Psych

Behavorism.pdf

Further Readings

Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: W. H. Freeman.

Beck, H.P., Levinson, S., & Irons, G. (2009). Finding Little Albert: A journey to John B. Watson's infant laboratory. American Psychologist, 64(7), 605-614.

Bouton, M. E. (2014). Why behavior change is difficult to sustain. Preventive Medicine, 68, 29-36. http://dx.doi .org/10.1016/j.ypmed.2014.06.010

Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of personality: Reconceptualizing situations, dispositions, dynamics, and invariance in personality structure. Psychological Review, 102(2), 246-268.

Skinner, B. F. (1974). About behaviorism. New York, NY: Knopf.

Watson, J.B. (1930). Behaviorism. New York, NY: W.W. Norton.

BEHAVIORISM

In its most general sense, behaviorism is the point of view that the appropriate subject matter of psy- chology is behavior and that the appropriate methods for studying it are grounded in the stan- dard observational methods of the natural sci- ences. Although this statement appears to be admirably succinct, a review of the history of psychology suggests behaviorism has been inter- preted in several different ways over the years. This entry examines these different interpretations to better understand behaviorism.

Classical S-R Behaviorism

A frequently cited date for the inception of behav- iorism is 1913, when John B. Watson published a controversial article titled "Psychology as the Behaviorist Views It." Watson's version of behav- iorism is called classical S-R behaviorism and asserts that psychology should study the responses (R) that are "called out" by environmental stimuli or situations (S).

For Watson, this relation involves overt, explicit activity such as the movement of arms and legs, as well as covert, implicit activity such as changes in heart rate or other visceral functions. The activity could be either unlearned or learned. Unlearned

Behaviorism 89

meant that certain responses were called out by certain stimuli right from birth. Learned meant the responses were called out by previously neutral stimuli after those stimuli had been recently or frequently associated with stimuli that originally evoked the response.

Classical behaviorism may be contrasted with the dominant point of view in psychology at the time, according to which the appropriate subject matter of psychology was mental life-for example, the sensations, images, and feelings that contribute to "consciousness." The appropriate method for studying that subject matter was introspection, understood as the rigorous, contemplative description of one's own mental phenomena. For his part, Watson zealously sought to reformulate psychology as a natural science concerned with the prediction and control of behavior. He did not see behaviorism as merely a methodological approach but rather as an entire worldview.

For example, in Behaviorism, Watson argued that mental life qua mental was nothing but a fic- tion engendered by cultural forces such as religion. Watson clearly recognized that people do think. However, he argued that thinking could be con- strued as a behavioral rather than a mental process, consisting largely but not exclusively of subvocal speech. His explanations of such traditional con- cerns as thinking and images were not particularly sophisticated by today's standards. Nevertheless, they can be understood today as honest efforts to formulate an effective science of behavior based on naturalistic principles as he understood them.

Despite Watson's extensive efforts, many classi- cal behaviorists continued to accept the existence of mental phenomena as distinct from behavioral. However, they interpreted behaviorism as meaning that the mental could be safely ignored. The assumption was that the resulting explanations were just as good as those appealing directly to mental variables. Mental variables could be included in philosophy or religion, but science required observability and measurement that pro- duced agreement about its explanatory concepts. Therefore, if psychology was going to be consid- ered a member in good standing of the scientific community, psychologists should not speak directly about the mental but only about what was observ- able. This approach was later called methodological behaviorism.

90 Behaviorism

Mediational S-0-R Neobehaviorism

By around 1930, many psychologists began to view the various forms of classical S-R behavior- ism as inadequate. The principal problem was that behavior was richer and more flexible than seemed to be recognized by the S-R formulation, even if unobservables were incorporated in Watson's terms of 1925. To overcome such apparent limita- tions, psychologists began to argue that by modify- ing the accepted S-R formulation somewhat, they could admit such unobservables as mental after all.

Their first move was to conceive of the mental variables as inside the organism in some sense. There, the mental variables were inferred to medi- ate the relation between S and R. The concept of mediation means that the S was inferred to initiate some state or process inside the organism and the mediating state or process then caused the R. Variations in the mediating state produced variations in response, thereby accommodating the richness and flexibility of behavior. More formally, the mental variables were designated as "organis- mic" variables and symbolically represented as 0, resulting in an S-0-R formulation. Initially, the mediators were designated as theoretical concepts and later as hypothetical constructs.

The mediationists' second move was to appeal to the principle of operationism, which had been developed by the physicist P. W. Bridgman in 1927. According to operationism, psychologists should appoint some publicly observable measure as evidence or justification for the mental phenomenon in question so that it could be agreed upon. As a result, mental variables could be known by inference, and talk of them was justified by the data from the operations entailed in their supposed measurement. Importantly, an emphasis on publicly observable measures meant that the approach could be considered scientific because psychologists were speaking only indirectly about mental phenomena instead of directly as in introspection, and their inferences about the mental phenomena were supported by data.

