Thursday, 7 April 2011

Draft commentary on Pies, Thommi and Ghaemi

Ronald Pies, Sairah Thommi and Nassir Ghaemi,
(Professor of Psychiatry and Lecturer on Bioethics & Humanities, SUNY Upstate Medical University; and Clinical Professor of Psychiatry, Tufts University School of Medicine; Research Assistant, Mood Disorders Program, Tufts Medical Center; and Professor of Psychiatry, Tufts University School of Medicine; Director of the Mood Disorders Program, Tufts Medical Center)
have written a paper called ‘Getting It from Both Sides: Foundational and Anti-Foundational Critiques of Psychiatry’ for the AAPP Bulletin commentary process.

Their papers begins:
“Modern-day psychiatry has been the target of numerous social, philosophical and scientific critiques over the past century, sometimes lumped together as manifestations of “anti-psychiatry.”  The aim of the present paper is to place the critics of psychiatric theory and practice in the broader framework of two philosophical traditions: logical positivism and post-modernism. Even more broadly, we want to distinguish two “meta-categories” of philosophical discourse, which we call “Foundational” and “Anti-Foundational.”  To oversimplify greatly, logical positivism may be considered a subset of foundational philosophies; and post-modernism, a subset of anti-foundational philosophies.  We make the latter claim, fully aware that the term “post-modernism” is subject to many interpretations; is sometimes considered vague to the point of meaninglessness; and is, in some ways, more a literary and cultural attitude than a well-articulated philosophical position.  Nevertheless, as a particular subtype of anti-foundational philosophy, post-modernism remains a useful heuristic term in understanding various critiques of psychiatry.
The burden of this paper will be to outline the historical roots of foundational and anti-foundational philosophies; describe how these philosophies have provided the basis for a “double-barreled” assault on modern-day psychiatry; and finally, to adumbrate very briefly why both kinds of attacks on psychiatry are generally unfounded.  First, however, we need to provide at least a notional idea of what the term “antipsychiatry” encompasses.”

My very rough draft commentary runs – today, at least – as follows:

Why taxonomise anti-psychiatry?
Of all disciplines, psychiatry is particularly keenly aware of the importance of a good taxonomy. Whilst in some scientific disciplines the explicit focus is on explanatory theories and there is only implicit attention to the taxonomies they presuppose, in psychiatry, getting the taxonomy right is one of the key foci of intellectual endeavour. This attention has helped reveal different virtues of taxonomies. Thus until recently, the key virtue aimed at for the DSM taxonomy has been reliability: roughly, the non-collusive agreement in applications of the taxonomy in classificatory judgements.
For DSM V, the key aim is, we are told, validity. But even validity can be subdivided. It might mean, for example, any of these or others:
·         Face validity: the extent to which a classification appears to be of relevant features (which has consequences for the acceptability of tests to test users ad subjects [Rust and Golombok 1989: 78]).
·         Construct validity: roughly, the extent to which it relates to underlying theory. Kendell articulates this thus: ‘the demonstration that aspects of psychopathology which can be measured objectively… do in fact occur in the presence of diagnoses which assume their presence and not in the presence of those which assume their absence’ [Kendell 1975: 40]. Anastasi says it is ‘the extent to which the test may be said to measure a theoretical construct or trait’ [Anastasi 1968: 114].
·         Predictive validity: the extent to which the classification allows us to predict future properties.
·         Content validity: ‘the demonstration that the defining characteristics of a given disorder are indeed enquired into and elicited before that diagnosis is made’ [Kendell 1975: 40].
So it is appropriate in thinking about a proposed taxonomy of forms of criticism of psychiatry – forms of anti-psychiatry – to examine the intellectual virtue of the proposal in something of the same spirit as critical reflection of psychiatric taxonomy itself. In this case, my concern is not so much whether anti-psychiatry can be divided into forms which are foundational and forms which are anti-foundational (although I do have worries about quite how this is proposed). It is rather whether we learn anything from doing that. I am not sure that we do.
An initial requirement for the taxonomy
I can illustrate one challenge to a taxonomy of this sort by looking at the business the paper sets itself. Called ‘Getting it from both sides’ it says:
The burden of this paper will be to outline the historical roots of foundational and anti-foundational philosophies; describe how these philosophies have provided the basis for a “double-barreled” assault on modern-day psychiatry…
A key theme is that psychiatry is criticised from both sides of a distinction. This sets up a particular kind of expectation about the significance of the duality that forms the taxonomy which I will illustrate indirectly.
