Learning and Adaptation: The Role of Experience PDF

CHAPTER 7 CHAPTER OUTLINE Learning and Adaptation: The Role of Experience Focus on Neuroscience: The Neuroscience of Fear Conditioning ADAPTING TO THE ENVIRONMENT How Do We Learn? The Search for Mechanisms Habituation and Sensitization Applications of Operant Conditioning CLASSICAL CONDITIONING...

CHAPTER 7 CHAPTER OUTLINE Learning and Adaptation: The Role of Experience Focus on Neuroscience: The Neuroscience of Fear Conditioning ADAPTING TO THE ENVIRONMENT How Do We Learn? The Search for Mechanisms Habituation and Sensitization Applications of Operant Conditioning CLASSICAL CONDITIONING: ASSOCIATING ONE STIMULUS WITH ANOTHER Pavlov’s Pioneering Research Basic Principles Applications of Classical Conditioning Applications: Learning, Virtual Reality, and Therapy BIOLOGY AND LEARNING Constraints on Classical Conditioning: Learned Taste Aversions Are We Biologically Prepared to Fear Certain Things? Constraints on Operant Conditioning: Animals That “Won’t Shape Up” Learning and the Brain COGNITION AND LEARNING OPERANT CONDITIONING: LEARNING THROUGH CONSEQUENCES Thorndike’s Law of Effect Skinner’s Analysis of Operant Conditioning Antecedent Conditions: Identifying When to Respond Consequences: Determining How to Respond Shaping and Chaining: Taking One Step at a Time Generalization and Discrimination Schedules of Reinforcement Escape and Avoidance Conditioning Insight and Cognitive Maps Cognition in Classical Conditioning Cognition in Operant Conditioning Frontiers: Animal Cognition OBSERVATIONAL LEARNING: WHEN OTHERS PAVE THE WAY Bandura’s Social-Cognitive Theory Research Foundations: Using Social-Cognitive Learning Theory to Prevent AIDS: A National Experiment A man who carries a cat by the tail learns something he can learn in no other way. —Mark Twain What are the issues here? What do we need to know? Where can we find the information to answer these questions? About one in five air passengers experience some degree of fear when they step aboard an airplane. Their heart and breathing rates increase. Their palms become sweaty, and their arousal levels are high. For 3 percent of air travellers, the arousal is so high that they are in a state of panic, even when sitting in the airport parking lot. These individuals have aviophobia. The fear and panic stems from a variety of sources: media reports, unusual sounds aboard the aircraft, turbulence, and a general loss of control. Some take to driving all the way across Canada to avoid air travel completely. In an effort to overcome this problem, people have turned to drugs, hypnosis, and audio self-help recordings. The self-help recordings have limited success, hypnosis works for some, and drugs, while reducing anxiety, do not really cure the problem. But recently, a new program run by Marc-Antoine Plourde of Montreal is showing a lot of promise. Plourde, a captain with Air Canada, runs the DePlour Training Centre where individuals can sign up for a two-day course to reduce their fear. The course involves seminars on pilot training, aircraft design, and maintenance; a ride in a flight simulator; plus a graduation “liberty flight” from Montreal to Toronto. Together with a licensed therapist, Plourde explains the science of flight and the psychology of fear. During a demonstration of aircraft design, Plourde uses a model of an Airbus A330 to show how the wings are engineered to bend. However, he bends them a little too much (on purpose), breaking off the wing much to the astonishment (and shock) of the participants. Their worst fear realized, Plourde explains how this could not happen in the real world. Does the program work? Plourde claims a 94 percent satisfaction rate and notes that only four of 500 people have failed to take the liberty flight. Apparently, dealing with the emotional anxiety is more important than stressing airplane safety. As Captain Tom Bunn who runs a similar program in the United States notes, “People tell us that they know that flying is safer than driving, but my car doesn’t fall 30 000 feet [9100 metres] from the sky.” R eflect for a moment on how much of your behaviour is learned: telling time, getting dressed, driving, reading, using money, playing sports, and so on. Beyond such skills, learning affects our emotional reactions, perceptions, and physiological responses. Through experience, we learn to think, act, and feel in ways that contribute richly to our individual identity. Learning is a process by which experience produces a relatively enduring change in an organism’s behaviour or capabilities. The term capabilities highlights a distinction made by many theorists: “knowing how,” or learning, versus “doing,” or performance. For example, experience may provide us with immediate knowledge (e.g., you may receive instructions on how to perform a skill), but in science we must measure learning by actual changes in performance. In this chapter, we explore basic learning processes. The first, habituation and sensitization, involve a change in behaviour that results from repeated exposure to a single stimulus. Next, we explore two forms of conditioning that involve learning associations between events, and hence have often been referred to as associative learning. Classical conditioning occurs when two stimuli become associated with each other. For example, seeing a dog and being bitten become associated such that one stimulus (seeing a dog) now triggers a new response (fear). In operant conditioning, we learn to associate our responses with specific consequences. For example, we learn that smiling at others is followed by a friendly greeting. The study of associative learning has been central to the study of learning in psychology. Indeed throughout much of the history of psychology, “learning” was used to mean associative learning. Our examination of learning then considers observational learning, in which we learn by watching others behave. Finally, we consider the role of cognition in conditioning. 1. Historically, how have behaviourists defined learning? ADAPTING TO THE ENVIRONMENT From the moment we are born, we encounter changing environments, each with its unique challenges. Some challenges affect survival, such as acquiring food and shelter. Others do not, such as deciding where to go on a date. But no matter the challenge, learning makes it possible for us to adapt to it. In fact, we can view learning as a process of personal adaptation to the ever-changing circumstances of our lives. How Do We Learn? The Search for Mechanisms Many of the key principles that we will discuss reflect important discoveries by the behaviourists (Bolles & Beecher, 1988). Within psychology, behaviourists focused on how organisms learn, examining the processes by which experience 2. What role does the environment play in personal and species adaptation? 