Originally presented 5-12-79 as "The Practice of Personhood in PSI, Keller's Personalized System of Instruction" at a conference sponsored by The Center for Personalized Instruction, Georgetown University, Washington, D.C.

edited 8/17/11

B. F. Skinner's Concept of Person
1999 Edward G. Rozycki


(Our previous speaker,) Professor Heslep undertakes an exegesis of B. F. Skinner's writings with the intent of uncovering a coherent notion of what it is to be a person. This concept is intended to enhance the theoretical base of PSI, the Personalized System of Instruction. Skinner's work is examined because he has been a major proponent of theories that the founders of PSI understand their system to have derived from.

I have little quarrel with Heslep's exposition. I would suggest, however, that he is looking in the wrong place. There may be some historical relationship between PSI and Skinner's theories. One would do better, nonetheless, to look at PSI's present practices to discern what conception of the person is operative in them, rather than to worry over the the philosophical musings of B.F. Skinner. To make an analogous point: we would not expect to explicate the practices of modern Mormonism by researching the Dead Sea scrolls; yet, when it comes to explicating the practice of PSI, Skinner's scrolls are, at the very least, dead.

Because I have, I fear, greatly understated my case, I will review the considerations upon which it is built. Thereafter I will lay out what I take to be the fundamental assumption about personhood operative in PSI.


There are two prima facie grounds for dismissing Skinner as a source of a coherent account of the person in PSI. The first is that operant theory has, in the considered judgment of sympathetic observers, failed to deliver anything even near to what Skinner has promised. The second is that Skinner has rightly been criticized for persistent incoherence in his notions of reinforcer, stimulus and behavior - which have been his primary concerns. Thus there is little reason to expect a minor aspect of his work, his conception of the person, to be more coherent.

Let me expand these points. In his presidential address to the Division of General Psychology, APA, entitled, "The Decline and Fall of the Laws of Learning" (1), Wilbert J. McKeachie reviews the literature on the application of Skinnerian theory and concludes that "the research evidence...demonstrates that each point enunciated by Skinner" (in his Technology of Teaching) "is untrue - at least in as general a sense as he believed. (2) McKeachie offers an explanation for the popularity of Skinnerian psychology despite its inadequacy: a) it is simple; b) educators are looking for magic; and c) Skinner is, at least, persuasive.

Those who buy his approach find that the basic ideas are simple to apply and work often enough to maintain their enthusiasm. This is no small matter. Anyone who can take discouraged, dispirited teachers, mental health aides or prison officials and revive their hope and vigor has done a great deal. Probably no one thing is more important in education than the teacher's enthusiasm and energy.(3)
I take this to be a diplomatic sop to pacify the more zealous Skinnerians who might otherwise find McKeachie's paper to leave little for an honest psychologist to hold to. But as a long time secondary school teacher I reject McKeachie's condoning Skinner's offering his nostrums to a clientele supposedly too ignorant to tell snake oil from real medicine. I can attest, also, as a teacher of teachers, that the failure of Skinner's promises have generated an extreme wariness towards any systematic approach to educational problems (4). Even as psychologists decry the invasion of our schools by astrologers and assorted gurus, they fail to recognize that, in consequence of the let-down over Skinnerian theory, teachers regard them, too, the academically-trained psychologists, as one more variety of occult practitioner.

I cite McKeachie not as the sole example of a critic of Skinner, but rather as a prominent and yet sympathetic one. Those of us who have had to modify our token economies, for example, in the face of real-world exigencies could no doubt hold forth at length on Skinner's ambiguities in such crucial concepts as stimulus property, response, and reinforcer. Lacking sympathy to leaven the severity of our judgment, we need but few abrupt words to accurately describe Skinner's educational offerings, for example, that boring device, the teaching machine.

As a college teacher, I have found one inadvertant contribution by Skinner to pedagogy: his Beyond Freedom and Dignity - an informal measure of an interesting type. A student's grasp of the relevant issues, and his or her knowledge of the relevant psychological literature is almost inversely proportional to his or her enthusiasm for the arguments of the book. But let's leave this hypothesis for someone's dissertation.


It is perhaps unfair to charge Skinner with incoherence, for if he is right, the notion of coherence is irrelevant. A sequence of utterance may be more or less effective than another, but that is an empirical matter. Coherence is not a matter of the effect of utterances, but of the relations among them by virtue of their structure and content. But such constructs, structure and content, are inadmissable in Skinner's universe for they rest on the postulation of inner processors and other mediating variables that Skinner assures us are at best superfluous. Now, every other school kid carries a pocket calculator containing hundreds of inner processors and mechanisms which mediate between its inputs and outputs. Yet , Skinner, as late as the early seventies, thinks it is a criticism of rival theories that they invent mediators between stimulus and response: "Information theory ran into ... (this) ... problem when an inner 'processor' had to be invented to convert input into output."(5) If you find yourself dismayed by the almost quaint perverseness of this statement, calm yourself! Just tell yourself that it's an utterance, a mere utterance; its meaning is superfluous.

Of course, one might reclaim a certain kind of consistency for Skinner by looking at his works developmentally and then conceding that he has given up his own theory; thus we wait in vain for answers to questions such as:

Other psychologists, even other behaviorists, might have a reasonable go at these questions. But that peculiarly restricted form of behaviorism that is Skinner's (6) cannot deal with them. Thus the relevance of Skinner to wider issues in psychology, education and social policy is a baseless presumption.

