Learning refers to the acquisition, maintenance, and change of an organism's behavior as a result of lifetime events. The behavior of an organism is everything it does, including private and covert actions like thinking and feeling (see section on assumptions in this chapter). An important aspect of human learning concerns the experiences arranged by other people. From earliest history, people have acted to influence the behavior of other individuals. Rational argument, rewards, bribes, threats, and force are used in attempts to promote learning or change the behavior of people. Within society, people are required to learn socially appropriate ways of doing things. As long as a person does what is expected, no one pays much attention. As soon as a person's conduct substantially departs from cultural norms, other people get upset and try to force conformity. All societies have codes of conduct and laws that their people have to learn; people who break moral codes or civil laws face penalties ranging from minor fines to capital punishment. Clearly, all cultures are concerned with human learning and the regulation of human conduct.
Theories of learning and behavior have ranged from philosophy to natural science. When Socrates was told that new discoveries in anatomy proved that bodily movement was caused by the arrangement of muscles, bones, and joints, he replied, “That hardly explains why I am sitting here in a curved position talking to you” (Millenson, 1967, p. 3). About 2300 years later, the philosopher Alfred North Whitehead asked the famous behaviorist B. F. Skinner a similar question. He said, “Let me see you account for my behavior as I sit here saying, ‘No black scorpion is falling upon this table’” (Skinner, 1957, p. 457). Although there was no satisfactory account of behavior in the time of Socrates, the science of behavior is currently addressing such puzzling questions.
Human behavior has been attributed to a great variety of causes. The causes of behavior have been located both within and outside of people. Internal causes have ranged from metaphysical entities like the soul to hypothetical structures of the nervous system. Suggested external causes of behavior have included the effect of the moon and tides, the arrangement of stars, and the whims of gods. Some of these theories of behavior remain popular today. For example, the use of astrological forecasts is even found in modern corporations, as demonstrated in the following passage taken from The Economist:
Is astrology the ultimate key to competitive advantage? That is what Divinitel, a French company specializing in celestial consulting, claims. For FFr350 ($70) a session, the firm's astrologers offer advice on anything from the timing of takeovers to exorcisms. For the busy executive, Divinitel's galaxy of services can be reached via Minitel, France's teletext system. The firm is even planning a flotation on France's over-the-counter stock market in March 1991.
So who is daft enough to pay for such mystical mumbo-jumbo? About 10% of French businesses are, according to a study by HEC, a French business school. A typical client is the boss of a small or medium-sized company who wants a second, astrological opinion on job applicants. The boss of one plastics company even uses Divinitel's advice on star signs to team up salesmen.
(“Twinkle, twinkle,” p. 91, January 1991)
The trouble with astrology and other such primitive and mystical accounts of behavior is that they are not scientific. That is, these theories do not hold up to testing by scientific methods. Over the last century, a scientific theory of learning and behavior has developed. Behavior theory states that all behavior is due to a complex interaction between genetic influence and environmental experience. The theory is based on observation and controlled experimentation, and it provides a natural-science account of the learning and behavior of organisms, including humans. This book is concerned with such an account.
The experimental analysis of behavior is a natural-science approach to understanding behavior regulation. Experimental analysis is concerned with controlling and changing the factors affecting the behavior of humans and other animals. For example, a behavioral researcher in a classroom may use a computer to arrange corrective feedback for a student's mathematical performance. The relevant condition that is manipulated or changed by the experimenter may involve presenting corrective feedback on some days and withholding it on others. In this case, the researcher would probably observe more accurate mathematical performance on days when feedback was presented. This simple experiment illustrates one of the most basic principles of behavior—the principle of reinforcement.
The principle of reinforcement (and other behavior principles) provides a scientific account of how people and animals learn complex actions. When a researcher identifies a basic principle that governs behavior, this is called an analysis of behavior. Thus, the experimental analysis of behavior involves specifying the basic processes and principles that regulate the behavior of organisms. Experiments are then used to test the adequacy of the analysis.
Experimental analysis occurs when, for example, a researcher notices that more seagulls fly around a shoreline when people are on the beach than when the beach is deserted. After checking that changes in climate, temperature, time of day, and other conditions do not affect the behavior of the seagulls, the researcher offers the following analysis: People feed the birds and this reinforces flocking to the beach. When the beach is abandoned, the seagulls are no longer fed for congregating on the shoreline. This is a reasonable guess, but it can only be tested by an experiment. Pretend that the behavior analyst owns the beach and has complete control over it. The experiment involves changing the usual relationship between the presence of people and food. Simply stated, people are not allowed to feed the birds, and food is placed on the beach when people are not around. Over time, the behavior analyst notes that there are fewer and fewer seagulls on the beach when people are present, and more and more gulls when the shoreline is deserted. The behaviorist concludes that people regulated coming to the beach because the birds were fed, or reinforced, for this behavior only when people were present. This is one example of an experimental analysis of behavior.
Although experimental analysis is the fundamental method for a science of behavior, contemporary researchers prefer to describe their discipline as behavior analysis. This term implies a more general scientific approach that includes assumptions about how to study behavior, techniques to carry out the analysis, a systematic body of knowledge, and practical implications for society and culture (Ishaq, 1991).
Behavior analysis is a comprehensive approach to the study of the behavior of organisms. Primary objectives are the discovery of principles and laws that govern behavior, the extension of these principles across species, and the development of an applied technology for the management of behavior. In the seagull example, the underlying principle is called discrimination. The principle of discrimination states that an organism will respond differently to two situations (e.g., presence or absence of people) if its behavior is reinforced in one setting but not in the other.
The principle of discrimination may be extended to human behavior and social reinforcement. You may discuss dating with Carmen, but not Tracey, because Carmen is interested in such conversation while Tracey is not. In a classroom, the principle of discrimination can be used to improve teaching and learning. The use of behavior principles to solve practical problems is called applied behavior analysis and is discussed at some length in Chapter 13.
As you can see, behavior analysis has a strong focus on environment—behavior relationships. The focus is on how organisms alter their behavior to meet the ever-changing demands of the environment. When an organism learns new ways of behaving in reaction to the changes that occur in its environment, this is called conditioning. The two basic kinds of conditioning are called respondent and operant.
A reflex is behavior that is elicited by a biologically relevant stimulus. When a stimulus (S) automatically elicits (→) a stereotypical response (R), the S → R relationship is called a reflex. This behavior and the reflex relationship had survival value in the sense that those animals that quickly and reliably responded to particular stimuli were more likely than other organisms to survive and reproduce. To illustrate, animals that startle and run to a sudden noise may escape a predator, and the startle reflex may provide an adaptive advantage over organisms that do not run, or run less quickly to the noise. Thus, reflexes are selected across the history of the species. Of course, different species of organisms exhibit different sets of reflexes.
Respondent conditioning occurs when a neutral or meaningless stimulus is paired with an unconditioned stimulus. For example, the buzz of a bee (neutral stimulus) is paired with the pain of a sting (unconditioned stimulus). After this conditioning experience, a buzzing bee usually causes people to escape it. The Russian physiologist Ivan Petrovich Pavlov made explicit this form of conditioning at the turn of the 20th century. He observed that dogs salivated when food was placed in their mouths. This relation between the food stimulus and salivation is an unconditioned reflex, and it occurs because of the animals' biological history. However, when Pavlov rang a bell just before feeding the dogs, they began to salivate at the sound of the bell. In this way, a new feature (sound of the bell) came to control the dogs' respondent behavior (salivation). As shown in Figure 1.1, a respondent is behavior that is elicited by the new conditioned stimulus.
