buylotrogold

It is frequently said that behaviorism ignores feelings, states of mind, and consciousness, but this is far from true. Since these issues are currently popular, it seemed reasonable to briefly outline the Skinnerian position on these topics, at least as I have come to understand it. It is important to differentiate between two common uses of the term consciousness. I will refer to these as consciousness in the general sense and consciousness in the restricted sense. With regard to the former, we tend to say that an organism is conscious simply when it moves around and responds to stimuli. The latter is more difficult to characterize and is frequently thought to be subsumed under the former by non-behavioristic philosophers. The latter meaning is closely allied with the notion of self-awareness. We humans are clearly self-aware - this is evident, of course, since we talk about ourselves and our feelings. But perhaps more compelling is our own phenomenological world; although the philosopher may conjure up a zombie which only appears to describe its feelings but has none, it is easy to convince ourselves that we are not such a zombie. I do not wish to belabor the point here since most readers of this NG will see the kinds of philosophical issues that are being raised. Let me just close this section by reiterating that, to a large extent, non-behavioristic philosophers tend to conflate these two issues; if a rat squeaks and jumps when foot-shock is applied, for example, it is conscious and it feels pain. If it could be taught to tell us it is in pain we would see that it has a phenomenological world similar to our own or, at least, has a phenomenological world. At the heart of the issue is whether or not feelings are to be viewed as causes of behavior. Another way to cast the issue is to ask this question: When we train an operant discrimination (visual, for example), are we training the animal to report what it already sees or are we establishing the seeing itself? VERBAL RESPONSES UNDER DISCRIMINATIVE CONTROL OF ONE'S OWN BEHAVIOR In the behavioristic view we learn to respond to the world discriminatively by having commerce with it. In a typical laboratory setting, we reinforce a response in the presence of a stimulus and withhold reinforcement when the stimulus is not present. The well-known outcome is that the organism comes to respond at a high rate when the stimulus is present and little, or not at all, when it is not present. The non-behavioristic interpretation is that we have simply provided the necessary motivation and skills required for the organism to report what it already (subjectively) sees. The behavioristic interpretation is that the contingencies of reinforcement just described produce the seeing. There appears to be no way to empirically differentiate these two interpretations; since we must arrange contingencies to train the 'reporting response' the door is opened to the interpretation that the seeing itself is the result of the contingencies. In what follows, the latter interpretation is assumed and its ramifications are examined. It follows from the above that there are at least two separate contingencies of reinforcement necessary to produce de_script_ions of our own operant behavior: the first is the contingency that results in the operant itself (response A), and the second is the contingency which produces discriminative control of the verbal operant (response by response A. That is, response A (or stimuli generated by response A) is the discriminative stimulus controlling response B. The nonsocial world may be responsible for response A, but it is the social world (the verbal community ) which arranges the contingency responsible for response B. It follows that response A may occur without the response generated by the contingency arranged by the verbal community (response . An organism which is not part of a verbal community may emit an operant but not be aware of its behavior. This, then, is the difference between consciousness in the general sense and consciousness in the restricted sense. VERBAL RESPONSES UNDER DISCRIMINATIVE CONTROL OF PRIVATE EVENTS According to Skinner, a small part of the world is enclosed within the skin. This portion of the world is not to be accorded a special _meta_physical status by virtue of its locus. Nonetheless, special problems are raised. If we are made aware of features of the world by contingencies of reinforcement, the issue of how the verbal community establishes such contingencies with respect to private stimuli comes to the fore. The issue is similar, as Skinner pointed out, to the case where a color-blind person tries to train color discriminations without special instrumentation. A color-blind person does not know when to reinforce the response red or when to withhold reinforcement. Skinner's contribution was to recognize how the verbal community arranges discriminative contingencies when it does not have direct access to the (private) stimuli which come to control verbal responses. There are four ways. THE FOUR WAYS A verbal response de_script_ive of private events may arise if the verbal community has access to 1) collateral responses which accompany the private event, or 2) public stimuli which accompany the event. Thus, we reinforce the response hurts if we see a child limping or holding their knee or when we observe tissue damage such as a scraped knee. A third possibility 3) arises because behavior may be executed at a scale undetectable publically; the discrimination may be established when behavior occurs publically, but the verbal response may be controlled when the behavior described occurs at a level undetectable by an observer. Finally, a verbal response may be emitted 4) when the event described shares features in common with a publically observable event which already controls a verbal response. Skinner writes, ...we describe internal states as 'agitated,' 'depressed,' 'ebullient,' and so on, in a long list. Responses in this class are all _meta_phors... CONSCIOUSNESS IN NON-HUMAN ORGANISMS It follows from the foregoing that membership in the human species is neither necessary, nor sufficient, for consciousness in the restricted sense. A human raised by wolves (if such a thing is possible) would behave, but it would not be aware that it does so. Conversely, a non-human organism exposed to the requisite history may be aware, in however a limited extent, of its own behavior or states of its own body. Indeed, the laboratory preparation known as drug-discrimination provides just such an example. The term drug discrimination is a misnomer since the organism discriminates, not the drug or injection but, rather, the behavioral effects produced by the drug injection or the resultant states of its body. Similarly, we may establish an operant via a contingency of reinforcement and arrange a separate contingency by which the organism comes to emit a second response under discriminative control of the first. A PERSONAL NOTE Though the foregoing is somewhat brief, disorganized and, in places, redundant, I have chosen not to edit and re-edit my verbal behavior. Were I to have done so, the piece would be much longer and would probably not appear here - it would be sent to a peer-reviewed journal instead. But this seems, to some extent, a waste of time. I have said little that has not already been said before, and said more eloquently, by Skinner. If, however, I had said simply, Hey, why don't those of you interested in consciousness read some Skinner? what would be the reaction of most of the members of this newsgroup? After all, don't we all know that behaviorism is dead? How many of you who are critical of behaviorism or, more apropos to the current Zeitgeist, how many of you who simply ignore behaviorism could have written this as an undergraduate lecture instead of the standard misrepresentation? Respectfully and cordially, Glen