Concepts of Disease and Health

Readers of this blog should be interested in a recent entry in the Stanford Encyclopedia of Philosophy (SEP) on the "concepts of disease and health." It is written by Dominic Murphy, author of a very important book, Psychiatry in the Scientific Image (2006). The latter offers a "qualified defense of the medical model, which says psychiatry is a branch of medicine dedicated to uncovering the neurological basis of disease entities." In other words, for Murphy, psychiatry is best understood as "clinical cognitive neuroscience."

As things stand today, Murphy explains, psychiatry relies on four criteria to validate a diagnosis: natural history (of a syndrome: its characteristic clinical course and outcome); family history (including genetics); differential response to treatment; and laboratory findings. The problem with these criteria, notes Murphy, as emblematically enshrined in the DSM, is a failure to give pride of place to "causal information," apart from the lack of explicit reference to requisite normative considerations (e.g., a model of rationality). We are left, in other words, with the picture of valid syndrome on the order of a "collection of symptoms that cluster together" but without the kind of information essential to a medical (or biological) model of psychiatric disease/disorder. Put differently, "The concepts of clinical phenomenology are notoriously vague, imprecise, unquantified. By limiting the data gathered in diagnosis to the salient and easily identifiable signs and symptoms of clinical phenomenology, the DSM-IV-TR [fourth ed., revised text] scheme ignores a wide range of other data about mental functioning that can be gathered by psychometric techniques and by methods used in cognitive science and neuroscience."

While I'm not fond of Murphy's conception of psychiatry and do not share his philosophy of mind assumptions about the nature of the "mental" or the inherent liabilities of "folk psychology" for that matter, there is much of value in this book whether or not one agrees with the (3) theses (p. 11) he sets out to defend. As in the aforementioned SEP entry, Murphy provides us with excellent conceptual clarification of such terms as "objectivism" and "constructivism" and the projects of "naturalization" and "mechanization" as they apply to psychiatry and psychology. And he makes a devastating critique of the DSM-IV's concept of mental disorder, including the fact that it "permits exclusively behavioral criteria for diagnosis, leading to a failure in many cases to distinguish mental disorders from 'problems in living'--fully comprehensible, indeed wholly expected, reactions to stressful circumstances." Whatever one thinks of Murphy's "medical model" for psychiatry, I think it's hard to quibble with his contention that psychiatry "need[s] both a theory of rationality and a theory of the mechanisms underpinning rational thought and behavior, so that departures from both can be assessed, explained, and treated." Murphy argues that, while there is not a convincing mechanistic model of reason along, say, computational lines, the prospects for a naturalistic rationalization project look better: "But agreement on the naturalistic rationalization project will require equilibrium between folk theory, various sciences, and naturalistic philosophy of mind and epistemology." In a review of the book, Luc Faucher well summarizes Murphy's frank and important conclusion about the implications for a purely scientific model of psychiatry:

"According to Murphy, one can naturalize reason (that is, one can provide a story concerning the possible evolved epistemic function of rationality), but not mechanize it. The consequence of this is that "[t]he assessment and treatment of these disorders [like delusion or alcoholism] usually requires a inexplicit model of what good reasoning is. If rationality is a normative notion, the model must be in part normative. Since this model is normative, it undermines the pretensions of the medical model to purely positive science" (153). However Murphy reminds us that this concession is not an endorsement of the anti-psychiatry's position of people like Szasz, because it applies only to a limited number of mental illnesses, and the norms in question are not social norms, but rather rationality norms."

Patrick S. O'Donnell

Comments

[Daniel S. Goldberg]

Much to chew on here. I have to say, I'm not a fan of either of his notions (neither the medical model nor the qualified defense).

You know that I am partial to Searle's account of mind, which places subjectivity in a primary and unavoidable role in construcing mental phenomena. I've also recently turned to phenomenology, and for a nice introduction to the application of phenomenology to problems of psychiatry, I find Eric Matthews' recent book to be a nice start.