Respondent conditioning is one way that organisms meet the challenges of change in their environments. A grazing animal that becomes conditioned to the sound of rustling grass and runs away is less likely to become a meal than one that waits to see the predator. All species that have been tested, including humans, show this kind of conditioning. In terms of human behavior, many of what we call our likes and dislikes are based on respondent conditioning. When good or bad things happen to us, we usually have an emotional reaction. These emotional responses can be conditioned to other people who are present when the positive or negative events occur (Byrne, 1971). Thus, respondent conditioning plays an important role in our social relationships—determining, to a great extent, how we feel about our friends as well as our enemies.
FIG. 1.1 Simple respondent conditioning. In a reflex for a dog, food in the mouth produces salivation. Next, a bell rings (new stimulus) just before feeding the dog; after several pairings of bell and food the dog begins to salivate at the sound of the bell.
Operant conditioning involves the regulation of behavior by its consequences. B. F. Skinner called this kind of behavior regulation operant conditioning because, in a given situation or setting (SD), behavior (R) operates on the environment to produce effects or consequences (Sr). Any behavior that operates on the environment to produce an effect is called an operant. During operant conditioning, an organism emits behavior that produces an effect that increases (or decreases) the frequency of the operant in a given situation (Skinner, 1938, p. 20). In the laboratory, a hungry rat in a chamber may receive food if it presses a lever when a light is on. If lever pressing increases in the presence of the light, then operant conditioning has occurred (see Figure 1.2).
Most of what we commonly call voluntary, willful, or purposive action is analyzed as operant behavior. Operant conditioning occurs when a baby smiles to a human face and is picked up. If smiling to faces increases in frequency because of social attention, then smiling is an operant and the effect is a result of conditioning. In a more complex example, pressing a sequence of buttons while playing a video game will increase in frequency if this response pattern results in hitting a target. Other examples of operant behavior include driving a car, talking on the phone, taking notes in class, walking to the corner store, reading a book, writing a term paper, and conducting an experiment. In each case, the operant is said to be selected by its consequences.
FIG. 1.2 Simple operant conditioning. In an operant chamber, lever pressing produces food for a hungry rat. The consequences of lever pressing (presentation of food) increase its frequency in that setting. In another example, a baby smiles to a human face and is picked up. The consequence of smiling (social attention) increases the frequency of this behavior in the presence of human faces.
B. F. Skinner (1938) viewed psychology as the study of the behavior of organisms. From this point of view, psychology is a subfield of biology. The main organizing principle of contemporary biology is evolution through natural selection (Dawkins, 1996). Skinner generalized this concept to a broader principle of selection by consequences. Selection by consequences applies at three levels: (1) the selection of characteristics of a species (natural or Darwinian selection); (2) the selection of behavior within the lifetime of an individual organism (selection by operant conditioning); and (3) the selection of behavior patterns (practices, traditions, rituals) of groups of human beings that endure beyond the lifetime of a single individual (cultural selection). In all three cases, it is the consequences arranged by the environment that select for (or against) the frequency of genetic, behavioral, and cultural forms (see Chapter 14).
Selection by consequences is a form of causal explanation. In science we talk about two kinds of causation: immediate and remote. Immediate causation is the kind of mechanism studied by physics and chemistry: the “billiard ball” sort of process where we try to isolate a chain of events that directly result in some effect. For example, chemical reactions are explained by describing molecular interactions. In the study of behavior, an immediate causal explanation might refer to the physiology and biochemistry of the organism. For example, the bar pressing of a rat for food or a gambler playing roulette each could involve the release of endogenous opiates and dopamine in the hypothalamus (Shizgal & Arvanitogiannis, 2003).
In contrast, remote causation is typical of sciences like evolutionary biology, geology, and astronomy. In this case, we explain some phenomenon by pointing to remote events that made it likely. Thus, the causal explanation of a species characteristic (e.g., size, coloration, exceptional vision, etc.) involves the working of natural selection on the gene pool of the parent population. An evolutionary account of species coloration, for example, would involve showing how this characteristic improved the reproductive success of organisms in a given ecological environment. That is, natural selection for coloration explains the current frequency of the characteristic in the population.
On the behavioral level, the principle of selection by consequences is a form of explanation by remote causation called functional analysis. When a rat learns to press a lever for food, we explain the rat's behavior by pointing to its past consequences (the function of behavior). Thus, the current frequency of bar pressing is explained by the contingency between bar pressing and food in the past. The rat's behavior has been selected by its history of reinforcement.
Both immediate and remote causal explanations are acceptable in science. Behavior analysts have emphasized functional analysis and selection by consequences (remote causation), but are also interested in direct analysis of physiological and neurochemical processes (immediate causation). Ultimately, both types of causal explanation will provide a more complete account of learning and behavior.
Behavior analysis is becoming more involved with the scientific analysis of the brain and nervous system (neuroscience). That is, researchers who primarily study the behavior of organisms are often interested in the brain processes that participate in the regulation of behavior (see special issue on Relation of Behavior and Neuroscience (2005) in Journal of the Experimental Analysis of behavior, 84, 305–667). For example, brain mechanisms (neurons or groups of neurons) obviously participate in the regulation of behavior (bar pressing) by its consequences (food). Describing how neurons code for, and respond to, reinforcement is an important and exciting addition to a behavior analysis—telling us more about what happens when behavior is shaped by its effects or consequences (Fiorillo, Tobler, & Schultz, 2003).
At the practical level, knowing the reinforcement contingencies (arrangements of the environment) for lever pressing is sufficient to allow us to predict and control the rat's behavior. That is, we can get the rat to increase or decrease its lever pressing by providing or denying food reinforcement for this behavior—there is no need to look at neural systems. But, we gain a more complete account of how a rat's behavior increases when the action of neurons (and neural systems) is combined with the analysis of behavior. For example, in some cases, it may be possible to “sensitize” or “desensitize” a rat to the behavioral contingencies by drugs that activate or block the action of specialized neurons (e.g., Bratcher, Farmer-Dougan, Dougan, Heidenreich, & Garris, 2005). Research at the neural level could, in this way, add to the practical control or regulation of behavior by its consequences (Hollerman & Schultz, 1998).
Neural processes also may participate as immediate consequences (local contingencies) for behavior that had long-range benefits for organisms (distal contingencies, as in evolution and natural selection) (Tobler, Fiorillo, & Schultz, 2005). The so-called neural basis of reward involves the interrelationship of the endogenous opiate and dopamine systems (as well as other neural processes) (Fiorillo et al., 2003). For example, rats that are food restricted and allowed to run in wheels increase running over days—up to 20,000 wheel turns. Wheel running leads to the release of neural opiates that reinforce this behavior (Pierce, 2001). If wheel running is viewed as food-related travel, one function of neural reinforcement is to promote locomotion under conditions of food scarcity. The long-range or distal contingency (travel produces food: travel → food) is proximally supported by the release of endogenous opiates (physical activity → release of endogenous opiates) that “keep the rat going” under conditions of food scarcity (e.g., famine, drought, etc.). We shall see that this process also plays a key role in human anorexia in later chapters.
The integration of the science of behavior with neuroscience (behavioral neuroscience) is a growing field of inquiry. Areas of interest include the effects of drugs on behavior (behavioral pharmacology), neural imaging and complex stimulus relations, choice and neural activity, and the brain circuitry of learning and addiction. We will examine some of this research in subsequent chapters in sections that focus on behavior analysis and neuroscience (FOCUS ON sections) or in sections that emphasize applications (ON THE APPLIED SIDE sections).
