1 Some Basic Assumptions Learning Objectives After reading this chapter, you should be able to do the following: • Analyze the claim that our behavior is determined by our heredity and environment, citing evidence for the role of our brains in controlling behavior and examples of how the environment can influence behaviors. • Explain the two main methods that have been proposed to study the laws of learning: introspection and experiments. • Explain the importance of manipulating only one independent variable at a time in an experiment, and how this can result in experiments seeming artificial and progress being slow. • Contrast behavioral and cognitive approaches to whether psychologists should study the role of the mind in determining behavior. • Explain the seeming paradox of why psychologists, whose goal is to understand human behavior, have often studied animals instead, and discuss the ethical issues raised by using animals in research. • Define learning and, in particular, a subcategory called associative learning, and explain the forms of associative learning that we will focus on here, classical and operant conditioning. lie6674X_01_c01_001-036.indd 1 3/14/12 4:18 PM Chapter 1: Some Basic Assumptions CHAPTER 1 As you already know from the title, this textbook will be about learning, which we can define as our capacity to change our behavior based on experience. We’ll be focusing on three particular kinds of learning: classical conditioning, reinforcement, and punishment. (We’ll define all three later.) For each one, we’ll try to understand the basic processes involved, but we’ll also be looking at some important practical questions. If you want to encourage a boy to study, for example, should you offer him a reward for good grades? If the answer is yes, what kind of reward would be most effective? Conversely, if you wanted to stop a girl from stealing, should you punish her when she does so? Psychologists have been studying questions like these for more than a century, but you’ll discover throughout this text that they still disagree, sometimes vehemently, about the answers. Why? Why should it be so difficult to discover the facts about learning? Or, even more fundamentally, what is a fact? Of course everyone knows what a fact is; it’s something that everyone knows to be true. Or is it? Was it a fact that the earth was flat because everyone before Columbus believed it to be so? Or that the earth was the center of the universe before Copernicus and Galileo moved it into orbit around the sun? If we cannot be sure of the truth in cases as seemingly obvious as these, how much more difficult must it be when the truth is more obscure, and when experts can’t agree among themselves? If one “scientist” claims that the moon is composed of blue cheese, and another that it is clearly a brownie, how are we to decide which of their views is correct? In older sciences, such as physics and chemistry, disputes over scientific facts are less obvious. Over the years, basic concepts such as the existence of the atom and the law of gravity have become firmly established. Only after considerable training are new initiates to the profession gradually introduced to the ambiguities and uncertainties of current research. In psychology, which is a relatively new science, the dividing line between “old established facts” and “new controversial hypotheses” is less clear, and there is no comforting bedrock of certainty and accomplishment to support students feeling overwhelmed by conflicting claims. Consider the use of corporal punishment: Is it an effective and ultimately humane way to eliminate a child’s harmful behavior, or is it a relic of our primitive past? There is evidence to support both views, and it can be more than a little frustrating to try to analyze the arguments of each side. In their attempts to resolve such disagreements—to decide what is a fact and what is not— psychologists have relied on several assumptions. These assumptions are now so widely accepted that psychologists rarely question them, but this does not necessarily mean that they are correct. It is perhaps worth emphasizing in advance that the assumptions we will examine in this chapter really are assumptions and are not universally accepted, even among psychologists. You should approach them with a healthy skepticism and form your own views as to their validity. The better you understand these assumptions, however, the better you will understand why research has followed the paths that we will trace in subsequent chapters. One purpose of this chapter, then, is to examine the methodological assumptions that have guided psychological research: why psychologists rely on experiments to understand behavior, and the logic that guides researchers in designing these experiments. lie6674X_01_c01_001-036.indd 2 3/14/12 4:18 PM Section 1.1 Is Behavior Lawful? CHAPTER 1 Before considering how to do research, however, we will begin by focusing on an even more fundamental issue: Why study behavior in the first place? 1.1 Is Behavior Lawful? The most fundamental assumption underlying research into the laws of learning is that there are such laws—if behavior were random, there would be little point in trying to discover the laws that govern it. To clarify the issue, let us begin by defining more precisely what we mean by a law. Within science, a law is essentially a statement of the form “If A, then B.” That is, if some condition A exists, we predict that event B will occur. The statement “The sun rises every morning,” for example, predicts that if it is morning, then the sun will rise. The assertion that behavior is lawful, therefore, is essentially a claim that behavior is in principle predictable: The same set of conditions will always produce the same behavior. Determinism Versus Free Will Most of us believe that at least some aspects of behavior are predictable. However much we might dislike some powerful bully, for example, we don’t usually walk over to him and punch him in the nose, because we know very well that his reaction will not be random but intensely and unpleasantly predictable. Opinion varies, however, as to the extent of this predictability. Determinism At one extreme, some believe that all behavior is predictable. According to the doctrine of determinism, human behavior is entirely dictated by heredity and environment (as used here, the word “environment” refers to past experiences as well as present environment). Your decision to go to college, for example, was probably influenced by factors such as the educational background of your parents, the grades you received at school, the economic advantages of a degree, and so forth. According to determinism, these factors made it inevitable that you would eventually choose to go to college, whether or not you were consciously aware of their influence. Dramatic advances in physics and chemistry have accustomed us to the idea that nature is inherently orderly, even though our lie6674X_01_c01_001-036.indd 3 The orbit of the moon is an example of an outcome that’s determined by a set law of physics. It always follows the same path, according to a precise law of nature. Determinists believe that behavior is also fully lawful, shaped by our environment and heredity. 3/14/12 4:18 PM Section 1.1 Is Behavior Lawful? CHAPTER 1 ignorance sometimes makes it appear random. But is the behavior of a living organism just as lawful, just as determined, as the orbit of the moon or the boiling point of water? Are we really just pawns in the grasp of environmental and genetic forces beyond our control? Free Will Within Western civilization, strict determinism has generally been rejected. In this view, humans are fundamentally free: We all have the power to determine our actions. This free will makes each of us responsible for our behavior and provides the basis for our concepts of morality and responsibility. Aside from some formal religious teachings, most of us resist the idea that we are only insignificant links within a causal chain. We are decidedly not like billiard balls hurtling through space, propelled by forces we cannot resist. Why, then, do many research psychologists still believe in determinism? The reasons are complex, and in the following sections we will consider some of them. As you read this material, some of the arguments might strike you as more philosophical than psychological, and you might wonder why a psychology textbook devotes so much attention to this issue. The answer is that a belief in determinism plays an important role in guiding psychological research. If you carry out a study to find a lawful relationship and your effort fails, you are much more likely to persist in the effort if you are convinced that such laws really exist. As a result, many of the most crucial discoveries about learning and memory have been made by psychologists with a stubborn, even fanatical, belief that behavior is lawful. (See, for example, the discussions of Pavlov in Chapter 2 and Ebbinghaus in Chapter 8.) In the sections that follow, we will conThe doctrine of free will says that when offered a choice, sider some arguments that have we can select whichever option we want, unconstrained by led to this belief. external forces. Neural Determinism One line of evidence supporting the determinist view comes from our growing understanding of the brain’s role in determining behavior. We will discuss the mechanisms involved in more detail in Chapter 2, but in essence the brain consists of a vast network of interconnected cells called neurons, and the transmission of electrical signals through these cells determines our behavior. When we see a friend, for example, the light falling on receptors located at the back of the eye produces electrical signals, and these are transmitted by a series of neurons to the cortex (a region of the brain) and ultimately to the muscles that cause us to raise our hand in greeting or to move our lips to say hello. If physicists are correct, and the behavior of all particles in the universe is lawful, then the transmission of these electrical impulses through our brain must also be lawful. (Indeed, neurophysiologists already lie6674X_01_c01_001-036.indd 4 3/14/12 4:18 PM Section 1.1 Is Behavior Lawful? have a good understanding of the chemical processes that govern the transmission of electrical signals through neurons, and how the arrival of a signal at a neuron’s terminal leads to the release of chemicals, which then activate the next neuron in the chain.) If these assumptions are correct—if our brains control our behavior, and if the brain’s operations are lawful— then it logically follows that our behavior must also be lawful. If the operation of any system is lawful, then the output of that system must also be lawful. CHAPTER 1 According to neural determinism, the transmission of electrical signals through nerve cells in the brain determines our behavior. We will call this argument neural determinism. Its first assumption—that the firing of neurons is governed by laws of physics and chemistry that govern all other materials—is widely accepted, at least within science. The second assumption, though—that our brains control all aspects of our behavior—is more controversial. We will therefore focus on this second assumption, examining the brain’s role in three fundamental aspects of behavior: movement, emotion, and thought. Movement Our physical movements are controlled by the transmission of electrical impulses though our neurons. When we move an arm, for example, the movement is caused by the contraction of muscles within the arm, which are in turn controlled by neurons. Neurons are connected to every muscle in the body, and electrical impulses arriving at the terminals of these neurons trigger the release of chemicals that initiate muscular contractions. If these neural connections are damaged—for example, if the spine is damaged in an automobile accident so that neural messages can no longer be transmitted from the brain to certain muscles—we lose the ability to control those muscles. Similarly, the tremors seen in Parkinson’s disease are due to degeneration of neurons in one of the regions of the brain that helps control movement. Administering a drug that restores the functioning of the affected neurons can help treat Parkinson’s. This