Does operational definition of operant and reinforcer define independent variables? No (7), although Skinner thinks they do. Is what the layman calls a reward really a reinforcer? Skinner might wish it were.(8) Are the internal states of the organism functionally epiphenomenal? Skinner's yes is just wishful thinking and on the role of wishful thinking in science I will quote B.F. Skinner:

Science is a willingness to accept facts even when they are opposed to wishes. Thoughtful men have perhaps always known that we are likely to see things as we want to see them instead of as they are, but thanks to Sigmund Freud we are today much more clearly aware of "wishful thinking." The opposite of wishful thinking is intellectual honesty -- an extremely important possession of the successful scientist.(9)

The fundamental assumption made by practitioners of PSI about persons is that persons can be practitioners, exercisers of options. Not every practice is appropriate to PSI. By way of contrast, on Skinner's theory, no persistently instantiated response type lacks control by some reinforcer. All practice must be accord with Skinner's theory! PSI's theory, or any of its competitors may be contravened, but not Skinner's.

Now, we may want other people, or ourselves, even, to behave in one manner and yet we may reinforce incompatible behavior. On Skinner's account, this merely means that our wants are functionally inoperative on the desired behavior. Why do Skinner and disciples think that wanting to make one's wants functionally operative -- by, say, availing oneself of Skinnerian theory -- is itself a functionally operative want?

This is a crucial point. I will recast it. Why should wanting to make our wants functionally operative, i.e. to control behavior, be a more functionally operative want than any of the many dismissed by Skinner as not operative?

Skinner's sleight-of-locution has been to insinuate control, functionally operative wants, to within the individual - to within his readers, at least -- while otherwise professing externality of control -- accessible, of course, to his readers. But if we take his account seriously, we have no options. What we do, we do. And our wants may or not be functionally operative in this doing. Pure luck.

Should the practitioners of PSI introduce torture into their programs? What could this consideration mean on Skinner's account? All those wantings and deliberations and weighings and speculations that bear on rational practice as we understand it, are, for Skinner, either part of a functional chain, or not. And if not, then what?

One theorist has suggested that the question, "What ought I to do?" should be understood by the Skinnerian as "What will reinforce me?" But this is a hopelessly circular question since it is increased frequency of a response which defines a reinforcer for its response class.(10) Nor can knowledge be invoked without conceding the usefulness of certain "mentalistic" constructs long anathematized by Skinner.

Now it would be easy enough to go on to show that PSI makes certain assumptions about the validity of a myriad of so-called "mentalistic" constructs on the basis, for example, that it utilizes a concept of language, which certainly does not merely refer to a congeries of environmental stimulus-properties.(11) Another major factor is PSI's attention to individual difference among dimensions not conceivable on a Skinnerian model. But time is short and I would merely point out that PSI can be accommodated to any of the concepts of personhood laid out so well by Professor Heslep.

I would issue only one caveat. Skinner's truncated behaviorism with its assumptions of accessible controlling variables maintains his popularity among would-be social engineers and moral know-it-alls who don't want to bother themselves with ethical questions. Skinner is properly recognized as a pioneer in the development towards a scientific psychology. But his theories are obsolete; their social policy extrapolations are atavisms. It would indeed be tragic if the practitioners of PSI, who have come a long way from where Skinner left them off, misconstrue their obligations to his proper appreciation by furthering that unhappy Skinnerian legacy that rationalizes one's pursuits beyond considerations of freedom and dignity.


(1) Wilbert J. McKeachie, "The Decline and Fall of the Laws of Learning" Educational Researcher, March 1974, pp.7-11.

(2) McKeachie, p.10

(3) McKeachie, p.10

(4) Cf. Meryl E. Englander, "Educational Psychology and Teacher Education," Phi Delta Kappan, March 1976, pp. 440-443

(5) B. F. Skinner, Beyond Freedom and Dignity, (New York: Bantam, 1972) p. 15.

(6) Cf. Kenneth W. Spence, "The Postulates and Methods of 'Behaviorism"' in Herbert Feigl and May Brodbeck (eds.) Readings in the Philosophy of Science (New York: Appleton, 1953) p. 571. Skinner notwithstanding, Behaviorism does not require that controlling variables be accessible to the experimenter.

(7) Cf. Karl Schick, "Operants," Journal of the Experimental Analysis of Behavior, May 1971,15/3, pp. 413-423.

(8) Cf. Edward G. Rozycki, "Rewards, Reinforcers and Voluntary Behavior" Ethics, 84/1, Oct 1973, pp. 38-47. Also, "More on Rewards and Reinforcers" Ethics, 84/4, July 1974, pp. 354-358.

(9) B. F. Skinner, Science and Human Behavior (New York: Free Press, 1953) p.12. But see EG Rozycki (1995), Critical Review of B.F. Skinner's Philosophy. 

(10) Cf. Max Hocutt, "Skinner on the Word, 'Good": a naturalistic semantics for ethics" Ethics 87/4. July 1977, pp., 319-338.

(11) Schick, p. 422 argues that Skinner defines 'mand' in terms of properties to which it could not be conditioned.