When organisms were faced with unpredictable and changing environments in their evolutionary past, natural selection favored those individuals whose behavior could be conditioned. Organisms who condition are more flexible, in the sense that they can learn new requirements and relationships in the environment. Such behavioral flexibility must reflect an underlying structural change of the organism. Genes code for the anatomical and physiological characteristics of the individual. Such structural physical changes allow for different degrees of functional behavioral flexibility. Thus, differences in the structure of organisms based on genetic variation give rise to differences in the regulation of behavior. Processes of learning, like operant and respondent conditioning, lead to greater (or lesser) reproductive success. Presumably, those organisms that changed their behavior as a result of experience during their lives survived and had offspring—those that were less flexible did not. Simply stated, this means that the capacity for learning is inherited.
The evolution of learning processes had an important consequence. Behavior that was closely tied to survival and reproduction could be influenced by experience. Specific physiological processes typically regulate behavior related to survival and reproduction. However, for behaviorally flexible organisms, this control by physiology may be modified by experiences during the lifetime of the individual. The extent of such modification depends on the amount and scope of behavioral flexibility (Baum, 1983). For example, sexual behavior is closely tied to reproductive success and is regulated by distinct physiological processes. For many species, sexual behavior is rigidly controlled by genetically driven mechanisms. In humans, however, sexual behavior is also influenced by socially mediated experiences. It is these experiences, not genes, that come to dictate when sexual intercourse will occur, how it is performed, and who can be a sexual partner. Powerful religious or social controls can make people abstain from sex. This example illustrates that even the biologically relevant behavior of humans is partly determined by life experience.
Behavior analysts recognize and promote the importance of biology and evolution but focus more on the interplay of the organism and its environment. To maintain this focus, the evolutionary history and biological status of an organism are examined as part of the context of behavior (see Morris, 1988, 1992). This contextualist view is seen in B. F. Skinner's analysis of imprinting in a duckling:
Operant conditioning and natural selection are combined in the so-called imprinting of a newly hatched duckling. In its natural environment the young duckling moves towards its mother and follows her as she moves about. The behavior has obvious survival value. When no duck is present, the duckling behaves in much the same way with respect to other objects. Recently it has been shown that a young duckling will come to approach and follow any moving object, particularly if it is the same size as a duck—for example, a shoebox. Evidently survival is sufficiently well served even if the behavior is not under the control of the specific visual features of a duck. Merely approaching and following is enough.
Even so, that is not a correct statement of what happens. What the duckling inherits is the capacity to be reinforced by maintaining or reducing the distance between itself and a moving object [italics added]. In the natural environment, and in the laboratory in which imprinting is studied, approaching and following have these consequences, but the contingencies can be changed. A mechanical system can be constructed in which movement toward an object causes the object to move rapidly away, while movement away from the object causes it to come closer. Under these conditions, the duckling will move away from the object rather than approach or follow it. A duckling will learn to peck a spot on the wall if pecking brings the object closer. Only by knowing what and how the duckling learns during its lifetime can we be sure of what it is equipped to do at birth.
(Skinner, 1974, pp. 40–41)
The duckling's biological history, in terms of providing the capacity for reinforcement by proximity to a duck-sized object, is the context for the regulation of its behavior. Of course, the anatomy and physiology of the duck allow for this capacity. However, the way the environment is arranged determines the behavior of the individual organism. Laboratory experiments in behavior analysis identify the general principles that govern the behavior of organisms, the specific events that regulate the behavior of different species, and the arrangement of these events during the lifetime of an individual.
Early behaviorists like John Watson (1903) used the terminology of stimulus-response (S—R) psychology. From this perspective, stimuli force responses much like meat in a dog's mouth elicits (or forces) salivation. In fact, Watson based his stimulus-response theory of behavior on Pavlov's conditioning experiments. Stimulus-response theories are mechanistic in the sense that an organism is compelled to respond when a stimulus is presented. This is similar to a physical account of the motion of billiard balls. The impact of the cue ball (stimulus) determines the motion and trajectory (response) of the target ball. Although stimulus-response conceptions are useful for analyzing reflexive behavior and other rigid response patterns (i.e., Newtonion mechanics), the push-pull model is not as useful when applied to voluntary actions or operants. To be fair, Watson talked about “habits” in a way that sounds like operant behavior, but he lacked the experimental evidence and vocabulary to distinguish between respondent and operant conditioning.
It was B. F. Skinner (1935, 1937) who made the distinction between two types of conditioned reflex, corresponding to the difference between operant and respondent behavior. In 1938, Skinner introduced the term operant in his classic book, The Behavior of Organisms. Eventually, Skinner rejected the mechanistic (S—R) model of Watson and based operant conditioning on Darwin's principle of selection. The basic idea is that an individual emits behavior that produces effects, consequences, or outcomes. Based on these consequences, those performances that are appropriate increase—becoming more frequent in the population or class of responses for the situation; at the same time, inappropriate forms of response decline or become extinct. Julie Vargas is the daughter of B. F. Skinner and a professor of behavior analysis. She has commented on her father's model of causation:
Skinner's paradigm is a selectionist paradigm not unlike Darwin's selectionist theory of the evolution of species. Where Darwin found an explanation for the evolution of species, Skinner looked for variables functionally related to changes in behavior over the lifetime of an individual. Both explanations assumed variation; Darwin in inherited characteristics, Skinner in individual acts. Skinner, in other words, does not concern himself with why behavior varies, only with how patterns of behavior are drawn out from the variations that already exist. In looking at the functional relationships between acts and their effects on the world, Skinner broke with the S-R, input-output transformation model.
(Vargas, 1990, p. 9)
Skinner recognized that operants are selected by their consequences. He also noted that operant behavior naturally varies in form and frequency. Even something as simple as opening the door to your house is not done exactly the same way each time. Pressure on the doorknob, strength of pull, and the hand that is used change from one occasion to the next. If the door sticks and becomes difficult to open, a forceful response will eventually occur. This response may succeed in opening the door and become the most likely performance for the situation. Other forms of response will occur at different frequencies depending on how often they succeed in opening the door. Thus, operants are selected by their consequences.
Similarly, it is well known that babies produce a variety of sounds called “babbling.” These natural variations in sound production are important for language learning. When sounds occur, parents usually react to them. When the infant produces a familiar sound, parents often repeat it more precisely. Unfamiliar sounds are usually ignored. Eventually, the baby begins to produce sounds (we say talk) like other people in their culture or verbal community. Selection of verbal behavior by its social consequences is an important process underlying human communication (Skinner, 1957).
Although much of the basic research in the experimental analysis of behavior is based on laboratory animals, contemporary behavior analysts are increasingly concerned with human behavior. The behavior of people occurs in a social environment. Society and culture refer to aspects of the social environment, the context, that regulate human conduct. One of the primary tasks of behavior analysis is to show how individual behavior is acquired, maintained, and changed through interaction with others. An additional task is to account for the practices of the group, community, or society that affect an individual's behavior (Lamal, 1997).
Culture is usually defined in terms of the ideas and values of a society. However, behavior analysts define culture as all the conditions, events, and stimuli arranged by other people that regulate human action (Glenn, 1988; Skinner, 1953). The principles and laws of behavior analysis provide an account of how culture regulates an individual's behavior. A person in an English-speaking culture learns to speak in accord with the verbal practices of that community. People in the community provide reinforcement for a certain way of speaking. In this manner, a person comes to talk like and share the language of other members of the public and, in doing so, contributes to the perpetuation of the culture. The customs or practices of a culture are therefore maintained through the social conditioning of individual behavior.