type of evidence illustrates the point that movement is controlled by the nervous system. Emotion In a similar way, our emotions are controlled by specific regions of our brains that are active. One early experiment demonstrating the brain’s role in emotions was reported by James Olds and Peter Milner (1954), who found that delivering a tiny electrical current to certain areas of a rat’s brain seemed to produce pleasurable sensations in the rat, as the rat would press a lever as often as 2,000 times per hour to turn on the current. A neurosurgeon, Dr. Robert Galbraith Heath (1963), reported similar effects in humans. One of his patients suffered from narcolepsy, a debilitating condition in which sufferers lie6674X_01_c01_001-036.indd 5 3/14/12 4:18 PM Section 1.1 Is Behavior Lawful? CHAPTER 1 fall asleep at inappropriate times, such as in the middle of a conversation. In an effort to help this patient stay awake, Heath and his colleagues implanted small electrodes into several areas of the patient’s brain. They provided the patient with a control panel that he could use to stimulate these areas. He described one of the buttons on this panel as his “happy button,” saying it gave him a drunk feeling, and another stimulated sexual arousal. Heath also described an episode in which the experimenter initiated stimulation while a patient was exhibiting agitated violent psychotic behavior. The effect was dramatic: “Almost instantly his behavioral state changed from one of disorganization, rage, and persecution to one of happiness and mild euphoria. . .He was unable, when questioned directly, to explain the sudden shift to his thoughts and feelings.” (Heath, 1963, p. 575) Drugs such as alcohol, heroin, and Ecstasy work in a similar way, although usually not quite so dramatically. By altering chemical activity in the brain, they can profoundly change the emotions we experience. Thought The suggestion that our brains also control what we think is perhaps the most controversial of these claims. Early evidence for the brain’s role in thought came from studies of epilepsy reported by a Canadian neurosurgeon, Wilder Penfield. Epileptic seizures are triggered by abnormal activity in one small region of the brain, and in severe cases it is important to identify the precise region involved so that it can be removed. One way to do this is to remove part of the skull and use electrodes to stimulate various parts of the brain while the patient is conscious; the patient can then report when he or she experiences the sensations that normally precede the seizures, so that the surgeon can remove the region that produces these feelings. (This technique might sound gruesome, but the scalp is anesthetized first, and because there are no pain receptors in the brain, the patient suffers no discomfort.) Penfield discovered that stimulation of some areas would give rise to specific thoughts or images. Depending on the area stimulated, patients reported hearing someone calling their name, feeling like they were waiting at a station for a train, or hearing music. If the stimulation was stopped, the sensation would cease, but it would often return if the same spot was stimulated again (Penfield, 1958). Activity in particular cortical areas thus seemed to control what thoughts a patient experienced. For many years Penfield’s findings stood almost alone, as few psychologists had the surgical skills, or the access to patients, to repeat his studies. More recently, though, a variety of techniques has been developed to monitor brain activity without surgical intervention, and this emerging research has confirmed Penfield’s findings. One example comes from a recent study by Sheth, Sandkühler, and Bhattacharya (2009). They gave participants difficult problems to solve, and asked them to press a key the instant they thought of the solution. The experimenters used EEG recordings to monitor activity in participants’ brains as they worked. (EEG stands for electroencephalogram; in which electrodes placed on the scalp record electrical activity occurring inside the brain.) The recordings revealed that activity in certain areas of the brain preceded solution—the heightened activity was only observed shortly before solution. This finding was particularly striking because the electrical activity was observed up to eight seconds prior to the moment when participants lie6674X_01_c01_001-036.indd 6 3/14/12 4:18 PM Section 1.1 Is Behavior Lawful? CHAPTER 1 became consciously aware of the solution. As in Penfield’s studies, conscious experience seemed to be the product of preceding neural activity. In summary, it appears as if every aspect of our behavior—movement, emotion, and thought—depends on the transmission of electrical impulses within our brains. If this neural activity is lawful, then the behavior that it controls must also be lawful. Examples of Lawful Behavior The neural determinism argument claims that behavior must be lawful, but that is not quite the same as proving that behavior is lawful. There is considerable evidence that our heredity and environment can strongly influence our behavior, and we will look now at three examples. Obedience Our first example comes from Stanley Milgram’s classic research on obedience (Milgram, 1963, 1974). Milgram was horrified by the behavior of German soldiers who participated in the murder of millions of people in concentration camps during World War II. Though some of those involved may have been evil, many seemed to be ordinary soldiers obeying orders, no matter how vile those orders were. Milgram designed an experiment that he hoped would allow him to study obedience in the laboratory. He told participants that he was studying the effects of peer-delivered punishment on learning. Their task was to administer an electric shock to a partner whenever the partner made an error on a memory problem. The intensity of the shock was controlled by a series of 30 switches, ranging from 15 to 450 volts, and the experimenter instructed subjects to increase the intensity of the shock after every error by the partner, who was in an adjoining room. (Unknown to the subject, the partner was actually in on the experiment and never received any shocks, but just acted as if they occurred.) Milgram had hoped to use the highest shock intensity his subjects were willing to administer as a measure of their obedience to authority—in this case, a scientist in a white lab coat. The astonishing result, which Milgram had not anticipated, was that there were essentially no limits to his subjects’ obedience. Sixty-five percent continued to administer shocks even when their partners pounded on the wall and refused to answer any questions and when the switch on the shock control panel was labeled “450 volts . . . Danger: Severe Shock.” His subjects became extremely upset as the experiment continued, some laughing hysterically and pleading with the experimenter to let them stop, but almost all continued to administer shocks when ordered to do so. Was the extraordinary behavior of the participants in Milgram’s study a product of the experiment’s artificial conditions and thus atypical (Baumrind, 1964)? The behavior of German soldiers during World War II argues against this view. Another poignant example comes from the Vietnam War in the 1960s. In one of the most notorious incidents of that war, a group of American soldiers entered a small Vietnam lie6674X_01_c01_001-036.indd 7 3/14/12 4:18 PM CHAPTER 1 Section 1.1 Is Behavior Lawful? village, My Lai, in search of enemy soldiers. When the American soldiers failed to find the enemy, they obeyed the orders of their officers and slaughtered everyone in the village, killing hundreds of defenseless women and children. Although the terrible conditions of war undoubtedly played a major role in producing such behavior, Milgram’s research suggests that obedience is not confined to such situations. We are more sensitive to social control—to the opinions of our parents, our friends, even our neighbors—than we sometimes realize (see also Cialdini, 1993). Child Abuse A further example of how powerfully our environment can influence our behavior comes from studies of children who are physically or sexually abused. Approximately two thirds of children who are abused develop serious symptoms, rangThe 1968 My Lai massacre, in which several ing from anxiety and bed-wetting to depression hundred Vietnamese civilians were killed, is and self-destructive behavior (Kendall-Tackett, an example of individuals obeying authority Williams, & Finkelhor, 1993). One of the saddest to the point of inflicting harm on others. of these after-effects is that many of these victims are much more likely to abuse their own children. Kaufman and Zigler (1987) reviewed the many studies in this area and concluded, “approximately one-third of all individuals who were physically abused, sexually abused, or extremely neglected will subject their offspring to one of these forms of maltreatment” (p. 190). Conversely, most adults who abuse children were themselves abused as children. In one typical study, Kasper and Alford (1988) studied 125 men who had sexually abused children and found that approximately 85% were themselves abused. The experience of abuse can profoundly influence a child’s present and future behavior. Aggression One of the consequences of childhood abuse is a 50% increase in the probability that boys will behave violently when they become adults. But that statistic also indicates that not all boys who are abused become violent. Why do some boys become violent but not others? One possible answer is genetics. Animal research has shown that an enzyme called monamine oxidase A (MAOA) plays an important role in reducing aggression, and that a single gene regulates production of this enzyme. Perhaps the reason that some abused boys are more likely to become violent is that they lack this inhibitory gene. In one study, males who had been abused as children and lacked the MAOA gene were found to be roughly six times more likely to be convicted of violent crimes than were males without these predisposing factors. In other words, just two factors—a history of abuse and the absence of a single gene—were enough to almost completely determine how these boys would behave when they became adults (see also Miles & Carey, 1997). lie6674X_01_c01_001-036.indd 8 3/14/12 4:18 PM Section 1.1 Is Behavior Lawful? CHAPTER 1 A second, perhaps surprising, determinant of aggression is nutrition. Several studies have shown that poor nutrition is associated with a wide range of violent and criminal behaviors, and that improving nutrition can substantially reduce this behavior. The Feeling of Freedom Findings such as these pose a puzzle: If our behavior is influenced so strongly by our heredity and environment, how is it that in our everyday lives we do not experience any sense of being controlled? When you decide what clothing to wear or what to eat for lunch, you have no sense of compulsion that you must act in a certain way; quite the contrary, you freely decide. How can our behavior be determined if we constantly feel so free? The answer proposed by determinists is that although we may feel free in such situations, we are still controlled. We are just less aware of the forces influencing us. Advertising A classic example of how we can be influenced without realizing it is advertising. Most of us believe that we are not influenced by advertisements—we insist that we base our decisions solely on evidence. Some research, however, suggests that we are all more susceptible to advertising than we realize. In one study on this point, Smith and Engel (1968) showed 120 men a picture of an automobile. For half the subjects, the photograph showed only the car, whereas for the other subjects the car was shown with a sexy redheaded woman standing in front of it. After examining the picture, participants were asked to evaluate the car on several dimensions. Those who saw the car with the attractive female rated the car as significantly more appealing and better designed. They also estimated it to be more expensive (by an average of $340), faster, and less safe. When the authors later asked a subset of the participants if their ratings had been influenced by the presence of the model, 22 out of 23 denied it. One