One way to fail to meet the expectation would be to propose a taxonomy of forms of anti-psychiatry based on the position in the alphabet of the first letter of the first author of the attack. If this were a simple duality – of first half versus second half – then (given the names ‘Foucault’ and ‘Szasz’, eg.), psychiatry would come under fire from both sides, a ‘double-barreled’ assault on modern-day psychiatry if you like. In that hypothetical case, there would, however, be no significance (such as, perhaps, a bitter irony) that psychiatry were attacked from both sides. Although the distinction is in one sense perfectly valid, and cuts anti-psychiatry at the joints of author nomenclature, it fails something like construct validity. It fits no deeper theory of anti-psychiatry.
So one test of the proposed taxonomy is that it does have significance. The ideas of foundationalism and anti-foundationalism should shed light on the nature of the anti-psychiatry in a way that mere surnames do not (because of the ‘arbitrariness of the signifier’ as Postmodernism, in particular, has taught us to say).
But there is a worry from a potential response to this. Suppose that from anti-foundationalist premises, a form of anti-psychiatrist were justified, it followed logically. And from foundationalist premises, another form of anti-psychiatrist were similarly justified. Then on the assumption that either foundationalism or anti-foundationalism is true, some form of anti-psychiatry would be justified come what may. So, as far as a defence of psychiatry goes, we had better hope that the relation of significance (between the category and anti-psychiatry) is not implication.
Whatever the kind of significance turns out to be, at the very least, some kind of light should be shed on anti-psychiatry by seeing it in the context of the taxonomy.
What is foundationalism?
The paper suggests that foundationalism has two key aspects. One is a traditional epistemological notion. Knowledge is based on a foundation (of experience, or belief) which is not itself (inferentially) dependent on anything else.
Logical positivism in its various forms is a modern-day expression of the foundational world-view...  [It] essentially held that all knowledge is based on logical inference grounded in observable fact...
Foundationalism, in this traditional epistemological sense, is usually held to contrast with forms of holism which deny that any of our beliefs, such as perceptual beliefs, are privileged and instead each is potentially subject to revision. A belief in the theory dependence of observation is one reason to support holism in the philosophy of science.
The other aspect is expressed in this way:
In simplest terms, foundational philosophies and philosophers hold that we can reliably describe a coherent, objectively-measurable “reality” or “truth,” whether one considers the world as a whole, or specific aspects of it, such as the classification of disease.  Anti-foundational philosophies and philosophers deny this claim, asserting that there are no objectively demonstrable “truths”—only various “perspectives” or “narratives” that cannot be privileged as uniquely or objectively “true.” 
Whilst this does contain an epistemological element in the claim that our descriptions can be reliable, the main thrust is ontological. It is the idea that foundationalists hold that there is an independent world to serve as a standard for truth. Elsewhere they describe this as ‘a confidence — some might say, a faith — that the world and its constituents actually exist’. This aspect would often be described as a form of (ontological) realism which contrasts with the anti-foundationalists’ claim that there are no objectively demonstrable truths.
(It may not be quite realism in a standard form, however, since the most obvious opposition to ontological realism is idealism, whilst the authors take Berkeley to be a foundationalist: ‘Berkeley effectively dispensed with the concept of material substance, but most certainly was a foundational philosopher: he merely argued that the “foundation” of reality consisted of ideas in the mind of God!’ But even in this case, the idea may be that whatever the substrate of the world, it is independent of claims made about it. That serves as a test of truth and thus stands in contrast with the anti-foundationalists’ mere interplay of narrative.)
There is some danger in combining both these aspects – epistemological and ontological – under a single term which can be illustrated by a philosopher mentioned in the paper: Quine. In his famous paper ‘Two dogmas of empiricism’ Quine explicilty rejects the idea of foundations when he rejects the’ dogma of reductionism’ which is the ‘supposition that each statement, taken in isolation from its fellows, can admit of confirmation or infirmation at all’. [Quine 1953: 41] But he continues ‘My counter suggestion… is that our statements about the external world face the tribunal of sense experience not individually but only as a corporate body.’ So whilst he rejects privileged epistemological foundations he does not reject the idea that our beliefs answer to something independent of us.
This is significant because that combination of ideas is the dominant view held by philosophers and self-conscious scientists alike. Epistemological foundationalism is dead. No observation is thought to be free of its theoretical context and thus, like any scientific statement, is fallible. But rejecting that view does not commit one to a denial that our beliefs answer to a world largely independent of us, nor to the embrace of mere shifting narratives.
Given that the taxonomy is offered, not for philosophy as a whole, but rather for anti-psychiatry, it might be that no anti-psychiatrist fails to combine the appropriate epistemological and ontological views. But if the taxonomy is to shed light on anti-psychiatry, such correlations should be explicit and subject to explanation rather than hidden in the taxonomy.
