Learning to tell apples from oranges

Recognizing images requires sorting them into the correct perceptual categories. Without categories, apples could not be discriminated from oranges. A fundamental question is whether categories are innate or learned, and if learned, how well such discriminations generalize. A recent finding by Notman et al. demonstrates that certain categorical discriminations are learned fast but vanish for slightly rotated stimuli. Rather than higher "cognitive" learning, this indicates involvement of "early" mechanisms, which cannot generalize across different stimulus orientations.     

Repeated acquisition and discrimination reversal in the squirrel monkey (Saimiri sciureus)

Repeated acquisition and discrimination reversal tasks are often used to examine behavioral relations of, respectively, learning and cognitive flexibility. Surprisingly, despite their frequent use in cognitive neuroscience and behavioral pharmacology, variables that control performance under these two tasks have not been widely studied. The present studies were conducted to directly investigate the controlling variables in nonhuman primates. Squirrel monkeys were trained with a touchscreen variant of the repeated acquisition task in which a novel pair of S⁺/S^? stimuli was presented daily. Subjects learned to discriminate the two stimuli (acquisition) and, subsequently, with the contingencies switched (reversal). Results indicate that rates of both acquisition and reversal learning increased across successive sessions, but that rate of reversal learning remained slower than acquisition learning, i.e., more trials were needed for mastery. Subsequent experiments showed this difference between the rate of learning novel discriminations and reversal was reliable for at least 5 days between acquisition and reversal and notwithstanding the interpolation of additional discriminations. Experimental analysis of the S⁺/S^? elements of the tasks revealed that the difference in the rate of learning could not be attributed to a relatively aversive quality of the S^? or to a relatively appetitive quality of the S⁺, but, rather, to contextual control by the S⁺/S^? stimulus complex. Thus, if either element (S⁺ or S^?) of the stimulus complex was replaced by a novel stimulus, the rate of acquisition approximated that expected with a novel stimulus pair. These results improve our understanding of fundamental features of discrimination acquisition and reversal.     

Why are artificial polymorphous concepts so hard for birds to learn?

Stimulus sets defined in terms of artificial polymorphous concepts have frequently been used in experiments to investigate the mechanisms of discrimination of natural concepts, both in humans and in other animals. However, such stimulus sets are frequently difficult for either animals or humans to discriminate. Properties of artificial polymorphous stimulus sets that might explain this difficulty include the complexity of the individual stimuli, the unreliable reinforcement of individual positive features, attentional load, difficulties in discriminating some stimulus dimensions, memory load, and a lack of the correlation between features that characterizes natural concepts. An experiment using chickens as subjects and complex artificial visual stimulus sets investigated these hypotheses by training the birds in discriminations that were not polymorphous but did have some of the properties listed above. Discriminations that involved unreliable reinforcement or high attentional load were found to approach the difficulty of polymorphous concept discriminations, and these two factors together were sufficient to account for the entire difficulty. The usual kind of artificial polymorphous concept may not be a good model for natural concepts as they are perceived and discriminated by birds. A RULEX account of natural concept learning may be preferable.     

Why are artificial polymorphous concepts so hard for birds to learn?

Stimulus sets defined in terms of artificial polymorphous concepts have frequently been used in experiments to investigate the mechanisms of discrimination of natural concepts, both in humans and in other animals. However, such stimulus sets are frequently difficult for either animals or humans to discriminate. Properties of artificial polymorphous stimulus sets that might explain this difficulty include the complexity of the individual stimuli, the unreliable reinforcement of individual positive features, attentional load, difficulties in discriminating some stimulus dimensions, memory load, and a lack of the correlation between features that characterizes natural concepts. An experiment using chickens as subjects and complex artificial visual stimulus sets investigated these hypotheses by training the birds in discriminations that were not polymorphous but did have some of the properties listed above. Discriminations that involved unreliable reinforcement or high attentional load were found to approach the difficulty of polymorphous concept discriminations, and these two factors together were sufficient to account for the entire difficulty. The usual kind of artificial polymorphous concept may not be a good model for natural concepts as they are perceived and discriminated by birds. A RULEX account of natural concept learning may be preferable.     

A discrimination analysis of training-structure effects on stimulus equivalence outcomes.

