aspects of the discrimination training procedure as: (1) the amount of prior nondifferential exposure to the positive stimulus before the discrimination was instituted, and (2) the rapidity with which the negative stimulus was introduced (whether progressively or abruptly) directly influenced the amount of behavioral contrast produced. This occurred independently of the number of errors made by a pigeon during acquisition of the discrimination. In a series of control experiments, substitution of a red keylight for the dark key during extinction resulted in greater behavioral contrast, while an increase to 3 min in the duration of the green keylight associated with reinforcement attenuated the behavioral contrast effect.     

Failure to find evidence of stimulus generalization within pictorial categories in pigeons

Pigeons' key pecks were reinforced in the presence of pictures from one of two categories, cats or cars. A single picture associated with reinforcement was used in Experiment 1, and 20 pictures from the same category were associated with reinforcement in Experiment 2. Pigeons then were presented with novel test pictures from the training category and from the other, previously unseen, category. During Session 1 of testing, pigeons pecked no more often at pictures from the reinforced category than at pictures from the previously unseen category. When pigeons were trained with pictures associated with reinforcement or its absence from different categories in Experiment 3, differential responding to novel pictures from different categories appeared during Session 1. These findings argue against a process of automatic stimulus generalization within natural categories and in favor of the position that category distinctions are not made until members of at least two categories are compared with one another.     

The effects of amount of training per reversal on successive reversals of a color discrimination

Three groups of pigeons were trained on a red-green discrimination in which the stimuli were alternately presented in a multiple schedule of reinforcement. The discrimination was reversed 24 times. Groups were given 1, 2, or 4 hr of training on each discrimination. Increasing the length of training had two principal effects on reversal performance: it increased the rate of extinction of responding to one of the stimuli and increased the rate of reacquisition of responding to the other. The latter effect involved both an increase in reacquisition of responding to a positive stimulus within reversals and an increase in recovery of responding to the previous negative stimulus between reversals. Improvements in performance of each group over the series of reversals were qualitatively similar to the two effects of length of training on each discrimination, and were analogous to effects obtained in other studies involving overtraining and successive reversals of simultaneous discriminations.