Another objective is to account for the evolution of cultural practices. Behavior analysts suggest that the principle of selection (by consequences) also occurs at the cultural level. Cultural practices therefore increase (or decrease) based on consequences produced in the past. A cultural practice of making containers to hold water is an advantage to the group because it allows for the transportation and storage of water. This practice may include making and using shells, hollow leaves, or fired-clay containers. The cultural form that is selected (e.g., clay jars) is the one that proves most efficient. In other words, the community values those containers that last the longest, hold the most, and are easily stored. Thus, people manufacture clay pots and the manufacture of less efficient containers declines.
Behavior analysts are interested in cultural evolution because cultural changes alter the social conditioning of individual behavior. Analysis of cultural evolution suggests how the social environment is arranged and rearranged to support specific forms of human behavior. On a more practical level, behavior analysts suggest that the solution to many social problems requires a technology of cultural design. B. F. Skinner (1948) addressed this possibility in his utopian book, Walden Two. Although this idealistic novel was written some five decades ago, contemporary behavior analysts are conducting small-scale social experiments based on Skinner's ideas (Komar, 1983). For example, behavioral technology has been used to manage environmental pollution, encourage energy conservation, and regulate overpopulation (Glenwick & Jason, 1980).
B. F. Skinner (1904–1990) was the intellectual force behind behavior analysis. He was born Burrhus Frederic Skinner on 20 March 1904 in Susquehanna, Pennsylvania. When he was a boy, Skinner spent much of his time exploring the countryside with his younger brother. He had a passion for English literature and mechanical inventions. His hobbies included writing stories and designing perpetual-motion machines. He wanted to be a novelist and went to Hamilton College in Clinton, New York, where he graduated with a degree in English. After graduating from college in 1926, Skinner reported that he was not a great writer because he had nothing to say. He began reading about behaviorism, a new intellectual movement, and as a result went to Harvard in 1928 to learn more about a science of behavior. Skinner earned his master's degree in 1930 and his PhD the following year.
Skinner (Figure 1.3) began writing about the behavior of organisms in the 1930s when the discipline was in its infancy, and he continued to publish papers until his death in 1990. During his long career, Skinner wrote about and researched topics ranging from utopian societies, the philosophy of science, teaching machines, pigeons that controlled the direction of missiles, air cribs for infants, and techniques for improving education. Many people considered him a genius, while some were upset by his theories.
FIG. 1.3 B. F. Skinner. © B. F. Skinner Foundation, Cambridge, MA. Published with permission.
Skinner was always a controversial figure. He proposed a natural-science approach to human behavior. According to Skinner, the behavior of organisms, including humans, was determined. Although common sense suggests that we do things because of our feelings, thoughts, and intentions, Skinner stated that behavior resulted from genes and environment. This position bothered many people who believed that humans have some degree of self-determination. Even though he was constantly confronted with arguments against his position, Skinner maintained that the scientific facts required the rejection of feelings, thoughts, and intentions as causes of behavior. He said that these internal events were not explanations of behavior; rather, these events were additional activities of people that needed to be explained:
The practice of looking inside the organism for an explanation of behavior has tended to obscure the variables which are immediately available for a scientific analysis. These variables lie outside the organism in its immediate environment and in its environmental history. They have a physical status to which the usual techniques of science are adapted, and they make it possible to explain behavior as other subjects are explained in science. These independent variables [causes] are of many sorts and their relations to behavior are often subtle and complex, but we cannot hope to give an adequate account of behavior without analyzing them.
(Skinner, 1953, p. 31)
One of Skinner's most important achievements was his theory of operant behavior. The implications of behavior theory were outlined in his book, Science and Human Behavior (1953). In this book, Skinner discussed basic operant principles and their application to human behavior. Topics include self-control, thinking, the self, social behavior, government, religion, and culture. Skinner advocated the principle of positive reinforcement and argued against the use of punishment. He noted how governments and other social agencies often resort to punishment for behavior control. Although punishment works in the short run, he noted that it has many negative side effects. Positive reinforcement, Skinner believed, is a more effective means of behavior change—people act well and are happy when behavior is maintained by positive reinforcement.
People have misunderstood many of the things that Skinner has said and done (Catania & Harnard, 1988; Wheeler, 1973). One popular misconception is that he raised his children in an experimental chamber—the so-called baby in a box. People claimed that Skinner used his daughter as an experimental subject to test his theories. A popular myth is that this experience drove his child crazy. His daughter, Julie, was confronted with this myth and recalls the following:
I took a class called “Theories of Learning” taught by a nice elderly gentleman. He started with Hull and Spence, and then reached Skinner. At that time I had read little of Skinner and I could not judge the accuracy of what was being said about Skinner's theories. But when a student asked whether Skinner had any children, the professor thought Skinner had children. “Did he condition his children?” asked another student. “I heard that one of the children was crazy.” “What happened to his children?” The questions came thick and fast.
What was I to do? I had a friend in the class, and she looked over at me, clearly expecting action. I did not want to demolish the professor's confidence by telling who I was, but I couldn't just sit there. Finally, I raised my hand and stood up. “Dr. Skinner has two daughters and I believe they turned out relatively normal,” I said, and sat down.
(Vargas, 1990, pp. 8–9)
In truth, the “box” that Skinner designed for his children had nothing to do with an experiment. The air crib was an enclosed cot that allowed air temperature to be adjusted. In addition, the mattress cover could be easily changed when soiled. The air crib was designed so the child was warm, dry, and free to move about. Most importantly, the infant spent no more time in the air crib than other children do in ordinary beds (Skinner, 1945).
Although Skinner did not experiment with his children, he was always interested in the application of conditioning principles to human issues. His many books and papers on applied behavioral technology led to the field of applied behavior analysis. Applied behavior analysis is concerned with the extension of behavior principles to socially important problems. In the first issue of the Journal of Applied Behavior Analysis, Baer, Wolf, and Risley (1968) outlined a program of research based on Skinner's views:
The statement [of behavior principles] establishes the possibility of their application to problem behavior. A society willing to consider a technology of its own behavior apparently is likely to support that application when it deals with socially important behaviors, such as retardation, crime, mental illness, or education. Better applications, it is hoped, will lead to a better state of society, to whatever extent the behavior of its members can contribute to the goodness of a society. The differences between applied and basic research are not differences between that which “discovers” and that which merely “applies” what is already known. Both endeavors ask what controls the behavior under study. … [Basic] research is likely to look at any behavior, and at any variable which may conceivably relate to it. Applied research is constrained to look at variables which can be effective in improving the behavior under study.
(p. 91)
One area of application that Skinner wrote about extensively was teaching and learning. Although Skinner recognized the importance of behavior principles for teaching people with learning disabilities, he claimed that the same technology could be used to improve our general educational system. In his book, The Technology of Teaching, Skinner (1968) offered a personalized system of positive reinforcement for the academic performance of students. In this system, teaching involves arranging materials, designing the classroom, and programming lessons to shape and maintain the performance of students. Learning is defined objectively in terms of answering questions, solving problems, using grammatically correct forms of the language, and writing about the subject matter.
An aspect of Skinner's history that is not generally known is his humor and rejection of formal titles. He preferred to be called “Fred” rather than Burrhus and the only person who called him Burrhus was his close friend and colleague Fred Keller, who felt that he had prior claim on the name Fred (he was a few years older than Skinner). One of Skinner's earliest students, C. B. Ferster, tells about a time early in his acquaintance when Skinner tried to get Ferster to call him “Fred.” The story goes (Ferster, personal communication to Paul Brandon) that one day Ferster walked into the living room of Skinner's house to see Skinner seated on the sofa with a large sign around his neck saying “FRED.”