respondent claimed, “I don’t let anything but the thing itself influence my judgments. The other is just propaganda.” Another commented, “I never let myself be blinded by advertising; the car itself is what counts.” Thus, although the model’s presence clearly altered the participants’ ratings of the One classic study on advertising found that products portrayed car, virtually none believed that in images with attractive women were much more likely to appeal to male consumers. they had been affected. Sexual Attraction Another illustration of how the environment can influence us without our realizing it comes from research on sexual attraction. Why is it that we are sexually attracted to some individuals but not to others? Psychologists are still in the early stages of trying lie6674X_01_c01_001-036.indd 9 3/14/12 4:18 PM Section 1.1 Is Behavior Lawful? CHAPTER 1 to understand attraction, but some interesting evidence has begun to emerge. One early study, by Dutton and Aron (1974), was carried out in an unusual setting for a psychology experiment—a deep river gorge in British Columbia. There were two ways of crossing the river: a narrow, wobbly footbridge located some 230 feet above rapids, or a much more substantial wooden bridge only 10 feet above a small rivulet. Men were approached as they crossed either bridge by an attractive woman who asked if they would answer some questions for a research project. When the interview was over, she gave the men her telephone number in case they later had any questions. The real purpose of the study was to measure sexual attraction—would the men later phone to ask for a date? Many did, but the study’s striking finding was that the proportion asking for a date depended on where the interview took place: Half the men interviewed after crossing the rickety bridge later phoned for a date, compared with only 12% of those interviewed after crossing the solid bridge. On the surface this result might seem bizarre—why should the location of the interview determine whether men think a woman is attractive? Dutton and Aron, however, had predicted precisely this result on the basis of a theory of emotion previously proposed by Schachter and Singer (1962). We will not review the theory in detail, but in essence it proposes that all emotions are characterized by similar states of physiological arousal—increased heart rate, rapid breathing, and so on. Schachter and Singer argued that we therefore need to rely on environmental cues to help us identify what emotion we are experiencing. When the men experienced strong arousal when crossing the high bridge and then encountered the attractive interviewer, they would have unconsciously thought, “Aha, it must be her beauty that is making me feel so excited.” And believing that they were attracted to her, they were more likely to ask her for a date. They would have felt that the decision to do so was entirely free, but they were being influenced by factors of which they were unaware. Political Attitudes A third example comes from research on how people decide what political party to support. We normally assume that we make decisions as important as these by evaluating the positions of the different parties, but as with the two previous examples, research suggests that our choices can be powerfully influenced without our being aware of it happening. One example comes from research by an Israeli psychologist, Ran Hassin, who asked both Israelis and Americans questions about their political beliefs. The questions were presented on a computer screen, with a picture of their national flag flashed on the screen very briefly before each question appeared. Each presentation of the flag lasted less than 1/50th of a second and was followed by a jumbled set of lines called a pattern mask. Previous research had shown that masking stimuli presented under these conditions effectively erase preceding stimuli before subjects can become aware of them. The procedure is sometimes referred to as subliminal presentation—limen being the Latin word for threshold, so “subliminal” indicating that the stimulus remains below the threshold of consciousness. The experimenters interviewed participants afterward to determine if they had been aware of the flags. None had. So, did the presence of the flag influence participants’ political views, even though they were not aware of it? lie6674X_01_c01_001-036.indd 10 3/14/12 4:18 PM Section 1.2 How Should We Discover Any Laws? One study found that Israelis who viewed a subliminal image of their flag exhibited increased prejudice toward Palestinians; similar results were found in a study in which an American flag was used to influence Americans’ preference for a presidential candidate. Results such as these support the notion that our political attitudes may be shaped by factors other than conscious choice. CHAPTER 1 It did. Hassin et al. (2009) found that subliminal exposure to the Israeli flag increased Israelis’ feelings of prejudice toward Palestinians, and they found comparable effects in a U.S. study. This study, conducted in 2008, asked citizens whether they intended to vote for Barack Obama or John McCain; subliminal exposure to an American flag while answering increased support for McCain and decreased support for Obama. (See also Ballew & Todorov, 2007; Rutchick, 2010.) So, Is Behavior Lawful? It seems clear that our heredity and environment do influence much of our behavior, ranging from whom we find attractive to what political parties we support. The fact that our behavior is influenced, however, does not necessarily mean that it is entirely determined. Even under the most intense environmental pressure, we possibly still retain some freedom to choose. Consider again the effects of sexual abuse on children. We have seen that roughly one third of children who are abused go on to become abusers as adults. This also means that two thirds of these children do not abuse as adults. Proponents of free will can thus argue that even under the most terrible pressures, each of us retains some capacity to choose our own path. In the end, it is unlikely that the debate between free will and determinism will ever be resolved conclusively. Not even the most optimistic determinist believes that we will ever be able to predict every aspect of a person’s behavior—we would have to know every law and record every moment of a person’s life to be able to calculate the cumulative impact of all his or her experiences. Given that we can never fully predict behavior, it will always be possible for believers in free will to argue that we have an essential inner freedom, whereas determinists will claim that a belief in free will only reflects limitations in our current state of knowledge. It is doubtful we will ever know whether behavior is completely lawful. The evidence we have reviewed, however, suggests that environment and heredity play a powerful role. 1.2 How Should We Discover Any Laws? If behavior is lawful, at least to some degree, how can we determine the laws that dictate it? lie6674X_01_c01_001-036.indd 11 3/14/12 4:18 PM Section 1.2 How Should We Discover Any Laws? CHAPTER 1 Introspection If we want to understand why people behave as they do, one obvious approach is to ask them; that is, to have them carefully observe their thoughts and feelings as they behave and then report them, a technique called introspection. We are all introspective on occasion, and literature abounds with references to people “searching their souls” in an attempt to understand themselves. The first systematic application of this concept, however, was the work of a late nineteenth-century German psychologist, Wilhelm Wundt. The essence of Wundt’s technique was simple: Subjects were exposed to a stimulus and then asked to report the sensations aroused by it. Learning to report those sensations accurately required long and arduous training. It was important, for example, that a subject report not only what he or she saw (such as a chair), but the exact sensations the object elicited, the quality and intensity of these sensations, how they changed over time, and so on. This is surprisingly difficult. Suppose, for example, that you were shown a piece of black coal next to a piece of white paper, and that the coal was illuminated by a bright light while the paper was in shadow. If you were asked which was darker, you would almost certainly say the coal, even if physically it was reflecting far more light. (If you looked at the two samples through two holes in a screen, so that you didn’t know which was which, the coal would appear far brighter to you.) The problem is that we often perceive what we expect to see, rather than what is actually there. Wundt’s subjects underwent extended training to overcome this and similar errors. Once the observers were properly trained, Wundt hoped to use their reports to analyze the complex patterns of human thought into their constituent elements and then discover the laws by which these elements are combined to produce the richness and variety of mental life. Though the rigorous demands of Wundt’s technique seem daunting, the underlying logic has great intuitive appeal. If we want to understand the processes of learning, what better way than by studying these processes within our own minds? Yet, despite its obvious attractions, introspection gradually fell into progressively greater disrepute, until eventually it almost disappeared from psychology. One reason for this collapse was that even as Wundt was painstakingly beginning to train his subjects, a Viennese physician named Sigmund Freud was developing his revolutionary theories—theories that would ultimately destroy the rationale for introspection. The Influence of Sigmund Freud Freud exposed for the first time the world of the unconscious, its primitive swirl of emotions hidden behind powerful defensive barriers. This metaphor of subterranean forces had devastating implications for introspection, because it attacked its foundational premise: a faith in the accessibility of all thought to conscious analysis. If consciousness was only the visible tip of the mental iceberg, with vast areas of the mind concealed under defensive barriers, then introspection could provide only an incomplete and fragmented account of why we behave as we do. Freud’s theories were the first to suggest that there might be limits to the power of conscious analysis, but it seems likely that these limits would have become apparent eventually, with or without Freud. Consider, for example, what happens when you try to prove a geometry lie6674X_01_c01_001-036.indd 12 3/14/12 4:18 PM Section 1.2 How Should We Discover Any Laws? CHAPTER 1 theorem. You may struggle for minutes or even hours, searching for a solution, when suddenly the correct answer occurs to you. What happened exactly? How did you suddenly go from being confused to knowing the correct answer? Clearly, some important mental processes intervened between these two states, but, introspectively, your mind is a blank slate from which the correct solution spontaneously emerged. Our inability to trace the processes involved in acts such as these suggests limits to the usefulness of introspection in analyzing complex thought and learning. The existence of the unconscious means that introspection can play only a limited role in helping us understand behavior. Still, you might think that it could at least help us to understand conscious processes. Yet even here, serious doubts soon arose over whether observers’ reports were accurate. One example concerned a phenomenon known as “imageless thought.” Sigmund Freud was the first to propose that much of human behavior was driven by the subconscious mind. Some psychologists believed that the meaning of any word was simply the image that it produced—the meaning of the word “chair,” for example, would be the image that comes to mind when you think about this word. This approach seems plausible when we consider concrete nouns such as chair, but what of more abstract words such as “truth” or “meaning”? Do these words also produce images? One influential introspectionist, Oswald Kulpe, reported that when he and his colleagues introspected, they could not detect any trace of an image while thinking of such words. Another leading introspectionist, however, insisted