Thursday, 4 February 2010

On wishing for a paradigm shift

I need to draft a short commentary (say 2,500 word) for the AAPP bulletin which has as its focus, in the words of Jim Phillips, ‘a couple recent articles on DSM-V - and the conflict they have generated - by Allan Frances. Dr. Frances, primary architect of DSM-IV, published articles in Psychiatric Times in June and October of this year criticizing the process of developing DSM-V. His June paper generated a lively exchange, with critiques, responses, and counter-responses by some of the major players.’

And so again in accord with my original motive for this blog (to tell my students about putting ships in bottles), I'll pop growing versions of it in this (thus changing) post.

On wishing for a paradigm shift

Tim Thornton

One aspect of the recent discussion of the development of DSM V has been whether it should aim to express, or perhaps bring about, a paradigm shift. Allan Frances has described the initial optimism about the possibility of such a change as ‘absurdly premature’. He writes:

The DSM-V goal to effect a “paradigm shift” in psychiatric diagnosis is absurdly premature. Simply stated, descriptive psychiatric diagnosis does not now need and cannot support a paradigm shift. There can be no dramatic improvements in psychiatric diagnosis until we make a fundamental leap in our understanding of what causes mental disorders. The incredible recent advances in neuroscience, molecular biology, and brain imaging that have taught us so much about normal brain functioning are still not relevant to the clinical practicalities of everyday psychiatric diagnosis. The clearest evidence supporting this disappointing fact is that not even 1 biological test is ready for inclusion in the criteria sets for DSM-V. [Frances 2009: 2]

In the first half of this short note, I will draw out some of the consequences of this on a broadly Kuhnian picture of science. Central to this picture is the connection between paradigms and the meaning of theoretical terms and hence the connection between changing paradigms and the consequent incommensurability of the meanings of terms across time. It is this that helps to support Kuhn’s theoretical scepticism about whether sciences can be said to progress. And this in turn calls into question whether it can be rational to wish for a paradigm shift.

In the second half, I address a related but more substantial point. Whilst the wish for a paradigm shift typically reflects optimism about the developments of neuroscience (notwithstanding the connections just summarised), psychiatry aims to use its technical innovations to relieve human distress. An improved psychiatry should thus be better able to address issues of distress. But if so its understanding of distress - which guides diagnosis, treatment, management and shared plans for recovery - had better remain closely wedded to the self-understanding of those it is supposed to help. And if so, any plans for a paradigm shift threatens to disconnect technical psychiatry from the understanding of human agents which should guide it.

Paradigms incommensurability and progress

The widespread use of the word ‘paradigm’ in the description of scientific change is the fault of Thomas Kuhn’s Structure of Scientific Revolutions. Although he is said to have used it in at least 21 different ways, one basic idea is central [Masterson 1970].

Kuhn argues that scientific activity falls into two kinds. In the main, scientists are engaged in ‘normal science’. This comprises the articulation and application of stable dominant theories and meta-theoretical assumptions to new areas. Kuhn calls this background the dominant paradigm. During such periods, no serious attempt is made to refute or even defend the theoretical background which is instead simply presupposed. But these stable periods of normal science are punctuated by brief periods of revolutionary theory change. Sparked both by the accumulation of anomalous results and by the development of rival theories or even rival meta-theoretical assumptions, the dominant orthodoxy is cast aside and a new theory or set of theories put in its place. Only during these revolutionary periods is the truth of what will become the new scientific background called into question.

Thus whilst during periods of normal science, some measure of progress can be based on an increasing ability to solve recognised puzzles against the background of a stable paradigm, that measure does not apply over periods of revolutionary change since a change of paradigm changes what is regarded as a potentially soluble puzzle.

In fact, however, a broadly Kuhnian view makes the idea of progress across a paradigm shift even more difficult. This follows from his, at the time, influential view of the meaning of theoretical terms. Like other philosophers and historians of science, Kuhn reacted against an influential view of the meaning of theoretical terms taken from the Logical Empiricists of the 1930s [Feigl 1970]. On that older view, theories could be judged against the standard of theoretically neutral observations and that separation was supposedly maintained by the independence of observation from theoretical language. Although theoretical terms were grounded in the observational predictions they collectively inferentially warranted, observational terms were thought to be definable antecedently.

A group of arguments towards the end of the twentieth century undermined that distinction between theory and observation (establishing instead the ‘theory dependence of observation’). Kuhn concluded that the holism that had been thought to apply to theoretical terms - albeit a holism constrained from the outside by their implications for observations claims - must apply to theoretical and observational terms collectively. But without a stable set of neutral observation claims against which to judge them, the new holism seemed to imply that a change of overall theory would change the context and hence the meaning of all theory-observation terms. As a result this seemed to suggest that there was no standard by which to compare overall theories across a paradigm change since different paradigms defined different scientific languages leaving no resources for a translation manual. And thus paradigm change was incommensurable and there could, in principle, be no content to the idea that science progresses.

