What changes in cognitive therapy? The role of tacit knowledge structures

Vol VI, No. 2, 2006 Comments (0)

E. Thomas DOWD
Kent State University, USA

Theoretical considerations-New ideas

Rarely has knowledge within cognitive psychology influenced the development and practice of cognitive behavior therapy. This article explores the integration of the contributions of the cognitive psychology areas of implicit learning, tacit knowledge structures, and encoding processes with producing or retarding change in cognitive behavior therapy. Differences are discussed between understanding something and knowing something. Implications for the practice of cognitive therapy, indeed all psychotherapy are discussed. A clinical example is provided.

Keywords: tacit knowledge, implicit learning, cognitive encoding, cognitive therapy

Pages: 141-148

Correspondence concerning this article should be addressed to:

E. Thomas Dowd, Department of Psychology, Kent State University, Kent, Ohio 44242, USA. Phone – 330.672.7664

This article is based on a symposium presentation at the 6th International Congress of Cognitive Psychotherapy, Gothenburg, Sweden, June 2005


The fact that cognitive psychotherapy results in psychological and behavior change is undeniable – many studies have demonstrated that to be true. On the other hand, cognitive psychotherapy has not been shown to be consistently more effective than alternative therapies. In a recent short review of the comparative outcome literature, Dowd (2004a) concluded that the evidence for the greater efficacy of cognitive therapy is mixed, especially when compared with specific alternative treatments rather than with placebo or generic and supportive types of therapy. There is some evidence, however, that cognitive therapy may be especially useful in preventing relapse or with certain specific problems such as panic disorder. Whisman (1999) argued that the evidence is equivocal regarding whether cognitive change precedes or follows symptom change.
Why is this happening? Why, despite our best efforts, can we not develop cognitive therapy procedures that are more effective that other targeted treatments? Why can we not even agree whether cognitive changes cause behavioral changes or results from behavior change (e.g. Whisman, 1993) or whether cognitive change precedes or follows cognitive change (Whisman, 1999)? Why, as Imber and his colleagues (Imber et al, 1990) discovered, did neither cognitive therapy nor other therapies, produce the specific results that were predicted by their theoretical orientations? Is cognitive therapy even necessary at all? Jacobson and his colleagues (1996) argued it is not necessary and that behavioral activation is enough. Likewise, Feeney, Hembree, and Zoellner (2003) argued that the research literature has not demonstrated that adding additional interventions (such as cognitive therapy) to exposure therapy produced better outcomes than exposure alone.
These questions cannot be answered in one article but a partial answer is possible that might account for some of these findings. To do so I suggest that the fundamental conceptual model of cognitive behavior therapy be expanded to include changes in implicit knowledge structures as well as changes in explicit knowledge. More specifically I propose an integration of certain features of tacit or implicit learning in cognitive psychology with the theory of cognitive change. As David and Szentagotai (2005) have pointed out, cognitive psychology and cognitive behavioral psychotherapy developed independently and rarely has the former influenced the latter. This article is an attempt to do that in a limited area of knowledge.
Years ago, Hobbs (1962) suggested – contrary to the wisdom of that day – that cognitive change (insight) is the result of behavior change, not the cause of it. Clients, he said, changed their behavior and then reflected cognitively about what that new behavior meant and its implications. Later, and in the same vein, Bandura (1977) said, “On one hand the mechanisms by which human behavior is acquired and regulated are increasingly formulated in terms of cognitive processes. On the other hand, it is performance-based procedures that are proving to be the most powerful for effecting psychological changes” (p. 191). Or, in more popular terms, “It is easier to act your way into a new way of thinking than to think your way into a new way of acting.” Kelly, Dowd, and Duffey (1983) found that using behavioral change methods first and cognitive change methods second was more effective in treating depression than using cognitive change methods first and behavioral change methods second. So, what is changing – and how?