Experiments designed to establish stimulus equivalence classes frequently produce differential outcomes that may be attributable to training structure, defined as the order and arrangement of baseline conditional discrimination training trials. Several possible explanations for these differences have been suggested. Here we develop a hypothesis based on an analysis of the simple simultaneous and successive discriminations embedded in conditional discrimination training and testing within each of the training structures that are typically used in stimulus equivalence experiments. Our analysis shows that only the comparison-as-node (many-to-one) structure presents all the simple discriminations in training that are subsequently required for consistently positive outcomes on all tests for the properties of equivalence. The sample-as-node (one-to-many) training structure does not present all the simple discriminations required for positive outcomes on either the symmetry or combined transitivity and symmetry (equivalence) tests. The linear-series training structure presents all the simple discriminations required for consistently positive outcomes on tests for symmetry, but not for symmetry and transitivity combined (equivalence) or transitivity alone. Further, the difference in the number of simple discriminations presented in comparison-as-node training versus the other training structures is larger when the intended class size is greater than three or the number of classes is larger than two. We discuss the relevance of this analysis to interpretations of stimulus equivalence research, as well as some methodological and theoretical implications.     

Parallel processing of multidimensional stimuli

A reaction time (RT) task in which S is required to judge pairs of stimuli either the same or different was used in three experiments in order to determine the processing mode employed in complex discriminations. The results of all experiments indicated that stimulus dimensions are processed by parallel but not necessarily independent analyzers.     

S+ versus S− Fading in prompting procedures with autistic children

The purpose of this investigation was to assess the relative effectiveness of two prompt-fading procedures for teaching difficult visual discriminations to autistic children. Both prompt procedures involved within-stimulus fading where manipulation occurred on the relevant component of the discrimination. One procedure involved fading first along the S+ stimulus, while holding the S? stimulus constant. The other procedure involved fading first along the S? stimulus, while holding the S+ stimulus constant. Eight autistic children were each taught two discriminations, one by each of the prompt procedures. Results indicated that for all but one child, the discriminations were acquired significantly faster, with fewer errors, when the S+ stimulus was faded first. These findings are related to the literature on the effects of stimulus novelty on selection and learning.     

Calibration of sensory and cognitive judgments: Two different accounts

The purpose of this study was to test predictions of two recent theories of realism of confidence. Ecological approaches to realism of confidence in one's general knowledge (Gigerenzer et al. , 1991; Juslin, in press a ; Björkman, in press) predict good calibration or, in the case of poor cognitive adjustment, overconfidence, within the cognitive domain. The subjective distance theory of confidence in sensory discriminations (Björkman et al. , 1992) predicts a pervasive underconfidence bias for sensory discriminations. Empirical data are reported showing that: (a) Calibration for sensory judgments is considerably poorer than calibration for well adapted cognitive judgements, a difference that can be entirely traced to underconfidence in the sensory domain. (b) While an initial overconfidence bias in the cognitive domain is removed by outcome feedback, the bias observed in sensory discriminations is unaffected even by a prolonged feedback session. It is suggested that the nature of confidence in sensory discriminations is different from the nature of confidence in cognitive judgments.     

Primate Numerical Competence: Contributions Toward Understanding Nonhuman Cognition

Nonhuman primates represent the most significant extant species for comparative studies of cognition, including such complex phenomena as numerical competence, among others. Studies of numerical skills in monkeys and apes have a long, though somewhat sparse history, although questions for current empirical studies remain of great interest to several fields, including comparative, developmental, and cognitive psychology; anthropology; ethology; and philosophy, to name a few. In addition to demonstrated similarities in complex information processing, empirical studies of a variety of potential cognitive limitations or constraints have provided insights into similarities and differences across the primate order, and continue to offer theoretical and pragmatic directions for future research. An historical overview of primate numerical studies is presented, as well as a summary of the 17‐year research history, including recent findings, of the Comparative Cognition Project at The Ohio State University Chimpanzee Center. Overall, the archival literature on number‐related skills and counting in nonhuman primates offers important implications for revising our thinking about comparative neuroanatomy, cross‐species (human/ape) cognitive similarities and differences, and the evolution of cognition represented by the primate continuum.     

Emergent simple discrimination in children: Transfer of stimulus control under non-reinforced conditions

Previous research on humans suggests that simple discriminations may emerge if both stimuli, B1 and B2, are compounded with the stimuli of a previously trained discrimination, A1 (S+) and A2 (S-), and responding to the compounds, B1A1 and B2A2, is reinforced. Two questions were addressed. First, do simple discriminations also emerge if (a) only one stimulus, B1, is compounded with a training stimulus, A1 (S+) or A2 (S-); or with both training stimuli, A1 (S+) and A2 (S-); and (b) neither B1 nor B2 is compounded with the training stimuli? Second, do simple discriminations emerge if reinforcement in the presence of the AB compounds is withheld? Normal preschool children served as subjects. The study consisted of six experiments. Transfer occurred in all experiments regardless of whether both test stimuli, one test stimulus, or none of the test stimuli were compounded with the training stimuli under non-reinforced conditions. The results can be described by the following rules: Respond to any stimulus that includes a component of a “correct” stimulus of a previous discrimination. Otherwise, respond away from the stimulus that incorporates a component from an “incorrect” stimulus of a previous discrimination. The implications of data for sensory pre-conditioning and language-based accounts are discussed.     