that even the most abstract words produced images if you introspected carefully enough. In the case of meaning, for example, he reported seeing “the blue-gray tip of a kind of scoop which has a bit of yellow about it (probably a part of the handle) and which is digging into a dark mass of what appears to be plastic material” (Titchener, 1915, p. 519). This image had its origins, he suggested, in injunctions from his youth to “dig out the meaning” of Latin and Greek phrases. Each side insisted that the other was wrong, and there was no way to resolve their disagreement. The realization that much of the mind’s functioning is unconscious, coupled with the difficulty of reliably observing even those areas that ostensibly are conscious, eventually led to the abandonment of introspection as a scientific technique. The Experimental Method As the limitations of introspection became clear, psychologists turned to experimentation to discover the causes of behavior. In outline, the experimental method is very simple: We deliberately change some aspect of the environment to see if it affects behavior. Suppose, for example, that you are a clinical psychologist and want to find a treatment for depression. The traditional approach is psychotherapy—talking to a client about his lie6674X_01_c01_001-036.indd 13 3/14/12 4:18 PM Section 1.2 How Should We Discover Any Laws? CHAPTER 1 or her experiences to discover the emotional causes of behavior—but you believe that depression is caused not by emotions but by chemical imbalances in the brain. The easiest way to address this, you think, would be a regimen of vigorous daily exercise to increase the flow of serotonin, a brain chemical that is known to enhance mood. To see if this approach would be superior to psychotherapy, you could run an experiment, offering one group of patients psychotherapy while training another to exercise daily. If the exercise group improved more, this would support your view that exercise was a more effective treatment. The aspect of the environment that is altered (in this case, the therapy) is called the independent variable, and the behavior that is measured (in this case, depression) is called the dependent variable. If there is a consistent relationship between them—for example, environmental condition A is always followed by behavior B—this is called a law. One Thing at a Time! If experimentation were really so simple, discovering the laws of behavior would be easy: All we would have to do is manipulate our independent variables, observe their effects, and combine the resultant laws into a comprehensive account of behavior. The problem is that we must manipulate only one independent variable at a time. If several independent variables changed simultaneously, it would be impossible to say which one was responsible for the resulting behavior. The obvious solution, to ensure that only one aspect of the environment changes during an experiment, turns out to be impossible in many situations. Consider again our depression example. On the surface, it seems a very simple experiment—the only difference between the groups was which treatment they received. In fact, however, the groups could have differed in other ways. Suppose, for example, that the experiment participants all viewed psychotherapy as a much more plausible treatment than exercise; members of the psychotherapy group would then have a stronger belief that they would improve through treatment, and this belief could have alleviated their feelings of depression. In other words, the groups differed not only in which treatment they received, but also in how hopeful they were that it would work, and the feeling of hope might have produced the improvement. Even if the subjects’ expectations of improvement were the same in the two groups, the experimenters’ expectations could be different. Perhaps previous research had suggested that psychotherapy was more likely to be effective, so that the experimenters expected participants receiving this treatment to improve more. If so, the greater improvement in this group might be due not to the treatment but to the fact that the experimenters expected them to improve more. Clever Hans How could an experimenter’s expectations affect a subject’s evaluation of a picture? We know very little about the underlying processes, but some evidence suggests that subtle cues from the experimenter are involved. lie6674X_01_c01_001-036.indd 14 3/14/12 4:18 PM Section 1.2 How Should We Discover Any Laws? CHAPTER 1 One of the classic examples of such cues is the case of Clever Hans. (See Pfungst, 1965.) Hans was a horse that lived on a German farm at the turn of the century. Hans wasn’t an ordinary horse: He was the only horse in Germany that could add! When asked the sum of two plus two, for example, Hans would slowly begin to tap the ground, one, two, three, four . . . and then stop. Nor was this simply a trick he had memorized, because he could add virtually any numbers, and Clever Hans amazed the masses with his mathematical skill. it didn’t matter who asked the Ultimately, he was found to be less of a math genius and more question. He was equally profiof a master of observation. cient at subtraction and, incredibly, multiplication and division. An obvious explanation for his prowess was some sort of signal from his master, but when a blue-ribbon panel of experts convened to investigate Hans’s extraordinary skill, they found that Hans performed equally well in his master’s absence. In a brilliant series of experiments, German psychologist Oskar Pfungst eventually discovered the explanation for Hans’s apparent genius. Pfungst found that Hans’s accuracy was reduced if the person who asked the question didn’t know the correct answer. Furthermore, the farther away the questioner stood, the less accurate was Hans’s answer. Finally, putting blinders around Hans’s eyes destroyed his performance. Clearly, Hans could answer questions only if he could see someone who knew the correct answer. But what visual cues could the questioner provide? The answer, Pfungst discovered, was that questioners tilted their heads slightly forward as they finished their questions, and this was Hans’s cue to begin tapping. As the tapping approached the correct answer, the observers tended to straighten up in anticipation, and this slight tensing was Hans’s cue to stop. Hans was extraordinarily sensitive to such cues, responding to the raising of eyebrows or even the dilation of nostrils, and Pfungst was eventually able to control Hans’s tapping completely by producing these cues. Hans was an extraordinary horse, but his genius lay in his powers of observation rather than any arithmetic ability. Let us now return to our depression experiment. Suppose that we redesigned our experiment to ensure that the experimenter who ran the study expected both groups to improve equally, and we again found substantially greater improvement in the exercise group. Now, at long last, would we have proved that increasing exercise (and thus serotonin) is an effective treatment for depression? Yet again, the answer is no. Why not? Because, even if we can’t identify alternative explanations, that doesn’t prove there aren’t any. The blueribbon panel was unable to find any plausible explanation for Hans’s performance, but that didn’t prove that he was a genuine math wizard. An experiment, in other words, can never prove that a particular explanation is correct because it is always possible that some alternative explanation will eventually be found. lie6674X_01_c01_001-036.indd 15 3/14/12 4:18 PM Section 1.2 How Should We Discover Any Laws? CHAPTER 1 Experiments can support a particular explanation, and with sufficient evidence our confidence in that explanation can become very high—few now doubt the existence of gravity—but it is important to remain open to the possibility of other explanations. The Nature of Scientific Progress We started with a seemingly simple experiment, but the more we analyzed it, the more complex it became. This is always the case. The goal of the experimental method is to change only one independent variable at a time, but this ideal can rarely be fully realized. We can control for the effects of particular factors, such as subject and experimenter expectations, but there are always changes that we cannot control—fluctuations in humidity, the occurrence of sunspots, the death of an earthworm in China! This in turn has important implications for the nature of scientific progress. Slow. . . One such implication concerns the pace at which science sometimes progresses. A popular image of science has the scientist in a white lab coat advancing through rigorous analyses. In practice, scientific progress is often much more confused and halting. As we have seen, it is impossible to control for all possible variables; we can only control for those variables that seem important. Our notions of what variables are important, however, are often wrong. For example, in 19th century England, one of the most dangerous things a woman could do was have a baby in a hospital. Many thousands of women died every year after giving birth. When Joseph Lister suggested that doctors could prevent these deaths if they washed Scientific progress can move at a slow pace and is more their hands before delivering a confusing and halting than some people think. baby, his proposal was greeted with incredulity: How could having a doctor wash his hands with boiled-down animal fat (soap) prevent a woman from dying? We now understand, thanks to the germ theory of disease, that this action helps prevent infection. At the time the idea was first proposed, however, it seemed preposterous. Similarly, in the case of Clever Hans, few would have believed beforehand that a horse could be so sensitive to human body language. There is thus a built-in catch-22 to scientific progress: To discover scientific laws, you must control all important variables, but you can only identify the variables that are important if you already know the laws! This problem is not insurmountable. We just have to plug away, identifying important variables as best we can in experiments that may initially lack important controls. This bootstrapping process means that progress will initially be slow and frustrating as we struggle to identify the important variables. lie6674X_01_c01_001-036.indd 16 3/14/12 4:18 PM CHAPTER 1 Section 1.2 How Should We Discover Any Laws? Artificial A second implication of our analysis concerns the inherent artificiality of experiments. To isolate the effects of one variable, you need to hold others constant, but the more you control the environment, the less like real life it becomes. The underlying strategy is summarized in Figure 1.1: You start with a complex environment and, by analysis, try to break it into simpler environments so you can study the effects of constituent elements (A, B, C, and so on) one at a time. Then, once you have determined the effects of each variable on its own, you use the method of synthesis to recombine them, studying what happens when two or more variables act together (AB, ABC, and so on). The scientific method thus proceeds by first analyzing complex environments into simpler ones, then gradually returning to the more complex environment that was the original focus of interest. Figure 1.1: Breaking down a complex environment complex environment A B complex environment C synthesis analysis A A B B C C B A B A B C A C A C simple environment B C In trying to understand a complex environment, experimenters begin by examining the effect of each variable individually (analysis); it is then possible to study how they act in combination (synthesis). The analysis step requires highly simplified (and thus artificial) situations, so that only one variable is allowed to change at a time; in the synthesis stages the goal is to move back toward more complex (and thus realistic) environments. In psychology, most research is still analytical, with the result that it is very easy to feel depressed by its artificiality. After all, what does the behavior of a student in an artificial laboratory setting have to do with real life? The answer lies in the assumptions we have been tracing in this chapter. If behavior is lawful, and if the best way to discover those laws is through