Life Coaching

Communication Coaching

Coach Bios





 For more information, contact:

   Gary Schouborg, PhD

   (925) 932-1982














Schouborg, Gary (2014).

"Reducing Uncertainty: Science and Psychotherapy”.



Reducing Uncertainty: Science and Psychotherapy


Gary Schouborg


Abstract. The APA proposed ideal of the therapist as scientist-practitioner depends on the relevance of science to psychotherapy. A helpful way to address that issue is to back into it by starting with an idealized scenario where all therapeutic interventions are completely intuitive, inexplicable, and effective. Consequently, empirical science is necessary only for confirming that healing has occurred, and then that the therapists were the cause. As successive scenarios become increasingly more like the real world, empirical science becomes increasingly necessary not only to differentiate among different results, but also to understand why the therapists differ among themselves in their success rate, in the specific problems they can heal, and in the degree and speed of improvement they can achieve. This perspective illuminates the strengths and weaknesses of an account by Sundararajan of a representative critique of evidence-based therapy (EBT) by humanistic-phenomenological therapy (HPT).

Backing into the Issue

Scenario 1

The proposed ideal of the therapist as scientist-practitioner depends on the relevance of science to psychotherapy. A helpful way to address that issue is to back into it by starting with an idealized Scenario 1. Suppose that all therapists can completely heal all clients of any psychological ill by using completely intuitive methods, whose nature neither the therapists nor anyone else can identify. Would we not say that the therapists were practicing art, not science, since the latter involves identifying at least some causes of the results? In any case, would we even care how we characterize the therapists, as long as they completely cured every psychological ill presented to them? Would government agencies, insurance companies, or individual clients care as long as the therapists could achieve their results at a cost that everyone could afford? Even academic departments would not care, if the therapists could teach their skills to students by merely having them hang around and intuitively learn to respond to clients just as the therapists do.

Where does science come into Scenario 1? Even in this idealized circumstance, science is necessary to establish that all the clients are completely healed. Even if the therapists intuitively know that they themselves have completely healed their clients and that they can continue to do so for any client they come across, those of us outside their circle do not have that luxury. We must plod along by explicitly describing each illness and establishing criteria that tell us that it has indeed been completely healed. In short, to know whether the therapists are real and complete healers, we must do science.

Furthermore, we must do more science still if we are to know if the therapists are reliable healers. Maybe they are merely enjoying an extraordinary run of luck. To conclude that they are reliable healers, we must have reason to believe that there are some factors in their intuitive processes that produce healing in a way that assures us that they will continue to heal in the future. That involves understanding at least three factors: the nature of the healing processes, the therapists’ cognitive abilities, and their motivation. The healing processes must be of such a nature that they will continue to heal future illnesses; the therapists’ cognitive abilities must be able to continue to produce those healing processes; and their motivation must be to continue their healing. Understanding each factor requires yet further scientific study. Again, the therapists themselves know the answer to all these questions; but we who fund them or seek their help do not.

Suppose that the therapists’ intuitions defy unpacking. Still, we can reasonably conclude that the longer their complete success, the less likely it is merely an amazing string of luck and the more likely it is that there are efficacious cognitive and behavioral factors in their practice, however unknowable they may be. That conclusion is itself a scientific inference about causes, effects, and probabilities.

Scenario 2

Now consider Scenario 2, where therapists practice completely intuitively and heal completely; but they vary in the time they take to achieve their results. Here, all the possibilities for scientific study in Scenario 1 apply, with the addition of determining the various healing times. 

Scenario 3

In Scenario 3, therapists practice completely intuitively but achieve the mixed results that we currently see in the real world. Here, all the possibilities for scientific study in Scenario 2 apply, with the addition of determining the specific ills treated, degrees of healing, and healing times.

Scenario 4

Scenario 4 adds to Scenario 3 the ability of researchers to identify at least one therapeutic process employed. Here, all the possibilities for scientific study in Scenario 3 apply, with the addition of determining the possible causal nexus between results and identified processes.

Engineer v. Tinkerer :: Coke v. Gourmet Wine

Sundararajan’s account

In addressing the current scientist-practitioner debate, Sundararajan (2002) applies a distinction between engineer and tinkerer that arises from the humanistic-phenomenological therapy tradition (HPT). Her aim is to critique what she considers to be a dichotomy between science and practice that arises from the evidence-based therapy (EBT) tradition.

She informs us that the distinction between engineer and tinkerer begins with Levi-Strauss (1966), Bourdieu (1990) later explaining that the engineer applies "scientific truth”—abstract and a-temporal concepts—to achieve a pre-conceived goal, whereas the tinkerer achieves “practical truth” in responding to living, continually changing contexts.