228 CHAPTER SEVEN influences behaviour. Behaviourists assumed that there are laws of learning that apply to virtually all organisms. For example, each species they studied— whether birds, reptiles, rats, monkeys, or humans— responded in predictable ways to patterns of reward or punishment. Behaviourists treated the organism as a tabula rasa, or blank tablet, upon which learning experiences were inscribed. Most of their research was conducted with nonhuman species in controlled laboratory settings. Behaviourists explained learning solely in terms of directly observable events and avoided speculating about an organism’s unobservable “mental state.” The concept of learning calls attention to the importance of adapting to the environment. Whereas evolution focuses on species’ adaption across many generations, learning represents a process of personal adaptation. The resurgence of the cognitive perspective, an interest in biological factors, and the emergence of cross-cultural psychology also have expanded our understanding of learning. Cognitive and biological factors play important roles in learning (Dickinson, 1997; Shanks, 2010). As we have seen and will continue to explore in upcoming chapters, crosscultural research highlights the important impact that culture has on what we learn—from social customs (norms) and beliefs, to our most basic perceptions of the world and ourselves (Figure 7.1; Super & Harkness, 1997). Culture’s impact is not surprising, given that learning represents adaptation to the environment and culture is the human-made part of our environment (Herskovits, 1948). And yet, the learning mechanisms that foster this adaptation are universal among humans and, in some cases, occur across countless species. Habituation and Sensitization 3. What is habituation, and what is its adaptive significance? Imagine that you are sitting in a room studying. You notice that the clock makes an audible sound as the second-hand moves to each new notch. Over time, as the tick occurs again and again, you notice it less and less, and eventually you do not register it at all. Now imagine that you hear a rustling sound outside your window at night. Although you think it must just be the wind in the trees, the sound startles you. You move to the window and experience an increase in arousal. The sound repeats and generates a stronger response in you, as you become increasingly fearful. These two examples illustrate the simplest forms of learning and demonstrate a change in behaviour because of the repeated presentation of a single stimulus. Habituation is a decrease in the strength of response to a repeated stimulus. It may be the simplest form of learning and occurs across species, FIGURE 7.1 People in different cultures learn different behaviours to adapt to their environment. Even the same general skill will take different forms depending on unique environmental features and demands. ranging from humans to dragonflies and sea snails (Glanzman, 2009). University of British Columbia psychologist Catherine Rankin has even demonstrated habituation in a microscopic worm (a nematode) that has only 302 neurons (Lau et al., 2013). Habituation serves a key adaptive function. You do not need to constantly respond to the pressure of clothing on your skin, to the sound of the ventilation system, or to the hum of distant traffic. If an organism responded to every stimulus in its environment, it would rapidly become overwhelmed and exhausted. By learning not to respond to uneventful familiar stimuli, organisms conserve energy and can attend to other stimuli that are important. Habituation also plays an important role in enabling scientists to study behaviour. Whether observing animals in the wild or schoolchildren, a researcher’s mere presence may initially disrupt participants’ natural responses. Thus, before collecting data, observers often allow people and animals to habituate to their presence. Learning and Adaptation: The Role of Experience Habituation is different from sensory adaptation, which we discussed in Chapter 5. Sensory adaptation refers to a decreased sensory response to a continuously present stimulus. Habituation, on the other hand, is a simple form of learning that occurs within the central nervous system. You may habituate to a stimulus, but that sensory information is still available if it becomes relevant. For example, you habituate to the feeling of your clothing against your skin. That tactile information has been presented continuously with no important consequences, so you no longer notice it. If, however, there is reason to become aware of skin sensations, perhaps because of a wasp or a mosquito in your vicinity, you suddenly become keenly aware of all the light touches on your skin that a few seconds ago had shown habituation. Sensitization is an increase in the strength of response to a repeated stimulus. For example, if a loud tone is sounded, an organism will show the startle reflex: They will orient to the sound; their muscle tension increases rapidly; and they jump and may vocalize. With repeated presentation of a loud tone, the startle response increases in intensity (Donahoe & Palmer, 1994). Have you ever touched a metal object, such as a door handle, and received a static electric shock? If you then touch another metal object and receive a second shock, you will jump a little more, pull your hand back a little more quickly, and show a slightly stronger emotional reaction (the words you call out may also change). Each shock elicits a stronger response; that is, you have shown sensitization. Like habituation, sensitization is found across a wide range of species, even among animals with very simple nervous systems (Cai et al., 2011). Sensitization tends to occur to strong or noxious stimuli (Donahoe & Palmer, 1994), and its purpose is to increase responses to a potentially dangerous stimulus. CLASSICAL CONDITIONING: ASSOCIATING ONE STIMULUS WITH ANOTHER Life is full of interesting associations. Do you ever hear songs on the radio or find yourself in places that instantly make you feel good because they’re connected to special times you’ve had? When you smell the aroma of popcorn or freshly baked cookies, does it make your mouth water or stomach growl? These examples illustrate a learning process called classical conditioning, in which an organism learns to associate two stimuli (e.g., a song and a pleasant event), such that one stimulus (the song) comes to produce a response (feeling happy) that originally was produced only by the other stimulus (the pleasurable event). Like habituation and sensitization, classical conditioning is a basic form of learning that occurs in mammals, birds, reptiles, fish, sea snails, and even insects (Kandel & Hawkins, 1992; Watanabe et al., 2008). Classical conditioning involves learning an association between stimuli. Its discovery dates back to the late 1800s and an odd twist of fate. Pavlov’s Pioneering Research In the 1860s, Ivan Pavlov was studying theology in a Russian seminary and preparing for the priesthood when his career plans unexpectedly changed. A new government policy allowed the translation of Western scientific publications into Russian. Before long, Pavlov read Darwin’s theory of evolution and In Review • Learning is a process by which experience produces a relatively enduring change in an organism’s behaviour or capabilities. Learning is measured by changes in performance. • Learning involves adapting to the environment. Historically, behaviourists focused on the processes by which organisms learn, and ethologists focused on the adaptive significance of learning. Today, these two perspectives have crossed paths, and more attention is paid also to how mental processes and cultural environments influence learning. 229 • Habituation is a decrease in the strength of a response to a repeated stimulus. It may be the simplest form of learning. • Habituation allows organisms to attend to other stimuli that are more important. • Sensitization is an increase in the strength of a response to a repeated stimulus. • Sensitization increases an organism’s response to potentially dangerous stimuli. 