Adolescent learners’ characteristics PDF

Preview

Summary

A that included those behaviors? How does the fixedness A Salience Theory of Learning that inheres in response reinforcement give way to the emergence of new behaviors and concepts, such as DUANE M. RUMBAUGH1, JAMES E. KING2, MICHAEL J. complex musical compositions and groundbreaking BERAN3, DAVID A...

Description

A A Salience Theory of Learning DUANE M. RUMBAUGH1, JAMES E. KING2, MICHAEL J. BERAN3, DAVID A. WASHBURN4, KRISTY GOULD5 1 Language Research Center, Georgia State University, Atlanta, GA, USA 2 University of Arizona, Tucson, AZ, USA 3 Language Research Center, Georgia State University, University Plaza, Atlanta, GA, USA 4 Department of Psychology, Georgia State University, University Plaza, Atlanta, GA, USA 5 Luther College, Decorah, IA, USA

Synonyms Comparative cognition; Emergent learning; Rational behaviorism

Definition The perspective that responses are elicited by stimuli to which they have become associated or learned because they are reinforced remains strongly entrenched in psychological thought. Just what reinforcers are and how they operate, perhaps as agents that bond responses to stimuli, are unresolved issues. The most generally accepted definition of a reinforcer is that it is an event that increases the probability that a response will reoccur if it is reinforced. But that definition is circular and does not explain how reinforcement works. Here, we outline a perspective on learning called Salience Theory that offers a process by which learning occurs across instances of stimulus pairings and the resultant sharing of response-eliciting processes that occur.

Theoretical Background Despite its popularity and robust history, this stimulus– response–reinforcement formulation has inherent weaknesses. How can radically new and creative behaviors suddenly occur in the absence of a training history

that included those behaviors? How does the fixedness that inheres in response reinforcement give way to the emergence of new behaviors and concepts, such as complex musical compositions and groundbreaking inventions? The Salience Theory of Learning proposes a new approach to account for the origins of novel, unexpected, and even intelligent behaviors and new abilities (e.g., competence in speech comprehension). It reformulates reinforcement and how it has its effects upon learning and behavior in terms of its salience, stimulus strength, and response-eliciting properties. Indeed, it formulates the contributions, the impact of all stimuli in the formation of our basic unit of learning, amalgams, in precisely those same terms. We know that what is trained via specific reinforcement of specific responses does not necessarily constrain what is learned (Rumbaugh and Washburn 2003). How can this be according to Reinforcement Theory? What the subject learns might well be far more complex and even qualitatively different from the behavior that one specifically reinforced. For instance, a rhesus monkey (Macaca mulatta) was trained with reinforcement over the course of several months and thousands of training trials to control a joystick with its foot in a complex interactive computer task. Use of a hand was precluded, hence never trained. Only in a later test was the monkey given its very first opportunity to use either its hand or foot to do the task. Now, since all reinforced training had been with its foot, use of its hand should have been at most a remote probability. Yet, when given a choice, the monkey promptly used its hand, scoring significantly better than it had ever done with its foot. Clearly, this finding is inconsistent with Reinforcement Theory. That the monkey used its hand might be said to reflect its massive and prior history in the use of its hand for fine manipulations of objects and foods, but that does not answer the stronger question of how the monkey knew how to use its hand skillfully.

N. Seel (ed.), Encyclopedia of the Sciences of Learning, DOI 10.1007/978-1-4419-1428-6, # Springer Science+Business Media, LLC 2012