Even a cursory glance at the latter shows, in my view, the inadequacies of either a medical or a chiefly rationalistic account of mind or of psychiatry. Arguably, one of the reasons we have gotten into so very much trouble in psychiatric medicine both historically and recently (re SSRIs, conflicts of interest, etc.) is because of our unwillingness to expand conceptions of mental illness beyond medical and rational models.

There's little doubt that the limits of such models also play a fairly significant role in the violations performed on mentally disabled bodies during the modern era.

[Patrick S. O'Donnell]

I like the book because it happens to be the most sophisticated defense of a model for psychiatry that neither of us happen to be drawn to. I think it is important to have before us the most articulate conception of that with which we disagree and the fact that this particular medical model is "biopsychosocial" puts it heads above others of its type. Finally, it is undoubtedly true that at least some forms of mental illness have biological and neurological causes and thus this model will help us discover and treat those types.

It just so happens that I've been immersing myself in the psychoanalytic literature of late (of largely neo-Freudian vintage) and remain convinced that the more severe forms of mental illness frequently require the resources of both psychiatry and psychology, indeed, many who have resorted to a psychopharmalogical approach to mental illness have done so in combination with an appreciation of the corresponding need for psychological therapy of one kind or another.

[Daniel S. Goldberg]

I agree completely that it is important to engage robust models with which one disagrees, and I appreciate your bringing this to my attention.

it is undoubtedly true that at least some forms of mental illness have biological and neurological causes and thus this model will help us discover and treat those types.

I agree with this, but it is also somewhat question-begging, because the thorny philosophical issue is what exactly we mean when we say "biological and neurophysiological causes." If we mean neurochemical and neurophysiological, that seems to me to be trivially true, but Matthews's point is largely that it is patently absurd to reduce the phenomena of mental illness for a being-in-the-world to these substrates, which is what so much of psychiatric and scientific reductionism tends to do.

Also, the risk of reifing mental illness only through neuroscientific and neurophysiological phenomena is a big risk here. Depression is quite real even if we "see" no organic 'causes for it,' as is pain. (This doesn't mean neurophysiological substrate is not a nec. c for such phenomena; of course they are, but such phenomena are not reducible to such substrate. Searle makes this point explicitly in his conception of mind).

Furthermore, there is no evidence that such substrates are nec. and sufficient for the experience of mental illness; given the tendencies to reductionism and scientism in Western culture, the move from arguing that such substrate is nec. to arguing that it is nec. AND sufficient is rarely justified but frequently made, IMO.

I'm less convinced that a psychopharmacological approach is truly needed for many kinds of mental problems and illnesses, especially given what we are finding out about SSRIs and the incredible power of the placebo effect.

[Patrick S. O'Donnell]

I suppose, without having read him, that I would have to heartily disagree with Matthew about the worth of reductionism, particularly when identifying such causes results in a cure or progress in the treatment of mental illness. Some forms of mental illness, as I implied above, are perhaps more amenable to such reductionism than others, so I agree that even if there is not "organic" cause for depression, it's no less a mental health issue (hence my mention of psychoanalysis, which does not commit itself to such organic causes). Risks of reification are, again, perhaps unavoidable, so assuming such risks with lucidity goes some measure in guarding against them.

Murphy no where argues that these "substrates" are necessary AND sufficient conditions (hence the 'psychosocial' aspects of his model).

Even the best critics of psychopharmacology and pharmeceutical marketing practices have seen fit to document its successes (e.g., chlorpromazine alleviating chronic psychosis; the antipsychotic properties of reserpine; and haloperidol for Tourette's syndrome): see, for instance, David Healy's The Creation of Psychopharmacology (2002). Incidentally, he notes that "SSRIs are in fact more efficacious in treating OCD [obsessive-compulsive disorder], social phobia, and even premature ejaculation than in treating depression."!!!

[Daniel S. Goldberg]

You think the phenomena of mental illness for a subject "is" reducible to neurobiological substrate? That what it is like for someone to experience, say, schizophrenia or depression is nothing but the firing of neurons or the condition of electrochemical impulses?