4. What is sensitization, and why would you want to sensitize to the repeated presentation of a stimulus? 230 CHAPTER SEVEN (b) (a) FIGURE 7.2 (a) Ivan Pavlov (the man with the white beard) is shown here with colleagues and one of his canine subjects. (b) In his early research, Pavlov measured salivation by using a simple device similar to the one shown here. In later research, a collection tube was inserted directly into the salivary gland. other works, which sparked in him a strong interest in the sciences (Windholz, 1997). Pavlov became a renowned physiologist, conducting research on digestion in dogs that won him the Nobel Prize in 1904. To study digestion, Pavlov presented various types of food to dogs and measured their natural salivary response (Figure 7.2). But as often occurs in science, Pavlov was about to make an accidental but important discovery through astute observation. He noticed that with repeated testing, the dogs began to salivate before the food was presented, such as when they heard the footsteps of the approaching experimenter. Further study confirmed Pavlov’s observation. Dogs have a natural reflex to salivate to food but not to tones. Yet when a tone or other stimulus that ordinarily did not cause salivation was presented just before food powder was squirted directly into a dog’s mouth, the sound of the tone alone soon made the dog salivate. Pavlov’s research team rigorously studied this process for decades, and this type of learning by association came to be called classical or Pavlovian conditioning (Pavlov, 1928). Many psychologists regard Pavlov’s discovery as “among the most important in the history of psychology” (Dewsbury, 1997). But why all the fuss about dogs salivating to tones? This question raises a major point about basic scientific research. As noted in Chapter 2, what is paramount is the underlying principle being demonstrated, not the specific findings. Classical conditioning performs a key adaptive function; classical conditioning alerts organisms to stimuli that signal the impending arrival of an important event. As Pavlov noted, if salivation could be conditioned, so might other bodily processes, including those affecting susceptibility to disease and mental disorders. Basic Principles What factors influence the acquisition and persistence of conditioned responses? Let us examine some basic principles of conditioning. Acquisition Acquisition refers to the period during which a response is being learned. Suppose we wish to condition a dog to salivate to a tone. Sounding the tone initially may cause the dog to perk up its ears and stare at us oddly, but not to salivate. At this time, the tone is a neutral stimulus because it does not elicit (i.e., trigger) the salivation response (Figure 7.3). Now, if we place food in the dog’s mouth, the dog will salivate. This salivation response to food is reflexive—it’s what dogs do by nature. Because no learning is required for the food to produce salivation, the food is called an unconditioned stimulus (UCS) and salivation is an unconditioned response (UCR). Next the tone and the food are paired—each pairing is called a learning trial. After several learning trials, when the tone is presented by itself, the dog salivates even though there is no food. Through association, the tone has become a conditioned stimulus (CS) and salivation has become a conditioned response (CR). Table 7.1 offers a quick reference to these classical conditioning terms. Notice that we have two terms for salivation: UCR and CR. When the dog salivates to food, this UCR is a natural, unlearned (unconditioned) reflex. But when it salivates to a tone, this CR represents a learned (conditioned) response. Learning and Adaptation: The Role of Experience 231 Before conditioning During conditioning Tone No salivation response Unconditioned Stimulus (UCS) (food powder) Unconditioned Response (UCR) (salivation) Conditioned Stimulus (CS) (tone) + Unconditioned Response (UCR) (salivation) Conditioned Stimulus (CS) (tone) Conditioned Response (CR) (salivation) Unconditioned Stimulus (UCS) (food powder) After conditioning FIGURE 7.3 In classical conditioning, after a neutral stimulus such as a tone is repeatedly associated with food (unconditioned stimulus), the tone becomes capable of eliciting a salivation response. During acquisition, a CS typically must be paired multiple times with a UCS to establish a strong CR (Figure 7.4). Pavlov also found that a tone became a CS more rapidly when it was followed by greater amounts of food. Indeed, when the UCS is intense and aversive conditioning may require only one CSUCS pairing (Richard et al., 2000). Someone who is in a car accident may develop a fear of cars or driving as a result of a single accident (Taylor et al., 2002). In the example of a fear of driving because of an accident, the stimulus (riding in a car) becomes a CS after only one pairing with an intense UCS (an emotionally and physically painful crash). Fear was TABLE 7.1 the UCR, and it can become the CR triggered by the sight of cars or the thought of driving in a car (Taylor et al., 2002). The sequence and time interval of the CS-UCS pairing also affect conditioning. Learning usually occurs most quickly with forward short-delay pairing: The CS (tone) appears first and is still present when the UCS (food) appears. In forward trace pairing, the tone would come on and off, and afterward the food would be presented. In forward pairing, it is often optimal for the CS to appear no more than two or three seconds before the UCS (Klein & Mowrer, 1989). Forward pairing has adaptive A Quick Guide to Classical Conditioning Term Abbreviation Description Example Unconditioned Stimulus UCS Food Conditioned Stimulus CS Unconditioned Response UCR Conditioned Response CR A stimulus that innately elicits a response A stimulus that gains value through learning A reflexive, unlearned response to an innately important stimulus A response elicited by a stimulus whose importance depends on past learning The sight of your favourite restaurant Salivation in response to food Feeling hungry when you see your favourite restaurant 5. How do you create a conditioned salivation response in a dog? 6. Under what circumstances are CRs typically acquired most quickly? 