(Redirected from Behavior analysis
Behaviorism (or behaviourism) is an approach to psychology based on the proposition that behavior is interesting and worthy of scientific research. Within that broad approach, there are different emphases. Some behaviorists argue simply that the observation of behavior is the best or most convenient way of investigating psychological and mental processes. Others believe that it is in fact the only way of investigating such processes, while still others argue that behavior itself is the only appropriate subject of psychology, and that common psychological terms (belief, goals, etc.) have no referents and/or only refer to behavior. Those taking this point of view sometimes refer to their field of study as behavior analysis or behavioral science rather than psychology.
J. B. Watson
Early in the 20th century, John B. Watson argued in his book Psychology from the Standpoint of a Behaviorist for the value of a psychology which concerned itself with behavior in and of itself, not as a method of studying consciousness. This was a substantial break from the structuralist psychology of the time, which used the method of introspection and considered the study of behavior valueless. Watson, in contrast, studied the adjustment of organisms to their environments, more specifically the particular stimuli leading organisms to make their responses. Most of Watson's work was comparative, i.e., he studied the behavior of animals. Watson's approach was much influenced by the work of Russian physiologist Ivan Pavlov, who had stumbled upon the phenomenon of classical conditioning (learned reflexes) in his study of the digestive system of the dog, and subsequently investigated the phenomena in detail. Watson's approach emphasized physiology and the role of stimuli in producing conditioned responses - assimilating most or all function to reflex. For this reason, Watson may be described as an S-R (stimulus-response) psychologist.
Watson's behaviorist manifesto persuaded most academic researchers in experimental psychology of the importance of studying behavior. In the field of comparative psychology in particular, it was consistent with the warning note that had been struck by Lloyd Morgan's canon, against some of the more anthropomorphic work such as that of George Romanes, in which mental states had been freely attributed to animals. It was eagerly seized on by researchers such as Edward L. Thorndike (who had been studying cats' abilities to escape from puzzle boxes). However, most psychologists took up a position that is now called methodological behaviorism : they acknowledged that behavior was either the only or the easiest method of observation in psychology, but held that it could be used to draw conclusions about mental states. Among well-known twentieth-century behaviorists taking this kind of position were Clark L. Hull, who described his position as neo-behaviorism, and Edward C. Tolman, who developed much of what would later become the cognitivist program. Tolman argued that rats constructed cognitive maps of the mazes they learned even in the absence of reward, and that the connection between stimulus and response (S->R) was mediated by a third term - the organism (S->O->R). His approach has been called, among other things, purposive behaviorism.
Methodological behaviorism remains the position of most experimental psychologists today, including the vast majority of those who work in cognitive psychology – so long as behavior is defined as including speech, at least non-introspective speech. With the rise of interest in animal cognition since the 1980s, and the more unorthodox views of Donald Griffin among others, mentalistic language including discussion of consciousness is increasingly used even in discussion of animal psychology, in both comparative psychology and ethology; however this is in no way inconsistent with the position of methodological behaviorism.
B.F. Skinner and radical behaviorism
B.F. Skinner, who carried out experimental work mainly in comparative psychology from the 1930s to the 1950s, but remained behaviorism's best known theorist and exponent virtually until his death in 1990, developed a distinct kind of behaviorist philosophy, which came to be called radical behaviorism. He also claimed to found a new version of psychological science, which he called behavior analysis or the experimental analysis of behavior.
Definition of radical behaviorism
Skinner was influential in defining radical behaviorism, a philosophy codifying the basis of his school of research (named the Experimental Analysis of Behavior, or EAB.) While EAB differs from other approaches to behavioral research on numerous methodological and theoretical points, radical behaviorism departs from methodological behaviorism most notably in accepting treatment of feelings, states of mind and introspection as existent and scientifically treatable. This is done by identifying them as something non-dualistic, and here Skinner takes a divide-and-conquer approach, with some instances being identified with bodily conditions or behavior, and others getting a more extended 'analysis' in terms of behavior. However, radical behaviorism stops short of identifying feelings as causes of behavior. Among other points of difference were a rejection of the reflex as a model of all behavior and a defense of a science of behavior complementary to but independent of physiology.
