Why behavioral approaches continue to fail: A review of Sidman’s Coercion and its Fallout

Why behavioral approaches continue to fail: a critical review of Murray Sidman’s Coercion and its Fallout. (2001) Boston, MA: Authors Cooperative, Inc. 300 pp.

First, an introductory remark: I noted in my article, The Missing Theory (Wynne, 2004), that I do not distinguish between the various kinds of behavioral approaches to dealing with unwanted behavior. For me, behavior modification, behavior therapy, and applied behavior analysis are essentially identical; I refer to them all as behavioral approaches, as interventions that deal with a person’s actions as they present themselves: what the person is doing, in what situations is s/he doing it, and what are the presumed consequences of his/her actions?

The presence in discussions of human conduct, whether in education, business management, or in clinical efforts, of the term behavioral objectives is all that one needs to recognize the presence of the behavioral approach. This point of view takes the position that it is “behavior,” and only behavior, which is important. I contend that this approach is narrow and, ultimately, ineffective. Murray Sidman, one of B.F. Skinner’s most distinguished colleagues, and still active in the field of applied behavior analysis, ostensibly approaches coercion from this perspective.

One notices immediately that Coercion and its Fallout has no list of references and less than a half-dozen entries in its Index of other people: B.F. Skinner (without which it seems no work of applied behavior analysis is ever considered complete), Fred Keller (Skinner’s long-time colleague), H.L. Cohen and J. Filipczak, authors of A New Learning Environment (based on operant principles), Sigmund Freud (of whom more in a moment) and, in passing, Bobby Fischer of chess fame, and tennis player Martina Navratilova. In the lack of references, Coercion and its Fallout is very much like Skinner’s Science and Human Behavior (Skinner, 1953) written about fifty years earlier. Indeed at a substantive level Coercion and its Fallout adds little to that earlier work.

For this and other reasons the work has a highly abstract quality: it rarely makes contact with any sort of data. For example, although Sidman touches upon terrorism at several points throughout the book, he conflates all terrorist movements. The Tamil Tigers, Al Qaeda, the IRA, Hamas, Hezbollah, and the ETA are not differentiated. That is to say, all are considered equally amenable to understanding and effective intervention through the techniques of applied behavior analysis.

One such technique is withholding reinforcement in the form of press coverage (p. 285). We are led to believe that terrorist attacks could be diminished if the media stopped covering them. Is Sidman serious? Does he really think that publicity constitutes that much of any terrorist’s agenda—especially the agenda of the suicide bomber him/herself for whom, of course, there are no palpable consequences contingent on his/her murderous act? Before venturing into a discussion of counterterrorism, Sidman might have at least acquainted himself with the writings of Max Boot, Thomas X. Hammes, Jon Nagl, or Martin van Creveld.

The cardinal principle of applied behavior analysis is that all behavior is controlled by contingencies in the environment. These contingencies can either be naturally occurring or arranged by one person to control the behavior of another. As a radical behaviorist (Wynne, 2007) I have no quarrel with the first of these; it is the second that is problematic for two reasons.

First there is the question of who decides when the controlling relation has become coercive (or whether it was coercive ab initio).  This question cannot be decided on an a priori basis, except in such obvious inherently coercive cases as slavery, extortion, armed robbery, and blackmail—relationships, however temporary, that Sidman does not list in his index. He does mention marriage, presumably to include divorce, child custody, and other arrangements dealt with in domestic relations courts. He might also have told us whether he considers laws requiring a child to attend a public school in a particular district, or rules within psychology departments requiring students to participate as subjects in experiments, as coercive or as examples of benign control.

On p. 36 Sidman says that “Control by positive reinforcement is noncoercive . . .” This is patently not the case, as Skinner himself noted in his discussion in Beyond Freedom and Dignity (1971, p. 39). Consider this example from my stint as a US Air Force experimental psychologist in the late 1960s. At the human factors laboratory where I was assigned one study was designed to test the limits of human capacity for aerial combat. Enlisted men were given the opportunity to participate in an experiment in which they sat in a seat which dropped them onto a hard unyielding surface.

