The New Behaviorism: Perfect Grammar or Poor Poetry?

Hayes, et. al 2004, DBT, FAP, and ACT: How Empirically Oriented are the New Behavior Therapy Technologies?

Several things excited me about this article. Specifically, it was very helpful to see how well-reasoned and promising some of the “new behavior therapies” may be, and how much they may, indeed, not just be “old wine in a new bottle,” due to a renewed focus on non-observable behaviors, in vivo practice/training methods, transportability methods, and clearer models of change. To wit:

-In the description of more comprehensive treatment packages such as DBT, it is exciting to hear about the sequential focus beginning with behavioral training and continuing through “living skills” such as employment and education. One of the biggest problems in the long-term effectiveness of any treatment is the lack of explicit tools to apply it to functional settings in an ongoing way. It seems to me that the addition of such functional modules to existing treatments would be an interesting imperative to put before practitioners of existing ESTs to see how their therapeutic principles apply.

-There were some interesting methods of ensuring agreeable treatment fidelity mentioned in this article. For instance, in comparing CBT to ACT, “the same primary therapist was trained both by Beck and Hayes in their particular form of intervention.” (p. 45) This is a good idea (to have the treatment originators of the competing treatments BOTH be involved in the training process), and, when feasible, can go a long way towards resolving some of the recursive arguing from both sides that can emerge in a comparison study.

-More personally, I was excited to see some of the principles employed in my Spotlight Program reflected (and deemed important) by these new behavior therapies. For instance, the value of in vivo examination (and reframing) because “it is easier to deal with actual relevant behavior within session than with a mere description of the behavior.” (p. 44) is a fundamental approach I’ve been espousing for years, and it is nice to see that others have been adopting it as well. In general, the notion that these approaches are “often more experiential than didactic,” (p. 36) is an exciting discovery for me in light of my experience with other purported behavioral approaches.

An important theme implicit in this article is the connection of the theoretical model of change in a given treatment to the specific disorder it is intend to treat and the consequent projected outcomes. In many of the meta-analyses we’ve read, there have been awfully impressive outcomes reported about effectiveness with a range of disorders. However, there is often not a clear picture of why the researchers figured their treatment might work with these disorders in the first place, and, concurrently, why certain specific outcome measures were chosen. A good instance of where this issue is addressed is Hayes’ et. al’s recount of how ACT was used for patients with positive psychotic symptoms. This particular treatment “targeted acceptance of the private experience of symptoms, defusion from these symptoms, the importance of distinguishing one’s self from the content of one’s thoughts, and the role of … action.” (p. 46) The authors then report that the best outcome in this case was lower levels of rehospitalization at follow-up, but say that, “paradoxically,” more ACT participants admitted to symptoms than TAU controls, and ACT participants showed lower levels of believability. In light of the purported goals of ACT in this case, this is not paradoxical at all. It seems that part of the proposed treatment mechanism is about accepting and owning up to such symptoms (while simultaneously disassociating from them personally), and, importantly, is not trying to extinguish them from existing altogether. As such, it is completely in line with the theory of change underlying ACT in this case that such patients would be happy to report their symptoms if asked, but would not be sufficiently perturbed by them to require subsequent hospitalization. Though the question remains as to whether extinguishing the underlying symptoms might be an important therapeutic adjunct (or alternative), it is important to note the clear relationship here between the theorized change mechanism in ACT, the chosen disordered population, and the subsequent measured (and, in this case, achieved) outcome. This seems a far superior (and more scientifically valid) method of approaching treatment research than that of haphazardly applying a treatment to a disorder and hoping that it sticks, somehow.

Jacobson, Martell, & Dimidjian, 2001, Behavioral Activation Treatment for Depression: Returning to Contextual Roots

It is also consistent with the very approach that, according to Jacobson et. al, 2001, led to the rise of Behavior Activation (in the second article we read) in the first place. This radical behavioral intervention is predicated on the notion that the cognitive components of cognitive therapy for depression are extraneous, and that comparable results can be obtained from focusing primarily on the adaptive behaviors that CT prescribes. Much is compelling to me about BA, despite myself. For instance, I am very partial to the notion of what Jacobson et. al call “focused activation.” Also known as “person-specific motivators” (as used in the Spotlight Program) this idea of using non-generic reinforces is – given its commonsense value -- surprisingly spare in much of the behavioral literature to which I’ve been exposed. That BA embraces it is further evidence of the evolution in thinking that the “new behavior therapies” demonstrate. The contextual approach (which is very much like an environmental or narrativist approach) is very much “old wine,” but is nonetheless compelling in its own right.

That said, I had a few problems with the sort of radical behaviorism espoused by the authors. First of all, the “distinctly behavioral model of depression” (p. 258) that is presented is somewhat problematic. Certainly, there are individuals and circumstances for whom contextual factors are a cause and support of depression. For such individuals (who, it could strongly be argued, may be the majority of those afflicted with major depressive disorder), getting at the events that have led to the particular episode(s) of depression can be fruitful, and doing so in the sophisticated way that BA employs can be valuable for many. However, there are those for whom even a diathesis-stress model of depression is tenuous – for whom a clear-cut cause of the specific depressive cycle is elusive at best. For these individuals, the notions upon which BA is predicated may be more frustrating than helpful – at least at first. This, indeed, may be the problem: for such people, the later steps of BA could be useful, but getting to them in the first place may be quite insurmountable.

Second, it seems ironic that an approach so grounded in visible behavioral processes can be so contingent upon a relational component. Specifically, the authors note that “establishing a collaborative foundation is a critical part of effectively presenting the model and conducting effective BA in general.” (260) This idea is present throughout much of the article, and is a circuitous way of saying: therapeutic alliance really matters here. That this is the case is fairly straightforward and evident: you can’t get a patient to buy into a model that requires so much of them (in terms of both belief and action) without getting them to trust you and want to work collaboratively with you as the therapist. However, this idea (very much relating to the content of the patients thoughts and feelings about the therapist) seems to run counter to the behavioral underpinnings of BA (that it is all context over content). This discrepancy is never reconciled through the article.

Finally, my largest concern emerges from the overall feeling I have about radical behavioral treatments in general. My position is this: they are not inherently wrong; there is certainly (and necessarily, a priori) a behavioral component to both disorders and the capacity to change them. This is true of the psychological condition in general. However, to borrow from my old mentor and esteemed colleague Dr. Karen Levine, to say this is much akin to saying that all sentences have grammar in them. It is definitely true, but it may be beside the point. I can write many sentences about old men in boats, lecherous gentlemen, and obscure neurological conditions, but this does not mean those sentences will be comparable to those written by Hemingway, Nabokov, or Sacks. Emotional content, form, artistry, word choice, and a host of other factors differentiate my sentences from theirs, even though one could properly say that both our form (grammatical structure) and content (topic) were identical. So it goes for behaviorism: I am not convinced that, in getting at the form that necessarily underlies psychological being and action, we have stumbled upon the best or even truest model of change, even if it is the most universal. This, I guess, is what comparative outcome and mechanism studies are all about: does one form of therapeutic prose outstrip another? Do the different forms of form perform differentially – just as Faulkner resonates more with some than Fitzgerald, does Rogers do likewise relative to Beck? Or, for that matter, is all we need the therapeutic equivalent of Strunk & White to cure what ails us?