Meichenbaum and Gilmore (1982) divided cognitive processes into three parts. Much of the problem may be that cognitive therapies of all varieties have devoted themselves largely to fostering change in two of these; cognitive contents (what we think – the specific automatic thought we have) and cognitive processes (how we think – the distorted processing errors, such as overgeneralization and all-or-nothing thinking identified by both Beck and Ellis). Both of these cognitive activities can be seen as examples of understanding-based “cold cognitions” (David & Szentagotai, 2005), in which cognitive representation of relevant circumstances is primary.
In the process, however, we have downplayed or reduced the importance of the third aspect, cognitive structures (the tacit representational network of associated beliefs and rules-for-living that are developed over time), although these are considered core cognitive beliefs (J.S. Beck, 1995), Early Maladaptive Schemas (Young, Klosko, & Weishaar, 2003), or the “Musts” and “Shoulds” of REBT (Ellis & Dryden, 1997). Cognitive structures involve not only “cold” cognitive activities but most importantly “hot cognitions” which involve emotionally-loaded appraisals of events (David & Szentagotai (2005). But these cognitive structures themselves are much more complex than commonly supposed and are examples of tacit or implicit knowledge – knowledge and appraisals that are not accessible to consciousness but that guide and direct our behavior and decision-making processes nonetheless. And it is very difficult to change cognitive structures because they were developed early in life – pre-verbally and non-verbally. How do you discuss something for which the client has no words? This type of knowledge may be similar to unconscious processing as described by psychodynamic therapists, although without the conceptual baggage associated with that term.
Have you ever had a client who said something like, “Yes, I understand that my thoughts are irrational – but I just don’t really believe (or feel) it.” What that means is that we have changed the cognitive contents, perhaps the cognitive processes, but not the cognitive structures – especially the “hot cognitions” aspect of these structures. It is what the psychoanalysts might have meant by intellectual insight rather than emotional insight. The sense or feeling of “correctness” is not present.
There are several aspects of the tacit knowledge contained in cognitive structures that are important to know (Dowd & Courchaine, 2002; Lewicki, Czyzewska, & Hill, 1997). First, tacit knowledge is acquired largely from experience in the environment in which it will be needed. It is best acquired in an environment that fosters learning from good models. It can often be best acquired in unusual situations. Second, tacit knowledge is very conservative and is not easily changed – accounting perhaps for the resistence many or most clients have toward significant change, even if they say they want it. As Dowd (2004b) has discussed, cognitive structures, especially those involving core concepts of personal meaning and self-identity, are protected from unusually rapid change because such changes can threaten our very sense of self and lead to personality disorganization. Third, tacit knowledge is phylogenetically older than and holds evolutionary primacy over explicit knowledge, particularly in affective or emotional learning. Fourth, and this is especially important, it occurs through the tacit detection of co-variation of features or events in the environment. Co-variation is essentially associational learning as originally described by Wolpe (1990) in the therapeutic process. What this means is that neurologically advanced organisms (such as humans) notice things that occur together and then tacitly assume that they belong together and represent reality; in other words, these co-variations develop into “inferential rules” about the what the world is and the way it operates. The problem is that these co-variations may be accidental, incidental, or not really true. Because of our advanced neurological structure, we can notice co-variations that are extremely subtle or are not really there. Learning in this way can occur very quickly and it does not take many repetitions for it to occur. These encoding rules can also be developed from a surprisingly small number of instances of consistent evidence; it does not take much for an inferential rule to develop. Cultural assumptions are examples of tacit knowledge as is the knowledge of our native language. Fifth, tacit or implicit learning occurs much faster than explicit learning and it is more effective in learning complex and multi-faceted tasks. Indeed, explicit learning often gets in the way of or interferes with implicit learning (Dowd & Courchaine, 2002).