232 CHAPTER SEVEN 24-hour rest 24-hour rest 15 Drops of saliva elicited by CS Acquisition (CS-UCS pairings) Extinction (CS alone) First spontaneous recovery (CS alone) 10 Second spontaneous recovery (CS alone) 5 0 2 4 6 8 10 12 14 16 18 20 22 Trials 2 4 6 8 2 4 6 8 FIGURE 7.4 The strength of the CR (salivation) increases during the acquisition phase as the CS (tone) and the UCS (food) are paired on each trial. During the extinction phase, only the CS is presented, and the strength of the CR decreases and finally disappears. After a rest period following extinction, presentation of the CS elicits a weaker CR (spontaneous recovery) that extinguishes more quickly than before. value because the CS signals the impending arrival of the UCS. Typically, presenting the CS and UCS at the same time (simultaneous pairing) produces less rapid conditioning, and learning is slowest, or does not occur at all, when the CS is presented after the UCS (backward pairing). To summarize, classical conditioning usually is strongest when there are repeated CS-UCS pairings, the UCS is more intense, the sequence involves forward pairing, and the time interval between the CS and UCS is short. Extinction and Spontaneous Recovery 7. Explain the key factor in producing extinction of a CR. 8. Explain the adaptive significance of stimulus generalization and discrimination. If the function of classical conditioning is to help organisms adapt to their environment, then there must be a way of eliminating the CR when it is no longer appropriate. Fortunately, there is. If the CS is presented repeatedly in the absence of the UCS, the CR weakens and eventually disappears. This process is called extinction, and each presentation of the CS without the UCS is called an extinction trial. When Pavlov repeatedly presented the tone without the food, the dogs eventually stopped salivating to the tone (Figure 7.4). Occasional re-pairings of the CS (e.g., tone) and the UCS (e.g., food) usually are required to maintain a CR. Even when a CR extinguishes, this does not mean that all traces of it are erased. If someone has been conditioned to respond to a specific location with fear, perhaps because that location was the scene of an accident, repeated exposure to that location (CS exposure) with no aversive consequences (no UCS), will lead to an extinction of the fear CR. However, if that person encounters that location again after a break, he or she will show a fear response. The extinguished CR, although weakened, has reappeared. This reappearance is called spontaneous recovery, which is defined as the reappearance of a previously extinguished CR after a rest period and without new learning trials. As Figure 7.4 shows, the spontaneously recovered CR usually is weaker than the initial CR and extinguishes more rapidly in the absence of the UCS. The phenomenon of spontaneous recovery is why practical applications of extinction, such as treatment of phobias or other anxiety disorders, require multiple sessions. The abnormal CR, such as fear, may appear to have undergone extinction, but it will return in the future. With each set of extinction trials, the CR is progressively weakened, and with sufficient extinction training, spontaneous recovery is weak enough that it is not a problem. Generalization and Discrimination Pavlov found that once a CR is acquired, the organism often responds not only to the original CS, but also to stimuli that are similar to it. The greater the stimulus similarity, the greater the chance that a CR will occur. A dog that salivates to a medium-pitched tone is more likely to salivate to a new tone slightly different in pitch, than to a low- or high-pitched tone. Learning theorists call this stimulus generalization: Stimuli similar to the initial CS elicit a CR (Figure 7.5). Learning and Adaptation: The Role of Experience Response (drops of saliva) 15 Original tone (CS) 10 5 0 400 800 1200 1600 Stimulus tone (hertz) 2000 FIGURE 7.5 A stimulus generalization curve. An animal will salivate most strongly to the CS that was originally paired with the UCS. Progressively weaker conditioned responses occur as stimuli become less similar to the CS, as seen here with tones of lower or higher frequencies (pitch). just prior to sounding the tone but do not present any food. After repeated square-tone pairings, the square will become a CS and elicit salivation by itself (Figure 7.6). This process, discovered by Pavlov, is called higher-order conditioning: A neutral stimulus becomes a CS after being paired with an already established CS. Typically, a higher-order CS produces a CR that is weaker and extinguishes more rapidly than the original CR. The dog will salivate less to the black square than to the tone, and its response to the square will extinguish sooner. Higher-order conditioning greatly expands the influence of conditioned stimuli and can affect what we come to value, like, fear, or dislike (Hussaini et al., 2007). For example, a child may value a gold star because that gold star was previously paired by social recognition and praise from the teacher. Applications of Classical Conditioning Stimulus generalization serves critical adaptive functions. An animal that ignores the sound of rustling bushes and then is attacked by a hidden predator will become alarmed by the sound of a rustling bush in the future (assuming it escapes). If stimulus generalization did not occur, then the next time the animal heard rustling it would become alarmed only if the sound were identical to that preceding the earlier attack. This absence of alarm does not contribute to the animal’s survival. Through stimulus generalization, the animal develops an alarm response to a range of rustling sounds. Some will be false alarms, but safe is better than sorry. To prevent stimulus generalization from running amok, organisms must be able to discriminate (i.e., detect) differences between stimuli. An animal that becomes alarmed at every sound would exhaust itself from stress. It must learn to distinguish irrelevant sounds from those that may signal danger. In classical conditioning, discrimination is demonstrated when a CR (such as an alarm reaction) occurs to one stimulus (a sound) but not to others. Organisms can be taught, through conditioning, to behaviourally discriminate two stimuli that were initially treated the same way. Pairing the CS with the UCS combined with pairing similar stimuli with no consequence leads to a narrowing of response to the specific CS and a loss of generalized responses to other similar stimuli. Higher-Order Conditioning Imagine that we have exposed a dog to repeated tone-food pairings, and the tone is now a wellestablished CS that elicits a strong salivation response. Next, suppose that we present a neutral stimulus, such as a black square, and the dog does not salivate. Now, we present the black square Pavlov’s belief that salivation was merely the tip of the classical conditioning iceberg has proven correct. Conditioning principles help us understand many diverse human behaviours and problems. Acquiring and Overcoming Fear Pavlov’s discoveries enabled early American behaviourists to challenge Freud’s psychoanalytic view of the causes of anxiety disorders, such as phobias. To Before higher-order conditioning Black square No salivation (neutral stimulus) During higher-order conditioning Black square + Salivation (CR) (CS1) After higher-order conditioning Black square Salivation (CR) (CS2) FIGURE 7.6 Once a tone has become a conditioned stimulus that triggers salivation, we can now use it to condition a salivation response to a new neutral stimulus—a black square. The tone is the CS1. The black square becomes the CS2. 233 9. Explain the process of higher-order conditioning. 234 CHAPTER SEVEN 10. How does classical conditioning explain fear acquisition? 