Sundararajan adds that Merleau-Ponty (1962), Dreyfus (1982), and Searle (1983) explain how novice practitioners differ from experts in being so dependent on rules that they are are unable to make the increasingly nuanced responses required for ever more complex and shifting conditions. She further cites Bernet: the expert’s “awareness via the body’s feelings, is much more rapid, and much more person- and context-specific, than cognitively processed information such as learned appraisal skills” (pp. 38-39, quoting Bernet (1996, p. 5), my italics). She concludes that insistence on EBTs therefore mistakes psychotherapy with engineering. She draws a parallel between EBTs in psychotherapy and the chemical testing of wines, which is useful only for identifying “the Cokes of the wine world” (p. 46) but cannot capture finer qualities.

Sundararajan draws an important difference between an arguably predominant form of EBT and more intuitive therapeutic approaches. However, her view raises both theoretical and practical issues: 1) it exaggerates the difference between engineer and tinkerer; 2) it overlooks rules as explanatory concepts, not just explicit instructions; 3) it fails ever to address the Motorola executive’s concern that she uses to introduce the debate; but 4) it legitimately brings to our attention that therapeutic practitioners cannot be dismissed as lacking expertise simply because current scientific methods cannot explain and measure what they do.

Theoretical problem 1

Sundararajan claims that therapeutic tinkering is “a unique way of knowing” (p. 35; my italics), distinctive in not being “applied theory,” a concept she never directly explains.  However, she implies that no empirical scientific theory is relevant to therapeutic tinkering, since the latter has its own “logic of practice … radically different from empirical science and technology” (p. 35, my italics). Thus, therapeutic tinkering could never be “evidence-based”: measuring to what degree an intervention is successful in achieving a pre-determined therapeutic goal. On Sundararajan’s account, therapeutic clients are constantly shifting participants, so that therapeutic goals “depend on the opportunities” (p. 36, quoting Jacob (1977, p. 1164) that arise during a session.

Sundararajan, then, begins by chaining empirical science and EBTs to rigid, pre-determined goals and offering a “radically different” alternative. However, the only problem she identifies is that because clients are constantly shifting participants, a therapeutic tinkerer can have no pre-determined goals, but must be opportunistic. At best, this brings to our attention that some EBTs may be simplistic in framing the therapeutic process; but it fails to show that clients have no goals against which successful therapy can be measured. In other words, as it stands, her account does not necessarily require a “radically different” approach, only a more sophisticated one commensurable with empirical science. After all, bridges are not simply deduced from general scientific principles; the engineer as well as the artist or expert therapist must ultimately address conditions unique to the particular bridge being built, ensuring that it will continue to stand up under constantly changing physical conditions. Admittedly, the therapist faces a more mercurial challenge; but Sundararajan identifies nothing that shows it to be “radically different” rather than just more complex than the engineer’s.

Sundararajan rightly points out that therapeutic goals may shift in the moment as issues spontaneously emerge. Again, the engineer may face this challenge as well when faced with unexpected challenges. Yes, the goal remains of building a bridge; but the therapist sustains similar continuity through a set of therapeutic goals, shifting from one to another as opportunities arise. It is not as if the practitioner shifts from reducing emotional suffering to repairing a herniated disk. There is no reason in principle why EBT cannot measure success in terms of a generalized goal—for example, reducing suffering—or a more specific goal picked from an appropriate set or range.

Theoretical problem 2

The second theoretical problem is that Sundararajan overstates the difference between engineer and tinkerer by conflating the phenomenology of rule following with the science of it. It is uncontroversial that expert response is too complex and too rapid to follow rules consciously; but that does not prevent empirical scientists from discovering complex rules that drive expert responses below expert awareness. Intuition is not necessarily the absence of rule following, only the absence of following rules consciously. Whether expert therapeutic rules exist is an empirical question; they cannot be decreed impossible a priori just because experts are unaware of them. Again, nothing in her account requires any “radically different” approach from EBTs, only a more sophisticated framing of the therapeutic process and more sophisticated ways of measuring outcomes, where possible.

