When does social learning become cultural learning?

Developmental research on selective social learning, or ‘social learning strategies’, is currently a rich source of information about when children copy behaviour, and who they prefer to copy. It also has the potential to tell us when and how human social learning becomes cultural learning; i.e. mediated by psychological mechanisms that are specialized, genetically or culturally, to promote cultural inheritance. However, this review article argues that, to realize its potential, research on the development of selective social learning needs more clearly to distinguish functional from mechanistic explanation; to achieve integration with research on attention and learning in adult humans and ‘dumb’ animals; and to recognize that psychological mechanisms can be specialized, not only by genetic evolution, but also by associative learning and cultural evolution.     

Decomposing serial learning: What is missing from the learning curve?

Our current understanding of serial learning relies on the form of the learning curve and on changes in the serial position curve over repeated study-test trials (Ward, 1937). The averaging of data that produces these functions obscures the detailed history of memory for individual items over the course of study-test trials. Extending Tulving’s (1964) analysis of free recall learning, we present a new analysis of serial learning that tracks the acquisition and forgetting of item and order information at the level of individual items. Applying this analysis to two large data sets on serial list learning allows us to discern among hypotheses that are indistinguishable solely on the basis of the learning curve.     

Word learning in dogs?

In a recent paper, Kaminski, Call and Fischer report pioneering research on word-learning in a dog. In this commentary we suggest ways of distinguishing referential word use from mere association. We question whether the dog is reasoning by exclusion and, if so, compare three explanations - learned heuristics, default assumptions, and pragmatic reasoning - as they apply to children and might apply to dogs. Kaminski et al.'s work clearly raises important questions about the origins and basis of word learning and social cognition.     

Constraints on word learning?

Recent studies that propose constraints on word learning are reviewed. Theoretical implications of constraints hypotheses are examined in the light of data from other prior and current studies of children's word learning at different developmental points. It is concluded that there is no evidence for strong internal constraints on the acquisition of words by young children of the kind proposed. process are suggested as more adequate explanations, based on the full research record.     

The “pure-study” learning curve: The learning curve without cumulative testing

The customary assumption in the study of human learning using alternating study and test trials is that learning occurs during study trials and that test trials are useful only to measure learning. In fact, tests seem to play little role in the development of learning, because the learning curve is similar even when the number of test trials varies widely (Tulving, Journal of Verbal Learning and Verbal Behavior 6:175-184, 1967). However, this outcome seems odd, because other research has shown that testing fosters greater long-term learning than does studying. We report three experiments addressing whether tests affect the shape of the learning curve. In two of the experiments, we examined this issue by varying the number of spaced study trials in a sequence and examining performance on only a single test trial at the end of the series (a "pure-study" learning curve). We compared these pure-study learning curves to standard learning curves and found that the standard curves increase more rapidly and reach a higher level in both free recall (Exp. 1) and paired-associate learning (Exp. 2). In Experiment 3, we provided additional study trials in the "pure-study" condition to determine whether the standard (study-test) condition would prove superior to a study-study condition. The standard condition still produced better retention on both immediate and delayed tests. Our experiments show that test trials play an important role in the development of learning using both free-recall (Exps. 1 and 3) and paired-associate (Exp. 2) procedures. Theories of learning have emphasized processes that occur during study, but our results show that processes engaged during tests are also critical.     

The simultaneous learning effect: why does simultaneous task learning improve retention?

Despite the prevalence of encoding variables that have been shown to influence the rate of learning, very few affect the rate of forgetting of verbal material. However, when a list of words is learned simultaneously with other lists, the rate of forgetting is markedly lower than that of single-task learning. Although the magnitude of this simultaneous learning effect is large compared with typical list learning effects, it has received little empirical attention. Experiments 1 and 2 tested the hypothesis that the simultaneous learning effect is the result of a differential contribution of short-term memory during encoding. The results showed that the advantage for simultaneous task learning was obtained even under conditions that minimized the potential effects of short-term memory. Experiment 3 revealed that the simultaneous learning effect was larger when the specific items learned simultaneously were the same on each learning trial than when they were different. This finding supported a cuing explanation of the effect: the items from the other lists act as retrieval cues during delayed recall.     

