Mark E. Bouton - US grants

"Associative learning" is a basic learning process in which animals (including humans) learn to associate different events or concepts. Dr. Bouton studies it in animal conditioning experiments in which rats learn that one stimulus, such as a tone or a light, is associated with the occurrence of another stimulus, such as a food pellet or a mild foot-shock. This project will study how the retrieval of associations already stored in memory can influence learning and behavior in such situations. Dr. Bouton's previous research indicates that the "context," the stimuli present in the background whenever learning and remembering occur, is crucial in the retrieval of stored associations. It also suggests that retrieval of old memories by the context is especially important when an animal is learning, or has recently learned, a conflicting association. Such conflict is present when therapists treat human anxiety disorders; new learning is usually introduced to conflict with old. One implication of Dr. Bouton's research is that the older memory is not destroyed by conflict; its retrieval after therapy may cause relapse. A series of experiments with rats will examine the role of context in a variety of learning situations. An important goal is to develop and test a general new theory that will integrate research on different learning and memory phenomena in both animals and humans. Another goal is to compare and integrate the effects of "contexts" provided by the physical environment (e.g., features of the room), by the passage of time, and by the occurrence of other recent events. The work will provide new insight into the basic mechanisms of associative learning and memory, and will have implications for the control and treatment of human emotional disorders.

Memories often require the pulling together of different elements of information. "Associative learning" is the process by which humans and other animals learn to connect, or associate, different elements, events, or concepts. A fundamental form of this process is classical conditioning. Rats will learn that one stimulus, such as a tone or a light, is associated with the occurrence of another stimulus, such as a food pellet or a mild footshock; the tone or light will come to evoke behavior, such as a mild fear, provided the subject has learned to associate that stimulus with footshock and then remembers the stimulus-footshock association. This project will investigate memory and behavior when stimuli are associated with conflicting events or stimuli (e.g., shock and food, or food and no food). Previous research has shown that memory and behavior in this situation is strongly determined by the current "context," stimuli that are present in the background, which can select between competing associations. Behavior evoked by stimuli associated with multiple events can also change as the memories age over time. The project will investigate how time and context influence memory and performance, and also how the effects of time and context might interact. By studying the effects of time and context on behavior over a range of related situations, the research will provide insight into how ambiguous information is stored, retrieved, and then expressed in behavior and emotional performance. The research will have its primary impact on theories of learning, memory, and acquired motivation. It will also be relevant to research on the neurobiology of learning and memory, which has recently begun to identify biological processes in cells of the brain that may correspond to the learning and memory processes investigated here. The results will also have implications for cognitive-behavioral therapies of human behavior disorders (e.g., drug dependence and anxiety), which often introduce new learning designed to conflict with cognition and behavior based on previous learning.

Nontechnical Abstract Context Dependence of Learning and Memory Processes Memory is usually best when the act of remembering takes place in a presence of the background stimuli ( contextual stimuli ) that were also present when learning occurred. However, this is not always true; memory for some types of material is not very context- dependent. This project seeks to understand why some memories are context-dependent and others are context-independent. It will study the role of contextual stimuli in classical conditioning, a fundamental example of associative learning that is connected with many aspects of cognition and behavior in animals and humans. Rats will learn that one stimulus, such as a tone or a light or a distinctive flavor, signals the occurrence of another stimulus, such as a food pellet or a mild footshock or malaise. The signal comes evoke behaviors that indicate that learning (and also remembering) have taken place. The project will test the effects of changing the context on various forms of this kind of learning. Some, but not all, are context-dependent. For example, if a signal is associated with different stimuli (e.g., shock and food) at different times, the second association, but not the first, is typically context-dependent. The project will test several ideas of why this should be so. It will also study several different kinds of cues that can provide a context. This research will have major impact on theories of learning, memory, and acquired motivation. It will also be relevant to research on the neurobiology of learning memory. And the results may help us design better cognitive- behavioral therapies of human behavior disorders (e.g., drug dependence and anxiety), which may be successful or not depending on whether the learning they involve is context- independent.

DESCRIPTION (provided by applicant): The general goal of this project is to understand how time influences Pavlovian learning, the fundamental learning process in which humans and animals learn to associate signals with biologically significant events. The study of conditioning is important in part because it is deeply involved in the development (and treatment) of many mental health problems, including (but not limited to) drug dependence and anxiety disorders. Many of the proposed experiments, which will be conducted with laboratory rats, will distinguish between two different ideas about how Pavlovian learning works: A view that emphasizes the importance of learning "trials" (occasions on which the signal occurs either with or without the significant event) and another view that ignores trials and emphasizes the organism's perception of time between presentations of the significant event. These ideas and others will be contrasted in number of experiments. Many experiments will investigate the effects of trial spacing in acquisition, whereas others will investigate the effects of trial spacing during learning treatments (extinction and counter conditioning) designed to replace old learning with new learning. All will uncover conditions that "optimize" learning by making it strong and durable. Other experiments will compare the effects of the passage of time and contextual change, and others will investigate the psychological processes involved in discriminating different temporal intervals. The results will increase our understanding of the influence of time and memory on conditioning, and will help reveal the interrelationships between timing and associative learning. Many of the results will also increase our understanding of the causes of relapse after therapy and suggest ways to help prevent or minimize that relapse.

DESCRIPTION (provided by applicant): This project is concerned with understanding extinction, the loss of learned performance that occurs when a Pavlovian signal or an instrumental action is repeatedly presented without its reinforcer. Extinction is a naturally-occurring process of behavior change, as well as a tool used in clinical treatments designed to eliminate unwanted thoughts, emotions, and behaviors in humans. Although it is tempting to assume that extinction erases the original learning, extinguished performance readily recovers, and several recovery effects (e.g., renewal, reinstatement, rapid reacquisition, spontaneous recovery, and resurgence) indicate that the original learning may be largely intact. In addition, because these effects can be interpreted as effects of changing the background or context, they suggest that extinction results from new inhibitory learning that is especially sensitive to the context in which it is learned. The goal of this project is to seek an integrated understanding of extinction as it is revealed in these and other response-recovery processes. It will focus especially on the extinction of instrumental (operant) learning, because principles of operant learning are crucial for understanding a range of behavior problems-such as smoking, drug abuse, and eating and overeating-in which voluntary contact with reinforces plays an essential role. The experiments will involve rats as subjects. One set will test new methods for reducing the renewal effect (in which extinguished behavior relapses when the context is changed after extinction) and resurgence (in which a behavior has been extinguished and replaced by a second behavior relapses when the replacement behavior is itself extinguished). Another set will examine ways to inhibit rapid reacquisition (in which an extinguished action rapidly spirals into relapse when action-reinforcer pairings are reintroduced) and test the effects of hunger as a contextual stimulus influencing relapse. A third set will analyze the extinction-enhancing effects of administering D-cycloserine (a partial agonist of a brain receptor that is thought to play a role in learning), as well as new hypotheses about how to enhance the generalization of extinction to new contexts. A fourth set will analyze the extinction of sequences or chains of behavior in which the subject must purchase (or procure) access to the reinforcer before she can consume (or take) it. The results will increase our understanding of extinction, a fundamental behavioral and clinical phenomenon, and develop new ways to help promote extinction learning so as to minimize lapse and relapse.