I'm guessing you don't, and I think that is mostly what Matthews was trying to say. There is a great deal we need to understand about the patient's subjective experiences to heal them, as opposed to merely treating a disease. Far too much psychiatric practice does the latter as opposed to the former, and this in large part traceable to the dominance of the medical, rationalistic models.

Last point: I'm not articulating a Szaszian or Illichian rejection of pharmaceuticals. Some drugs are very helpful for some mental illnesses. But many are either not or are unknown, and the behaviorial pharmacologism in Western culture, accompaned with the social, technical, and material resources that operationalize the use of these drugs has created an environment in which we rush to apply powerful pharmaceuticals to all manner of mental problems and illnesses -- or medicalize social phenomena into diseases that feature powerful drug treatments -- when there is often excellent evidence that other, less invasive or less risky treatments may be just as if not more effective in the long run.

[Patrick S. O'Donnell]

I did not say nor think the *phenomenology* (or subjective experience) of mental illness is reducible to its substrate but think, for nosological and treatment reasons, reduction is absolutely essential to the determination of "neurobiological" causality for at least some forms of mental illness. And attending to the causal questions has more than a little bearing on the nature of the phenomenology of mental illness. The medical treatment of a disease may be a necessary condition for the healing process, whereas ignoring the questions of scientific causality may often guarantee little or no hope for attaining a condition of mental health. In short, I see no necessary conflict between (scientific) reductionism in some cases and appreciation of the phenomenology of mental illness.

I well appreciate your last point and have made it myself many times over the years.

[Daniel S. Goldberg]

reduction is absolutely essential to the determination of "neurobiological" causality for at least some forms of mental illness.

Assessment and understanding of neurobiological substrate is always a good idea. Reduction of any disease to such phenomena is never a good idea. (Perhaps we have different senses of what "reduction" means here?) There is little doubt that the vast majority of resources allocated to mental illness as to science & medicine are devoted to the former -- partly because they may lead to drug treatments, whereas facilitation of empathy or qualitative treatment methods, let alone public health and prevention, generally do not -- which is itself a product of such reductionism.

I have no hesitation in declaring scientific and medical reductionism to be unhelpful or worse in a variety of different contexts related to human health and illness, and especially mental illness. It does not follow, of course, that scientific and clinical method is of no help in ameliorating human suffering from mental illness, but simply questions what is and what ought to be taken to qualify as the practice of science and clinical medicine in healing such patients.

The medical treatment of a disease may be a necessary condition for the healing process, whereas ignoring the questions of scientific causality may often guarantee little or no hope for attaining a condition of mental health.

(my turn!!) I did not say or imply that we should ignore questions of scientific causality. I still think that you are question-begging, however, because one node of our debate here is to query what we mean when we say "scientific causality." To attribute the causes of mental illness solely to neurobiological substrate is a deeply misleading, unhelpful, and stigmatizing way of conceiving of disease causality, though one that I understand neither you nor Murphy endorse. So what exactly does it mean to term an attribution the "scientific" causality of mental illness?

All well and good -- but far too many do tacitly or explicitly make just such an endorsement, and while there is nothing in models of psychiatric practice that requires such a model, there is, IMO, little basis for arguing that dominant practices today do not embody just such a reductionism, to the detriment of those most vulnerable and in need of help, never mind the healers themselves.

[Patrick S. O'Donnell]

What you mean by and what I intend by "reductionism" appear to be two different things, or at least we are using it in different senses: "Assessment and understanding of neurobiological substrate is always a good idea." That, as I understand it, is a reduction in the context of trying to account for (explain) mental illness. As Elster clearly points out, there's "premature," "crude," and "speculative" forms of reductionism. Yet, "Overwhelmingly, the history of science show that reductionism is a progressive and antireductionism an obstructionist force in science." So, if we value science (sans scientism) and the mechanisms of causality related to a "neurobiological substrate" in the accounting for a particular mental illness, we are necessarily implicated in the process(es) of reduction. Social science is sometimes reduced to psychology, biology is often reduced to chemistry and, sometimes, psychology needs to be reduced to biology (genetics, physiology, developmental biology, etc.). Notice that the reduction is for a limited and specific purposes: here for the puroposes of a causal explanation.