well-controlled experiments, then eventually the principles discovered in these artificial settings will help us understand behavior in the more complex conditions of the real world. lie6674X_01_c01_001-036.indd 17 3/14/12 4:18 PM Section 1.3 Behavioral and Cognitive Approaches CHAPTER 1 1.3 Behavioral and Cognitive Approaches One reaction to the discovery that introspection is severely limited as a tool for studying the mind was the emergence of a new approach within psychology called behaviorism. Behaviorists such as John B. Watson argued that because it is not possible to study the mind accurately, psychologists should instead focus on visible or overt behavior. Rather than studying subjective feelings such as hunger, we should study visible behavior such as eating. By focusing on behavior that could be observed, behaviorists hoped that psychologists would at least be able to agree on their data, thereby allowing theories to be evaluated by solid evidence rather than by the eloquence or prestige of opposing theorists. If you want to understand the effects of rewards on children, for example, behaviorists argued that you should present rewards and observe their effects, not speculate about what the children might be thinking. Skinner’s Radical Behaviorism Although all behaviorists agree on the importance of objective observations of behavior, they disagree concerning what attention, if any, should be paid to mental states. B.F. Skinner developed one influential approach (for example, Skinner, 1953). Like Watson before him, Skinner believed that the goal of psychology should be practical, helping with problems such as the best ways of rearing children, coping with phobias, and making education more enjoyable as well as more effective. All these problems involve changing behavior, and the only way to change anyone’s behavior is to change his or her environment. If you want to convince someone to vote for a particular political party, for example, you could try to do so by talking to him about the party’s merits, but your words would then represent a change in the person’s environment, and you would be hoping that this would lead to a consequent change in his voting behavior. If you want to change people’s behavior, therefore, you must understand the environmental conditions that determine this behavior. B.F. Skinner was particularly influential because of his advocacy of behaviorism, with its practical focus on discovering the environmental causes of behavior, rather than speculating about the mind. lie6674X_01_c01_001-036.indd 18 To pursue this point, suppose that a friend tells you that she has decided to vote for candidate X for president. Why did she reach this decision? The most obvious explanation is that she likes the candidate. Skinner did not doubt the existence of feelings such as liking, but he argued that it is a mistake to explain behavior in terms of such feelings; we must go on to ask why people have these feelings. In this case, perhaps your friend heard a speech the candidate made and was impressed by it. If so, Skinner believed that we should view this speech as the cause of her voting behavior, rather than the feelings that followed. 3/14/12 4:18 PM Section 1.3 Behavioral and Cognitive Approaches CHAPTER 1 Figure 1.2: A summary of Skinner’s views Skinner environmental conditions mental states behavior cognitive psychology B.F. Skinner recommended ignoring the mind and focusing on the direct relationship between environmental conditions and behavior. Figure 1.2 summarizes Skinner’s views. It outlines a sequence in which an environmental condition gives rise to a mental state, which in turn produces behavior. To change this behavior, we must understand the environmental conditions that produce it, and Skinner argued that we should study the relationship between environmental conditions and behavior directly. It is not easy to ignore the mind—our private worlds consist of our own thoughts and feelings, so we are inevitably fascinated by the possible thoughts and feelings of others—but Skinner argued that psychologists must focus on the environmental determinants of behavior if they want to be able to help people. Cognitive Approaches A rather different approach to the role of the mind in psychology emerged from cognitive psychology. (Cognition refers to the processes involved in thinking; cognitive psychologists try to understand these processes.) Cognitive psychologists are also behaviorists, in that they believe that most operations of the brain are unconscious. If you ask people their telephone number, for example, they will probably all respond immediately, but if you then ask them how they managed to retrieve this information, they will probably just stare at you. Most of the brain’s activity occurs at an unconscious level, so introspection is very limited in what it can tell us. Cognitive psychologists thus agree with other behaviorists that introspection is of limited value, but they argue that this need not prevent us from studying cognitive processes. Consider that physicists, for example, cannot directly observe the existence of atoms, but this has not stopped them from developing theories about the properties of atoms and other invisible particles, and these theories have led to discoveries that have transformed our world. Similarly cognitive psychologists believe that an understanding of how the lie6674X_01_c01_001-036.indd 19 3/14/12 4:18 PM Section 1.4 The Use of Animals CHAPTER 1 mind works would inevitably lead to better methods of education, more effective treatments for people with problems, and so on. In essence, the disagreement between Skinner’s approach and that of cognitive psychologists concerns the desirability of theories about the mind. Both sides agree that the primary data of psychology must be objective observations of behavior. They disagree, however, about whether there is a useful role for theories about the mind. Skinnerians argue that because behavior is ultimately determined by environmental conditions, our effort should go into studying the effect of these conditions on behavior. Cognitive psychologists argue that because cognitive processes form a critical part of the causal chain that controls behavior, a deeper understanding of these processes will inevitably lead to practical applications. The Current Approach The approach taken in this text can be seen as a blend of the Skinnerian and cognitive approaches. We think Skinner was right to emphasize the importance of environmental conditions in determining behavior. His emphasis on the practical application of learning principles played a crucial role in encouraging applied research, and we will devote considerable attention to the applications that resulted. We think that Skinner was also right to recognize the potential hazards in speculating about the mind—it is all too easy to attribute behavior to invisible mental states, without any evidence that these states really exist. However, we also agree with cognitive psychologists that these dangers can be avoided if theories are stated clearly enough that they lead to testable predictions, and that under these circumstances they can substantially enrich our understanding of behavior. Suppose, for example, that a theorist proposed that reading involves three cognitive processes. If there were no way to test the truth of this claim, the theory would be useless. On the other hand, if the theory led to testable predictions, and these were confirmed, then our understanding of these processes might allow us to identify which process is impaired in different individuals, and thereby help us to develop treatments tailored to their individual problems. 1.4 The Use of Animals Having decided to study the laws of behavior, and to do so through careful experimentation, we now come to the question of what species to study. If our goal is to understand human behavior, the answer might seem obvious: We should study humans. Given the clarity of this logic, why have psychologists sometimes studied animals instead? The Advantages of Using Animals The reasons that psychologists study animals are complex, but all are rooted in the problems of experimental control discussed earlier. We said then that one crucial problem lie6674X_01_c01_001-036.indd 20 3/14/12 4:18 PM Section 1.4 The Use of Animals in psychological research is to manipulate only one independent variable at a time while holding all others constant. One way to do this is to exert stringent control over the environment, so as to minimize how many variables change during an experiment. Another, less obvious, is to study an organism with a simpler nervous system, so that fewer variables are likely to affect it and it is thus easier to manipulate just one. Each of these strategies is easier to implement in research involving animals, and we will consider each in turn. CHAPTER 1 There are several advantages to using animals in psychological experiments. The use of animals limits the number of variables and allows psychologists to study an organism with a simpler nervous system. Control of the Environment For both moral and practical reasons, it is much easier to control an animal’s environment than it is to control a human’s. For example, one problem of considerable importance in human behavior concerns the effects of a child’s early environment on her or his development. Freudians have long argued that the first years of life are crucial in determining personality. More recently, educators have suggested that early sensory and social deprivation is an important factor in the poor school performance of some children (particularly from underprivileged homes) and have urged governments to invest in compensatory child-care programs for young children. How are we to determine whether the role of early experience is really so crucial and, if so, which aspects are most important? To determine the importance of early sensory experience, should we run controlled experiments in which half the children are reared normally while the other half are permanently confined to a barren environment, devoid of all stimuli? Similarly, to determine the importance of a mother’s role in a child’s normal development, should we compare children reared with their mothers and children reared in isolation? Such experiments would hardly be humane or practical. The questions involved are significant, with serious implications for the future structure of our schools and even our families, but the experiments necessary to answer such questions are clearly unacceptable. Using animals as subjects, however, psychologists have conducted experiments to answer these questions, with often fascinating results. Harry Harlow, for example, reported a series of experiments with infant rhesus monkeys. When taken away from their mothers immediately after birth and reared in isolation, these infants became highly neurotic: They spent much of their time huddled in corners, rocking back and forth and sucking their thumbs. Furthermore, this pattern of disturbed behavior persisted into adulthood, and most of the isolated monkeys were unable to function normally in a group, or even to mate. lie6674X_01_c01_001-036.indd 21 3/14/12 4:18 PM Section 1.4 The Use of Animals CHAPTER 1 These studies supported the critical role of early experience in social development, and in later experiments Harlow and others isolated some important variables. The presence of the mother, for example, is not necessarily critical; infants taken away from their mothers but reared with other infants showed significantly less disturbance (Harlow & Harlow, 1965). Another finding, with poignant social implications for men, is that male rhesus monkeys, which normally play an insignificant role in child rearing, can, if necessary, replace the mother with no apparent ill effects to either child or father (Mitchell & Brandt, 1972). A similar line of experiments has examined the role of early sensory experience in the development of rats. Some rats were reared in “enriched” environments that included other rats and a variety of toys, platforms, colors, and sounds; other rats were reared in “deprived” environments The use of rhesus monkeys allowed Harry that lacked these stimuli. Animals reared in the Harlow to demonstrate the importance enriched environments