The challenge was to determine the extent to which the human spine could tolerate the stresses of acceleration and deceleration involved in explosive-powered aircraft ejection seats. The volunteers were required to sign a waiver of their right to sue should they suffer serious injury, and they were paid, if my memory serves, $55.00 per month for their participation. This represented an increase of more than 20% of the monthly salary of these enlisted people, and itself could go a long way toward buying a month’s groceries at the base commissary for a family of three or four people.

I argue that this is coercive positive reinforcement. To be sure, the participants were voluntary, but what does that term mean in a context of barely subsistence wages and the need to support a family? Further examples may be found in the textile factories surrounded by security fences (to keep workers in) allegedly now operated by multi-national corporations on Saipan.

The question whether an arrangement is coercive can only be answered by acknowledging that there is no such a thing as benign control, positive or aversive, without ongoing consent, and that only the person without the means to resist the arrangement may decide whether or not it is coercive.

The second reason that arranged contingencies are problematic deals directly with an issue that Sidman only skirts: what is mental illness or, better, exactly what are we talking about when we use that term? He never offers us any sort of understanding beyond the usual applied behavior analytic cant: “[I]n the end we see mental illness in conduct . . .” (p.180). Indeed he uses the term mental illness, mental illness, without apology.

He suggests weakly that psychiatrists and psychologists, using statistical criteria, equate the abnormal with the unusual. While the unusual is certainly one of the criteria used, there is also the issue of dependence, and so long as the unusual occurs in partnership with financial independence, it is likely to be overlooked. But we should not permit such concerns to divert us from the main point: all that these considerations involve is the labeling of unwanted conduct. (It should also be noted that none of the editions of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders [DSM] has contained any statistics.)

Applied behavior analysts approach all unwanted behaviors as the product of unspecified positive contingencies which are not investigated and verified before new contingencies are arranged and imposed on the person in question. There is nothing in the applied behavior analytic manual of operations which suggests that the person’s history, not only of positive consequences but also of trauma and the family values and rules regarding how such trauma are to be dealt with, should be examined before anything else is done.

Consider for example the thinking behind the school program that required all teen-age girls to carry a 10 lb bag of sugar in a backpack presumably to simulate the difficulty of caring for a baby. The emphasis here is primarily, if not entirely, on the presumed aversive contingencies, but is it not important to consider the historical factors leading to teen pregnancy? Perhaps these factors are what applied behavior analysts are talking about when they speak of setting events or establishing operations, but that is not clear to me. In any case, the same argument could be made against programs to reduce teen drinking, dropping out of school, and a host of other problems.

Applied behavior analysis begins consistently, as Sidman’s discussion demonstrates, with the experimental paradigm of a rat in an operant chamber. (Much of B.F. Skinner’s experimental work was actually done with pigeons, but the problem remains the same: neither of these species talks.)  I argue that the rat analogy prevents behavior analysts from considering the language of the so-called mentally ill. It is not so much the conduct of a person that gets him/her labeled paranoid schizophrenic, but the sorts of things that he/she says.

Skinner insisted that there was no fundamental difference between verbal behavior (language) and non-verbal behavior, saying that talk was subject to fundamentally the same environmental contingencies as non-verbal behavior. Perhaps, but I suggest that the question has not been researched much since Verbal Behavior (Skinner, 1957). I have pointed out elsewhere (Wynne, 2004) that one well-received book on research methods in applied behavior analysis (Johnson and Pennypacker, 1980) made essentially no reference to how research on this topic might be pursued. (For a more recent defense of the applied behavior analytic approach to language, see Leigland, 2007.)

Sidman demonstrates his detachment from the field of mental illness in another significant way: he construes the Freudian terms reaction formation, sublimation, projection, displacement, and regression into the terms of applied behavior analysis. This verbal exercise in translation was for me a déjà vu experience: once again I was sitting in an undergraduate abnormal psychology class longer ago than I care to divulge in this paper (though one may estimate just how long ago that was by my college’s quite-accepted rule that wearing jeans to class by a man was inappropriate, and by a woman unthinkable). I think I can safely say that in well over thirty years in the mental health business I have never, repeat never, heard any of these Freudian constructs used in discussions of the conduct of any so-called mentally ill person. Sidman’s abstract analysis only reinforces (you should pardon the expression) my belief that applied behavior analysis has, at least thus far, taught us nothing about so-called mental illness.

