Comments on Vic's Comments on Doug's Comments

• A Favor-----Please by Doug W., 12/5/98
  • (...)
    
    • Uncertainty by Vic Comello, 3/20/99
      
      

CAUSAL INDUCTION OR PROBLEM-SOLVING ALGORITHM?

Preface

In the spirit of wanting to incorporate Doug's worthy cautions and reservations about the limitations of psychological theorizing when we face considerable doubt and confusion about the potential influence of so many variables affecting human beings, I would like to comment on Vic's comment on Doug's comment.

DOUG AS CRITIC OF INDUCTIVE LIMITS

Contrary to Vic's assessment, I have not understood Doug to be searching for a "guaranteed-by-God complete answer." Instead I have understood the thrust of all of Doug's postings to BOL to be a voice on behalf of skepticism advising us to not presume to have found complete answers in our psychological theories of therapy because these theories may have many unacknowledged limitations.

Thus, rather than be unwilling to grub around with imperfect data, Doug has warned of the possibility of presumptuous illusions that may result from ending grubbing prematurely. In fact, I see Doug's recommendation of the work of Robert Langs as further evidence of his urging therapists to learn a method allowing much closer attention to patients' unconscious complaints about how therapists may often presume to know more than they can know.

BUT: CMT AND VIC'S DREAM THEORY DO "WORK"

Although a causal theory of human motivation, such as CMT or Vic's dream theory, may be used (along with certain ancillary assumptions) to devise interventions (such as interpretations) that "work" in that the interventions are followed by patient improvements, such successful outcomes are not alone probative of any theory's causal hypotheses about human motivation.

However, if repeated, such successful outcomes may authorize the specific theory-assumption-intervention reasoning process as a serviceable problem-solving algorithm.

THOERIES THAT "WORK" VERSUS CAUSAL INDUCTIONS

Thus it is important to note that I refer above to two important uses of rationality that are distinguishable :

(1) inductive inferences (i.e., theories) of human motivation regarded as true after an assessment of probative evidence and reasoning about the causal structure of human nature, and

(2) guidelines and procedures (i.e., algorithms) that "work" in the sense of being serviceable instructions about how to solve practical human problems.

GOOD INDUCTIONS, NOT ALGORITHMS, NEED TO RULE OUT RIVAL EXPLANATIONS

A theory that purports to be an account of the causal structure of human motivation is well supported by evidence and/or reason to the extent that rival causal explanations are excluded as implausible. Doug is concerned, if I follow him, to show that very often psychological theorists falsely believe they have ruled out rival explanations and they become even more convinced that this is so just because they have many followers. Rather than going about the intellectual task of excluding rival explanations, some theorists, Doug warns, would rather rule out the human voices proposing plausible rival explanations. Thus I think one can see why Doug regards Langs's work as so important. It is a corrective to encourage therapists to closely attend to at least one critical voice, the patient's, that will express in symbolic language many critiques of a therapist's mistaken ideas about what the patient needs.

VIC EMPHASIZES THE PROBLEM-SOLVING VALUE OF HIS DREAM THEORY

Vic, if I understand him, is concerned to show the effectiveness of his dream theory, and of CMT, as good problem-solving algorithms. Thus he argues that these theories serve as good problem-solving guidelines. As serviceable algorithms, these are theories that can be said to "work" regardless of whether they present true assertions about human motivation.

But what I wish were more clear to me is whether Vic also wants to further improve upon CMT in a way that would do something other than affect CMT's usefulness as an algorithm for therapeutic intervention. I understand Vic as also wanting to improve upon CMT as a theory about the causal structure of human motivation. (I welcome any corrections by Vic in my supposition here about his aims.)

A question naturally arises at this point. Question: Can the problem-solving effectiveness of a theory certify the validity of its causal hypotheses about human motivation?

My answer is no: such effectiveness is insufficient reason to believe in a theory's causal hypotheses.

Vic has advised us often that he has no objection to how CM therapist proceed in the conduct of an individual therapy. Thus he has not recommended therapists change what they do with patients. If no changes are made in the rules for therapeutic practice with Vic's proposed extensions of CMT, then his extensions, I would assume, are efforts directed at improving CMT as a true set of inductive inferences about the causal nature of human motivation rather than at improving the theory's usefulness as a problem-solving algorithm.

However, as I wrote above, I don't believe one can use a theory's effectiveness as probative evidence of its validity as a set of causal assertions. No matter how much evidence exists that a theory provides good problem-solving guidelines and no matter how logically elegant and internally consistent the theory, clinical evidence from individual therapy in session-by-session process analyses cannot provide sufficient evidence to rule out many rival explanations that Doug reminds us may exist.