In the later part of his life, Skinner worked with Margaret Vaughan (Skinner & Vaughan, 1983) on positive approaches to the problems of old age. Their book, Enjoy Old Age: A Program of Self-Management, is written for the elderly reader and provides practical advice on how to deal with daily life. For example, the names of people are easy to forget and even more so in old age. Skinner and Vaughan suggest that you can improve your chances of recalling a name by reading a list of people you are likely to meet before going to an important occasion. If all else fails “you can always appeal to your age. You can please the friend whose name you have forgotten by saying that the names you forget are always the names you most want to remember” (p. 52).
Skinner, who held the A. E. Pierce Chair in Psychology, officially retired in 1974 from Harvard University. Following his retirement, Skinner continued an active program of research and writing. Each day he walked 2 miles to William James Hall, where he lectured, supervised graduate students, and conducted experiments. Eight days before his death on 18 August 1990, B. F Skinner received the first (and only) citation given for outstanding lifetime contribution to psychology from the American Psychological Association. Skinner's contributions to psychology and a science of behavior are documented in a recent film B. F. Skinner: A Fresh Appraisal (1999). Murray Sidman, a renowned researcher in the experimental analysis of behavior, narrated the film (available from the bookstore of the Cambridge Center for Behavioral Studies, www.behavior.org).
Contemporary behavior analysis is based on ideas and research that became prominent at the turn of the 20th century. The Russian scientist Ivan Petrovich Pavlov discovered the conditional reflex (a reflex that only occurs under a particular set of conditions such as the pairing of stimuli), and this was a significant step toward a scientific understanding of behavior.
Pavlov (Figure 1.4) was born the son of a village priest in 1849. He attended seminary school in order to follow his father into the priesthood. However, after studying physiology he decided on a career in the biological sciences. Although his family protested, Pavlov entered the University of St. Petersburg, where he graduated in 1875 with a degree in physiology. After completing his studies in physiology, Pavlov was accepted as an advanced student of medicine. He distinguished himself and obtained a scholarship to continue his studies of physiology in Germany. In 1890, Pavlov was appointed to two prominent research positions in Russia. He was Professor of Pharmacology at the St. Petersburg Medical Academy and Director of the Physiology Department. For the next 20 years, Pavlov studied the physiology of digestion, and in 1904 he won the Nobel Prize for this work, the year that B. F. Skinner was born.
FIG. 1.4 Ivan Petrovich Pavlov. Reprinted with permission from the Archives of the History of American Psychology, The University of Akron.
Ivan Pavlov worked on the salivary reflex and its role in digestion. Pavlov had dogs surgically prepared to expose the salivary glands in the dogs' mouths. The animals were brought into the laboratory and put in restraining harnesses. As shown in Figure 1.5, food was then placed in the dogs' mouths and the action of the salivary glands was observed.
The analysis of the salivary reflex was based on prevailing notions of animal behavior. At this time, many people thought that animals, with the exception of humans, were complex biological machines. The idea was that a specific stimulus evoked a particular response in much the same way that turning a key starts an engine. In other words, animals reacted to the environment in a simple cause—effect manner. Humans, on the other hand, were seen as different from other animals in that their actions were purposive. Humans were said to anticipate future events. Pavlov noticed that his dogs began to salivate at the sight of an experimenter's lab coat before food was placed in the animal's mouth. This suggested that the dogs anticipated food. Pavlov recognized that such a result challenged conventional wisdom.
Pavlov made an important observation in terms of the study of behavior. He reasoned that anticipatory reflexes were learned or conditioned. Further, Pavlov concluded that these conditioned reflexes were an essential part of the behavior of organisms. Although some behaviors were described as innate reflexes, other actions were based on conditioning that occurred during the animal's life. These conditioned reflexes (termed conditional reflexes in Pavlov, 1960) were present to some degree in all animals but were most prominent in humans.
FIG. 1.5 A dog in the experimental apparatus used by Pavlov.
The question was how to study conditioned reflexes systematically. Pavlov's answer to this question represents a major advance in the experimental analysis of behavior. If dogs reliably salivate at the sight of a lab coat, Pavlov reasoned, then any arbitrary stimulus that preceded food might also be conditioned and evoke salivation. Pavlov replaced the experimenter's lab coat with a stimulus that he could systematically manipulate and reliably control. In some experiments, a metronome (a device used to keep the beat while playing the piano) was presented to a dog just before it was fed. This procedure resulted in the dog eventually salivating to the sound of the metronome. If a particular beat preceded feeding while other rhythms did not, the dog salivated most to the sound associated with food.
Although Pavlov was a physiologist and believed in mental associations, his research was directed at observable responses and stimuli. He discovered many principles of the conditioned reflex. These principles included spontaneous recovery, discrimination, generalization, and extinction. The later part of his career involved an experimental analysis of neurosis in animals. He continued these investigations until his death in 1936.
Pavlov's research became prominent in North America, and the conditioned reflex was incorporated into a more general theory of behavior by the famous behaviorist John B. Watson (Figure 1.6). Watson acknowledged Pavlov's influence:
I had worked the thing out [conditioning] in terms of Habit formation. It was only later, when I began to dig into the vague word Habit that I saw the enormous contribution Pavlov had made, and how easily the conditioned response could be looked upon as the unit of what we had been calling Habit. I certainly, from that point on, gave the master his due credit.
(Watson, personal communication to Hilgard & Marquis, 1961, p. 24)
Watson went on to argue that there was no need to make up unobservable mental associations to account for human and animal behavior. He proposed that psychology should be a science based on observable behavior. Thoughts, feelings, and intentions had no place in a scientific account and researchers should direct their attention to muscle movements and neural activity. Although this was an extreme position, Watson succeeded in directing the attention of psychologists to behavior-environment relationships.
Watson was a rebellious young man who failed his last year at Furman University because he handed in a final-exam paper “backwards.”1 He graduated in 1899, when he was 21 years old. After spending a year as a public-school teacher, Watson decided to further his education and was admitted to graduate studies at the University of Chicago. There he studied philosophy with John Dewey, the famous educator. He never really appreciated Dewey's ideas and later in his life commented, “I never knew what he was talking about then, and, unfortunately for me, I still don't know” (Watson, 1936, p. 274). While a graduate student at Chicago, he also studied psychology with James Angell and biology and physiology with Henry Donaldson and Jacques Loeb (Pauley, 1987). In 1903, he obtained his doctorate for research with laboratory rats. The experiments concerned learning and correlated changes in the brains of these animals.
FIG. 1.6 John Watson. Reprinted with permission from the Archives of the History of American Psychology, The University of Akron.
Watson (1903) published a book called Animal Education: An Experimental Study on the Psychical Development of theWhite Rat, Correlated with the Growth of Its Nervous System that was based on his doctoral research. The book demonstrated that Watson was a capable scientist who could clearly present his ideas. Ten years later, Watson (1913) published his most influential work in Psychological Review, “Psychology as the Behaviorist Views It.” This paper outlined Watson's views on behaviorism and argued that objectivity was the only way to build a science of psychology:
I feel that behaviorism is the only consistent logical functionalism. In it one avoids [the problem of mind-body dualism]. These time-honored relics of philosophical speculation need trouble the student of behavior as little as they trouble the student of physics. The consideration of the mind-body problem affects neither the type of problem selected nor the formulation of the solution of that problem. I can state my position here no better than by saying that I should like to bring my students up in ignorance of such hypotheses as one finds among other branches of science.