11. How is classical conditioning used in society to increase or decrease our arousal/attraction to stimuli? explain a snake phobia, no Freudian assumptions about hidden unconscious conflicts or repressed traumas are needed. Instead, the behaviourist view is that snakes have become a fear-triggering CS because of pairing with an aversive UCS (such as injury) and stimulus generalization. Does this explanation seem reasonable? It may, but it suffers from a serious limitation: Almost any explanation can seem plausible when it is provided after an event occurs. Behaviourist John B. Watson and his assistant Rosalie Rayner (1920) set out to obtain evidence that fear could be conditioned. They studied a number of infants, including, most famously, an 11-month-old infant named Albert, often referred to as Little Albert. One day, as Albert played in a hospital room, Watson and Rayner showed him a white rat. Albert displayed no sign of fear. Later, knowing that Albert was afraid of loud noises, they hit a steel bar with a hammer, making a loud noise as they showed Albert the rat. The noise scared Albert and made him cry. After several rat– noise pairings, the sight of the white rat alone made Albert cry. To examine stimulus discrimination and generalization, Watson and Rayner exposed Albert to other test stimuli. Albert displayed no fear when shown coloured blocks, but furry white or grey objects, such as a rabbit and a bearded Santa Claus mask, made him cry (Figure 7.7). By the time Albert left the hospital, he had not been exposed to any treatment designed to extinguish his fear. Unfortunately, we do not know what became of Albert. Thinking critically FIGURE 7.7 John Watson and Rosalie Rayner examine how Little Albert reacts to a Santa Claus mask. strong fear of rabbits. Jones, who acknowledged Watson and Rayner’s work, gradually extinguished Peter’s fear by using the procedure shown in Table 7.2. Her approach foreshadowed current behaviour therapies, discussed in Chapter 17. They are called exposure therapies because their basic goal is to expose the phobic patient to the feared stimulus (CS) without any UCS, allowing extinction to occur. Although psychologists still debate whether all phobias are learned (Coelho & Purkis, 2009), exposure therapy is effective in most cases. Mental imagery, real-life situations, or both can be used to present the phobic stimulus. Exposure therapies are highly effective and represent one of behaviourism’s important applied legacies (Hamm, 2009). Recently, clinical psychologists have used virtual reality (VR) as part of exposure therapy to successfully treat spider phobias, fear of flying, claustrophobia, fear of driving, and fear of heights (see the Applications feature; Krijn et al., 2004a; Krijn et al., 2004b; Rothbaum et al., 2006). WAS THE LITTLE ALBERT STUDY ETHICAL? Review boards that oversee research ethics did not exist in the 1920s. Would you have approved Watson and Rayner’s request to conduct the Little Albert study? Why or why not? Think about it, and then see the Answers section at the end of the book. TABLE 7.2 This table lists 10 of the 17 steps used by Mary Cover Jones to eliminate Peter’s fear of rabbits. Step No. Two other sources of evidence suggest that at least some fears are conditioned. First, laboratory experiments show that humans and other mammals become afraid of neutral stimuli that are paired with electric shock (Merz et al., 2010). Second, behavioural treatments partly based on classical conditioning principles are among the most effective psychotherapies for phobias (Vervliet & Elen, 2006). The key assumption is that if phobias are learned, they can be “unlearned.” In 1924, psychologist Mary Cover Jones successfully treated a boy named Peter, who had a Using Exposure Training to Reduce Fear 1 2 4 5 6 8 10 12 16 17 Peter’s Progress Rabbit anywhere in room triggers fear Rabbit 4 metres away tolerated Rabbit 1 metre away tolerated Rabbit close in cage tolerated Rabbit free in room tolerated Rabbit touched when free in room Rabbit allowed on tray of high chair Holds rabbit on lap Fondles rabbit affectionately Lets rabbit nibble his fingers Source: Adapted from Jones (1924). Learning and Adaptation: The Role of Experience 235 Applications LEARNING, VIRTUAL REALITY, AND THERAPY The most widely accepted theory for the acquisition of anxiety disorders, such as phobias, is that these disorders are acquired through classical conditioning. Exposure to an environmental stimulus (CS) is paired with an aversive event (UCS), and as a result, the originally neutral stimulus comes to elicit an emotional reaction of anxiety or fear (CR). If we acquire anxiety disorders through conditioning, then conditioning procedures should be effective at treating these disorders. The most commonly used and most effective therapies for anxiety disorders, such as specific phobias, are based on a classical conditioning model. These therapeutic approaches have been classified as exposure treatments because they all involve exposure to the phobic stimulus without aversive consequences. From studies of classical conditioning, we know that if a CS is presented repeatedly without any biologically important following event, the learned response will gradually diminish in strength. As discussed earlier, this is the process of extinction; with extinction training, the CS loses its value and the learned response (CR) is progressively weakened. The traditional exposure therapy approaches have involved presenting the client with either the real, phobic stimulus, or exposure to a series of stimuli that gradually get closer to, and more like, the phobic stimulus. Such procedures, especially when combined with relaxation training, are very effective at treating anxiety disorders such as snake and spider phobias, fear of flying, and public-speaking anxiety. Exposure therapy with gradual introduction of the phobic stimulus is the treatment of choice for specific phobias (Antony & Swinson, 2000; Garcia-Palacios et al., 2002; Marks, 1987). In a variant of exposure therapy, the client imagines exposure to the feared stimulus rather than confronting the real thing. Clinical research has found that although imaginal exposure can be successful, real-world exposure (referred to as in vivo exposure) is superior (Krijn et al., 2004a). Although exposure therapy is very successful, a surprising number of individuals with phobias do not seek treatment. Some estimates relay that as many as 40 percent of individuals with a specific phobia never seek treatment and the phobia continues to needlessly generate anxiety and disrupt their lives (Garcia-Palacios et al., 2002). Individuals with a phobia often avoid treatment because they fear having to confront the phobic stimulus, even though they know that the therapy will help them overcome that fear. Recent advances in computer and video technology have presented an innovative approach to treating anxiety disorders: the use of virtual reality (VR). VR uses real-time computer graphics and high-resolution three-dimensional visual displays, body tracking, sound, and, in some cases, other types of sensory input (e.g., tactile stimulation) to immerse clients in a computer-generated world. Via a computer, the therapist guides what happens within the client’s virtual world (see Figure 7.8). If you can overcome your fear of snakes, public speaking, or heights with exposure therapy, can that exposure take place in a virtual world? If VR is effective, then the application of exposure therapy would be more practical in some cases (e.g., fear of flying, fear of heights). This therapy may also be more appealing for individuals who have avoided or abandoned therapy because exposure to the real stimulus generates such intense fear or is impractical. Apart from presenting virtual stimuli within a virtual world, exposure therapy using VR progresses like any other graded exposure therapy. The initial stimulus or situation is only remotely like the fear-inducing stimulus, and as the client is able to confront the situation without anxiety, the stimulus is gradually moved closer and closer