Same/different concept learning in the pigeon: the effect of negative instances and prior adaptation to transfer stimuli

Interference of same/different learning by a spatial discrimination

Same/different learning by pigeons has long been of interest to experimental psychologists. In one of these procedures, matching-to-sample, responses to a sample stimulus result in the presentation of two comparison stimuli, one of which matches the sample, the other of which does not, and choice of the matching stimulus is reinforced. Evidence of a matching concept has been found when transfer has been found to new stimuli. Given the transfer results, it is surprising that acquisition of two matching tasks (or two mismatching tasks), has not been found to be any faster than one matching and one mismatching task (i.e., two compatible tasks do not appear to facilitate each other). In the present experiment, we asked if matching acquisition involving three colors would be retarded if the correct response to a fourth color was not matching but was spatial (e.g., if the sample is red choose the red comparison, if the sample is green choose the green comparison, if the sample is yellow choose the yellow comparison, but if the sample is blue choose the left comparison). We found that acquisition of this task was slower than acquisition of a four color matching task (i.e., when the sample was blue, the blue comparison was correct). The results suggest that there is an interaction among matching associations, such that common rules facilitate learning compared with having to learn an inconsistent (spatial) rule. This result provides further evidence of the development of a matching concept by pigeons.

Are There Differences in "Intelligence" Between Nonhuman Species? The Role of Contextual Variables

We review evidence for Macphail’s (1982, 1985, 1987)Null Hypothesis, that nonhumans animals do not differ either qualitatively or quantitatively in their cognitive capacities. Our review supports the Null Hypothesis in so much as there are no qualitative differences among nonhuman vertebrate animals, and any observed differences along the qualitative dimension can be attributed to failures to account for contextual variables. We argue species do differ quantitatively, however, and that the main difference in “intelligence” among animals lies in the degree to which one must account for contextual variables.

Matching and Oddity Learning in the Pigeon: Transfer Effects and the Absence of Relational Learning

Three experiments examined the extent to which pigeons trained on a matching or oddity discrimination with one pair of colours showed transfer when tested on a new matching or oddity discrimination with a new pair of colours. Experiment 1 examined the effects of key spacing and a delay procedure and replicated previous reports that in the transfer stage subjects given the same kind of problem (Non-shift condition) in general learn more rapidly than those given the opposite problem (Shift condition). However, this difference appeared only when pigeons given matching in both training and transfer stages were compared to those shifted from oddity to matching; it did not appear in birds transferred to oddity. Transfer was not significantly affected by key spacing or by the delay.

Experiments 2 and 3 examined transfer from a non-relational conditional discrimination based on one set of colours to a subsequent matching or oddity task based on two new colours. Both a comparison between the results of Experiment 1 and 2 and the corresponding within-experiment comparison from Experiment 3 showed that transfer from conditional training to matching was as great as from prior training on matching, while prior training on oddity produced negative transfer on shift to matching. It was suggested that this negative transfer occurs because pigeons trained on oddity have not learned to override an initial bias towards the odd stimulus in an array. Whatever the correct explanation; the present results provide no support for the claim that pigeons solve matching or oddity discriminations relationally.

Transfer of Relational Rules in Matching and Oddity Learning by Pigeons and Corvids

Three experiments compared the performance of pigeons and corvids when they were given the opportunity to transfer the relational rule underlying matching or oddity discriminations to new sets of stimuli. In the first, pigeons and jackdaws were initially trained either on a matching or on a non-relational conditional discrimination and then transferred to a new matching discrimination. In the second, pigeons and jays were trained on a series of three matching (or oddity) discriminations with three different pairs of colours and finally tested, either with the same or the reversed rule, on matching or oddity to line orientations. In the third, pigeons and rooks were trained to perform one response when two coloured panels were the same and a different response when the two colours were different and then transferred, either with the same or the reversed rule, to a new set of colour stimuli. All three experiments produced the same result: no evidence of transfer of the relational rule by pigeons, but substantial and significant transfer by corvids.