(Watson, 1913, p. 166)
In this paper, Watson also rejected as scientific data what people said about their thoughts and feelings. Further, he pointed to the unreliability of psychological inferences about another person's mind. Finally, Watson noted that the psychology of mind had little practical value for behavior control and public affairs.
Perhaps Watson's most famous experiment was the study of fear conditioning with Little Albert (Watson & Rayner, 1920). Little Albert was a normal, healthy infant who attended a day-care center. Watson and his assistant used classical-conditioning procedures to condition fear of a white rat. At first Little Albert looked at the rat and tried to touch it. The unconditioned stimulus was the sound of a hammer hitting an iron rail. This sound made Little Albert jump, cry, and fall over. After only six presentations of the noise and rat, the furry animal also produced the fear responses. The next phase of the experiment involved a series of tests to see if the child's fear reaction transferred to similar stimuli. Albert was also afraid when presented with a white rabbit, a dog, and a fur coat.
At this point, Watson and Rayner discussed a number of techniques that could be used to eliminate the child's fear. Unfortunately, Little Albert was removed from the day-care center before counter-conditioning could be carried out. In his characteristic manner, Watson later used the disappearance of Little Albert to poke fun at Freud's method of psychoanalysis. He suggested that as Albert got older, he might go to an analyst because of his strange fears. The analyst would probably convince Albert that his problem was the result of an unresolved Oedipal complex. But, Watson remarked, we would know that Albert's fears were actually caused by conditioning—so much for Freudian analysis.
Watson had many interests and he investigated and wrote about ethology, comparative animal behavior, neural function, physiology, and philosophy of science. Based on his controversial views and charisma, Watson was elected president of the American Psychological Association in 1915 when he was only 37 years old. After leaving Johns Hopkins University, he became successful in industry by applying conditioning principles to advertising and public relations (Buckley, 1989). Watson implemented the use of subliminal suggestion and the pairing of hidden symbols in advertising—techniques still used today.
Watson's behaviorism emphasized the conditioned reflex. This analysis focuses on the events that precede action and is usually called a stimulus-response approach. Another American psychologist, Edward Lee Thorndike (Figure 1.7), was more concerned with how success and failure affect the behavior of organisms. His research emphasized the events and consequences that follow behavior. In other words, Thorndike was the first scientist to systematically study operant behavior, although he called the changes that occurred trial-and-error learning (Thorndike, 1898).
Edward L. Thorndike was born in 1874 in Williamsburg, Massachusetts. He was the son of a Methodist minister and had no knowledge of psychology until he attended Wesleyan University. There he read William James's (1890) book, Principles of Psychology, which had a major impact on him. After reading the book, Thorndike was accepted as a student at Harvard, where he studied with William James. It is important to note that James's psychology focused on the mind and used the method of introspection (people's reports of feeling and thoughts). Thus, in contrast to John Watson, Thorndike was concerned with states of mind. In terms of contemporary behavior analysis, Thorndike's contribution was his systematic study of the behavior of organisms rather than his mental interpretations of animal and human behavior.
Thorndike was always intrigued with animal behavior. While at Harvard, his landlady became upset because he was raising chickens in his bedroom. By this time, James and Thorndike were good friends, and Thorndike moved his experiments to the basement of James's house when he could not get laboratory space at Harvard. He continued his research and supported himself by tutoring students for 2 years at Harvard. Then Thorndike moved to Columbia University where he studied with James McKeen Cattell, the famous expert on intelligence testing. Thorndike took two of his “smartest” chickens with him to Columbia, but soon switched to investigating the behavior of cats.
FIG. 1.7 Edward Thorndike. Reprinted with permission from the Archives of the History of American Psychology, The University of Akron.
At Columbia University, Thorndike began his famous experiments on trial-and-error learning in cats. Animals were placed in what Thorndike called a “puzzle box” and food was placed outside the box (Chance, 1999). A cat that struggled to get out of the box would accidentally step on a treadle, pull a string, and lift a latch. These responses resulted in opening the puzzle-box door. Thorndike found that most cats took less and less time to solve the problem after they were repeatedly returned to the box (i.e., repeated trials). From these and additional observations, Thorndike made the first formulation of the law of effect:
The cat that is clawing all over the box in her impulsive struggle will probably claw the string or loop or button so as to open the door. And gradually all the other non-successful impulses will be stamped out and the particular impulse leading to the successful act will be stamped in by the resulting pleasure [italics added], until after many trials, the cat will, when put in the box, immediately claw the button or loop in a definite way.
(Thorndike, 1911, p. 40)
Today, Thorndike's law of effect is restated as the principle of reinforcement. This principle states that all operants may be followed by consequences that increase or decrease the probability of response in the same situation. Notice that references to “stamping in” and “pleasure” are not necessary and that nothing is lost by this modern restatement of the law of effect.
Thorndike was appointed to the Teachers College, Columbia University as a professor in 1899, and he spent his entire career there. He studied and wrote about education, language, intelligence testing, comparison of animal species, the nature-nurture problem, transfer of training, sociology of the quality of life, and, most importantly, animal and human learning. Thorndike published more than 500 books and journal articles. His son (Robert Ladd Thorndike, 1911–1990) became a well-known educational psychologist in his own right and in 1937 joined the same department of psychology as his father. Edward Lee Thorndike died in 1949.
The works of Pavlov, Watson, Thorndike, and many others have influenced contemporary behavior analysis. Although the ideas of many scientists and philosophers have had an impact, Burrhus Fredrick Skinner (1904–1990) is largely responsible for the development of modern behavior analysis. In the “Focus on: B. F Skinner” section above, we described some details of his life and some of his accomplishments. An excellent biography is available (Bjork, 1993), and Skinner himself wrote a three-volume autobiography (1976, 1979, 1983). Here we will outline his contribution to contemporary behavior analysis.
Skinner was studying at Harvard during a time of intellectual change. He wanted to extend the work of Pavlov to more complicated instances of the conditioned reflex. Rudolph Magnus was a contemporary of Ivan Pavlov, and he had been working on the conditioning of physical movement. Skinner had read his book Korperstellung in the original German and was impressed with it. Skinner said, “I began to think of reflexes as behavior rather than with Pavlov as ‘the activity of the cerebral cortex’ or, with Sherrington, as ‘the integrative action of the nervous system’” (Skinner, 1979, p. 46).
The idea that reflexes could be studied as behavior (rather than as a reflection of the nervous system or the mind) was fully developed in Skinner's (1938) book, The Behavior of Organisms. In this text, Skinner distinguishes between Pavlov's reflexive conditioning and the kind of learning reported by Thorndike. Skinner proposed that respondent and operant conditioning regulated behavior. These terms were carefully selected to emphasize the study of behavior for its own sake. Pavlov interpreted reflexive conditioning as the study of the central nervous system, and Skinner's respondent conditioning directed attention to environmental events and responses. Thorndike's trial-and-error learning was based on unobservable states of mind, and Skinner's operant conditioning focused on the functional relations between behavior and its consequences. Both operant and respondent conditioning required the study of observable correlations among objective events and behavior.
Skinner soon talked about a “science of behavior” rather than one of physiology or mental life. Once stated, the study of behavior for its own sake seems obvious, but consider that most of us say that we do something because we have made up our mind to do it or, in more scientific terms, because of a neural connection in our brain. Most people accept explanations of behavior that rely on descriptions of brain, mind, intelligence, cognitive function, neural activity, thinking, or personality. Because these factors are taken as the cause(s) of behavior, they become the focus of investigation. Skinner, however, suggested that remembering, thinking, feeling, the action of neurons, etc. are more behaviors of the organism that require explanation. He further proposed that the action of organisms could be investigated by focusing on behavior and the environmental events that precede and follow it.