Exclusion performance and learning by exclusion in dogs

Responding by exclusion is a type of emergent repertoire in which an individual chooses an alternative by the apparent exclusion of other available alternatives. In this case it is possible to respond appropriately to an undefined stimulus (one that has not previously acquired discriminative functions) by excluding the defined alternatives. There is evidence of exclusion in humans and nonhuman animals, although learning as an outcome of exclusion does not always occur. This study aimed to investigate exclusion in visual simple discriminations and learning of new simple discriminations resulting from exclusion in four border collies. Subjects were trained to perform simple simultaneous discriminations between pairs of tridimensional objects, and were then tested for exclusion, novelty control and learning of new simple discriminations. All dogs successfully responded by exclusion, choosing an undefined stimulus displayed with an S‐. For three dogs, it was possible to conclude that these previously undefined stimuli acquired S+ functions, documenting learning of new simple discriminations. However, this required up to four exposures to exclusion trials with each pair of stimuli.     

Natural concepts in a juvenile gorilla (gorilla gorilla gorilla) at three levels of abstraction

The extent to which nonhumans are able to form conceptual versus perceptual discriminations remains a matter of debate. Among the great apes, only chimpanzees have been tested for conceptual understanding, defined as the ability to form discriminations not based solely on simple perceptual features of stimuli, and to transfer this learning to novel stimuli. In the present investigation, a young captive female gorilla was trained at three levels of abstraction (concrete, intermediate, and abstract) involving sets of photographs representing natural categories (e.g., orangutans vs. humans, primates vs. nonprimate animals, animals vs. foods). Within each level of abstraction, when the gorilla had learned to discriminate positive from negative exemplars in one set of photographs, a novel set was introduced. Transfer was defined in terms of high accuracy during the first two sessions with the new stimuli. The gorilla acquired discriminations at all three levels of abstraction but showed unambiguous transfer only with the concrete and abstract stimulus sets. Detailed analyses of response patterns revealed little evidence of control by simple stimulus features. Acquisition and transfer involving abstract stimulus sets suggest a conceptual basis for gorilla categorization. The gorilla's relatively poor performance with intermediate-level discriminations parallels findings with pigeons, and suggests a need to reconsider the role of perceptual information in discriminations thought to indicate conceptual behavior in nonhumans.     

Common control by compound samples in conditional discriminations

We tested whether teaching control by single stimulus samples in conditional discriminations would result in common control of two-stimuli compound samples, and vice versa. In Experiment 1, 5 participants were first taught four single-sample conditional discriminations. The first conditional discrimination was as follows: given sample stimulus P1, select comparison stimulus A1 and not A2; given sample P2 select comparison A2 and not A1. The second conditional discrimination was as follows: given sample P1 select comparison B1 and not B2; given sample P2 select B2 and not B1. Different sample stimuli (Q1 and Q2) were used in the third and fourth conditional discriminations. Moreover, A1 and B1 were presented together as comparisons, such that, if Q1 was presented as the sample, A1 was correct and B1 was incorrect; and if Q2 was presented as the sample, B1 was correct and A1 was incorrect. A2 and B2 were also presented as comparisons. When Q1 was presented, A2 was correct and when Q2 was presented B2 was correct. After training with these four single stimulus sample discriminations, participants were tested with compound PQ samples presented with A1, A2, B1, and B2 as comparisons. If common control were established by the PQ stimuli, a participant would select A1 when P1Q1 was presented, A2 when P2Q1 was presented, B1 when P1Q2 was presented, and B2 when P2Q2 was presented. Such common control by PQ samples occurred in 4 of 5 participants. In Experiment 2, 4 participants were given reverse training. They were first taught to select the A1, A2, B1, and B2 stimuli in response to the appropriate PQ combinations and then probed on the single stimulus sample discriminations. All 4 participants were successful on this probe. Experiments 3 and 4 investigated the effects of teaching additional conditional discriminations with novel stimuli on subsequent transfer from the single-sample discriminations to performance on the compound-sample conditional discrimination.     