Practical problem 1

Even if the two theoretical problems could be resolved, so that it could be established that expert therapists have a way of knowing that is incommensurable with empirical science and allows them to practice without rules, a practical one remains. If expert therapy really is incommensurable with engineering processes, so that measuring outcomes by empirical science is completely impossible, Sundararajan presents the rest of us with the equivalent of Scenario 2. All of us bereft of expert therapists’ skills have no way of knowing whether the experts enjoy “a unique way of knowing” and are actually healing their clients. In fact, we are really in Scenario 3, since she tells us only that experts transcend rules, which by itself does not tell us that they are experts—that is, that they know anything or are ever successful. Furthermore, she lays out no path for the rest of us to understand how often error occurs—if it does at all—and how it is corrected. Yes, the experts may intuitively catch their error and intuitively correct it; but how are the rest of us to understand how often it happens and how successfully?

At the start, Sundararajan raises the problem of EBTs in response to a Motorola executive who wants to know whether the therapies Motorola reimburses work (pp. 34-35). Intent on showing where EBTs inadequately address that concern, Sundararajan unaccountably never addresses it herself. Not being expert therapists, how are business executives to identify experts and know whether they are effective? The best that she comes up with is that EBTs provide Coke compared to expert therapies’ fine wine (p. 46). This leaves responsible executives with only one option: limit reimbursements to therapeutic Coke and let those who have the money pay for the luxury of fine wine therapies.

However, this Draconian conclusion follows only if engineers and tinkers are “radically different” rather than different in degree. Sundararajan provides no a priori reason to believe that empirical research on therapies cannot extend its reach to  increasingly finer wines even if never to the very finest.

Practical problem 2

Ironically, the other side of practical problem 1 is that EBT proponents also assume that EBTs and HPTs are incommensurable while increasingly drawing the opposite conclusion: what is outside EBT is not knowledge and should be excluded from professional psychological practice. However, if I am right that the two camps differ only in degree, then there is room for accommodation. There can be degrees of uncertainty in any particular approach’s effectiveness. Just as in medical practice, procedures whose effectiveness is more certain can be proportionately recommended and funded, whereas those whose effectiveness is more uncertain can be proportionately rated as elective and unfunded, or in extreme circumstances tried and funded as a last resort. This policy will give those in the HPT camp heartburn for two reasons. The legitimate one obtains if institutional powers too narrowly define evidence, thereby mistakenly excluding HPT methods that do in fact provide valid evidence for their effectiveness. The illegitimate heartburn arises from wanting to have it both ways: HPT methods transcend scientific assessment but should nevertheless be given the same respect as those subject to it.

The frustration of HPT proponents over such a policy is understandable, especially if EBT proponents misguidedly dismiss them as quacks; but it is unreasonable to equate therapies that have significantly reduced uncertainty with those that have not yet done so. Fortunately, if I am right about there being only a difference in degree between HPT and EBT, however wide, HPT proponents have an alternative to outrage: they can develop ways to reduce the public perception of uncertainty in their methods.


A helpful way to view the role of science in psychotherapy is to back into the issue by posing a scenario where all therapeutic interventions are completely intuitive, inexplicable, and effective.  As uncertainty increasingly enters into successive scenarios, empirical research becomes correspondingly necessary to reduce it. Although proponents of HPT and of EBT have argued for their mutual incommensurability, neither side has made its case. The fact that evaluation methods currently used on behalf of EBTs are often inadequate to assess the effectiveness of HPTs does not justify assuming a priori that more adequate methods can never be developed. This currently puts HPTs at a disadvantage for funding. However, that problem can be alleviated to the extent that they can develop ways of reducing uncertainty about their effectiveness. Unfortunately, the disadvantage will never be completely eliminated, if Sundararajan is right that at least some HPTs are incommensurable with empirical science. No economic system is plausibly going to be able to provide third-party funding for practices whose effectiveness is a wide-open uncertainty.


Bernet, M. (1996, August). Emotional intelligence: Components and correlates. Paper presented at the Annual Convention of the American Psychological Association, Toronto, Canada.

Bourdieu, P. (1990). The logic of practice (R. Nice, Trans.). Stanford, CA: Stanford University.

Dreyfus, H., & Dreyfus, S. (1982). Mind over machine. New York: Free Press.

Lévi-Strauss,C. (1966).The savage mind. Chicago: University of Chicago.

Merleau-Ponty,M.(1962). Phenomenology of perception (C. Smith, Trans.). London: Routledge Kegan Paul.

Searle,J. R. (1983). Intentionality.Cambridge,UK:Cambridge University.

Sundararajan, Louise. (2002). Humanistic Psychotherapy and the Scientist-Practitioner Debate: An "Embodied" Perspective. Journal of Humanistic Psychology, 42(2), 34-47. doi: 10.1177/0022167802422004 Retrieved from http://jhp.sagepub.com/content/42/2/34