developed larger brains of parental attachment to an organism’s (Rosenzweig, 1984), with considerably more mental and physical well-being. complex interconnections among their neurons (Turner & Greenough, 1985). These results suggest that early stimulation plays a critical role in the brain’s development and thus in our capacity for learning in later life. Simpler Systems One advantage of using animals as subjects, then, is that we can more easily control their environments and thus determine which variables are important. A related advantage is that it is easier to identify fundamental principles when we study simpler systems. Suppose, for example, that you wanted to understand the principles of electronics. You would find it easier to understand these principles if you first studied a transistor radio rather than a mainframe computer: The simpler the system, the easier it is to understand its operations. Thus, scientists were able to isolate the fundamental principles of genetics by first studying two less complex life-forms—the fruit fly and the pea—that possess simpler systems. Both of these organisms rely on fewer genes for development and function, and thus scientists could isolate the effects of these genes more easily. If scientists had first tried to understand the principles of genetics in a more complex system—the inheritance of intelligence in humans, for example, is almost certainly influenced by many thousands of genes—we would probably still have little or no understanding of the principles of genetics. The less complex the system, then, the easier it is to determine its fundamental principles. Determining the principles of behavior in animals, however, can help us to understand human behavior only if these principles are similar. Is this assumption justified? Are the principles of animal and human learning sufficiently similar that understanding animals can help us to better understand humans? lie6674X_01_c01_001-036.indd 22 3/14/12 4:18 PM Section 1.4 The Use of Animals CHAPTER 1 Are Animal and Human Behavior Similar? Since the 19th century, when Charles Darwin first explained the mechanism by which human and animal species evolved from common ancestors and were shaped by the same environmental forces, animals have been found to have important physiological similarities to humans. Indeed, despite the incredible diversity of animal species (there are now thought to be more than three million species, ranging in size from virtually invisible microorganisms to the mammoth blue whale), the underlying biological principles are surprisingly similar. Our understanding of human neurophysiology, for example, is built largely on the pioneering work of Hodgkin and Huxley on the giant squid. Similarly, our understanding of human vision is based on Hartline and Ratliff’s investigations of the eye of the horseshoe crab, a primitive species almost unchanged from primordial times. When we begin to examine species more closely related to humans, the similarities become even greater. The basic principles of digestion, vision, respiration, locomotion, and so forth are, for all practical purposes, identical across the various mammalian species, and it is because of this fundamental similarity that modern medicine has been able to advance so quickly. The drugs and surgical techniques on which our lives now depend were generally pioneered not with people, but with mice, monkeys, and the famous guinea pig. For post-Darwin scientists, animals and people were clearly similar, at least in physical construction. Behaviorally, on the other hand, this similarity was less obvious. Even if humans had once been apes, the argument went, they had long since begun a unique evolutionary path that left them the only animal capable of thought and symbolic communication. In recent years, however, evidence has accumulated that human beings are not unique even in these areas. Washoe The most striking evidence has come from research on language. Because chimpanzees are our closest relatives, many psychologists believed that if any animal could master the rudiments of language, it would be chimpanzees. Early attempts to teach chimpanzees to speak, however, met with little success. Hayes and Hayes (1951), for example, reared a chimpanzee named Vicki in their home, but after four years of effort Vicki had learned a grand total of only four words: mama, papa, cup, and up. Subsequent research on the anatomy of the chimpanzee vocal tract revealed that they are not physically capable of producing the full range of sounds required for speech. Vicki’s failure could have been In 1951, Hayes and Hayes attempted to teach a chimpanzee to because of this physical limita- speak, but the chimpanzee only learned a total of four words. tion rather than any deficiency Further research concluded that a chimpanzee may learn a in her intellectual capacity. language that does not require speech: American Sign Language. lie6674X_01_c01_001-036.indd 23 3/14/12 4:18 PM Section 1.4 The Use of Animals CHAPTER 1 To test this hypothesis, Allen and Beatrice Gardner set out to teach a chimpanzee to use a language that did not require speech—American Sign Language for the deaf. The subject for their study was a baby chimpanzee named Washoe, and the results were dramatic. By the time she was five, Washoe had learned more than 130 signs and was able to use them reliably in a variety of situations. The sign for dog, for example, was elicited by a wide variety of dogs, both living and in pictures, and even by the barking of a dog that could not be seen. Washoe also demonstrated some ability to combine signs; when she wanted a refrigerator opened, for example, she signed “open food drink” (Gardner, Gardner, & Van Cantfort, 1989). When the Gardners’ research was published, it provoked intense controversy (see, for example, Terrace, 1985; Pinker, 1994). To some degree, this was because of genuine problems in the methodologies used, but in some cases it probably also reflected disbelief that any animal was capable of language, a skill that for so long had been assumed to be uniquely human. At the heart of the controversy was whether the chimpanzees really understood the signs that they were using. If Washoe was hungry and made the sign for banana, did this mean that she understood what this sign meant, or was she simply repeating a movement that had been rewarded with food in the past? A chimpanzee making a sign might be behaving no more intelligently than a rat pressing a bar—both might simply be repeating behavior that had previously produced food. The key issue was what linguists call semanticity, whether when a word is used it is evoking some sort of mental representation of the named object. In our banana example, when Washoe saw this sign, did it evoke some representation of a banana in her brain? Because we cannot observe animals’ mental states, there will probably always be some level of doubt. Several studies published since the Gardners’ work, however, support the claim that chimpanzees understand the signs that they use. We will look at two examples. Our first, rather poignant example involves Washoe. After her period of active training ended, she became a mother at the age of 15. Her baby was ill at birth, and Washoe had to be anesthetized so that the infant could be removed for treatment. He recovered and was returned to her, but several weeks later he again became ill, so that a pediatrician again needed to anesthetize her. When she saw the needle, she began to scream and sign “My baby, my baby.” Sadly, the infant died. When Washoe saw her trainer the next day, her first sign was “Baby?” The trainer replied by signing “Baby gone, baby finished.” Washoe’s response was dramatic: Washoe dropped her arms that had been cradled in the baby sign position . . . broke eye contact and slowly moved away to a corner of the cage . . . She continued for the next several days to isolate herself from any interactions with the humans and her signing dropped off to almost nothing. Her eyes appeared to be vacant or distant. (Fouts, Hirsch, & Fouts, 1982, p. 170) This account is anecdotal and therefore must be treated with caution, but it is difficult to read it without feeling that Washoe had some understanding of the meaning of the signs that were used. (For a more formal test of understanding, see Savage-Rumbaugh, Rumbaugh, Smith, & Lawson, 1980.) lie6674X_01_c01_001-036.indd 24 3/14/12 4:18 PM Section 1.4 The Use of Animals CHAPTER 1 Kanzi An alternative strategy for bypassing the limitations of chimpanzee’s vocal cords was developed by Duane Rumbaugh of Georgia State University, and later continued in collaboration with his wife, Sue Savage-Rumbaugh. They, too, believed that it was a mistake to try to teach chimpanzees to speak, but instead of using sign language, they developed a new language using geometrical shapes (lexigrams) as words. The lexigrams were displayed on a keyboard linked to a computer, and subjects could choose words by pressing the appropriate symbol on the board. The chimpanzees trained in this program soon showed performances very similar to those of Washoe. One of the participants, a female named Lana, developed an intriguing ability to create novel word combinations. Some of the foods that she ate were not assigned lexigrams by the experimenters, and Lana therefore invented her own names to request them. When she wanted a cucumber, for example, she asked for “banana which-is green,” and she requested an orange by using the lexigrams for “apple which-is orange (color)” (Rumbaugh & Savage-Rumbaugh, 1994). Another participant was a bonobo chimpanzee named Kanzi. (Bonobos are one of two chimpanzee species.) Kanzi’s mother was one of the early participants in the program, but she proved to be a very slow learner and made little progress. Though Kanzi was present during his mother’s training sessions, the experimenters made no effort to train him. Nevertheless, when Kanzi was 2 years old, the experimenters discovered that he understood the meaning of the lexigrams that the experimenters had tried and failed to teach his mother. Simply by watching this training, he seemed to have worked out for himself what the symbols meant. The experimenters then initiated an active training program for Kanzi, and by the time he was 5½, his lexigram vocabulary had increased to 149 words. At this point, Kanzi astonished the experimenters for a second time when they realized that he had also learned to understand human speech. Again, simply by listening to the conversations of his Kanzi was an example of a chimpanzee trainers as they taught him to use the lexigrams, who astonished human researchers with Kanzi had learned the meaning of a number of his ability to understand human speech. English words and phrases. In one test of his abilities, he was placed in a room containing 12 objects and given verbal instructions about what to do with these objects. (The experimenter was located in an adjacent room behind a one-way mirror, to avoid inadvertently providing Kanzi with cues through gestures.) One of the instructions, for example, concerned a sponge ball that had eyes, a nose, and a mouth; Kanzi was told “Feed your ball some tomato.” Even though Kanzi had never been asked to do anything remotely like this, he immediately picked up the ball and tried to place a tomato in its mouth. To provide a baseline for comparison, Alia, the 2½-year-old lie6674X_01_c01_001-036.indd 25 3/14/12 4:18 PM Section 1.4 The Use of Animals CHAPTER 1 daughter of one of Kanzi’s caretakers, was also tested with the same set of instructions; Kanzi responded correctly to 74%, Alia to 66% (Savage-Rumbaugh, et al., 1993; SavageRumbaugh, Rumbaugh, & Fields, 2009). More than 100 years ago, Charles Darwin wrote that “The difference in mind between man and the higher animals, great as it is, certainly is one of degree and not of kind” (Darwin, 1871/1920, p. 128). The evidence on this point is not yet conclusive, but it is already clear that animals are far more intelligent than once believed. This does not mean that animals and humans are identical. Every species is unique, and it would be foolish to expect to gain a complete understanding of people from the study of pigeons or white rats. On the other hand, given that animals and humans have shared millions