to the phobic situation. One study of the effectiveness of VR to treat phobias involved clients with a spider phobia (Garcia-Palacios et al., 2002). FIGURE 7.8 In VR exposure therapy, the client wears a VR display helmet (and, in some cases, other means of delivering sensory stimulation) and body position sensors, and is connected to a computer that generates the virtual environment. A therapist controls the virtual environment that the client explores. VR exposure therapy has been found effective for treating phobias, such as spider phobias, and related anxiety disorders. continued 236 CHAPTER SEVEN Clients in this study had to meet a series of criteria, including the full diagnostic criteria for a specific phobia established by the American Psychiatric Association (DSM-IV, the American Psychiatric Association, 1994; see Chapter 17). During treatment, clients donned a VR helmet and visited a virtual kitchen. Gradually, over a series of trials, clients received increasing exposure to a virtual spider. For example, they initially saw a virtual spider at a distance, later they came within arm’s reach of a virtual spider, and eventually they were to touch the virtual spider. The goal of the VR exposure was to have the client hold a furry virtual tarantula within their cyber-kitchen and report low levels of anxiety. In a clever and creative twist, tactile feedback was provided by having the client’s real hand explore a model spider while their virtual hand explored the cyber-spider. Across the course of this study, members of a wait-list control group showed no change in their spider phobias. Those who had experienced virtual exposure showed significant and clinically meaningful improvement. Based on behaviour avoidance tests, measures of anxiety, a fear of spiders questionnaire, and a clinician’s rating of the phobia, VR exposure was effective in treating the spider phobia. A similar approach has been used with other phobias. For example, Rothbaum and colleagues (Rothbaum et al., 2006) compared VR, standard in vivo exposure therapy, and a waitlist control among a group of people with a fear of flying. Participants in this study met the diagnostic criteria (DSM-IV) for a specific phobia, flying phobia, or agoraphobia in which flying was the feared situation. Both the VR and standard exposure groups received anxiety management training and then a series of exposures to aircraft and flying. For the VR condition, clients sat on a chair with a real airline seatbelt, and had the virtual experience of sitting in an airplane, flight attendant announcements, takeoffs, landings, and flying in both calm and stormy conditions delivered through a VR helmet. Standard exposure therapy clients completed a typical series of exposures to real airports and airplanes. They received realworld exposure to preflight activities (e.g., they visited ticket counters and airport waiting rooms), watched airplanes, and sat on a stationary plane while imagining flying. Assessments done after the completion of therapy and at 6- and 12-month follow-ups found that the VR exposure and standard exposure therapy had a similar effect on attitudes toward flying, willingness to fly, anxiety ratings during a real flight, self-ratings of improvement, and patient satisfaction (Rothbaum et al., 2006). Most of the studies using VR have been conducted with specific phobias, and have included spider phobias, fear of flying, claustrophobia, fear of driving, and fear of heights (Krijn et al., 2004a; Krijn et al., 2004b; Rothbaum et al., 2006). There have also been clinical studies of the effectiveness of VR exposure therapy with other anxiety disorders such as posttraumatic stress disorder, fear of public speaking/social phobia, and agoraphobia (Krijn et al., 2004b). For these anxiety disorders, exposure therapy using VR has either been shown to be an effective treatment or has revealed promising preliminary results, which suggest that further study will indeed demonstrate that VR exposure is clinically useful (Krijn et al., 2004b). Theories of phobias and related anxiety disorders that emerged from learning theory led to learning-based treatments that use graded exposure to the anxiety-provoking stimulus or situation. These exposure therapies became the treatment of choice for a range of otherwise debilitating disorders. VR exposure to cyber-spiders, virtual airplanes, and computer-generated audiences are effective and provide a number of practical benefits, such as improved client compliance. A VR helmet is not just for gaming; it can be an effective adjunct to psychotherapy. Conditioned Attraction and Aversion Much of what attracts and pleasurably arouses us is influenced by classical conditioning. Consider sexual arousal. An outfit or the scent of a partner’s cologne can become a conditioned stimulus for arousal. Experiments show that pairing a neutral odour with pleasing physical massage increases people’s attraction to that smell (Baeyens et al., 1996) and that people become sexually aroused to stimuli after those stimuli have been paired with sexually arousing UCSs (Rachman & Hodgson, 1968). Classical conditioning also can decrease our arousal and attraction to stimuli. This principle is used in aversion therapy, which attempts to condition an aversion (a repulsion) to a stimulus that triggers unwanted behaviour by pairing it with a noxious UCS. To reduce an alcoholic’s attraction to alcohol, the patient is given a drug that induces severe nausea when alcohol is consumed. Aversion therapies yield mixed results, often producing short-term changes that do not last or do not generalize outside of the environment where the learning occurred (Garbutt, 2009). Conditioned attraction and aversion also play a role in attitude formation (Walther, 2002). Neutral stimuli can become attractive or unattractive by being paired with stimuli that already elicit positive or negative attitudes. Advertising executives are keenly aware of classical conditioning’s power. They carefully link products and company logos to cute animals, attractive and famous people, humour, “fuzzy-warm” family images, and most of all, to pleasurable interactions with the opposite sex (Figure 7.9). And it works, marketing experiments show that this Learning and Adaptation: The Role of Experience FIGURE 7.9 Advertisers attempt to classically condition favourable consumer attitudes to products by associating products with other positive stimuli, such as physically attractive models. approach creates favourable attitudes toward novel products (Priluck & Till, 2004). Conditioning also can create unfavourable attitudes toward a CS by pairing the CS with a negative or unpleasant UCS. At times, this principle can have beneficial effects. In a study of fourth- and ninth-grade schoolchildren, Laura Moore and her colleagues (1982) paired the concepts of smoking, drinking, and drug use with words having negative connotations. Experimental group children exposed to the pairings developed more negative attitudes toward these activities than did control group children who were not exposed. Behaviourists, such as John Watson, originally argued that an emotional reaction, whether it is fear or attraction, could be classically conditioned to any stimulus. We know now, however, that there are some constraints on learning. For example, it is easier to condition fear to some stimuli than others; we seem to be biologically prepared to easily learn