This approach is now called mediational S-0-R neobehaviorism, reflecting the new way to deal with the mental. Mediational neobehaviorism is also a version of methodological behaviorism, in the sense that observable behavior is taken as a proxy for unobservable causes inferred to be

operating elsewhere. Thus, neobehaviorists can avoid speaking directly of the mental and defend themselves against charges that they were being mentalistic themselves.

An important feature of mediational neobehav- iorism is antecedent causation. According to this concept, the cause of behavior is to be found and expressed in terms of some antecedent factor, in the tradition of the reflex. If an environmental stimulus could not be found, then some mediating internal or physiological state or process was inferred. If this could not be found, then some mediating mental or cognitive state or process was inferred. In every case, causal processes were inferred to follow a linear chain of events, the last of which was an independent contribution of the organism from a nonenvironmental dimension.

In American psychology, the mediational approach dominated learning theory during the second quarter of the 20th century and into the third quarter-for example, in 1956, in the work of Edward C. Tolman, Clark L. Hull, and Kenneth W. Spence. Tolman's mediating hypothetical con- structs tended to have a cognitive flavor, Hull's tended to have a physiological flavor, and Spence's a strictly logical flavor, although all three disputed that they were mentalists because their mediators were based on operationism rather than introspection.

The mediational approach has remained preva- lent in much of the rest of American psychology, such as in clinical work, social psychology, person- ality theory, and developmental psychology. Different researchers and theorists have different conceptions of the mediators, but the approaches remain mediational nonetheless. Sometimes the term behaviorism is used in the popular literature to indicate a form of psychology that doesn't speak directly of the mental. However, classical behavior- ism has actually been superseded by the media- tional S-0-R form of methodological behaviorism in which the direct talk concerns observable data, and mental antecedent causes are inferred but spo- ken of only indirectly through operational definitions.

Behavior Analysis

B. F. Skinner developed an entirely different inter- pretation of behaviorism beginning in the 1930s,

or roughly the same time as the mediational approach just described. Skinner's view is called behavior analysis. A good source is Skinner's Science and Human Behavior. The philosophy of science that underlies Skinner's behavior analysis is called radical behaviorism. Radical here means a thoroughgoing or comprehensive behavioral ori- entation rather than an extreme or fanatical com- mitment to observable variables strictly.

Behavior analysis recognizes that some behavior developed during the evolutionary history of the species. Often, this behavior is called innate. Nest building and song learning in birds are examples. Thus, in behavior analysis, not all behavior is learned, although behavior analysis does empha- size that important forms of behavior might depend critically on environmental circumstances during development.

Other behavior occurs according to the tradi- tional reflex model, where an antecedent stimulus elicits behaviors with food-elicited salivation in Pavlov's dogs, for example. Behavior analysis calls this elicited form of behavior respondent behavior. The process of conditioning is concerned with how respondent behavior develops to previously neu- tral stimuli, as when Pavlov's dogs salivated to a light or tone (called a conditioned stimulus) that Pavlov had presented with food (called an uncon- ditioned stimulus). According to behavior analysis, some human behavior is indeed respondent, as when a person becomes anxious when entering a dentist's office if he or she has experienced pain on a prior visit or when a person salivates in response to sights, sounds, or aromas associated with food preparation.

Still other behavior occurs as a function of its consequences, and thus behavior analysis goes far beyond classical S-R behaviorism, which disre- garded any influence of consequences. One previ- ous account that did emphasize the importance of consequences was E. L. Thorndike's law of effect. It posited a hypothetical connection between a stimulus setting and a response, which was inferred to be "stamped in" when the response produced a satisfying effect and "stamped out" when it pro- duced an annoying effect or discomfort.

However, behavior analysis goes beyond Thorndike's view by setting aside the various appeals to hypothetical causes and proposing the concept of operant behavior. According to this

Behaviorism 91

concept, it is sufficient to state the fact of the obser- vation: An organism emits some response, and the probability of the response increases when it is fol- lowed by a reinforcing consequence or decreases when it is followed by a punishing consequence. A prior stimulus serves by signaling the correlation between the response and its consequence in the relation called stimulus control. In other words, there is no need to appeal to hypothetical, mediat- ing processes to explain what is observed. Indeed, it might be harmful and misleading to do so.

To be sure, physiological processes participate in the observed behavioral relations, and neuro- science may tell what they are. However, it will do so by directly investigating those processes and not by inferring that a hypothetical process is at work.