Whenis category learning holistic?

Ward and Scott (1987) recently provided evidence consistent with the idea that category learning can occur analytically whether that learning takes place under intentional or incidental conditions and whether the learner is an adult or a young child. Kemler Nelson (1988) raised concerns about Ward and Scott’s conclusions as well as about the logic of some of their arguments. Kemler Nelson also attempted to strengthen the argument that incidental learning conditions induce a less strategic approach to tasks, which necessarily results in a holistic mode of processing. The present paper answers Kemler Nelson’s concerns. It also presents arguments in favor of the idea that incidental learning is neither necessarily holistic nor necessarily analytic. It is argued that tasks, whether intentional or incidental, encourage the participant to process particular types of information. If the task encourages or demands the processing of separate features, then that type of information will be stored whether or not the participant was explicitly asked to learn the category membership of presented items. There are varieties of both intentional and incidental category learning, and what the learner does with the presented material is more important than his/her intent to learn. Furthermore, it is argued that the apparently discrepant results discussed by Kemler Nelson and Ward and Scott are all interpretable as specific cases within a more general episodic trace model of processing.     

Processing abstract sequence structure: learning without knowing, or knowing without learning?

Constant interaction with a dynamic environment—from riding a bicycle to segmenting speech—makes sensitivity to the sequential structure of the world a fundamental dimension of information processing. Accounts of sequence learning vary widely, with some authors arguing that parsing and segmentation processes are central, and others proposing that sequence learning involves mere memorization. In this paper, we argue that sequence knowledge is essentially statistical in nature, and that sequence learning involves simple associative prediction mechanisms. We focus on a choice reaction situation introduced by Lee (1997), in which participants were exposed to material that follows a single abstract rule, namely that stimuli are selected randomly, but never appear more than once in a legal sequence. Perhaps surprisingly, people can learn this rule very well. Or can they? We offer a conceptual replication of the original finding, but a very different interpretation of the results, as well as simulation work that makes it clear how highly abstract dimensions of the stimulus material can in fact be learned based on elementary associative mechanisms. We conclude that, when relevant, memory is optimized to facilitate responding to events that have not occurred recently, and that sequence learning in general always involves sensitivity to repetition distance.     

Does learning a complex task have to be complex? A study in learning decomposition

Many theories of skill acquisition have had considerable success in addressing the fine details of learning in relatively simple tasks, but can they scale up to complex tasks that are more typical of human learning in the real world? Some theories argue for scalability by making the implicit assumption that complex tasks consist of many smaller parts, which are learned according to basic learning principles. Surprisingly, there has been rather sparse empirical testing of this crucial assumption. In this article, we examine this assumption directly by decomposing the learning in the Kanfer-Ackerman Air-Traffic Controller Task (Ackerman, 1988) from the learning at the global level all the way down to the learning at the keystroke level. First, we reanalyze the data from Ackerman (1988) and show that the learning in this complex task does indeed reflect the learning of smaller parts at the keystroke level. Second, in a follow-up eye-tracking experiment, we show that a large portion of the learning at the keystroke level reflects the learning even at a lower, i.e., attentional level.     

Rule learning by zebra finches in an artificial grammar learning task: which rule?