This is not the only way to view the meaning of theoretical terms and thus not the only view of possibility of comparing theories. But part of the force of the idea of a paradigm shift is that the change of world view is radical. Indeed Kuhn himself notoriously suggested that after such a shift, scientists inhabited a different world. Thus any less radical account of the consequences of theory change would undermine the point of the suggestive word ‘paradigm’.

This, however, suggests that, at the very least, there is something strange about wishing to usher in a new paradigm. Without a standard by which to judge progress across such a change, what rational motive is there to wish for such a change? That, however, is not my purpose in recalling the close connection between paradigms, meaning and incommensurability. The real issue concerns the application of these ideas to psychiatry in particular.

Neuroscience, human distress and the prospects of paradigm change

In order to develop my real concern I will return to (and re-quote) the passage from Frances I quoted at the start. The most obvious reason for thinking that psychiatry is awaiting a paradigm shift are developments at the hard science end of psychiatry. Even Frances mentions ‘incredible recent advances in neuroscience, molecular biology, and brain imaging’ when discussing others’ confidence in the possibility of a new paradigm. Frances himself argues that ‘descriptive psychiatric diagnosis does not now need and cannot support a paradigm shift’ but he goes on to say that there ‘can be no dramatic improvements in psychiatric diagnosis until we make a fundamental leap in our understanding of what causes mental disorders’ and that the absence of biological tests in diagnostic criteria suggests that this has not been reached. But that comment does not distance himself from what might seem a plausible aspiration for a bio-medical psychiatry. What is needed, on this assumption, is greater biological understanding of ‘what causes mental disorders’ and a sufficiently ‘fundamental leap’ in that might give us the hoped for paradigm shift.

But I think that there are two fundamental complexities that this view - a view Frances does not sufficiently distance himself from - ignores. The first is that, within psychiatry, the focus of neuroscientific, biological and brain imaging technology is mental pathology. Progress has been recently made in these areas and more progress is needed but, additionally, progress is also needed in determining not just what causes mental disorders but what they comprise. What is it, in other words, for something to be a mental disorder? There is no reason to think that an answer to this question can be provided by neuroscience, molecular biology, and brain imaging since, insofar as these can help shed light on psychopathology, one needs first to have decided the extension of that concept then to study its neurological and biological underpinnings. Given the conceptual complexity of the very idea of mental disorder, and that what is so classed is so contested, any leap forward in knowledge of brain mechanisms needs to go hand in hand with answers to that question.

The second complexity stems from the first. Suppose that innovations in neuroscience, molecular biology, and brain imaging were used to articulate a form of psychopathology on the basis of its neurological similarity to currently identified forms but which had no connection to any mental distress or suffering. That would not, I suggest, mark a triumph of neuroscientific psychiatry. Rather, it would amount to psychiatry losing its way by losing its connection to its particular subject matter.

This point suggests, however, a Winchian point [Winch 1958]. Psychiatry, unlike a more disinterested study of the brain, has an essential connection to human distress and suffering. But if so, it has an essential connection to the concepts with which we, as agents, make sense of ourselves. Whilst there seems no constraint imposed by the subject matter of much of natural science on the limits of conceptual innovation (as long as the concepts arrived at can still be understood by at least some scientists), the concepts of psychiatry need to retain some connection to those concepts in terms of which we ordinarily make sense of ourselves. Only so, can human experiences play at least some guiding role for psychiatric diagnosis, theorising and care .

If this is so, then the wish for a paradigm shift in psychiatry seems doubly misplaced. First, the connection to incommensurability undermines the rationality of the wish. Second, a change which did not sever the connection to the concepts we use to make sense of ourselves would not be a paradigm shift since the innovation would be merely partial leaving in place standards for rational assessment of the technical innovations. But a properly radical paradigm shift which rendered the pre- and the post- shift worldviews incommensurable would have to sever the connection to those grounding concepts and that could only be because psychiatry had lost its way.


Feigl, H. (1970) ‘The “orthodox” view of theories: Remarks in defense as well as critique’ in Radner, M. & Winokur, S. (eds.) Analyses of theories and methods of physics and psychology. Minnesota studies in the philosophy of science Vol. IV. Minneapolis: University of Minnesota Press: 3-16

Frances, A. (2009) ‘A Warning Sign on the Road to DSM-V: Beware of Its Unintended Consequences’ Psychiatric Times June 26

Masterson, M. (1970) ‘The Nature of a Paradigm’, in Lakatos, I. and Musgrave, A. (eds.) Criticism and the Growth of Knowledge, Cambridge: Cambridge University Press

Winch, P. ([1958] 1990). The Idea of a Social Science and its Relation to Philosophy, London: Routledge