For behavior analysis, the relation between (a) the circumstances in which behavior occurs, (b) the behavior itself, and (c) the consequences of the behavior is called the three-term contingency of reinforcement. A contingency is symbolically expressed as follows:

This expression says that a discriminative stim- ulus (S 0 ) sets the occasion (:) for a response (R) to produce(=>) a reinforcer (SK+). This contingency is the unit of analysis for operant behavior. In a stan- dard, nonhuman example, when a tone is present, a hungry rat presses a lever in the presence of the tone, and a small food pellet follows. The tone is the discriminative stimulus, the lever press is the response, and the food pellet is the reinforcer. Most human behavior is operant behavior. When the telephone rings, a person answers, and talking with a friend is the reinforcer.

The concept of operant behavior relies on the principle of selection by consequences, rather than antecedent causation. The principle of selection became prominent in biology with Darwin, and its application to the study of behavior is straightfor- ward. Just as evolution through natural selection selects organisms that have the morphological characteristics necessary to satisfy the environmen- tal demands for survival, so also does reinforce- ment select responses that have the behavioral characteristics necessary to satisfy the environmen- tal demands of contingencies.

92 Behaviorism

In this way, operant repertoires emerge and develop through the lifetime of the organism: Reinforcing consequences select responses that satisfy the contingencies in an organism's environ- ment; that is, those responses occur more often in the future. Some consequences reinforce operants without any special history, as does food for a hun- gry organism, water for a thirsty organism, and so on. Many, but not all, of these consequences are related to the organism's survival needs. In addi- tion, some consequences acquire their reinforcing effectiveness during the lifetime of an organism as a function of its experiences in the environment. Money is one example. Another is the interper- sonal attention a person receives in social interactions.

A further feature of behavior analysis is that it actually admits unobservable variables. However, it regards them as just as relevant and just as much a part of the behavioral dimension as observable phenomena rather than from a mental dimension. Thus, when people talk about their aches and pains, they are in principle talking about condi- tions that are just as physical as lights and tones are. The difference is that the aches and pains are covert and accessible only to one person, whereas lights and tones are overt and accessible to others.

Similarly, when people think, they are engaging in covert forms of behavior that do not differ from overt forms such as writing notes, reminders, or shopping lists. Covert forms of behavior serve the same function as overt: They contribute to stimu- lus control. In this way, private or covert behav- ioral events are not the same as introspection because the talked-about events, variables, and relations (a) occur in the same dimension as observable behavior; (b) function in the same way as observable events, variables, and relations; and (c) originate and influence subsequent behavior through interaction with environmental circum- stances during one's lifetime.

A summary of potentially helpful comparisons and contrasts follows:

1. Behavior is important as a scientific subject matter in its own right rather than merely as evi- dence to justify inferences about causes in another dimension. The study of behavior is a branch of biology but is not reducible to neuroscience. Neuroscience is an independent science concerned

with how the systems of the body work when an organism interacts with the world. The study of behavior is an independent science concerned with the relation between the behaving organism and the parts of the world with which the organism interacts. Neuroscience and behavior analysis thus provide complementary knowledge.

2. Behavior is a function of "selection by conse- quences" at the levels of phylogeny (species), ontogeny (individual organisms), and culture (soci- ety). The thesis of selection by consequences makes clear how behavior and the physiological systems that underlie it are adaptations selected through interaction with the environment at the three levels.

3. Elicited or respondent behavior may be dis- tinguished from emitted or operant behavior. Respondent behavior is a function of the presenta- tion of an eliciting stimulus. Operant behavior is a function of its consequences.

4. Explanations of behavior follow from analy- ses of the relations that cause the behavior.

a. Explanations of respondent behavior follow from analyses of the relations among stimuli that elicit the behavior.

b. Explanations of operant behavior follow from analyses of contingencies of reinforcement. These contingencies operate across time for the species to which the individual belongs, for individual organisms, and for the social-cultural group to which the individual belongs.

c. Explanations of behavior do not subscribe to mentalism. Mentalism consists in explaining behavior by attributing its cause to phenomena from a dimension that differs from the one in which behavior takes place. Neither the study of behavior nor neuroscience is important because they validate inferences about mental causes. Mental causes are "explanatory fictions." Appeals to these fictions are cherished for quite incidental and irrelevant sociocultural reasons that are unrelated to prediction and control.

5. Verbal behavior is regarded as operant behavior. It is to be given the same analysis as any other form of operant behavior-that is, in terms

of contingencies. In particular, it is not caused by underlying mental processes or mental phenomena.

6. Private behavioral events (verbal reports or covert operants) are important. The relevant ques- tions are (a) how they develop and (b) how they subsequently enter into the contingencies that influence subsequent behavior.