A hallmark of the human language faculty is the use of syntactic rules. The natural vocalizations of animals are syntactically simple, but several studies indicate that animals can detect and discriminate more complex structures in acoustic stimuli. However, how they discriminate such structures is often not clear. Using an artificial grammar learning paradigm, zebra finches were tested in a Go/No-go experiment for their ability to distinguish structurally different three-element sound sequences. In Experiment 1, zebra finches learned to discriminate ABA and BAB from ABB, AAB, BBA, and ABB sequences. Tests with probe sounds consisting of four elements suggested that the discrimination was based on attending to the presence or absence of repeated A- and B-elements. One bird generalized the discrimination to a new element type. In Experiment 2, we continued the training by adding four-element songs following a ‘first and last identical versus different’ rule that could not be solved by attending to repetitions. Only two out of five birds learned the overall discrimination. Testing with novel probes demonstrated that discrimination was not based on using the ‘first and last identical’ rule, but on attending to the presence or absence of the individual training stimuli. The two birds differed in the strategies used. Our results thus demonstrate only a limited degree of abstract rule learning but highlight the need for extensive and critical probe testing to examine the rules that animals (and humans) use to solve artificial grammar learning tasks. They also underline that rule learning strategies may differ between individuals.     

What is learning? On the nature and merits of a functional definition of learning

Learning has been defined functionally as changes in behavior that result from experience or mechanistically as changes in the organism that result from experience. Both types of definitions are problematic. We define learning as ontogenetic adaptation—that is, as changes in the behavior of an organism that result from regularities in the environment of the organism. This functional definition not only solves the problems of other definitions, but also has important advantages for cognitive learning research.     

What's social about social learning?

Research on social learning in animals has revealed a rich variety of cases where animals--from caddis fly larvae to chimpanzees--acquire biologically important information by observing the actions of others. A great deal is known about the adaptive functions of social learning, but very little about the cognitive mechanisms that make it possible. Even in the case of imitation, a type of social learning studied in both comparative psychology and cognitive science, there has been minimal contact between the two disciplines. Social learning has been isolated from cognitive science by two longstanding assumptions: that it depends on a set of special-purpose modules--cognitive adaptations for social living; and that these learning mechanisms are largely distinct from the processes mediating human social cognition. Recent research challenges these assumptions by showing that social learning covaries with asocial learning; occurs in solitary animals; and exhibits the same features in diverse species, including humans. Drawing on this evidence, I argue that social and asocial learning depend on the same basic learning mechanisms; these are adapted for the detection of predictive relationships in all natural domains; and they are associative mechanisms--processes that encode information for long-term storage by forging excitatory and inhibitory links between event representations. Thus, human and nonhuman social learning are continuous, and social learning is adaptively specialized--it becomes distinctively "social"--only when input mechanisms (perceptual, attentional, and motivational processes) are phylogenetically or ontogenetically tuned to other agents.     

Is statistical learning a mechanism?

Philosophers of science have offered several definitions of mechanism, most of which have biological or neuroscientific roots. In this paper, I consider whether these definitions apply equally well to cognitive science. I examine this question by looking at the case of statistical learning, which has been called a domain-general learning mechanism in the cognitive scientific literature. I argue that statistical learning does not constitute a mechanism in the philosophical sense of the term. This conclusion points to significant limitations in the scope of the mechanist philosophy when it comes to accounting for explanation in cognitive science.     

What is learned in sequential learning? An associative model of reward magnitude serial-pattern learning

A computational model of sequence learning is described that is based on pairwise associations and generalization. Simulations by the model predicted that rats should learn a long monotonic pattern of food quantities better than a nonmonotonic pattern, as predicted by rule-learning theory, and that they should learn a short nonmonotonic pattern with highly discriminable elements better than 1 with less discriminable elements, as predicted by interitem association theory. In 2 other studies, the model also simulated behavioral "rule generalization," "extrapolation," and associative transfer data motivated by both rule-learning and associative perspectives. Although these simulations do not rule out the possibility that rats can use rule induction to learn serial patterns, they show that a simple associative model can account for the classical behavioral studies implicating rule learning in reward magnitude serial-pattern learning.     