to fear stimuli such as heights, snakes, spiders, and bats. Similarly, it is relatively easy to condition an aversion to a taste by pairing a taste and an illness, but it is very difficult to condition a similar aversion to a visual stimulus by pairing a visual cue and an illness. We will return to this issue later in this chapter when discussing constraints on classical and operant conditioning. Beyond influencing fear, attraction, and aversion, classical conditioning also can affect our physical health. Allergic responses occur when the immune system overreacts and releases too many antibodies to combat pollen, dust, or other foreign substances (called allergens). When a neutral stimulus (such as a distinct odour) is repeatedly paired with a natural allergen (the UCS), it may become a CS that triggers an allergic CR (Irie et al., 2001). Research has demonstrated that classical conditioning procedures can contribute to the development of anticipatory nausea and vomiting that cancer patients can develop during chemotherapy and radiation therapy (Parker et al., 2006). Classical conditioning can even increase immune system functioning (Saurer et al., 2008). OPERANT CONDITIONING: LEARNING THROUGH CONSEQUENCES For all its power to affect our emotions, attitudes, physiology, and health, classical conditioning cannot explain how a dog learns to sit on command. Nor can it account for how we learn to drive cars, use computers, make friends, or be good citizens. Unlike salivating to a tone, these are not elicited responses automatically triggered by some stimulus. Rather, they are emitted (voluntary) responses, and they are learned in a different way. In Review • Classical conditioning involves pairing a neutral stimulus with an unconditioned stimulus (UCS) that elicits an unconditioned response (UCR). Through repeated pairing, the neutral stimulus becomes a conditioned stimulus (CS) that evokes a conditioned response (CR) similar to the original UCR. • The acquisition phase involves pairing the CS with the UCS. Extinction, the disappearance of the CR, occurs when the CS is presented repeatedly in the absence of the UCS. Sometimes, spontaneous recovery occurs after a rest period and the CS temporarily will evoke a response even after extinction has taken place. • Stimulus generalization occurs when a CR is evoked by a stimulus similar to the original CS. Discrimination occurs when a CR occurs to one stimulus but not another. • Once a stimulus (e.g., a tone) becomes a CS, it can now be used in place of the original UCS (food) to condition other neutral stimuli. This is called higher-order conditioning. • A wide range of bodily and psychological responses can be classically conditioned, including fears, sexual attraction, and positive and negative attitudes. Techniques based on classical conditioning are highly successful in treating fears and phobias. 237 238 CHAPTER SEVEN Thorndike’s Law of Effect While Pavlov was studying classical conditioning, American psychology student Edward L. Thorndike (1898) was exploring how animals learn to solve problems. He built a special cage, called a puzzle box, that could be opened from the inside by pulling a string or stepping on a lever (Figure 7.10). Thorndike placed a hungry animal, such as a cat, inside the box. Food was put outside, and to get it the animal had to learn how to open the box. The cat scratched and pushed the bars, paced, and tried to dig through the floor. By chance, it eventually stepped on the lever, opening the door. Performance slowly improved with repeated trials, and over time the cat learned to press the lever soon after the door was shut. Because performance improved slowly, Thorndike concluded that the animals did not attain “insight” into the solution. Rather, with trial-anderror, they gradually eliminated responses that failed to open the door, and became more likely to perform actions that worked. Thorndike (1911) called this process instrumental learning because an organism’s behaviour is instrumental in bringing about certain outcomes. He also proposed the law of effect, which stated that in a given situation, a response followed by a “satisfying” consequence will become more likely to occur, and a response Time to escape (seconds) 12. What evidence led Thorndike to propose the “law of effect?” 360 240 120 0 40 20 60 Trial FIGURE 7.10 Through trial and error, cats eventually learned to open Thorndike’s puzzle boxes to obtain food. Based on Thorndike, 1898, 1911. followed by an unsatisfying outcome will become less likely to occur. The law of effect became the foundation for the school of behaviourism. Skinner’s Analysis of Operant Conditioning Harvard psychologist B.F. Skinner was the leading American proponent of behaviourism throughout most of the 20th century. Skinner coined the term operant behaviour, meaning that an organism operates on its environment in some way; it emits responses that produce certain consequences. Operant conditioning (akin to Thorndike’s instrumental learning) is a type of learning in which behaviour is influenced by its consequences (Skinner, 1938, 1953). Responses that produce favourable consequences tend to be repeated, whereas responses that produce unfavourable consequences become less likely to occur. Through operant conditioning, organisms learn to increase behaviours that benefit them and reduce behaviours that harm them. Skinner designed a special chamber, called a Skinner box, to study operant conditioning experimentally. A lever on one wall is positioned above a small cup, and a food pellet automatically drops into the cup whenever a rat presses the lever (Figure 7.11). A hungry rat is put into the chamber and, as it moves about, it accidentally presses the lever. A food pellet clinks into the cup and the rat eats it quickly. We record the rat’s behaviour on a cumulative recorder, and find that it presses the bar more and more frequently over time. Today, a computer can be programmed to control the delivery of stimuli and reinforcer and to record the responses. Skinner identified several important types of consequences. For now, we focus on two: reinforcement and punishment. With reinforcement, a response is strengthened by an outcome that follows it. Typically, “strengthened” is operationally defined as an increase in the frequency of a response. The outcome (a stimulus or event) that increases the frequency of a response is a called a reinforcer. Food pellets are reinforcers because they increase the rat’s frequency of lever pressing. Once a response becomes established, reinforcers maintain it: The rat keeps pressing the lever because it continues to receive food. Punishment is the opposite of reinforcement; it occurs when a response is weakened by outcomes that follow it. Take our lever-pressing rat. Suppose we change things so that pressing the lever delivers a one-second electric shock, rather than food. If lever pressing decreases (which it will), then the electric shock represents a punisher: a consequence that weakens the behaviour. Notice that reinforcers and punishers are defined in terms of their observable Learning and Adaptation: The Role of Experience effects on behaviour. If the food doesn’t increase lever pressing, then for this particular rat it is not a reinforcer. Drum Paper direction IF antecedent stimuli (A) are present AND behaviour (B) is emitted, THEN consequence (C) will occur. IF I say “Sit” One response Pause in