Concept Learning and Learning Strategies

Concept learning and learning strategies of pigeons were manipulated in a matching-to-sample task. Groups of 4 pigeons responded either 0, 1, 10, or 20 times to a sample stimulus, and then chose between a matching comparison stimulus and a nonmatching comparison stimulus. Tests with unfamiliar arrangements of the three training stimuli showed that learning was not by if-then rules. Tests with novel stimuli showed that as the number of sample responses increased, learning about the configural pattern of each display gave way to more learning about the sample-comparison relationship and more concept learning. Pigeons making the most sample responses showed complete concept learning.

Concept learning set-size functions for Clark's nutcrackers

Same/Different abstract-concept learning by Clark's nutcrackers (Nucifraga columbiana) was tested with novel stimuli following learning of training set expansion (8, 16, 32, 64, 128, 256, 512, and 1024 picture items). The resulting set-size function was compared to those from rhesus monkeys (Macaca mulatta), capuchin monkeys (Cebus apella), and pigeons (Columba livia). Nutcrackers showed partial concept learning following initial eight-item set learning, unlike the other species (Magnotti, Katz, Wright, & Kelly, 2015). The mean function for the nutcrackers' novel-stimulus transfer increased linearly as a function of the logarithm of training set size, which intersected its baseline function at the 128-item set size. Thus, nutcrackers on average achieved full concept learning (i.e., transfer statistically equivalent to baseline performance) somewhere between set sizes of 64 to 128 items, similar to full concept learning by monkeys. Pigeons required a somewhat larger training set (256 items) for full concept learning, but results from other experiments (initial training and transfer with 32- and 64-item set sizes) suggested carryover effects with smaller set sizes may have artificially prolonged the pigeon's full concept learning. We find it remarkable that these diverse species with very different neural architectures can fully learn this same/different abstract concept, and (at least under some conditions) do so with roughly similar sets sizes (64-128 items) and numbers of training exemplars, despite initial concept learning advantages (nutcrackers), learning disadvantages (pigeons), or increasing baselines (monkeys).

Endpoint distinctiveness facilitates analogical mapping in pigeons

Analogical thinking necessitates mapping shared relations across two separate domains. We investigated whether pigeons could learn faster with ordinal mapping of relations across two physical dimensions (circle size & choice spatial position) relative to random mapping of these relations. Pigeons were trained to relate six circular samples of different sizes to horizontally positioned choice locations in a six alternative matching-to-sample task. Three pigeons were trained in a mapped condition in which circle size mapped directly onto choice spatial position. Three other pigeons were trained in a random condition in which the relations between size and choice position were arbitrarily assigned. The mapped group showed an advantage over the random group in acquiring this task. In a subsequent second phase, relations between the dimensions were ordinally reversed for the mapped group and re-randomized for the random group. There was no difference in how quickly matching accuracy re-emerged in the two groups, although the mapped group eventually performed more accurately. Analyses suggested this mapped advantage was likely due to endpoint distinctiveness and the benefits of proximity errors during choice responding rather than a conceptual or relational advantage attributable to the common or ordinal mapping of the two dimensions. This potential difficulty in mapping relations across dimensions may limit the pigeons' capacity for more advanced types of analogical reasoning. This article is part of a Special Issue entitled: Tribute to Tom Zentall.

Contrasting styles in cognition and behaviour in bumblebees and honeybees

Bumblebees and honeybees have been the subjects of a great deal of recent research in animal cognition. Many of the major topics in cognition, including memory, attention, concept learning, numerosity, spatial cognition, timing, social learning, and metacognition have been examined in bumblebees, honeybees, or both. Although bumblebees and honeybees are very closely related, they also differ in important ways, including social organization, development, and foraging behaviour. We examine whether differences between bumblebees and honeybees in cognitive processes are related to differences in their natural history and behaviour. There are differences in some cognitive traits, such as serial reversal learning and matching-to-sample, that appear related to differences between bumblebees and honeybees in foraging and social behaviour. Other cognitive processes, such as numerosity, appear to be very similar. Despite the wealth of information that is available on some aspects of bumblebee and honeybee cognition and behaviour, there are relatively few instances, however, in which adequate data exist to make direct comparisons. We highlight a number of phenomena, including concept learning, spatial cognition, timing, and metacognition, for which targeted comparative research may reveal unexpected adaptive variation in cognitive processes in these complex animals. This article is part of a Special Issue entitled: In Honor of Jerry Hogan.