The role of Encoding Algorithms plays a significant part in the acquisition and maintenance of tacit knowledge (Lewicki et al., 1997). All incoming stimuli must be encoded according to some classification system if they are to be meaningful and therefore noticed and used. These algorithms are responsible for filtering this information according to the “inferential rules” previously developed. It must be emphasized that people are not aware of these rules nor of the filtering process. The problem – if you want to look at it that way – is that the cognitive filter acts in such a way to let information in (stimuli) that supports the existing rules and categories and acts to keep out information (stimuli) that does not support or contradicts the existing rules and categories. Over time, therefore, these algorithms become self-perpetuating as the inferential rules become more elaborated and stronger because they are repeatedly confirmed. This process is especially pronounced in situations that are ambiguous – which most social situations are – where a variety of explanations might be plausible for a certain event. Over time, we gradually “see what we expect to see” and “find what we expect to find” – thus further strengthening these algorithms. Therefore, rather than “seeing is believing,” a more accurate saying might be “believing is seeing.” Tacit knowledge is demonstrated when people say, “But it’s obvious. That’s just reality!” “That’s just the way things are!” “Everyone knows that!”
But it gets worse! There is a phenomenon called “hasty encoding,” in which the individual uses very little supportive information before imposing a particular interpretation on the event – even if the co-variations which result in the inferential rules do not match very well. It is the best that can be done in a short time. Essentially, the individual substitutes speed for accuracy; what we might call “jumping to conclusions.” There appear to be individual differences in whether one uses a more internal (“hasty”) encoding style versus a more external encoding style, in which external data are used. Those who use an internal encoding style are more likely to use their pre-existing inferential rules and expectations and are less likely to use outside information in interpreting new information. They already know what they believe and don’t need any additional information. “Data be damned! I know what I think!” This phenomenon can be seen in many aspects of life.
Intuition can be seen as an example of implicit learning. Intuitive people have a sense of what is right or wrong – good or bad – appropriate or inappropriate – but they cannot describe in words why they think (or feel) that way. They often just seem to “know” without being quite able to describe how or why they “know.” They just know they do (know). Intuitive people may be those people who use a more internal encoding style (Lewicki, et al., 1997).
What are the implications for cognitive psychotherapy, indeed all psychotherapy, of these ideas? First, I want to emphasize that implicit and explicit knowledge processes are interactive systems; they do not function as parallel, non-associative systems. Thus it is possible for the output of an implicit cognitive process to become an explicit activating event (in REBT terms) which in turn can be evaluated explicitly. Explicit behavioral experiments may in turn generate “hot” cognitive appraisal processes. Second, it appears very important to use behavioral methods to change cognitions rather than the reverse. While a discussion and examination of cognitive distortions and the irrationality of beliefs and assumptions may be useful in changing cognitive contents and cognitive processes, it is less likely to change tacit cognitive structures. Behavior change can be seen as primary (Dowd, 2006) and in many cases it may lead to spontaneous cognitive change. Therapists should use “behavioral experiments” as much as possible, in which they encourage their clients to try new and different things and then to examine and interpret the results. Interestingly, this is consistent with Beck, Rush, Shaw, and Emery’s (1979) original argument that behavioral techniques are used in cognitive therapy for the purpose of changing cognitions. Over the years it seems to have been neglected. “Shame-attacking” or other REBT behavioral exercises (Ellis & Dryden, 1997) are also examples. Third, tacit knowledge is very resistant to change. Therefore, repetition is important in the therapeutic process. People usually don’t “get the point” the first time – or even the second or third time. Ideas that are very different from clients’ tacit rules and assumptions have a way of “disappearing from the head” quickly. Repeated exploration of important points, repeated homework assignments, and repeated writing down of important ideas are all very useful. Ask the client to keep a written