The quote above: "The medical treatment..." was not an accusation, i.e., I was stating what I understand to be a fact for puroposes of clarity in communication. To dismiss the process of "reduction" is to dismiss the delineation of causal mechanisms. I don't see the question-begging but I do discern a straw-man, to wit: Who "attributed the causes of mental illness solely to neurobiological substrate"? There are some mental illnesses that can and have been reduced to neurobiological mechanisms. That's NOT all forms of mental illness, nor does it rule out yet other forms of neurobiological substrate in conjunction with different levels or kinds of explanation: social, psychological, what have you. I don't find scientic causality, when successful, to be such a mystery, and Murphy provides a nice account of what he means by that in the context of psychiatry and mental illness (as Paul Thagard does for diseases of the body, as it were).

[Patrick S. O'Donnell]

erratum: "scientific causality"

[Daniel S. Goldberg]

If all you mean is that the application of scientific method requires a "reduction" in some sense, then I don't think we really disagree. Reductionism and reduction are not necessarily referring to the same behavior.

But I do not agree that we can use these processes to entirely specify a causal mechanism in a necessary and sufficient sense (which, by propositional logic, is actually what traditional notions of causation is meant to imply). If all you mean is that we can use scientific "reductions" to delineate certain kinds of 'necessary but not sufficient' causal processes and mechanisms -- i.e., neurobiololgical ones -- then again, I think we agree.

discern a straw-man, to wit: Who "attributed the causes of mental illness solely to neurobiological substrate"?

I think I expressly indicated to whom I attribute this:

"I understand neither you nor Murphy endorse."

"there is, IMO, little basis for arguing that dominant practices today do not embody just such a reductionism"

I do not see the straw man. I understand you and Murphy do not reduce mental illness to neurobiological causes. But my argument was and remains precisely that many do, and that in fact such reductionism (not reduction per se) fairly characterizes contemporary psychiatric practices in the West. That may or may not be correct, but I fail to see how that is a straw man.

There are some mental illnesses that can and have been reduced to neurobiological mechanisms.

Such as? What mental illnesses have been established to be caused by nothing but neurobiological mechanisms? No social or enviromental factors? Just neurobiology? Or perhaps we might agree that neurobiological mechanisms are modulated by social and cultural notions of meaning, which is precisely what neuroscientist Howard Fields argues? (Channeling Lewontin here, in his rejection of the dichotomy between biological and social factors).

If so, then again I doubt our disagreement is as sharp as it seems, but I do think yet again we'd have to discuss what it is we mean by suggesting that a neurophysiological substrate causes anything at all. Which is why I continue to think, if the propositions are not question-begging, they at the very least require a great deal more unpacking of what it is you mean when you refer to scientific causality.

And I disagree with you on whether scientific causality is a mystery. Sylvia Tesh and Christopher Hamlin explain quite nicely how infantile our notions of disease causality are (notice that the entire field of epidemiology, which I find extremely important and useful, essentially charts correlations. Even tobacco cannot be said to be a definitive "cause" of lung cancer, because plenty of people who smoke for decades do not develop lung cancer, and at leasdt 10% of all lung cancer patients have never smoked. What we have there is a particularly robust correlation, and we proceed to [quite properly] treat and enact public health policies on that basis). Hume documented quite well just how inductive correlations are extremely useful. And indeed they are. And that is part and parcel of the succes of science and clinical medicine.

But to suggest that disease causality is a relatively simple matter, is, I think, quite wrong. I suspect that some of our worst problems in thinking about disease -- as well as some of our worst habits of stigmatization -- occur both because we rely on a particularly linear, simplistic model of an excessively complex nonlinear system, and because we do not recognize the social and cultural factors that contribute to our causal attributions.