Skinner's behavioral focus was partially maintained and influenced by his lifelong friend, Fred Simmons Keller. Skinner and Keller attended Harvard graduate school together, and Keller encouraged Skinner to pursue a behavioral view of psychology. By 1946, Skinner had formed a small group of behaviorists at Indiana University. At this same time, Fred Keller and his friend Nat Schoenfeld organized another group at Columbia University (Keller, 1977, Chapters 2 and 6).
Although the number of behavior analysts was growing, there were no sessions on behavioral issues at the American Psychological Association annual meetings. Because of this, Skinner, Keller, Schoenfeld, and others organized their own conference at Indiana University. This was the first conference on the experimental analysis of behavior (see Figure 1.8). These new-style behaviorists rejected the extreme views of John B. Watson and offered an alternative formulation. Unlike Watson, they did not reject genetic influences on behavior; they extended the analysis of behavior to operant conditioning, and they studied behavior for its own sake.
These new behavior analysts found it difficult to get their research published in the major journals of psychology. This was because they used a small number of subjects in their experiments, did not use statistical analysis, and their graphs of response rate were not appreciated. By 1958, the group was large enough to start its own journal, and the first volume of the Journal of the Experimental Analysis of Behavior (JEAB) was published. As research accumulated, the practical implications of behavior principles became more and more evident, and applications to mental illness, retardation, rehabilitation, and education increased. In 1968, the Journal of Applied Behavior Analysis (JABA) was published for the first time.
By 1964, the number of behavior analysts had grown so much that the American Psychological Association established a special division. Division 25 is called The Experimental Analysis of Behavior and has several thousand members. Subsequently, the Association for Behavior Analysis (ABA) was founded in the late 1970s. This association holds an annual international conference that is attended by behavior analysts from many countries. The association publishes a journal of general issues called The Behavior Analyst.
FIG. 1.8 Photograph taken at the first conference on the experimental analysis of behavior held in 1946 at Indiana University. From left to right in front row: Dinsmoor, Musgrave, Skinner, Keller, Schoenfeld, Lloyd. Middle row: Ellson, Daniel, Klein, Jenkins, Wyckoff, Hefferline, Wolin. Back row: Estes, Frick, Anderson, Verplanck, Beire, Hill, Craig. From the Journal of the Experimental Analysis of Behavior, 5, 456. Copyright 1958, by the Society for the Experimental Analysis of Behavior, Inc. Reprinted with permission.
In addition to ABA, Robert Epstein, one of B. F. Skinner's last students, and past editor of the well-known magazine Psychology Today, founded the Cambridge Center for Behavioral Studies in 1981. The Cambridge Center is devoted to helping people find effective solutions to behavior problems (e.g., in education, business, and other applied settings). As part of this mission, the Center maintains an information web site (www.behavior.org) for the public, publishes books and journals, and sponsors seminars and conferences on effective behavior management in applied settings (e.g., the annual conference Behavioral Safety Now in the field of industrial safety).
A continuing issue in the field of behavior analysis is the separation between applied behavior analysis and basic research. During the 1950s and 1960s, no clear distinction existed between applied and basic investigations. This was because applied behavior analysts were trained as basic researchers. That is, the first applications of behavior principles came from the same people who were conducting laboratory experiments. [As an aside, Skinner's second book, after his basic text called The Behavior of Organisms (1938), was a novel describing the application of behavior principles in a Utopian community, Walden Two (1948).] The applications of behavior principles were highly successful, and this led to a greater demand for people trained in applied behavior analysis. Soon applied researchers were no longer working in the laboratory or reading the basic journals.
This separation between basic and applied research was first described by Sam Deitz (1978), who noted the changing emphasis from science to technology among applied behavior analysts (see also Hayes, Rincover, & Solnick, 1980; Michael, 1980; Pierce & Epling, 1980). Donald Baer (1981) acknowledged the technical drift of applied behavior analysis, but suggested that this was a natural progression of the field that may have positive effects.
Today the separation issue is not entirely resolved although much progress is apparent. Applied researchers are more in contact with basic research through the Journal of Applied Behavior Analysis. In addition to application research, this journal publishes applied articles based on modern behavior principles (e.g., Ducharme & Worling, 1994) as well as reviews of basic research areas (e.g., Mace, 1996). We have written this book assuming that an acquaintance with basic research is important, even for those who are primarily concerned with behavioral applications. Students can study this text for a basic grounding in behavioral science, or for a solid foundation in human behavior and application.
All scientists make assumptions about their subject matter. These assumptions are based on prevailing views in the discipline and guide scientific research. In terms of behavior analysis, researchers assume that the behavior of organisms is lawful. This means that it is possible to study the interactions between an organism and its environment in an objective manner. To carry out the analysis, it is necessary to isolate behavior-environment relationships. The scientist must identify events that reliably precede the onset of some action and the specific consequences that follow behavior. If behavior systematically changes with variation in the environmental conditions, then behavior analysts assume that they have explained the action of the organism. There are other assumptions that behavior analysts make about their science.
Contemporary behavior analysts include internal events as part of an organism's environment. This point is often misunderstood; internal functioning like an upset stomach, full bladder, and low blood sugar are part of a person's environment. Internal physical events have the same status as external stimuli such as light, noise, odor, and heat. Both external and internal events regulate behavior. Although this is so, behavior analysts usually emphasize the external environment. This is because external events are the only stimuli available for behavior change. The objective procedures of psychological experiments are giving instructions and observing how the person acts. From a behavioral view, the instructions are external stimuli that regulate both verbal and nonverbal behavior. Even when a drug is given to a person and the chemical alters internal biochemistry, the direct injection of the drug is an external event that subsequently regulates behavior. To make this clear, without the drug injection neither the biochemistry nor the behavior of the person would change.
Many psychological studies involve giving information to a person in order to change or activate cognitive processes. Thus, cognitive psychologists “… invent internal surrogates which become the subject matter of their science” (Skinner, 1978, p. 97); and cognitive psychology has been defined as “… an approach to scientific psychology—that encourages psychologists to infer unobservable constructs on the basis of observable phenomena” (Baars, 1986, p. ix). In the cognitive view, thoughts are used to explain behavior. The problem is that the existence of thoughts (or feelings) is often inferred from the behavior to be explained, leading to circular reasoning. For example, a child who peers out of the window about the time her mother usually comes home from work is said to do this because of an “expectation.” The expectation of the child is said to explain why the child peers from the window. In fact, there is no explanation because the cognition (expectation) is inferred from the behavior it is said to explain. Cognitions could explain behavior if the existence of thought processes were based on some evidence other than behavior. In most cases, however, there is no independent evidence that cognitions caused behavior and the explanation is not scientifically valid. One way out of this problem of logic is not to use thinking and feeling as causes of behavior. That is, thinking and feeling are treated as more behavior to be explained.
Most people assume that their feelings and thoughts explain why they act as they do. Contemporary behavior analysts agree that people feel and think, but they do not consider these events as causes of behavior. They note that these terms are more correctly used as verbs rather than nouns. Instead of talking about thoughts, behavior analysts point to the action word “thinking.” And instead of analyzing feelings as things we possess, the behavioral scientist focuses on the action of feeling or sensing. In other words, thinking and feeling are activities of the organism that require explanation.
Because feelings occur at the same time that we act, they are often taken as causes of behavior. Although feelings and behavior necessarily go together, it is the environment that determines how we act, and at the same time how we feel. Feelings are real, but they are the result of the environmental events that regulate behavior. Thus, a behavioral approach requires the researcher to trace feelings back to the interaction between behavior and environment.