of years of evolution, it would be surprising if there were not also similarities. Just as research with fruit flies and the pea made possible the extraordinary advances in genetics in the last century, so research on animals might help us in understanding our own behavior. Ethical Issues Because it is possible to exert much greater control of the environment in experiments on animals, such research has the potential to significantly enhance our understanding of basic processes. On the other hand, the very similarity of animal and human behavior that makes research on animals attractive also raises serious ethical issues. If animals are similar to us in intelligence, and presumably also in feelings, how can we justify confining them in cages and, in some cases, subjecting them to painful stimuli such as electric shocks? One view is that such research cannot be justified, because animals are living creatures that have just as much right to life and freedom as humans. This position is attractive in its strong value for all life, but few people hold it in its pure form. Suppose, for example, that you had a child who contracted rabies, and that the only way to save the child’s life required killing a mouse. Would you do it? Very few people faced with this dilemma would not choose to save the child, implicitly valuing a child’s life more than that of a mouse. Rightly or wrongly, then, most people do value human welfare more than that of animals, but this does not imply that animal life is worthless. Thus, the problem remains of deciding whether the benefits of particular experiments with animals outweigh the cost to the animals. To assess this, we need some method of quantifying both the benefits and the costs; in practice, though, this is difficult if not impossible. Suppose, for example, that we wanted to assess the cost to the subjects of an experiment on the effects of punishment. How could we decide how much pain a rat would experience if it were given an electric shock? What if we substituted a fish or a cockroach as the experimental subject? Do they also feel pain? If so, is it more or less than that experienced by the rat? If it is difficult to find an objective way to assess the costs of animal research, it can be equally difficult to assess its benefits. In our hypothetical rabies example, we assumed that killing the mouse would save the life of the child, but the benefits of experiments are rarely this predictable. Experiments that seem minor when they are performed can eventually have momentous theoretical and practical benefits. In a study by Comroe and Dripps (1977), for example, physicians were asked to rate the ten most important advances lie6674X_01_c01_001-036.indd 26 3/14/12 4:18 PM CHAPTER 1 Section 1.5 Learning: An Overview in cardiovascular and pulmonary medicine and surgery. A total of 663 studies were found to have been crucial in leading to these breakthroughs; 42% of them involved experiments that, at the time they were reported, seemed unrelated to the later clinical application. When doing basic research, it is difficult to predict what benefits might eventually be derived from enhanced understanding of a fundamental mechanism. In deciding whether a planned experiment is justifiable, then, it is difficult to assess either the costs to the animals used or the long-term benefits to humans. There are no simple guidelines; all we can say here is that an assessment of the benefits depends heavily on the validity of the assumptions discussed in this chapter. If behavior is lawful, if experimental research is the best way to discover these laws, and if animal and human behavior is similar in important respects, then research on animals might play an important role in increasing our understanding of human behavior. 1.5 Learning: An Overview Summarizing our discussion until this point, we have suggested that psychological research is based on several assumptions: that behavior is lawful, that the best way to discover these laws is through controlled experiments, and that research on animals can sometimes help us understand human behavior. Before proceeding to examine the research that has resulted from these assumptions, though, we need to address one final question—namely, what this book is about. Of course, you already know that it is about learning, but in this section we will examine more closely what we mean by this term. Learning Learning is a vast topic. It affects almost everything we do, from making friends to riding a bicycle to learning organic chemistry. As a result, it is impossible to cover every aspect of learning in a single course, and it has become customary to study different aspects in different courses: Courses on developmental psychology deal with one aspect, courses on educational psychology with another, courses on cognition a third, and so on. Within this division, courses on learning generally concentrate on a particular form of learning called associative learning. To explain what associative learning is, we will begin by examining what we mean by the broader term learning. Some stimuli always elicit the same reaction. If you accidentally touch a hot pan, for example, it will make you pull your hand back every time; if a sudden gust of wind hits you in the eye, it will make you blink every time. In cases like this, in which a stimulus always elicits the same response, we call the stimulus-response relationship a reflex. We can represent the way a reflex works in the following way, where S = stimulus, and R = reflex: Reflex: S R Our definition of a reflex requires that the stimulus always elicit the same response, but under some conditions the strength of a reflexive response can change with experience. lie6674X_01_c01_001-036.indd 27 3/14/12 4:18 PM CHAPTER 1 Section 1.5 Learning: An Overview Suppose, for example, that you were quietly studying in your room one day when you suddenly heard a deafening noise—a particularly loud burglar alarm in the building next door to you had just gone off. This sharp noise would almost certainly make you jump, a reaction that is known as the startle reflex. The first time it happened, your reaction would be very intense, but if it happened again five minutes later, you would probably react less strongly; a third repetition would produce even less of a reaction, and so on. This decrease in the strength of a reflex when the stimulus is repeated a number of times is called habituation; it is a common characteristic of reflexes, especially when the stimulus is repeated within a relatively short period of time. One experiment illustrating habituation was reported by Davis (1974). He placed rats in a cage that was mounted on springs, so that if they made a sudden movement he could measure the magnitude of this movement by measuring the movement of the floor. He then presented a loud tone to the rats a number of times. As shown in Figure 1.3, he found that the tone initially produced a strong startle response, but that the magnitude of this response decreased over successive presentations. Figure 1.3: Magnitude of the startle response Mean Startle Amplitude 40 30 20 10 0 0 2 4 6 8 10 Blocks of 10 Tones As evidenced by Davis in 1974, the startle response diminishes with repeated presentations of a tone. lie6674X_01_c01_001-036.indd 28 3/14/12 4:18 PM CHAPTER 1 Section 1.5 Learning: An Overview Why did the rats’ reactions to the tone habituate when it was presented repeatedly? One possibility is sensory fatigue—frequent presentations of the tone could have impaired the capacity of the sensory system to react. A closely related possibility is motor fatigue—as the rats responded to the tone on successive trials, they could have become progressively more tired and thus physically less able to respond. Although sensory and motor fatigue undoubtedly can occur, habituation is rarely caused by such fatigue. If, at the conclusion of his experiment, Davis had presented a test trial in which he had increased the intensity of the tone, he would almost certainly have observed an increase in the vigor of the startle response back to its original level. If so, the rats’ sensory and motor systems would clearly have been capable of producing a response, and the weakened responding on earlier trials could not have been caused by fatigue. Defining Learning Habituation, then, is not caused by fatigue in either the senses or the muscles; by elimination, it seems to involve some sort of change in the nervous system that links them. Specifically, habituation seems to involve learning that a potentially dangerous stimulus is not, in fact, dangerous, and thus can be safely ignored. Changes in behavior of this kind illustrate what we mean by learning. It is difficult to define learning precisely, but one simple definition would be a change in behavior due to experience. As sometimes happens with simple definitions, however, this one quickly presents a challenge. One problem, as we have already seen, is that there are some changes caused by experience that are really not what we mean by learning. If your behavior changed because you had not eaten for several hours, for example, that would hardly be an example of learning. Intuitively, what we mean by learning is experiences that result in the storage of information in our brains, information that alters our capacity to respond in the future. If you were taught to ride a bicycle, this would be an example of learning whether or not you later chose to use this skill. To capture the meaning of learning more precisely, we will redefine the term as a change in our capacity for behavior, as a result of particular kinds of experience. This definition is regrettably more cumbersome, but it comes closer to what we really mean when we talk about learning. Associative Learning In the case of habituation, learning occurs as a result of the presentation of a single stimulus (however, see Whitlow & Wagner, 1984). A more elaborate form of learning occurs when two events occur together and we learn about the relationship between them. If we use the symbol E1 to represent one event and E2 to represent the second event, then in associative learning we learn about the association or relationship between the two events: E1 lie6674X_01_c01_001-036.indd 29 E2 3/14/12 4:18 PM Section 1.5 Learning: An Overview CHAPTER 1 The two events could potentially be anything: a drop in air pressure warning of a storm to come; a television theme tune announcing the start of a TV show; a tone of voice signaling annoyance. Learning psychologists, however, have been particularly interested in instances of associative learning where the second event is biologically important—food, say, or bodily injury—and survival might depend on being able to predict this event. Suppose that a lion always visits a watering hole at 4:00 in the afternoon; if antelopes that lion), this also use this water could learn this stimulus-stimulus relationship (4:00 P.M. would allow them to avoid the area at this time and thereby prolong their lives. Or consider a related situation from the lion’s point of view: Suppose that whenever it stalks an antelope while remaining downwind of it, it is more likely to succeed. If it could learn this succulent antelope), then it too response-stimulus relationship (downwind stalking would enjoy a longer, more satisfying life. Classical Conditioning In those cases in which an important event is reliably preceded by a stimulus, the stimulus often comes to elicit the same behavior as the event it predicts. If the presentation of a light is repeatedly followed by a puff of air to the eye, for example, then the light on its own would eventually begin to elicit a blink. This is an example of classical or Pavlovian conditioning. Classical conditioning allows us to prepare for forthcoming events; in our eyeblink example, if we blink before the puff arrives, the lid closure can prevent particles from being blown into our eyes. Operant Conditioning When an important event follows a response rather than a stimulus, the result is often a change in the response’s probability, and this is called instrumental or operant conditioning. If your parents gave you a sports car every time you received an A for a course, you would probably increase the amount of time you spent studying. This example illustrates one of