Differential autoshaping to common and distinctive elements of positive and negative discriminative stimuli

The learning by hungry pigeons of a discrimination between two successively presented compound visual stimuli was investigated using a two-key autoshaping procedure. Common and distinctive stimulus elements were simultaneously presented on separate keys and either followed by food delivery, S+, or not, S−. The subjects acquired both between-trial and within-trial discriminations. On S+ trials, pigeons pecked the distinctive stimulus more than the common stimulus; before responding ceased on S− trials, they pecked the common stimulus more than the distinctive one. Mastery of the within-display discrimination during S+ trials preceded mastery of the between-trials discrimination. These findings extend the Jenkins-Sainsbury analysis of discriminations based upon a single distinguishing feature to discriminations in which common and distinctive elements are associated with both the positive and negative discriminative stimuli. The similarity of these findings to other effects found in autoshaping—approach to signals that forecast reinforcement and withdrawal from signals that forecast nonreinforcement—is also discussed.     

A comparison of methods for teaching receptive labeling to children with autism spectrum disorders

Many early intervention curricular manuals recommend teaching auditory-visual conditional discriminations (i.e., receptive labeling) using the simple-conditional method in which component simple discriminations are taught in isolation and in the presence of a distracter stimulus before the learner is required to respond conditionally. Some have argued that this procedure might be susceptible to faulty stimulus control such as stimulus overselectivity (Green, 2001). Consequently, there has been a call for the use of alternative teaching procedures such as the conditional-only method, which involves conditional discrimination training from the onset of intervention. The purpose of the present study was to compare the simple-conditional and conditional-only methods for teaching receptive labeling to 3 young children diagnosed with autism spectrum disorders. The data indicated that the conditional-only method was a more reliable and efficient teaching procedure. In addition, several error patterns emerged during training using the simple-conditional method. The implications of the results with respect to current teaching practices in early intervention programs are discussed.     

AN EXTENSION OF INCIDENTAL TEACHING PROCEDURES TO READING INSTRUCTION FOR AUTISTIC CHILDREN

In an extension of incidental teaching procedures to reading instruction, two autistic children acquired functional sight-word reading skills in the context of a play activity. Children gained access to preferred toys by selecting the label of the toy in tasks requiring increasingly complex visual discriminations. In addition to demonstrating rapid acquisition of 5-choice discriminations, they showed comprehension on probes requiring reading skills to locate toys stored in labeled boxes. Also examined was postteaching transfer across stimulus materials and response modalities. Implications are that extensions of incidental teaching to new response classes may produce the same benefits documented in communication training, in terms of producing generalization concurrent with skill acquisition in the course of child-preferred activities.     

Stimulus-outcome interactions during instrumental discrimination learning by rats and humans

The associative structure mediating goal-directed action was investigated using congruent and incongruent conditional discriminations. The stimulus was the same as the outcome in each component of the congruent discriminations, whereas the stimulus of one component of the incongruent discriminations was the same as the outcome of the other component. Humans, but not rats, learned the congruent discrimination more rapidly than the incongruent discrimination, a difference that the authors attribute to the fact that outcome-response associations caused response conflict in the incongruent discrimination. Moreover, responding was resistant to outcome devaluation following incongruent, but not congruent, training, suggesting that both humans and rats adopted a stimulus-response strategy to resolve the incongruent discrimination.     

A component model for stimulus variables in discrimination learning

A general function is derived describing the conditioning of a single stimulus component in a discriminative situation. This function, together with the combinatorial rules of statistical learning theory [5, 12], generates empirically testable formulas for learning of classical two-alternative discriminations, probabilistic discriminations, and discriminations based on the outcomes of preceding trials in partial reinforcement experiments.The researches on which this paper is based were facilitated by a grant from the National Science Foundation.     

Theoretical and methodological considerations for the study of imagery in animals

Transformation of a moving stimulus is an operational definition of imagery that is useful in animal research. This definition was developed from structural, functional, and interactive theories of imagery in humans. The historical roots of the definition are traced to behavioristic methodologies within cognitive psychology. An imagery procedure for animals was developed that tests this definition with extrapolated movement. Pigeons were reinforced for discriminating between a rectangle that rotated at constant velocity from a rectangle whose final location violated constant velocity. Imagery trials included a short-term memory component in which the task of the pigeon was to discriminate between extrapolated trajectories of rotation while the rectangle was absent. The evidence supports similarities between human and animal representations of movement. A visible transformation employed as a sample stimulus and followed by a retention interval provides a powerful procedure for measuring the interaction between perception and memory. An advantage of this paradigm is that it could be employed to investigate imagery at the neural level or to obtain comparative behavioral data from different species.