record of important ideas. Fourth, affective or emotional learning is especially likely to be resistant to explicit change techniques because it is more likely to be learned implicitly. The concept of mental contamination proposed by Wilson and Brekke (1994) has especial relevance here. They defined it as a phenomenon whereby a client (or anyone) is influenced by a cognition, emotion, or behavior that is the result of unconscious or uncontrollable mental processing. David (2003) argues that may be involved in cognitive restructuring and has proposed therapeutic techniques derived from the mental contamination model. Fifth, it is very difficult to discuss and describe ideas and concepts for which we have no words. It may be more useful to use non-verbal techniques such as body work, imagery, meditation, mindfulness, and hypnosis (Dowd, 2004c; Mahoney, 2003), especially where powerful emotions are present. It is interesting that some of the newer forms of cognitive psychotherapy incorporate these techniques in some way, such as Mindfulness Meditation in the treatment of depression (Segal, Williams, & Teasdale, 2002). Sixth, because tacit knowledge is very resistant to change, it is important that the therapist not introduce too much interpretation discrepancy from the clients’ existing ideas and concepts (Claiborn & Dowd, 1985) at first. Research has shown that the discrepancy of therapist interventions from client assumptions and understanding makes more of a difference than the content of those interpretations. Too little discrepancy will result in no change. Too much discrepancy may result in the client disregarding the therapist’s interpretations or leaving therapy entirely. Seventh, because tacit cognition tends to influence behavior automatically without conscious processing, it may be helpful to teach clients to use conscious, more effortful processing instead (Palfai & Wagner, 2004). Training clients to use secondary information that is not their typical response strategy can be useful. Specifically, training them to pause before acting makes it more likely they will be able to use secondary information (Wallace & Newman, 2004). There may well be much wisdom in the saying, “Count to ten….” Meichenbaum (1977) taught impulsive children to talk out loud to themselves about using different behavioral strategies as a way of slowing them down and thereby reducing their impulsivity.
Let me provide a clinical example of what I mean by these ideas. I once had a male client whose view of women was that they were fickle creatures who played with men’s emotions as if it were a game. He had difficulty seeing women as truly interested in long-term relationships and as people who could be emotionally hurt just like he could. He had grown up almost entirely with boys and because he had very little experience with girls, he had difficulty seeing women as people at all. They were mysterious creatures. He understood, after numerous therapy sessions, that women were really people just like him – with the same hopes, fears, and wishes – but he had great difficulty feeling and understanding this at a core level. In other words, he understood (involving “cold cognitions) – but he did not know (involving “hot cognitions”). Only after repeated interactions and experiences with women – not just the therapist and him talking about women – was he really able to know what they were like and also to know that all women were a little different from each other and perhaps not much different from himself. There are of course examples of women who think they know all about men (“Men are only interested in one thing!”) from their early and limited experience. In both cases these early ideas are very difficult to change because later experiences are processed and interpreted using the older inferential rules. Thus, we tend to see what we expect to see and to find what we expect to find and we can then say, “I knew it all along!”
It is important to emphasize that this process is not usually deliberate. There is a strong human tendency to arrive at our conclusions first (what we want to believe or want to be true) and then look for supporting evidence afterwards. This helps preserve our core cognitive structures of identity and personal meaning. We all do it; the only difference is how much we do it and how aware we are of doing it.
Different techniques and different assumptions are necessary in changing tacit cognitive structures (e.g. David, 2003). We cannot assume that changes in cognitive contents and cognitive processes will automatically and eventually lead to tacit cognitive structural changes, although they might. We must plan and target our interventions directly at these structures – knowing that it is not easy – but that it is necessary if change is to endure.


  • Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84,191-215.
  • Beck, J.S. (1995). Cognitive therapy: Basics and beyond. New York: Guilford.
  • Beck, At., Rush, A.J., Shaw, B.F., & Emery, G. (1979). Cognitive therapy of depression. New York: Guilford.
  • Claiborn, C.D., & Dowd, E.T. (1985). Attributional interpretations in counseling: Content vs. discrepancy. Journal of Counseling Psychology, 32, 188 196.
  • David, D. (2003). Rational Emotive Behavior Therapy (REBT): The view of a cognitive psychologist. In W. Dryden (Ed.), Theoretical developments in REBT. Brunner/Routledge.
  • David, D., & Szentagotai, A. (2005). Cognition in cognitive behavioral psychotherapies: Toward an intragetive model. Clinical Psychology Review,
  • Dowd, E.T. (2006, June). Erkenntnisse, die ich bei der Kognitiven Verhaltenstherapie gewonnen habe (Insights which I have learned from Cognitive Behavior Therapy). Presentation at the Institute for Psychological Therapy, University of Leipzig.
  • Dowd, E.T. (2004a). Cognition and the cognitive revolution in psychotherapy: Promises and advances. Journal of Clinical Psychology, 60, 415-428.
  • Dowd, E.T. (2004b). What makes people really change? (And why is it so difficult?). Constructivism in the human sciences, 9, 59-66.
  • Dowd, E.T. 2004c). Expanding the cognitive therapy model: Imagery, meditation, and hypnosis. Journal of Cognitive Psychotherapy: An International Quarterly, 18, 351-360.
  • Dowd, E.T., & Courchaine, K.E. (2002). Implicit learning, tacit knowledge, and implications for stasis and change in cognitive psychotherapy. In R.L. Leahy & E.T. Dowd (Eds.), Clinical advances in cognitive psychotherapy: Theory and application. New York: Springer.
  • Ellis, A., & Dryden, W. (1997). The practice of Rational-Emotive-Behavior Therapy. New York: Springer.
  • Feeney, N.C., Hembree, E.A., & Zoellner, L.A. (2003). Myths regarding exposure therapy for PTSD. Cognitive and Behavioral Practice, 10, 85-90.
  • Hobbes, N. (1962). Sources of gain in psychotherapy. American Psychologist, 17, 741-747.
  • Imber, S.D. et al. (1990). Mode-specific effects among three treatments for depression. Journal of Consulting and Clinical Psychology, 58, 352-359.
  • Kelly, F.D., Dowd, E.T., & Duffey, D.K. (1983). A comparison of cognitive and behavioral intervention strategies in the treatment of depression. British Journal of Cognitive Psychotherapy, 1, 51-58.
  • Jacobson, N. et al. (1996). A component analysis of cognitive-behavioral treatment for depression. Journal of Consulting and Clinical Psychology, 64, 295-304.
  • Lewicki, P, Czyzewska, M, & Hill, T. (1997). Nonconscious information processing and personality. In D.C. Berry (Ed.), How implicit is implicit learning? Oxford: Oxford University Press.
  • Mahoney, M.J. (2003). Constructive psychotherapy: A practical guide. New York: Guilford.
  • Meichenbaum, D. (1977). Cognitive-behavior modification. New York: Plenum.
  • Meichenbaum, D., & Gilmore, B. (1982). The nature of unconscious processes: A cognitive-behavioral perspective. In K.S. Bowers & D. Meichenbaum (Eds.), The unconscious reconsidered. New York: Wiley.
  • Palfai, T., & Wagner, E.F. (2004). Introduction to current perspectives on implicit cognitive processing in clinical disorders. Cognitive and Behavioral Practice, 11, 135-138.
  • Segal, Z.V., Williams, J.M.G., & Teasdale, J.D. (2002). Mindfulness-based therapy for depression: A new approach to preventing relapse. New York: Guilford.
  • Wallace, J.F., & Newman, J.P. (2004). A theory-based treatment model for psychopathy. Cognitive and Behavioral Practice, 11, 178-189.
  • Whisman, M.A. (1993). Mediators and moderators of change in cognitive therapy. Psychological Bulletin, 114, 248-265.
  • Wilson, T.D., & Brekke, N. (1994). Mental contamination and mental correction: Unwanted influences on judgements and evaluations. Psychological Bulletin, 116, 117-142.
  • Wolpe, J. (1990). The practice of behavior therapy (4th edition). Pergamon.
  • Young, J.E., Klosko, J.S., & Weishaar, M.E. (2003). Schema therapy: A practitioner’s guide. New York: Guilford.

Leave a comment


Metal Music Videos psychotherapy directory resources wordpress consulting travel blog