A harbor seal can transfer the same/different concept to new stimulus dimensions

We investigated the formation of an abstract concept of same/different in a harbor seal by means of a two-item same/different task. Stimuli were presented on a TFT monitor. The subject was trained to respond according to whether two horizontally aligned white shapes presented on a black background were the same, or different from each other, by giving a no-go or go response. Training comprised of four stages. First, the same/different task was trained with two shapes forming two same problems (A-A and B-B) and two different problems (A-B and B-A). After the learning criterion was reached, training proceeded with new pairs of shapes. In the second experimental stage, every problem was presented just five times before new problems were introduced. We showed that training to criterion with just two shapes resulted in item-specific learning, whereas reducing the number of presentations to five per problem led to the formation of a same/different learning set as well as some restricted relational learning. Training with trial-unique problems in the third stage of this study resulted in the formation of an abstract concept of same/different which was indicated by a highly significant performance in transfer tests with 120 novel problems. Finally, extra-dimensional transfer of the concept was tested. The harbor seal showed a significantly correct performance on transfer tests with 30 unfamiliar pattern and 60 unfamiliar brightness same/different problems, thus demonstrating that the concept is not restricted to the shape dimension originally learned, but can be generalized across stimulus dimensions.

Visually guided decision making in foraging honeybees

Honeybees can easily be trained to perform different types of discrimination tasks under controlled laboratory conditions. This review describes a range of experiments carried out with free-flying forager honeybees under such conditions. The research done over the past 30 or so years suggests that cognitive abilities (learning and perception) in insects are more intricate and flexible than was originally imagined. It has become apparent that honeybees are capable of a variety of visually guided tasks, involving decision making under challenging situations: this includes simultaneously making use of different sensory modalities, such as vision and olfaction, and learning to use abstract concepts such as “sameness” and “difference.” Many studies have shown that decision making in foraging honeybees is highly flexible. The trained animals learn how to solve a task, and do so with a high accuracy, but when they are presented with a new variation of the task, they apply the learnt rules from the earlier setup to the new situation, and solve the new task as well. Honeybees therefore not only feature a rich behavioral repertoire to choose from, but also make decisions most apt to the current situation. The experiments in this review give an insight into the environmental cues and cognitive resources that are probably highly significant for a forager bee that must continually make decisions regarding patches of resources to be exploited.

Formation of transitivity in conditional matching to sample by pigeons

Four homing pigeons were trained over 5 months in a zero-delay, "arbitrary" matching-to-sample procedure with sample and comparison stimuli presented on any of three response keys. Birds were also required to complete a fixed-ratio 10 requirement on both sample and comparison stimuli to terminate their presentation. The procedure resulted in the establishment of relations that were not specifically trained and that can be characterized by the property of transitivity in a stimulus equivalence context. This result was in contrast with the findings obtained from most previous research with nonhuman subjects.

Transfer of matching-to-figure samples in the pigeon

Three pigeons were trained on a modified six-key matching-to-sample procedure. The third peck on the figure-sample key (which presented a bird, hand, face, beetle, rabbit, fish, flower, or red hue, as the sample) lighted only one comparison key. Every three additional pecks on the sample lighted another comparison key, up to a maximum of five keys. Pecks on keys of matching figures produced grain. Pecks on nonmatching keys (mismatches) turned off all lights on the comparison keys and repeated the trial. Three figures were used during acquisition. The birds learned to peck each sample until the matching comparison stimulus appeared on one of three comparison stimulus keys, and then to peck that key. Later, five novel stimuli, employed as both sample and comparison stimuli, and two additional matching keys were added. Each bird showed matching transfer to the novel samples. The data suggest that the birds may have learned the concept of figure matching rather than a series of two-component chains or discrete five-key discriminations.