What is learned when concept learning fails?—A theory of restricted-domain relational learning

Two matching-to-sample (MTS) and four same/different (S/D) experiments employed tests to distinguish between item-specific learning and relational learning. One MTS experiment showed item-specific learning when concept learning failed (i.e., no novel-stimulus transfer). Another MTS experiment showed item-specific learning when pigeons’ novel-stimulus transfer decreased because they chose familiar training comparisons instead of matching novel comparisons. In 8-item and 3-item S/D tasks, pigeons and monkeys were accurate with unfamiliar training-stimulus pairings, stimulus inversions, and distorted stimuli, suggesting relational learning within a domain restricted to the training stimuli (i.e., no novel-stimulus transfer). In 32-item S/D tasks, pigeons with previous 8-item training showed less transfer than those without prior training, suggesting a carryover of restricted-domain relational learning. Pigeons shifted from 1024-item to 8-item S/D tasks showed reinstatement of restricted-domain relational learning. These findings are important in specifying which types of learning occur in these tasks, showing that subjects failing novel-stimulus transfer are not required to switch from item-specific to relational learning as a training set is expanded, and demonstrating that concept learning failure is not proof of item-specific learning.     

Does stimulus appearance affect learning?

We examined the learning process with 3 sets of stimuli that have identical symbolic structure but differ in appearance (meaningless letter strings, arrangements of geometric shapes, and sequences of cities). One hypothesis is that the learning process aims to encode symbolic regularity in the same way, largely regardless of appearance. Another is that different types of stimuli bias the learning process to operate in different ways. Using the experimental paradigm of artificial grammar learning, we provided a preliminary test of these hypotheses. In Experiments 1 and 2 we measured performance in terms of grammaticality and found no difference across the 3 sets of stimuli. In Experiment 3 we analyzed performance in terms of both grammaticality and chunk strength. Again we found no differences in performance. Our tentative conclusion is that the learning process aims to encode symbolic regularity independent of stimulus appearance.     

Does fun promote learning? The relationship between fun in the workplace and informal learning

Although research has demonstrated that workplace fun has important benefits, we have an incomplete understanding of the role of fun in the learning domain, especially informal learning. To address this need, the present study examined the influence of fun activities and manager support for fun on informal learning among 206 managers. Fun activities were significantly related to overall informal learning, but manager support for fun was not. Examination of the dimensions of informal learning found that manager support for fun was significantly related to learning from oneself, while fun activities were significantly related to learning from others and learning from non-interpersonal sources. Furthermore, a negative interaction between core-self evaluations and fun activities in predicting learning from oneself was found, suggesting that fun may not be beneficial for all individuals. The key practical implication is that organizations should consider fun as a viable strategy to promote informal learning beyond traditional learning supports. At the same time, organizations should consider the personality of their learners to ensure fun has its intended impact.     

Errorful and errorless learning: The impact of cue–target constraint in learning from errors

The benefits of testing on learning are well described, and attention has recently turned to what happens when errors are elicited during learning: Is testing nonetheless beneficial, or can errors hinder learning? Whilst recent findings have indicated that tests boost learning even if errors are made on every trial, other reports, emphasizing the benefits of errorless learning, have indicated that errors lead to poorer later memory performance. The possibility that this discrepancy is a function of the materials that must be learned—in particular, the relationship between the cues and targets—was addressed here. Cued recall after either a study-only errorless condition or an errorful learning condition was contrasted across cue–target associations, for which the extent to which the target was constrained by the cue was either high or low. Experiment 1 showed that whereas errorful learning led to greater recall for low-constraint stimuli, it led to a significant decrease in recall for high-constraint stimuli. This interaction is thought to reflect the extent to which retrieval is constrained by the cue–target association, as well as by the presence of preexisting semantic associations. The advantage of errorful retrieval for low-constraint stimuli was replicated in Experiment 2, and the interaction with stimulus type was replicated in Experiment 3, even when guesses were randomly designated as being either correct or incorrect. This pattern provides support for inferences derived from reports in which participants made errors on all learning trials, whilst highlighting the impact of material characteristics on the benefits and disadvantages that accrue from errorful learning in episodic memory.