The application of the taxonomy
Having set up the taxonomy, the authors apply it to particular critics of psychiatry. I will discuss just the first: Szasz. Responding to a recent summary by Szasz of his original argument they say:
A full-blown critique of this argument is beyond the scope of this paper.  However, it is instructive to note some of the key “properties” of Szasz’s claim: (1) It is based on an implicit assertion that “analytic truths” are not empirically falsifiable—a claim that Quine is at pains to challenge; (2) It appears to remove from the realm of scientific investigation the question of whether schizophrenia or bipolar disorder, for example, are diseases or illnesses; (3) It conflates the terms “disease”,“illness”, and “disorder” without any attempt to discern conceptual or clinical distinctions among them; and (4) It implies that there is a single, univocal “materialist-scientific definition of illness” to which one can appeal, and which then can be used unambiguously to compose an “analytic truth.”  Also note that the hyphenated term “materialist-scientific” implicitly suggests that science and “materialism”—roughly, the view that the only thing that exists is “matter”—are linked in some essential way.
In the context of a paper suggesting a categorisation of anti-psychiatry, I would expect that this list would demonstrate how Szasz fits his assigned place: foundationalism. And indeed, pace my worries about Quine, the first point does. Szasz is within a tradition of philosophy which accepts analytic truths and Quine, at least, has argued that this is an important part of foundationlism. Point 2 does not obviously exemplify the category but, perhaps, neither does it contradict it. Point 3 seems to lie simply outside the terms of the taxonomy. One might be guilty of this which ever side one belonged to. Likewise, point 4 does not seem to be an effect of or have anything to do with foundationlism.
That is a bit odd. Only the first point helps locate Szasz on the foundationalist side of the taxonomy.
There is then an argument against Szasz. In a paper outlining a taxonomy, such an argument is not the main business. But it may illustrate what we learn from applying the taxonomy and thus why the taxonomy is helpful. The central argument runs:
Szasz’s argument purports to rest upon an analytic statement—similar in kind to “All bachelors are unmarried males”—while implicitly drawing upon the historical and empirical claims of “materialist” science. Yet any putative “materialist-scientific definition of illness”—to the extent we can even specify one—did not arise ex nihilo or out of some syllogism; but rather, from specific empirical observations of cells, tissues and organs, by pathologists like Virchow and von Rokitansky.  Thus, Szasz’s argument that “mental illness is a metaphor” seems to us far from a straightforward “analytic” claim; rather, it appears to be a pseudo-analytic claim that depends critically on a huge body of historical, synthetic and empirical claims.
One way of approaching this argument is to think that it helps demonstrate the value of the taxonomy. If Szasz is a typical foundationalist and if typical foundationalists presuppose the analytic-synthetic distinction, but if that is an invalid distinction (as Quine has argued), then Szasz’ argument will fail and it will fail because he is a foundationalist. That would be a partial vindication for the taxonomy. Putting him into that camp helps shed light on why he is wrong.
But it is not clear that that is what the authors intend here. The comment that Szasz’ claim is ‘far from a straightforward “analytic” claim; rather, it appears to be a pseudo-analytic’ suggests that an analytic claim might be in perfectly good order. The problem is not so much that Szasz is appealing to the notion an analytic truth, rather, he is doing that badly. If so, the problem with Szasz’ anti-psychiatry is not that it is foundationalist but that it is bad foundationalism. But if that is the case, the taxonomy of anti-psychiatry into foundationalism and anti-foundationism does not seem to be carving the nature of anti-psychiatry at the right – significant, informative – joint.
It is also worth noting that if that is not the meaning of that phrase and that any appeal to analyticity is misguided (thus preserving the point of the taxonomy for the foundationalist side), the key architect of the downfall of analyticity is Quine whom the authors call an anti-foundationlist. So why would not the failure of foundationalist anti-psychiatry be a partial argument, at least, for the success of an anti-foundationalist variant? In fact in the later parts of the paper, anti-foundationalist anti-psychiatry is criticised on grounds which do not even mention analyticity. Thus no light is shed on criticisms of anti-foundationalists in virtue of their analyticity-eschewing position in the taxonomy. In either case, at least one side of the taxonomy will not be informative.
A different taxonomy?
I think that it is a mistake to hope that a binary opposition which locate forms of anti-psychiatry on both sides will, in itself, be very helpful. How could it? If a binary distinction exhausts logical possibilities – if everything is either in the one or the other category – then all the positions we can take will be in one or the other. All forms of philosophical view which support modern psychiatry will be located rubbing up against the views which oppose them. The taxonomy will not ehd light on the difference between the pro- and anti- view.
I think that a more fruitful approach is a taxonomy of approaches to the nature of mental illness itself. Here are two, related distinctions.
One key disagreement is whether mental illness in particular, or illness more generally, is essentially evaluative. Does the analysis of mental illness contain reference to values or not? Some philosophers and psychiatrists argue that at the heart of the idea of illness is something that is either bad for a sufferer or is a deviation from a social or moral norm. Both of these are evaluative notions and hence both are ‘values in’ views.