2

A

A Salience Theory of Learning

The answer to this stronger, more pointed, question is that the monkey somehow had learned about the task and principles of performing with precision by reinforced training with its foot. The effects of reinforced training were not limited to controlled use of the foot. The learning became more abstract and served the skillful use of its hand when that became an option in subsequent test. Reinforcement Theory dates back more than 100 years to E. L. Thorndike and Alexander Bain. Today, advocates of controlling and modifying behavior testify to the apparent effectiveness of reinforcement and its pragmatic effects. Although we do not deny the seemingly special power of reinforcement in the acquisition and control of behavior, we suggest that it has no special power apart from its salience, its strength as a stimulus, and its response-eliciting properties as it enters amalgam formation with other contiguous stimuli that originate either from the external environment or internally. It thus stands to reason that it also will be the salience of any given stimulus and the strength of its responseeliciting properties that will determine its impact in amalgam formation – but always relative to the salience strengths and the response-eliciting properties of other stimuli with which it might form other amalgams. The Salience Theory of Learning proposes an account of learning that does not have the extreme fixedness that inheres in the stimulus–response– reinforcement model, at least when the latter is taken at its face value. Behavior is too variable, too clever, too creative, and too versatile for it to be so constrained. Tolman (1948) observed this fact and concluded, as does Salience Theory, that expectancies and cognitions about what-leads-to-what emerge from the integration of past experiences including conditioning. Stimulus–response and stimulus–stimulus associations are posited to have a basic role in our Salience Theory of Learning and Behavior, but not as instantiated by reinforcement as historically defined. Rather, associations that are induced among reliably and contiguously associated events are held to generate new composites that we term amalgams – our basic units of learning in Salience Theory. Amalgams are neither habits nor bonds. Importantly, all of the contiguous events that enter into the formation of a given stream of amalgam formations are posited to share interactively their saliencies and their response-eliciting properties. Thus, amalgams are somewhat different from the

individual events that form them. Accordingly, amalgams might generate unique behavior, possibly apart from any single event that enters in their formation. Salience Theory (Rumbaugh et al. 2007) embraces behavioral parameters from heritable and stereotyped instincts through conditioning and on to the emergence of highly complex behaviors that are adaptive. It merits emphasis that complex behaviors can be so novel, so complex, that their emergence through selective shaping and reinforcement is virtually impossible. Those complex behaviors and skills are called emergents, and they constitute a category of behavioral adaptations distinctly separate from the well-known respondent (i.e., Pavlovian) and operant dichotomy proposed mid-twentieth century by B. F. Skinner. Thus, Salience Theory proposes a trichotomy of learned behaviors: Respondent conditioning, Operant conditioning, and Emergents. Each of these categories has its own distinctive protocols and defining attributes (Rumbaugh et al. 2007). Traditional learning theory has regarded both respondent and operant conditioning as contingent upon stimulus events that are in close temporal contiguity with the responses to be conditioned – the unconditional stimulus in the case of respondent conditioning and contingent reinforcement of the response in the case of operant conditioning. Rather than limiting emphasis to events that act solely upon responses, Salience Theory views organisms as constantly surveying their perceptual worlds as if foraging for stimuli that are important or salient along with other stimuli temporally or spatially contiguous with those salient stimuli. Thus, organisms are able to garner the resources needed to sustain life and learn adaptive behavior, while minimizing risk and conserving energy. Salient events, including those related to significant others (e.g., mothers, family members and cohorts) in a social group, provide the basis for observational learning from birth through maturity and the transmission of culture Salience Theory illuminates both the antecedents and the consequences of learned and emergent behaviors, as does Reinforcement Theory. The theory is eclectic; it includes many components that are parts of other theories. It does not reject any body of empirical evidence and intends no derision of the giants of our time (see Marx and Hillix 1987, for an overview of our roots).

A Salience Theory of Learning

On the Parsimony of Salience Theory Amalgam formation is similar to the initial learning stages of sensory preconditioning and classical conditioning. However, psychologists dating back to Thorndike have invoked a new process to explain instrumental conditioning, namely reinforcement. So there is an awkward discontinuity here going from the primitive association of two contiguous stimuli to the more complex operation of reinforcers. There is an old and extensive literature on the awkwardness of the reinforcement explanation, including the problem of explaining how an event following a response can affect its probability. Furthermore, reinforcement implies an evolutionary discontinuity in the learning process in which the primitive association of two stimuli is amended by the conceptually more complex operation of reinforcement. The advantage of the saliency approach is that both the evolutionary discontinuity and consequent lack of parsimony of the reinforcement approach disappear. There is just one fundamental process: amalgam formation. Amalgam formation underlies sensory preconditioning, classical conditioning, and most importantly instrumental conditioning and the formation of emergents. Amalgam formation at its most basic level occurs when a highly salient stimulus (i.e., the unconditional stimulus) and a less salient stimulus (i.e., the conditional stimulus) come into either temporal or spatial contiguity. The impact of the unconditional stimulus is so strong and dominant in classical conditioning that the conditional stimulus comes to serve as an approximation of it. Thus, the conditional stimulus accrues salience and response-eliciting properties approximating those of the unconditional stimulus. To a lesser yet measurable degree, the conditional stimulus shares its salience and response-eliciting attributes with the unconditional stimulus. In operant conditioning, the reward is the most salient stimulus event. Responseproduced stimuli produced by the correct response form an amalgam with the reward produced stimuli. The already existing high salience of the reward-related stimuli then accrues to the response-associated stimuli, thereby increasing the likelihood of the response. Thus, the subject learns how in a given situation it can obtain the resources for which it forages while minimizing risk and injury. As amalgams are incorporated into higherlevel templates, emergent behaviors become possible.