Reconstruction into the terms of applied behavior analysis is something that is seen quite often in the pages of “behavioral” journals. One such example is Nelson-Gray, et al., “The development and maintenance of personality disorders: A behavioral perspective” (2007). My objection to this paper goes beyond remarks I have made earlier (Wynne, 2004) that analysis must be followed by effective intervention in order to validate it. Here I argue that Nelson-Gray’s analysis begins with the explicit acceptance (a) of the DSM as an authoritative and scientific document (as opposed to a political statement of unwanted behaviors in the United States of the late twentieth and early twenty-first centuries) and (b) that there is such a thing as a personality disorder.

Nelson-Gray, et al., would have done better if they had confined themselves to talking about the unwanted conduct itself, although issues of coercion would still have remained. There is nowhere in their extensive review of the literature any mention of what consents were obtained, or if the subjects/patients/clients were paid for their participation in the various studies.

In their enthusiasm to implement their techniques, applied behavior analysts seem at the very least to have put patient rights on the back burner. The issue is all the more poignant because people who are desperate, either for financial reasons as in the case of the Air Force subjects, or because their child behaves in strange and distressing ways, as in, for example, autism, will sign their consent to almost any intervention.

Sidman says (p. 180): “Clearly, many factors can contribute to mental illness and any particular case requires consideration of all possibilities, social and individual, internal and external.” This sentence is breathtaking in its simplicity; indeed, it actually says nothing at all. But it is the use of the word internal that is most revealing—especially since it is being used by a behaviorist. What does Sidman allude to here—surely not the Skinnerian concept of the private event, which is in danger of becoming the behaviorist’s golden parachute to be used whenever the context appears to be failing to account for behaviors of interest? (I note with alarm the creeping cognitivism in the editorial statement of the on-line journal The Behavior Analyst Today. www.behavior-analyst-today.com)

Coercion and its Fallout does little to allay our fears regarding the prevalence of coercion in our lives. While I agree with Sidman that environmental control is the ultimate “prime mover” of human action, it is quite another thing to participate in that control in the name of investigative science or clinical intervention. Such participation, as Thomas Szasz has shown repeatedly, involves deprivation of liberty and personal degradation. That is why most of both the research and the clinical work done by applied behavior analysts has been done with participants who have had little or no freedom to refuse. Legally competent adults will not voluntarily participate in either (Wynne, 2004).

Sidman’s agenda is to expand a theory of behavior that continues to depend on laboratory animal studies (as well as on hypothetical translations) into the arena of human conduct in all of its complexities—from terrorism to mental illness. This agenda has failed, I believe, because it is arcane, constricted, and meaningful only to academics.


Buchanan, J.A., Christenson, A.M., Ostrom, C., and Hofman, N. (2007). Non-pharmacological interventions for aggression in persons with dementia: A review of the literature. The Behavior Analyst Today, 8, (4), 413-425.

Cohen, H.L., and Filipczak, J. (1971). A new learning environment. San Francisco: Jossey-Bass.

Johnston, J.M., and Pennypacker, H.S. (1980). Strategies and tactics of human behavioral research. Hillsdale, NJ: Lawrence Erlbaum.

Leigland, S. (2007).  Fifty years later: Comments on the further development of a science of verbal behavior. The Behavior Analyst Today, 8, (3), 336-346.

Nelson-Gray, R.O., Mitchell, J.T., Kimbrel, N.A., and & Hurst, R.M. (2007). The development and maintenance of personality disorders: A behavioral perspective. The Behavior Analyst Today, 8, (4), 443-482.

Skinner, B.F. (1953). Science and human behavior. NY: The Free Press.

Skinner, B.F. (1957). Verbal behavior. NY: Appleton-Century-Crofts.

Skinner, B.F. (1971). Beyond freedom and dignity. NY: Alfred A. Knopf.

Wynne, L. (2004). The missing theory: Why Behavior Analysis has little impact on voluntary adult outpatient services. Ethical Human Psychology and Psychiatry, 6 (2), 135-146.

Wynne, L. (2007). The missing theory II: Further musings on a new zeitgeist for clinical psychology. Ethical Human Psychology and Psychiatry, 9 (2), 87-98.

© Louis Wynne 2008