Others argue that it is, what I will call, a plainly factual matter. Typically, they argue that illness involves a failure of a biological function and function – and hence deviation from, or failure of, function – is a plainly factual, biological term couched in evolutionary theory. Of course, disagreement about the presence or absence of values in the analysis is just one aspect of the debate. It is a further question, for example, what follows from this for the objectivity of mental illness and the status of psychiatry as a science. For Szasz this is the basis of an argument against psychiatry. For Bill Fulford (and the authors of the paper), for example, it is not.
A second useful characterisation links the debate about mental illness to other debates in philosophy about the place in nature of problematic concepts. On this second construal, the question is whether mental illness can be naturalised. That is, can mental illness be accommodated within a satisfactory conception of the natural realm?
The most common form of philosophical naturalism is reductionism which attempts to show the place in our conception of nature of puzzling concepts by explaining them in terms of, and so reducing them to, basic concepts that are unproblematically natural. So on this second characterisation of the debate, a pressing question is whether, or to what extent, the concept of mental illness can be reduced to plainly factual concepts. If it cannot be naturalised, to what extent is it consistent with a scientific account of the world.?
What makes reductionism difficult is that different concepts can seem to behave quite differently from one another. Take, for example, a distinction drawn from the work of the philosopher Wilfrid Sellars, and repopularised by John McDowell [McDowell 1994] between the ‘realm of law’ and the ‘space of reasons’. Whilst the space of reasons concerns meaning-laden and normative phenomena that we take for granted in understanding minds, the realm of law concerns events that can be explained by subsuming them under natural scientific laws. In the philosophy of mind, reductionists attempt to show how the space of reasons can be completely explained using the resources of the realm of law. Anti-reductionists argue that the normativity of mental states and meanings – the fact that beliefs can rationalise and support one another, can be right or wrong – cannot be captured in terms, for example, of statistical laws of association.
In fact, value theorists in the debate about mental illness are making a similar point to anti-reductionists in the philosophy of mind. They argue that the very idea of mental illness is a normative notion – since values are normative and have a good versus bad dimension – and for that reason cannot be reduced to plainly factual or realm of law terms.
Using distinctions such as these provides tools for the diagnosis of assumptions made both by those who oppose and those who support modern psychiatry. In the paper, the authors criticise one ‘foundationalist’ anti-psychiatrist in this way:
Stevens’ foundational critique is built upon a scaffolding of selective quotes from a large cadre of mental health professionals... all in the service of showing that we cannot identify any biological abnormalities in any of the major psychiatric disorders; and that, absent such physical “causes,” these conditions cannot be considered bona fide diseases.
Whilst I agree with the criticism they go on to make, I do not think that it helps to characterise Stevens as a foundationalist. The quotation does not imply anything about his epistemological views nor would ontological realism about an objective world help shed light on his particular brand of anti-psychiatry.
What is picked out in the quotation is an assumption that he has made about what counts as real: biological abnormalities. And thus he counts as both ‘values out’ and reductionist in the two distinctions above. This is not yet to provide an argument against his position. But it does help outline the commitments he needs to maintain. He owes an account of why biological abnormalities are all that can count as real in this context. A biologically minded reductionist supporter of modern psychiatry will agree with Stevens on that metaphysical claim and will have to look elsewhere to disagree. But an anti-reductionist supporter of psychiatry can target that assumption. The assumption – and hence his reductionism – sheds light on his position.
In sum, I think it a mistake to aim a taxonomy at anti-psychiatry rather than at views of mental health and illness (or disease or disorder) in general. It will probably, at least, not be particularly helpful. If the aim is, additionally, to attempt to undermine anti-psychiatry, then it seems doubly mistaken because it cannot work. If – and this is a key assumption – we were to assume, from the perspective of a defender of contemporary psychiatry, that anti-psychiatry were simply misguided then a helpful analogy might be with Tolstoy’s views of unhappy families in Anna Karenina. There is little point in aiming to taxonomise arguments against psychiatry because whilst valid views in support of psychiatry are all alike; every unhappy anti-psychiatric argument is unhappy in its own way. But to evaluate that key antecedent assumption, we will need valid general taxonomies of views of health and illness and nature in general.
Anastasi, A. (1968) Psychological testing, New York: Macmillan
Kendell, R.E. (1975) The Role of Diagnosis in Psychiatry, Oxford: Blackwell
McDowell, J. (1994) Mind and World, Cambridge, Mass.: Harvard University Press
Quine, W.V.O. (1953) ‘Two Dogmas of Empiricism’ in From a Logical Point of View, Cambridge, Mass.: Harvard University Press
Rust, J. and Golombok, S. (1989) Modern Psychometrics, London: Routledge