The codes of man and beasts

Exposing the chimpanzee to language training appears to enhance the animal's ability to perform some kinds of tasks but not others. The abilities that are enhanced involve abstract judgment, as in analogical reasoning, matching proportions of physically unlike exemplars, and completing incomplete (external) representations of action. The abilities that do not improve concern the location of items in space and the inferences one might make in attempting to obtain them. Representing items in space and making inferences about them could be done with an imaginal code, but representing relations and judging the relations between them, as in analogies, requires a more abstract code. Language training cannot instill such an abstract code, but for species that have the code to start with, it may enhance the animal's ability to use it.

The meaning of representation in animal memory

A representation is a remnant of previous experience that allows that experience to affect later behavior. This paper develops a metatheoretical view of representation and applies it to issues concerning representation in animals. To describe a representational system one must specify the following: the domain or range of situations in the represented world to which the system applies; the content or set of features encoded and preserved by the system; the code or transformational rules relating features of the representation to the corresponding features of the represented world; the medium, or the representation's physical instantiation; and the dynamics, or how the system changes with time. In part because of the behaviorist assumption that the hypothetical, covert changes occurring in an organism during learning correspond to the overt physical changes that are observed, issues of representation in animal behavior have been largely ignored as irrelevant or misleading. However, it can be inferred that representations, acting as models of environmental regularities, operate at many levels of behavioral functioning, both cognitive and noncognitive. Objections to the use of this concept in explanations of animal behavior, based on the claim that it is indeterminate and on behaviorist considerations of parsimony, can be answered. Animal representations may be specialized in terms of tasks and species. Data from tasks involving spatial memory, delayed matching-to-sample, and sequence learning suggest some foundations for a general theory of animal representations.

Matching and oddity relational learning by pigeons (Columba livia): transfer from color to shape

Relational learning, as opposed to perceptual learning, is based on the abstract properties of the stimuli. Although at present there is no doubt that pigeons are capable of relational behavior, this study aims to further disclose the conditions under which it occurs. Pigeons were trained in an outdoor cage on a matching-to-sample or an oddity-from-sample task, with colored cardboard stimuli presented horizontally. The apparatus involved three sliding lids on which the stimuli were drawn and which, when displaced, revealed the reinforcement. The lids were either adjacent to each other or somewhat separated. Training sessions involved two colors, and test sessions six different colors (same dimension test), or six different shapes (different dimension test). One group of birds trained under the 'adjacent' condition failed when tested with new stimuli, but succeeded in both dimension tests after training under the 'separate' condition. Two other groups of birds succeeded in all tests after training under the latter condition. These results show that depending on procedural details, pigeons are or are not able to transfer from one visual dimension to another, thus extending previous related findings.

Two-item same-different concept learning in pigeons

We report the first successful demonstration of a simultaneous, two-item same-different (S/D) discrimination by 6 pigeons, in which nonpictorial color and shape stimuli were used. This study was conducted because the majority of recently successful demonstrations of S/D discrimination in pigeons have employed displays with more than two items. Two pairs of stimulus items were simultaneously presented on a touch screen equipped computer monitor. Pigeons were reinforced for consistently pecking at either the same (i.e., identical) or the different (i.e., nonidentical) pair of items. These pairs were created from combinations of simple colored shapes drawn from a pool of six colors and six shapes. After acquiring the discrimination with item pairs that differed redundantly in both the shape and the color dimensions, the pigeons were tested for transfer to items that varied in only one of these dimensions. Although both dimensions contributed to the discrimination, greater control was exhibited by the color dimension. Most important, the discrimination transferred in tests with novel colored, shaped, and sized items, suggesting that the mechanisms involved were not stimulus specific but were more generalized in nature. These results suggest that the capacity to judge S/D relations is present in pigeons even when only two stimuli are used to implement this contrast.

Learning Processes in Matching and Oddity: The Oddity Preference Effect and Sample Reinforcement

Eight pigeons learned either matching (to sample) or oddity (from sample) with or without reward for sample responding. The training stimuli were coarse-white, fine-black, or smooth-mauve gravels in pots with buried grain as the reinforcer. Oddity without sample reward was learned most rapidly, followed by matching with sample reward, oddity with sample reward, and matching without sample reward. Transfer was related to acquisition rate: The oddity group without sample reward showed full (equal to baseline) color and texture transfer; the matching group with sample reward showed partial texture transfer; other groups showed no transfer. Sample reward was shown to determine rate of acquisition of matching and oddity and the oddity preference effect. The results are discussed in terms of item-specific associations operating early in learning prior to any relational learning between sample and comparison stimuli.