Unless someone can devise better and more creative research methods than those now available, I don't see how an individual case study method is going to rule our rival explanations for favored causal hypotheses. CMT has done an amazingly good job of squeezing out every bit of probative mileage one can reasonably expect to find from their kind of case-study process analysis. And the theory of CM, fallible though its method may be--as is any method fallible--is a set of some very efficient inductive inferences and certainly more efficient and creative than any I could devise.

Though it is seldom acknowledged openly, wise intuition is in fact the basis for a good measure of the accuracy found in inductive conclusions based on sound evidence and reasoning, and such wise intuition is part of the success of every creative science. I just read of how Kary Mullis, Nobel Prize winner in chemistry, discovered the PCR formula (a formula that can be used to make a DNA needle into a DNA haystack) while driving his car one evening. He had to pull over along the side of the road to write it all down before it left him. His discovery revolutionized DNA research, and now leaves the world a very different place.

It is important to realize that there is no way to make causal inductive inferences that are anything other than good-better-best or bad-worse-worst guesses. Even the highly touted gold-standard of behavioral-research methodolgy, the controlled-experimental method, though it can rule out many more rival explanations than a correlation, and is preferred for that very reason, always leaves some variables left unaccounted for. And, contrary to popular opinion, the controlled-experiment, like the unfairly maligned correlation coefficient, cannot be used to prove causality. In fact, a controlled experiment is really another way of finding a correlation between dependent and independent variables. The controlled experiment has the advantage of reducing the number of rival variables that could account for the correlation but offers no superiority over a correlation coefficient in being able to prove causality. Correlation does not prove causation nor does anything else. Even though Doug and I agree that there remain rivals to CMT's causal hypotheses, I think I am less skeptical than Doug about the likelihood that CMT is a sound theory of human motivation. Thus, of course CMT may be in error, about survival guilt or about its other casual hypotheses and assumptions. But I doubt it less than Doug doubts it. So, though I'm skeptical, I'm less skeptical than Doug about CMT itself, and I'm more skeptical than Vic about Vic's proposed extensions of CMT. (However I am less skeptical of Vic's dream theory and also believe that my ostensible skepticism of his other extensions of CMT may be instead a case of limitations in my own intuition and reasoning cognitive modules since I am very impressed with the power of these same modules in Vic. Accordingly I am hopeful that Vic will provide a succint statement summarizing his key points as a a way to see if I have understood him.)

GOOD THEORIES OF KNOWLEDGE NEED A CONTINUUM FROM DOUBT TO CERTAINTY

If one aspires to a theory of knowlege that would erase all doubt about our inductive conclusions, it is likely, because doubt is ubiquitous, that one will be led into another black hole--the black hole of RADICAL skepticism. In general I share Doug's skepticism and would urge that it be extended to include skepticism about radical skepticism. Thus, I imagine myself a skeptic, too, as a "rational fallibilist" rather than as a radical skeptic.

As a rational fallibilist I adjudicate knowledge claims in the reason-and-evidence court and then accept some claims as true along a continuum of doubtful-to-certain acceptance. Claims reaching certain acceptance (or what amounts to certainty for all practical purposes) are based on direct observation, axiomatic or self-evident truths, and valid deductive (e.g., mathematical) conclusions using true premises. But inductive conclusions always come with some measure of doubt.

Among competing doubtful inductive claims in a set of causal hypotheses, I select for adoption as my beliefs (until they are shown to be too doubtful) a subset of least doubtful claims.

Sometimes I have no basis in reason or evidence to prefer one belief (within a selected subset) as more likely to be true than another, however. Even when I can select as most likely to be true one belief from within a subset, I understand that an assertion that this belief is a claim to true knowledge depends more heavily upon (1) my belief in the dubiety of a contradictory belief than upon (2) confidence in fallible reason or evidence on behalf of the adopted belief itself. That is, as a rational fallibilist I select beliefs by placing more trust in what we know to be false than in what we believe to be true.

In short, given that human knowledge in most things of importance is highly fallible, we do as best we can with what available reason and evidence can adjudicate until we find we are mistaken. But, unlike radical skeptics, I do not abandon beliefs because they come with some doubt unless that doubt rises to a level that convinces me something else is more likely to be true. And I know that there is no method can can erase all doubt about inductive inferences, causal or otherwise.

Accordingly, it is my opinion that CMT provides fallible evidence about elusive matters--evidence that convinces me that CMT is among the best that theories of psychotherapy have to offer because I find many other theories attempting to address these same elusive matters to be more fallible. Until reason or evidence convinces me otherwise, I adopt CMT as a worthy, but not certain, theory of the causal structure of human motivation. As a set of guidelines for therapeutic practice, CMT works very well and thus may be, as Vic is keen to insist, a more worthy problem-solving algorithm than a set of true causal assertions about human motivation. However, I believe CMT is equally worthy on both counts.

• Proof? by Vic Comello, 3/22/99