Pretend that you are in an elevator between the 15th and 16th floors when the elevator suddenly stops, and the lights go out. You hear a sound that appears to be the snapping of elevator cables. Suddenly, the elevator lurches and then drops 2 feet. You call out, but nobody comes to your rescue. After about an hour, the elevator starts up again, and you get off on the 16th floor. Six months later, a good friend invites you to dinner. You meet downtown, and you discover that your friend has made reservations at a restaurant called The Room at the Top, which is located on the 20th floor of a skyscraper. Standing in front of the elevator, a sudden feeling of panic overwhelms you. You make a socially appropriate excuse like, “I don't feel well,” and you leave. What is the reason for your behavior and the accompanying feeling?
There is no question that you feel anxious, but this feeling is not why you decide to go home. Both the anxious feeling and your decision to leave are easily traced to the negative experience in the elevator that occurred 6 months ago. It is this prior conditioning that behavior analysts emphasize. Notice that the behavioral position does not deny your feelings. These are real events. However, it is your previous interaction with the broken elevator that changed both how you feel and how you act.
You may still wonder why behavior analysts study overt behavior instead of feelings—given that both are changed by experience. The answer concerns the accessibility of feelings and overt behavior. Much of the behavior of organisms is directly accessible to the observer or scientist. This public behavior provides a relatively straightforward subject matter for scientific analysis. In contrast, feelings are largely inaccessible to the scientific community. Of course, the person who feels has access to this private information, but the problem is that reports of feelings are highly unreliable.
This unreliability occurs because we learn to talk about our feelings (and other internal events) as others have trained us to do. During socialization, people teach us how to describe ourselves, but when they do this they have no way of accurately knowing what is going on inside us. Parents and teachers rely on public cues to train self-descriptions. They do this by commenting on and correcting verbal reports when behavior or events suggest a feeling. A preschooler is taught to say, “I feel happy” when the parents guess that the child is happy. The parents may base their judgment on smiling, excitement, and affectionate responses from the child. Another way this training is done is that the child may be asked, “Are you happy?” in a circumstance where the parents expect the child to feel this way (e.g., on Christmas morning). When the child appears to be sad, or circumstances suggest that this should be so, saying “I am happy” is not reinforced by the parents. Eventually, the child says, “I am happy” in some situations and not in others.
Perhaps you have already noticed why reports of feelings are not good scientific evidence. Reports are only as good as the training of correspondence between public conditions and private events. In addition to inadequate training, there are other problems with accurate descriptions of feelings. Many of our internal functions are poorly correlated (or uncorrelated) with public conditions, and this means that we cannot be taught to describe such events accurately. Although a doctor may ask for the general location of a pain (e.g., abdomen), he or she is unlikely to ask whether the hurt is in the liver or the spleen. This report is simply inaccessible to the patient because there is no way to teach the correspondence between exact location of damage and public conditions. Generally, we are able to report in a limited way on private events, but the unreliability of such reports makes them questionable as scientific observations. Based on this realization, behavior analysts focus on the study of behavior rather than feelings.
Behavior analysts have also considered thinking and its role in a science of behavior. In contrast to views that claim a special inner world of thought, behavior analysts suggest that human thought is human behavior. Skinner (1974) stated that:
The history of human thought is what people have said and done. Symbols are the products of written and spoken verbal behavior, and the concepts and relationships of which they are symbols are in the environment. Thinking has the dimensions of behavior, not a fancied inner process which finds expression in behavior.
(pp. 117–118)
A number of behavioral processes, like generalization, discrimination, matching to sample, and stimulus equivalence (see later chapters), give rise to behavior that, in a particular situation, may be attributed to higher mental functions. From this perspective, thinking is treated as private behavior (see Moore, 2003, and Tourinho, 2006, on private events).
One of the more interesting examples of thinking involves private behavior or behavior only accessible to the person doing it. Thinking as private behavior is observed in a game of chess. We may ask another person, “What is the player thinking about?” A response like “She is probably thinking of moving the castle” refers to thinking that precedes the move itself. Sometimes this prior behavior is observable—the player may place a hand on the castle in anticipation of the move. At other times, behavior is private and cannot be observed by others. An experienced chess player may think about the game, imagining the consequences of moving a particular piece.
Presumably, this private behavior is overt when a person learns to play chess. For example, first the basic rules of the game are explained and a novice player is shown how the pieces move and capture. In moving the pieces from place to place, the player is asked to describe the relationships between the opposing chess pieces. This establishes the behavior of visualizing the layout of the board. As the player receives additional corrective feedback, visualizing the layout becomes more skillful. The person begins to see relationships and moves that were not previously apparent. During the first few games, new players are often given instructions like “Don't move your knight there, or you'll lose it.” Additionally, the player may be told, “A better move would have been …,” and a demonstration of the superior move is usually given. After some experience, the student is asked to explain why a particular move was made, and the explanation is discussed and evaluated. Eventually, the teacher stops prompting the player and encourages the person to play chess in silence. At this point, visualizing the layout of the board (e.g., white controls the center of the board) and describing the possible consequences of moves (e.g., moving the knight to this square will split the two rooks) becomes covert.
The function of thinking, as private behavior, is to increase the effectiveness of practical action. People can act at the covert level without committing themselves publicly. An advantage is that an action can be revoked if the imagined consequences are not reinforcing. In our example, the chess player considers the possible moves and the consequences that may follow given her/his past experience. Based on the covert evaluation, a player makes the move that appears to be best. Thus, the covert performance functions to prompt and guide overt action. Once the move is made, the player faces the objective consequences. If the move produces an advantage that results in checkmate, then thinking about such a move in similar circumstances is strengthened. On the other hand, a bad move weakens the consideration of such moves in the future. Overall, thinking is operant behavior controlled by its consequences. Thinking about a move that guides effective action is likely to occur again, while thinking that prompts ineffective performance declines.
In this section, we have discussed thinking as covert and private behavior. There are many other ways that the term thinking is used. When a person remembers, we sometimes talk about thinking in the sense of searching and recalling. Solving problems often involves private behavior that furthers a solution. In making a decision, people are said to think about the alternatives before a choice is made. The creative artist is said to think of novel ideas. In each of these instances, it is possible to analyze thinking as private behavior that is regulated by specific features of the environment. The remainder of this book discusses the behavioral processes that underlie all behavior, including thinking.
This chapter has introduced the idea that behavior often is acquired during one's lifetime as a result of experience. At birth we emit behavior with very little organized activity. However, as our behaviors cause consequences, some responses are strengthened while others are weakened. The consequences of behavior function to select and establish a behavior repertoire. Several prominent persons were introduced to illustrate the history of the science of behavior analysis. In particular B. F. Skinner was described as the major force behind the experimental and applied analyses of behavior, which is the topic of this book. This approach is related to biology in that behavior is considered to be a product of genes interacting with the organism's environment. Behavior analysis can be extended to the understanding of feelings and to complex behavior involving problem solving and thinking.
Applied behavior analysis
behavior
Behavior analysis
Behavior analyst
Behavioral neuroscience
Behaviorism
Conditioned reflex
Culture
Experimental analysis of behavior
Immediate causation
Law of effect
Learning
Operant
Operant conditioning
Private behavior
Reflex
Remote causation
Respondent
Respondent conditioning
Science of behavior
Selection by consequences
1. This description of John Watson is partially based on a paper by James Todd and Edward Morris (1986) on “The Early Research of John B. Watson: Before the Behavioral Revolution.”