the two subtypes of operant conditioning—reinforcement and punishment, that differ in whether the change in response is an increase or a decrease. In reinforcement, the consequence that follows a response is desirable and the effect is to strengthen it—the use of a reward to increase studying, for example. In punishment, on the other hand, the consequence is undesirable and the effect is to weaken the response. Children who burn their hands when touching a hot pan quickly learn not to repeat this behavior. As summarized in Figure 1.4, the essential distinction between classical and operant conair ditioning lies in whether an important event follows a stimulus (for example, light burn). As we shall see, both forms of puff) or a response (for example, touching pan conditioning play a major role in shaping our lives. This might not be obvious for classical conditioning because classical conditioning often occurs without our awareness (see Chapter 4). Also, the best-known conditioned responses are salivation and blinking, neither of which would probably make a “top 10” list of critical skills. However, classical conditioning also affects far more important aspects of our behavior, including emotions such as fear and sexual arousal, what foods we like, and the effects of drugs such as heroin and alcohol. Learning psychologists have been able to use an understanding of the processes involved to develop therapies for problems such as phobias, alcoholism, and bed-wetting. We will look at the principles of conditioning, and how they can be practically applied, in Chapters 2 and 3. We will also examine theories of conditioning in Chapter 5. lie6674X_01_c01_001-036.indd 30 3/14/12 4:18 PM Section 1.5 Learning: An Overview CHAPTER 1 Figure 1.4: Varieties of associative learning Associative learning involves the detection of relationships between events (E), where the events concerned can be responses (R) or stimuli (S), and the stimuli can be positive (SPOS) or negative (SNEG). The importance of reinforcement and punishment is probably more obvious, but even here we tend to underestimate significance. As in the case of classical conditioning, this is partly because we are not always aware of effects. Attention from others, for example, can be very reinforcing, and when parents and teachers pay attention to a child who is misbehaving they sometimes inadvertently reinforce the behavior they are trying to eliminate. Also, reinforcement and punishment sometimes appear ineffective because they are not used optimally. Improved understanding of the principles involved has allowed psychologists to develop techniques to reduce children’s misbehavior, to teach convicts to master a year’s worth of school in only a month, and to help autistic children to lead normal lives. We will look at the principles of reinforcement in Chapter 5 and their application in Chapter 6. lie6674X_01_c01_001-036.indd 31 3/14/12 4:18 PM Review Questions CHAPTER 1 Summary and Review In this chapter we reviewed some of the assumptions that underlie the research that we will discuss in later chapters. • • • Human behavior is lawful, as our environment and heredity jointly determine how we behave. The position that we have labeled neural determinism argues that behavior must be lawful. We also looked at examples in which environment and heredity strongly influence us, and evidence that such influence can occur without our awareness. Introspection is of limited value in helping us to understand people’s behavior. Experiments allow us to identify which aspects of our environment affect our behavior, but progress is often constrained by the need to manipulate only one aspect of the environment at a time. Researchers can more easily control the environment when designing and conducting experiments on animals, which makes it easier to identify the processes involved. An understanding of how animals learn has the potential to help us understand learning in humans. However, this similarity also raises ethical issues about the use of animals in experiments. Remember that these are all assumptions—it is important to understand the reasoning behind them, but you don’t have to accept them. • • We also introduced the term learning, which we defined as a change in our capacity for behavior as a result of particular kinds of experience. Our discussion of learning will focus on associative learning, which involves learning about the relationship between two events. In classical conditioning, we learn about the relationship between two stimuli; in operant conditioning (reinforcement and punishment), we learn about the relationship between a response and its consequence. Review Questions Research on memory has shown that one of the most effective techniques for remembering material you are studying is to review after reading it, both immediately and then again after a delay (e.g., Karpicke & Roediger, 2010). Reviews of this kind can be hard work, and the effort involved often seems unnecessary because immediately after reading a chapter it is still fresh in our memory, giving us the impression that we really know it well. Unfortunately, this impression can be seriously misleading—while the material may still be present in our temporary or short-term memory store, it may not yet have been transferred to our more permanent, long-term store. One of the best ways to ensure that you really will remember is to pause after each section that you read and try to recall it without looking back at the text, and then to review it again when you finish the chapter. Such reviews can help you to identify material which you didn’t understand as well as you thought, and also, make it much, much easier to retrieve the material later—for instance, to use a wildly hypothetical example, in an exam. lie6674X_01_c01_001-036.indd 32 3/14/12 4:18 PM Concept Check CHAPTER 1 You can use the review questions provided at the end of each chapter to help you in this. Try to answer the questions without looking back at the text, and then reread any material you had difficulty remembering. The more often you practice retrieving material from memory—for example, quizzing yourself the next day while eating lunch or having a snack—the more likely you are to remember that material in the long-term. Here are some review questions for this first chapter: 1. Is human behavior lawful? What are the arguments for and against this view? 2. In what ways do the views of Skinner and of cognitive psychologists differ? In what respects are they the same? 3. How do experiments control for unwanted variables? 4. What are the strengths and weaknesses of the experimental method? 5. Why do psychologists believe that the results of experiments carried out in highly artificial laboratory settings can tell us something about behavior in the real world? 6. What are the arguments for and against the use of animals in psychological research? 7. How does the text define learning? What are the main types of learning discussed and how is each defined? Concept Check 1. In experimental research, animals are frequently used as subjects because a. b. c. d. animals have a higher tolerance for pain in a controlled environment. animal behaviors are easier to manipulate in an open environment. animal behaviors are easier to manipulate in a controlled environment. animals have a lower tolerance for pain in an open environment. 2. Milgram’s research on obedience yielded unexpected results. He found a. 65% of the subjects continued to administer shocks regardless of the recipients’ level of pain. b. 65% of the subjects ceased to administer shocks due to the recipients’ level of pain. c. 35% of the subjects ceased to administer shocks due to the recipients’ level of pain. d. 35% of the subjects continued to administer shocks regardless of the recipients’ level of pain. 3. Kasper and Alford (1988) studied 125 men who had sexually abused children. Of these men, 85% were abused as children. Data from this study indicates that a. 85% of child sexual abusers are men. b. children who are sexually abused will likely abuse children when they are adults. c. 15% of the men did not sexually abuse children. d. 85% of men who sexually abuse children will not abuse their own children. lie6674X_01_c01_001-036.indd 33 3/14/12 4:18 PM CHAPTER 1 Key Terms 4. Harlow’s research with monkeys supported the notion that the environment plays a critical role in cognitive development. Newborn monkeys removed from their mothers and placed in a sterile environment exhibited a. b. c. d. normal social behavior and cognitive development. abnormal social behavior and normal cognitive development. abnormal social and cognitive development. normal social development and abnormal cognitive development. 5. Research using animals has ethical implications, such as weighing the benefits of the findings to the cost to the animal. The primary reason animals are used in research is a. b. c. d. animals are like humans. animals feel less pain. animals adapt easily to diverse environments. animal’s environment can be controlled. Answers: 1) c, 2) a, 3) b, 4) c, 5) d Key Terms associative learning A more elaborate form of learning that occurs when two events occur together and we learn about the relationship between them. behaviorism The view that psychology should focus on visible behavior rather than mental states. classical (Pavlovian) conditioning An increase in responding to a stimulus because of pairings of that stimulus with an important event such as food. cognitive psychology A branch of psychology that tries to understand the processes involved in thinking. Because so many of the processes occur at an unconscious level, cognitive psychologists use experiments to infer the nature of these processes, rather than trying to observe them directly. dependent variable The observable behavior that is measured during an experiment, to see if it is affected by the manipulation of the independent variable. lie6674X_01_c01_001-036.indd 34 determinism The view that all behavior is caused by either environmental or genetic factors. experimentation A method for uncovering the causes of behavior by changing one aspect of the environment (the independent variable) and observing its effect on some other aspect of behavior (the dependent variable). free will The belief that people have the power to determine their own actions, regardless of any external pressures. habituation The decrease in the strength of a reflex when a stimulus is repeated a number of times. independent variable The aspect of the environment that an experimenter changes during an experiment. introspection A person’s examination of his or her own thoughts or feelings. 3/14/12 4:18 PM Key Terms CHAPTER 1 law In science, a consistent relationship between independent and dependent variables such that the occurrence of some set of conditions A always leads to outcome B. operant (instrumental) conditioning A change in the probability of a behavior due to its having been followed by an important event. learning A change in the capacity for behavior due to particular kinds of experience. punishment A form of operant conditioning in which the likelihood of a behavior is reduced because it produces an aversive consequence. neural determinism The argument that the brain controls behavior, and that the physical and chemical processes involved in the transmission of neural signals are lawful; hence behavior must also be lawful. neurons Cells whose function is to transmit electrical signals. lie6674X_01_c01_001-036.indd 35 reinforcement A form of operant conditioning in which a behavior is strengthened because it produces a positive consequence. reflex A stimulus-response relationship in which a stimulus reliably elicits the same response innately, without prior experience. 4/9/12 8:09 AM lie6674X_01_c01_001-036.indd 36 3/14/12 4:18 PM 2 Foundations of Classical Conditioning Learning Objectives After reading this chapter, you should be able to do the following: • Understand the fundamental contribution made by the Associationist tradition to a theory of human behavior based on associations. • Identify and describe the main principles of association: frequency, intensity, and contiguity. • Describe Ivan Pavlov’s research with salivary conditioning in dogs...
Purchase answer to see full attachment