A

Although great gaps of knowledge need to be filled by neuroscience and continuing behavioral research, Salience Theory advances the consilience of psychology, biology, and neuroscience (Naour et al. 2009; Rumbaugh et al. 2007).

Instincts, Respondents, Operants, and Emergents We assume that organisms attend most closely to the most salient events in their perceived worlds and thus garner vital resources and minimize risk. To avoid circularity insofar as possible, we describe the major sources of natural and acquired facets of saliences. Briefly, they are as follows: Genes – sign stimuli; releasers (e.g., the pecking of the red dot on mother’s beak by gull chicks to induce her to regurgitate food) Stimulus intensity – pressure, pain, sharp roar, bright lights, strength of sign stimuli Past associations – conditioning; sensory preconditioning; classical and operant conditioning; conditioned reinforcers (Skinner 1938); and secondary reinforcers (Hull 1943) Principles of perceptual organization – (e.g., closure, clustering of similar stimuli, induced motion, uniqueness/novelty) Amalgams are posited as the basic units of learning. Metaphorically, they may be viewed as neural entries to a never-ceasing sequence of events and stimuli. In creating that record, salient events might serve as commands to “make an entry.” The brains of all species have become honed to make these entries in such a way that the likelihood of adaptation and survival are maximized. Salience Theory views the brain as generating an endless flow of amalgams that reflect experience as time flows on; the brain also organizes the amalgams into natural templates (e.g., readiness to learn different things within a general category.) and/or acquired characteristics (e.g., symbol-based, as with language, traffic signs, and language itself). Salience Theory posits that as the brain works to resolve for the best fit among the amalgams and the templates to which they are assigned that emergents and even new skills might be given birth (Rumbaugh et al. 2007; Savage-Rumbaugh et al. 1993; Tolman 1948). They, in turn, might enable the performance of familiar tasks in more efficient ways and facilitate novel

3

A

4

A

A Stability Bias in Human Memory

problem solving in an ever-changing environment. Thus, the accumulation of experience can contribute richly to the formation of a knowledge base. By contrast, from the perspective of basic Reinforcement Theory, what is the etiology of new and creative behaviors (i.e., emergents) that have no training history? Their etiology in Reinforcement Theory is unlikely to involve a history of specific reinforcement because all responses, either elementary or highly complex and novel (e.g., a highly-complex emergent), must occur at least once before reinforcement can affect its reoccurrence. Salience Theory outlines how this problem is obviated.

Important Scientific Research and Open Questions Might naturally salient stimuli not be blockable? Or might they be pre-potent or dominant as elements of compound CSs? Might they more readily enter into the formation of new amalgams than do arbitrary stimuli? What is the loading of naturally salient stimuli vs. neutral stimuli in amalgam formation? How strong must an arbitrary stimulus be to be equal in effect to naturally salient stimuli? Do the relative strengths of stimuli in sensorypreconditioning alter how they interrelate? Can symbols functionally substitute for physical stimuli in compound CSs? How does incorporation of a stimulus into one kind of amalgam impact its availability and function in other types of amalgams? Does amalgam formation predict flash-bulb memory instatement? Does amalgam formation account for perceptual discriminations as with new objects or in learning visual discrimination? How does Salience Theory align with neural activity/recordings of areas of the brain in various kinds of contexts?