Skinner's conceptual innovations
This essentially philosophical position gained strength from the success of Skinner's early experimental work with rats and pigeons, summarised in his books The Behavior of Organisms (1938) and Schedules of Reinforcement (1957, with C. B. Ferster). Of particular importance was his concept of the operant response, of which the canonical example was the rat's lever-press. In contrast with the idea of a physiological response, an operant is a class of structurally distinct but functionally equivalent responses. For example, while a rat might press a lever with its left paw or its right paw or its tail, all of these responses operate on the world in the same way and have a common consequence. Operants are often thought of as species of responses, where the individuals differ but the class coheres in its function--shared consequences with operants and reproductive success with species. This is a clear distinction between Skinner's theory and S-R theory.
Another crucial contribution was his clarification of the key concept of reinforcement, which had been introduced by Thorndike and used extensively by Hull, but seemed to be mired in issues of definitional circularity. Whereas Thorndike had tried to define reinforcement mentalistically, as a "satisfying state of affairs", and Hull had tried to define it physiologically, in terms of the reduction of a drive, Skinner defined it empirically: if an event was experimentally observed to increase the rate of response, it was then called a reinforcer for that particular animal at that time. Food, water, brain stimulation, sex, social contact, and reinforcing drugs are all reinforcers that have been used in operant research with animals. The issue of whether these stimuli were satisfying to the animal (Thorndike's definition) was thereby bypassed, and the issue of whether they involved the reduction of a drive was left open for empirical physiological investigation (and it was quickly realised that many do not).
Skinner's empirical work expanded on earlier research on trial-and-error learning by researchers such as Thorndike and Guthrie with both conceptual reformulations – Thorndike's notion of a stimulus-response 'association' or 'connection' was abandoned – and methodological ones – the use of the 'free operant', so called because the animal was now permitted to respond at its own rate rather than in a series of trials determined by the experimenter procedures. With this method, Skinner carried out substantial experimental work on the effects of different schedules and rates of reinforcement on the rates of operant responses made by rats and pigeons. He achieved remarkable success in training animals to perform unexpected responses, and to emit large numbers of responses, and to demonstrate many empirical regularities at the purely behavioural level. This lent some credibility to his conceptual analysis.
Radical behaviorism and language
As Skinner turned from experimental work to concentrate on the philosophical underpinnings of a science of behavior, his attention naturally turned to human language. His book Verbal Behavior (1957) laid out a vocabulary and theory for functional analysis of verbal behavior. This was famously attacked by the linguist Noam Chomsky, who presented arguments for the bankruptcy of Skinner's approach in the domain of language and in general. Skinner did not rebut the review, later saying that it was clear to him that Chomsky hadn't read his book (though subsequent rebuttals have been provided by Kenneth MacCorquodale and David Palmer, among others). Skinner's supporters claim Chomsky's consideration of the approach was superficial in several respects, but the appropriate subject for a study of language was a major point of disagreement. Chomsky (like many linguists) emphasized the structural properties of behavior, while Skinner emphasized its controlling variables.
What was important for a behaviorist analysis of human behavior was not language acquisition so much as the interaction between language and overt behavior. In an essay republished in his 1969 book Contingencies of Reinforcement, Skinner took the view that humans could construct linguistic stimuli that would then acquire control over their behavior in the same way that external stimuli could. The possibility of such instructional control over behavior meant that contingencies of reinforcement would not always produce the same effects on human behavior as they reliably do in other animals. The focus of a radical behaviorist analysis of human behavior therefore shifted to an attempt to understand the interaction between instructional control and contingency control, and also to understand the behavioral processes that determine what instructions are constructed and what control they acquire over behavior. Important figures in this effort have been A. Charles Catania and C. Fergus Lowe .