Degree of representation of the matching concept in pigeons (Columba livia)

In Experiment 1, 12 pigeons (Columba livia) were trained on a simultaneous matching-to-sample task with 2 stimuli and then tested with 2 novel stimuli. Half of the birds were trained with a fixed ratio schedule requirement of 1 (FR1) or 20 (FR20) pecks on the sample stimulus. None of the birds showed any evidence of concept-mediated transfer. In Experiment 2, 12 pigeons were trained with 3 stimuli and then tested with the same novel stimuli used in Experiment 1. Half of the birds in each group were trained with either an FR1 or FR20 requirement on the sample stimulus. Two of the FR20 birds showed high levels of transfer to the novel stimuli similar to that of monkeys in a previous study.

Categorization, concept learning, and behavior analysis: an introduction

Categorization and concept learning encompass some of the most important aspects of behavior, but historically they have not been central topics in the experimental analysis of behavior. To introduce this special issue of the Journal of the Experimental Analysis of Behavior (JEAB), we define key terms; distinguish between the study of concepts and the study of concept learning; describe three types of concept learning characterized by the stimulus classes they yield; and briefly identify several other themes (e.g., quantitative modeling and ties to language) that appear in the literature. As the special issue demonstrates, a surprising amount and diversity of work is being conducted that either represents a behavior-analytic perspective or can inform or constructively challenge this perspective.

The concepts of 'sameness' and 'difference' in an insect

Insects process and learn information flexibly to adapt to their environment. The honeybee Apis mellifera constitutes a traditional model for studying learning and memory at behavioural, cellular and molecular levels. Earlier studies focused on elementary associative and non-associative forms of learning determined by either olfactory conditioning of the proboscis extension reflex or the learning of visual stimuli in an operant context. However, research has indicated that bees are capable of cognitive performances that were thought to occur only in some vertebrate species. For example, honeybees can interpolate visual information, exhibit associative recall, categorize visual information and learn contextual information. Here we show that honeybees can form 'sameness' and 'difference' concepts. They learn to solve 'delayed matching-to-sample' tasks, in which they are required to respond to a matching stimulus, and 'delayed non-matching-to-sample' tasks, in which they are required to respond to a different stimulus; they can also transfer the learned rules to new stimuli of the same or a different sensory modality. Thus, not only can bees learn specific objects and their physical parameters, but they can also master abstract inter-relationships, such as sameness and difference.

Base rates versus sample accuracy: competition for control in human matching to sample

People often place undue weight on specific sources of information (case cues) and insufficient weight on more global sources (base rates) even when the latter are highly predictive, a phenomenon termed base-rate neglect. This phenomenon was first demonstrated with paper-and-pencil tasks, and also occurs in a matching-to-sample procedure in which subjects directly experience case sample (cue) accuracy and base rates, and in which discrete, nonverbal choices are made. In two nonverbal experiments, subjects were exposed to hundreds of trials in which they chose between two response options that were both probabilistically reinforced. In Experiment 1, following one of two possible samples (the unpredictive sample), either response was reinforced with a .5 probability. The other sample (predictive) provided reinforcement for matching on 80% of the trials in one condition but in only 20% of the trials in another condition. Subjects' choices following the unpredictive sample were determined primarily by the contingencies in effect for the predictive sample: If matching was reinforced following the predictive sample, subjects tended to match the unpredictive sample as well; if countermatching the predictive sample was generally reinforced, subjects tended to countermatch the unpredictive sample. These results demonstrate only weak control by base rates. In Experiment 2, base rates and sample accuracy were simultaneously varied in opposite directions to keep one set of conditional probabilities constant. Subjects' choices were determined primarily by the overall accuracy of the sample, again demonstrating only weak control by base rates. The same pattern of choice occurred whether this pattern increased or decreased rate of reinforcement. Together, the results of the two experiments provide a clear empirical demonstration of base-rate neglect.