[Patrick S. O'Donnell]

To reply to your question: "What mental illnesses have been established to be caused by nothing but neurobiological mechanisms? No social or enviromental factors? Just neurobiology?"

While there are very few mental illnesses according to Murphy that would fit this bill, Huntington's Disease seems to be "nothing but neurobiological" (actually, genetic) as it is a DSM-IV-TR diagnosis as a source of dementia. Murphy describes this as one of the few instances in which we can say with some confidence that genetics is a "fundamental explanation." Although "the entire causal story is not yet known, we do know that Huntington's Disease is associated with a gene, IT15 or 'huntingtin,' on the short arm of chromosome 4, characterized by an abnormally long CAG trinucleotide repeat." To be sure, "The gene is not fully penetrant, meaning that one can have and not show the symptoms. But this occurs only in rare cases and with low values of the repeat." Murphy uses this as an exemplar of "single-major-locus" (SML) conditions. He further writes that "One could, for instance, die of something else before the disease develops. However, the relations between alleles and SML conditions are determinate enough for us to call the genetic explanations fundamental in the sense that they identify genetic causes that, under normal conditions, suffice for the diseases." There's of course more to this specific example but the interested reader should see pp. 133-135 for the complete discussion.

Murphy notes the following concerning the aforementioned "fundamental explanation:"

"If we understand the fundamental explanation as the identification of the sole or chief cause of a condition, fundamental explanations rarely are obtainable, even when other things are equal. In psychiatry the paradigmatic fundamental explanations are genetic because single-gene disorders typically are conditions with which the gene has its customary effects almost regardless of the state of other parts of the system. But most mental illnesses do not fit this picture and therefore lack fundamental explanation. They result from interrelated causal processes at different levels of explanation and a different timescales."

It is also interesting to note that,"psychopathologies (like General Paralysis of the Insane) have historically disappeared from the province of psychiatry once we have learned of their biological etiology." In this case GPI was probably caused by tertiary syphilis.

Interestingly, Murphy argues that "In psychiatry, reduction makes matters more complicated, rather than simpler. Instead of supplanting higher-level generalizations with lower-level ones, it adds new, lower levels of explanation to an existing structure. This both extends our causal understanding of why the upper-level generalizations work and increases our knowledge of points in the system where it might be tweaked to change the outcomes. We cannot know what goes on in the brain until we figure out what happens at the molecular level, and the more causal relations we understand, the more opportunities we have for therapeutic interventions in a system."

Thankfully, Murphy reminds us that "There is no necessary tie between the medical model and an exclusively pharmacological approach to treatment. This is true on any account of the biology of mental illness. The factor that best explains a condition may not be the factor that is most easily manipulated in therapy."

[Patrick S. O'Donnell]

By the way, and for the record, I did not (or at least did not intend to) "suggest that disease causality is a relatively simple matter," but did say that I found scientific causality, WHEN (or where) SUCCESSFUL, not to be such a mystery. Paul Thagard's How Scientists Explain Disease (1999) is rather compelling evidence for that remark. His discussion does not at all rely on any linear or simplistic models (what respectable academic or scientific account these days does?) of causation. (He also has an interesting table on p. 61 that lists the many and complex 'criteria for causation.')

[Daniel S. Goldberg]

Fair enough. Re huntington's, you might find interesting if you have not read it, Pemberton & Wailoo's book entitled "The Troubled Dream of Genetic Medicine."

This exchange has been edifying, and I thank you for that. I suspect our disagreements are not quite as sharp as they appeared at times in this discussion.

[Ofer Zur, Ph.D.]

The DSM is a political not a scientific document. It pathologizes women, children, and minorities. It defines existentially normal behaviors as mental illnesses. It is a money making endeavor for psychiatry and other mental health professionals. It ‘dares’ to define what is normal and what is abnormal and who should be free or detained against their will. A detailed critical article about the DSM at http://www.zurinstitute.com/dsmcritique.html .
Ofer Zur, Ph.D.
Psychologist
Zur Institute