responding Not responding Pen reset at this point Pen direction Pen ABCs of Operant Conditioning Skinner’s analysis of operant behaviour involves three kinds of events: antecedents (A), which are stimuli that are present before a behaviour occurs; behaviours (B) that the organism emits; and consequences (C) that follow the behaviours. Thus, 239 Series of rapid responses Time AND my dog Jessie sits, THEN she gets a tasty treat. The relations between A and B, and between B and C, are called contingencies. Jessie’s behaviour of sitting is contingent on my saying “Sit.” The consequence of receiving food is then contingent on her response of sitting. Before exploring operant conditioning more closely, we wish to emphasize two points. First, keep in mind the key differences between classical and operant conditioning: • In classical conditioning, the organism learns an association between two stimuli—the CS and UCS (e.g., a tone and food)—that occurs before the behaviour (e.g., salivation). In operant conditioning, the organism learns an association between behaviour and its consequences. Behaviour changes because of events that occur after it. • Classical conditioning focuses on elicited behaviours. The conditioned response is triggered involuntarily, almost like a reflex, by a stimulus that precedes it. Operant conditioning focuses on emitted behaviours: In a given situation, the organism generates responses (e.g., pressing a lever) that are under its physical control. Second, although classical and operant conditioning are different processes, many learning situations involve both. When your dog hears the can opener, he will run to you, wagging his tail and salivating. The sound of dinner being prepared is a CS that automatically triggers a CR of salivation. It also is a signal to your dog that if he comes to you (an operant response) he will be reinforced by the desirable consequence of being fed. Thus, one stimulus (the sound of the can opener) can have classical as well as operant functions, which appear to be processed through different neural pathways in the brain (Schmajuk & Holland, 1998). FIGURE 7.11 With B.F. Skinner watching, a rat raises up and presses a lever in an operant experimental chamber (Skinner box). Pressing the lever turns on a light inside the chamber (notice the lever just below and to the right of the light). A food reinforcer is automatically delivered by the apparatus to the left of the box, and the rat’s performance is displayed on a cumulative recorder. Antecedent Conditions: Identifying When to Respond In operant conditioning, the antecedent may be a general situation or specific stimulus. Let’s return to our lever-pressing rat. At present, simply being in the Skinner box is the antecedent condition. In this situation, the rat will press the lever. Suppose we place a light on the wall above the lever. When the light is on, pressing the lever dispenses food, but when the light is off, no food is given. The rat will soon learn to press the lever only when the light is on. The light becomes a discriminative stimulus, a signal that a particular response will now produce certain consequences. Discriminative stimuli “set the occasion” for operant responses. Discriminative stimuli guide much of our everyday behaviour. If you are hungry, food on your plate 13. Identify two key differences between classical and operant conditioning. 14. Why are antecedent stimuli important in operant conditioning? 240 CHAPTER SEVEN is a discriminative stimulus to start eating. Classroom bells, the sight of your favourite restaurant, the words people speak to us, and the sight of a friend’s face are all discriminative stimuli that set the occasion for us to make certain responses. Consequences: Determining How to Respond Behaviour is governed by its consequences. Two major types of reinforcement strengthen responses, and two major types of punishment weaken them. Operant behaviour also is weakened by extinction. Figure 7.12 shows these processes. Positive Reinforcement Behaviour is reinforced by desirable outcomes. Being presented with a stimulus we find pleasing represents a desirable outcome. A rat receives food for pressing a lever. We receive praise for a job well done. This process is called positive reinforcement: A response is strengthened by the subsequent presentation of a stimulus. The stimulus that follows and strengthens the response is called a positive reinforcer. Food, drink, comforting physical contact, attention, praise, and money are common positive reinforcers. The term reward often is used as if it were synonymous with positive reinforcement. Behaviourists prefer the term positive reinforcement because it describes how consequences affect behaviour. In many instances, “rewards” do not function as positive reinforcers. Parents may “reward” a child with a new toy for cleaning her room, but if the child does not clean her room again, then the toy was not a positive reinforcer for that behaviour. Negative Reinforcement 15. How does negative reinforcement differ from positive reinforcement and from punishment? 16. Explain how operant extinction, positive punishment, and negative punishment differ. Receiving something pleasurable is a good outcome, but it’s only half of the story. Getting rid of something we find aversive—or avoiding something we anticipate will be aversive—also is a good outcome. We take Aspirin to relieve headaches, children clean up their rooms to stop their parents’ nagging, and we use umbrellas to avoid getting wet. This process is called negative reinforcement: A response is strengthened by the subsequent removal or avoidance of a stimulus (see Figure 7.12). The stimulus that is removed or avoided is called a negative reinforcer. Do not confuse negative reinforcement with punishment. Punishment weakens a response. Reinforcement—whether positive or negative— always means that a response is being strengthened (or maintained once it has reached full strength). Operant Extinction Operant extinction is the weakening and eventual disappearance of a response because it is no longer reinforced. When previously reinforced behaviours no longer pay off, we are likely to abandon and replace them with more successful ones. If pressing a lever no longer results in food pellets, the rat eventually will stop making this response. The degree to which non-reinforced responses persist is called resistance to extinction. Nonreinforced responses may stop quickly (low resistance), or they may keep occurring hundreds or thousands of times (high resistance). People who solicit charitable donations do not stop just because 100 passersby in a row fail to give money. As we examine later, resistance to extinction is strongly influenced by the pattern of reinforcement that has previously maintained the behaviour. Operant extinction can provide a good alternative to punishment as a method for reducing undesirable behaviours (O’Leary & Wilson, 1987). Positive Punishment Like reinforcement, punishment comes in two forms. One involves actively applying aversive stimuli, such as painful slaps, electric shock, and verbal reprimands. This is positive punishment, also called aversive punishment. A response is weakened by the subsequent presentation of a stimulus. Spanking or scolding a child for misbehaving are obvious examples, but so is a child’s touching a hot stovetop burner. The pain delivered by the burner makes it less likely that the child