Molar versus Molecular Behaviorism
Skinner's view of behavior is most often characterized as a "molecular" view of behavior, that is each behavior can be decomposed in atomistic parts or molecules. Molar behaviorists (e.g. Howard Rachlin ) argue that behavior can not be understood by focusing on events in the moment. That is, they argue that a behavior can be understood best in terms of the ultimate cause of history and that molecular behaviorist are committing a fallacy by inventing a ficticious proximal cause for behavior. Molar behaviorist argue that standard molecular constructs such as "associative strength" are such ficticious proximal causes that simply take the place of molar variables such as rate of reinforcement. Thus, a molar behaviorist would define a behavior such as loving someone as a exhibiting a pattern of loving behavior over time, there is no known proximal cause of loving behavior (i.e. love) only a history of behaviors (of which the current behavior might be an example of) that can be summarized as love.
Behaviorism in philosophy
Although behaviorism is commonly thought of as a psychological movement, most modern behaviorist's would agree that Behaviorism is a philosophy and upon which the science of the Experimental Analysis of Behavior is based. The basic premise of the behaviorist movement is that the study of behavior should be a natural science, such as chemistry or physics, without any reference to hypothetical inner states of organisms. Thus, behaviorism takes a functional view of behavior. Whether the science of the experimental analysis of behavior is rightly considered part of psychology, separate from psychology, or the only "true" psychology is still a topic debated by modern behaviorists.
There are points of view within analytic philosophy that have called themselves, or have been called by others, behaviorist. In logical behaviorism (as held, e.g., by Rudolf Carnap and Carl Hempel), psychological statements meant their verification conditions, which consisted of performed overt behavior. W. V. Quine made use of a variety of behaviorism, influenced by some of Skinner's ideas, in his own work on language. Gilbert Ryle defended a distinct strain of philosophical behaviorism, sketched in his book The Concept of Mind. Ryle's central claim was that instances of dualism frequently represented 'category mistakes,' and hence that they were really misunderstandings of the use of ordinary language.
It is sometimes argued that Ludwig Wittgenstein defended a behaviorist position, and there are important areas of overlap between his philosophy, logical behaviorism, and radical behaviorism (e.g., the beetle in a box argument). However, Wittgenstein was not a behaviorist, and his style of writing is sufficiently elliptical and allusive to admit of a range of interpretations. Equally contentious is an identification of Alan Turing as a behaviorist.
Leading developers of behaviorism (in rough chronological order):
Critics of Behaviorism
Dictionary of the History of Ideas: Behaviorism
References and further reading
- Chomsky, N. (1959). A review of B. F. Skinner's Verbal Behavior. Language, 35(1), 26-58. (on-line)
- Ferster, C. B., and Skinner, B. F. (1957). Schedules of reinforcement. New York: Appleton-Century-Crofts.
- Mills, John A., Control: A History of Behavioral Psychology, Paperback Edition, New York University Press 2000
- Skinner,B.F., Beyond Freedom & Dignity, Hackett Publishing Co, Inc 2002
- Skinner, B. F. (1938). The behavior of organisms. New York: Appleton-Century-Crofts.
- Skinner, B. F. (1945). The operational analysis of psychological terms. Psychological Review. 52, 270-277, 290-294.
- Skinner, B. F. (1953). Science and Human Behavior (ISBN 0029290406)
- Skinner, B. F. (1957). Verbal behavior. Englewood Cliffs, NJ: Prentice-Hall.
- Skinner, B. F. (1969). Contingencies of reinforcement: a theoretical analysis. New York: Appleton-Century-Crofts.
- Watson, J. B. (1913). Psychology as the behaviorist views it. Psychological Review, 20, 158-177. (on-line)
- Watson, J. B. (1919). Psychology from the Standpoint of a Behaviorist
Last updated: 05-07-2005 17:34:58
Last updated: 05-13-2005 07:56:04