7. Analytic and explanatory terms are generic in nature and are functionally defined. For example, responses and stimuli are members of classes of similar phenomena and are not individual, isolated entities. In addition, they are defined functionally in terms of the environmental relations in which they participate rather than in terms of their physi- cal properties as such.

8. The truth criterion in radical behaviorism is pragmatic-that is, successful working. To ask whether a statement is true is to assess how well and on what basis it promotes effective action, which in science means to predict and control natural events. A statement is a verbal discrimina- tive stimulus. It ultimately enters into contingen- cies of adaptation in which the reinforcer is a beneficial outcome in a person's interactions with the natural environment.

9. Cultures should actively encourage practices that increase, and should actively discourage prac- tices that decrease, the quality of life for their members. This is most reasonably implemented according to known behavioral principles.

Behavior Analysis Versus Mentalism and Methodological Behaviorism

As suggested earlier in paragraph 4(c), Skinner's behavior analysis is vigorously antimentalistic. However, there is much misunderstanding about the relation between behavior analysis and men- talism. Older forms of mentalism, such as those in Watson's time, claimed the mental could be known through introspection. Contemporary mentalism differs. It says the mental can be known through objective inference by following the experimental method. For contemporary mentalism, behavior is said to be explained through the appeal to the operating characteris- tics and capacities of antecedent causes in a

Behaviorism 93

dimension beyond the one in which behavior actually takes place.

Some terms that are often used in connection with this dimension are mental-cognitive, spiri- tual, and subjective-in short, the dimension of "mind." Some terms that are often used in connec- tion with these mental causes are acts, states, mechanisms, processes, entities, and structures. Contemporary mentalism charges that behavior- ism naively tries to explain behavior in terms of observable factors and thereby ignores or denies unobservable mental or cognitive factors that are necessary to explain behavior.

Thus, contemporary mentalism strongly dis- putes that behavior can be explained through an analysis of the relation between behavior and environmental circumstances at a descriptively consistent level for phylogeny, ontogeny, and cul- ture. Mentalism is the dominant orientation in culture as evidenced in cultural patterns of reli- gion and jurisprudence. When compared with behaviorism, mentalism is especially dominant in psychology.

For behavior analysis, if a mental explanation appears to be successful, there are behavioral rea- sons for that success, not mental reasons, and it is the job of psychology to formulate what those reasons are. In any event, the useful distinction is actually not between (a) behaviorism, which includes all forms ranging from classical S-R behaviorism to mediational neobehaviorism to behavior analysis, and (b) mentalism.

Rather, the useful distinction is between (a) behavior analysis and (b) other forms of psy- chology, such as the overt mentalism of cognitive psychology and methodological-behaviorist posi- tions ranging from classical S-R behaviorism to a mediational S-0-R neobehaviorism. Methodo- logical behaviorism is at least incomplete if it doesn't recognize private behavioral events and can be mentalistic in its own right if it appeals to mediating hypothetical constructs as mental causes, which it generally does. Behavior analysis is con- cerned about mediational approaches, mentalism, and methodological behaviorism because they uncritically reflect dominant social-cultural pre- conceptions, such as the dualistic preconceptions of "folk psychology."

For example, many people comfortably invoke "mind" as a cause of behavior, as when they say

94 Big Five Model and Lexical Trait Theory

that they "made up their minds" to stop at the market on the way home from work to purchase a loaf of bread. For behavior analysis, such an account fails to identify what led them to make up their minds. A more useful approach is to identify that purchasing a loaf of bread is relevant because there is no bread at home, that one has prior expe- rience with the market's operating hours and with purchasing bread at the market, and that the mar- ket is conveniently on the route from work to home. All these factors exist in one, behavioral, dimension.

Ultimately, behavior analysts are concerned about mentalism and methodological behaviorism on pragmatic grounds. Behavior analysts argue that a critical examination of those traditions reveals that they obscure and thus actively impede the search for important details about the genuinely relevant relations between behavior and environment. Moreover, they allay curiosity by urging fanciful "explanatory fictions" as causes. Finally, they misrepresent the facts to be accounted for, and they offer false assurances about the state of knowledge. Consequently, mentalism and methodological behaviorism seriously interfere with effective prediction, control, and explanation of behavio1; despite their claim to offer a superior form of psychology.

jay Moore

See also Applied Behavior Analysis; Behavior Therapies; Behavioral Development; Behavioral Economics; Behavioral Health; Behavioral Perspectives on Psychopathology; Behavioral Theories of Personality; Classical Conditioning; Habits; Human Agency; Law of Effect; Operant Conditioning; Pavlov's Legacy; Punishment; Reinforcement; Self-Control; Skinner's Legacy; Watson,]. B.