Acknowledgments Preparation of this article was supported by the National Institute of Child Health and Human Development grant HD-38051.

Cross-References ▶ Abstract Concept Learning in Animals ▶ Analogical Reasoning in Animals

▶ Animal Learning and Intelligence ▶ Comparative Psychology and Ethology ▶ Intelligent Communication in Social Animals ▶ Language Acquisition and Development ▶ Learning and Numerical Skills in Animals ▶ Linguistic and Cognitive Capacities of Apes ▶ Reinforcement Learning ▶ Theory of Mind in Animals

References Hull, C. L. (1943). Principles of behavior. New York: Appleton. Marx, M. A., & Hillix, W. A. (1987). Systems and theories in psychology (p. 312 ff.). New York: McGraw-Hill. Naour, P., Wilson, E. O., & Skinner, B. F. (2009). A dialogue between sociobiology and radical behaviorism. New York: Springer. Rumbaugh, D. M., & Washburn, D. A. (2003). Intelligence of apes and other rational beings. New Haven: Yale University Press. Rumbaugh, D. M., King, J. E., Washburn, D. A., Beran, M. J., & Gould, K. L. (2007). A Salience theory of learning and behavior with perspectives on neurobiology and cognition. International Journal of Primatology, 28, 973–996. Savage-Rumbaugh, E. S., Murphy, J., Sevcik, R. A., Brakke, K. E., Williams, S., & Rumbaugh, D. M. (1993). Language comprehension in ape and child. Monographs of the Society for Research in Child Development, 58(3–4), 1–242. Serial No. 233. Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. New York: Appleton-Century. Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological Review, 55, 189–208.

A Stability Bias in Human Memory NATE KORNELL Department of Psychology, Williams College, Williamstown, MA, USA

Definition Human memory is anything but stable: We constantly add knowledge to our memories as we learn and lose access to knowledge as we forget. Yet people often make judgments and predictions about their memories that do not reflect this instability. The term stability bias refers to the human tendency to act as though one’s memory will remain stable in the future. For example, people fail to predict that they will learn from future study opportunities; they also fail to predict that they will forget in the future with the passage of time. The stability bias

A Stability Bias in Human Memory

appears to be rooted in a failure to appreciate external influences on memory, coupled with a lack of sensitivity to how the conditions present during learning will differ from the conditions present during a test.

Theoretical Background All memories are not created equal. Some memories feel strong, vivid, and familiar; others feel shakier and less reliable. People are generally confident in the first type of memory but unsure about the second. Behavior reflects this difference; for example, most people only volunteer to answer a question in class if they feel confident about their response. The term metacognition refers to the process of making judgments about one’s cognition and, frequently, about one’s memory (Dunlosky and Bjork 2008). Metacognitive processes are used to distinguish accurate memories from inaccurate ones. A memory is only valuable to the degree that we can trust it, which makes metacognition vital in our day-to-day use of memory. Moreover, virtually all memory retrievals are associated with a feeling of certainty (or lack thereof). Thus, metacognition is a critical, and omnipresent, component of human memory. Metacognitive judgments are often accurate. For example, your memory of what you ate for breakfast today is probably more accurate than your memory of what you ate for breakfast on this date 11 years ago, and it probably feels more accurate as well. It would be natural to assume that metacognitive judgments are made on the basis of the memory being judged – that is, that when confidence is low, it is because a memory is weak. The empirical evidence suggests otherwise. Instead of being made based on memories themselves, metacognitive judgments appear to be made based on inferences about those memories. For example, if an answer comes to mind quickly and easily, people tend to judge that they know that answer well. This inference is usually correct. But it is an inference all the same, and when conditions are created that reverse this relationship – when answers that come to mind quickly are less memorable – people give high judgment of learning ratings to information that comes to mind quickly, not to information that is highly memorable (Benjamin et al. 1998). If metacognitive judgments are inferential, what is the basis of the inferences? Koriat (1997) put forward a highly influential framework that has successfully

A

accounted for a great deal of subsequent data. He proposed that three categories of cues influence metacognitive judgments. Intrinsic cues were defined as information intrinsic to the information being judged (e.g., the semantic relatedness of a question an...