Wednesday 23 February 2011

Adding philosophy to mental health research

I read today one of the many papers by my new line manager in the School of Health: the gifted and hard working Joy Duxbury, School lead on mental health research. It’s a co-authored piece looking at the defence of restraint and the rationale for training in methods of restraint in mental healthcare [Paterson & Duxbury 2007]. The main storyline, justified by a review of zillions of pieces of research, is that things are rather complicated. Thus, for example:

Training staff in restraint procedures can have a role to play in improving staff and service user safety, but only when used as part of broader integrated approaches that incorporate specific reference to the issues of power and inequality, and address the root causes of aggression and violence in how organizations provide services. [ibid: 542]

The paper is clear and persuasive in the view it takes, based on the literature reviewed, and I will add it to my masters programme reading list.

It is thus odd that it adds a philosophical manifesto at the start which runs thus:

There is ... a potential conflict between those advocating restraint reduction and those calling for staff to be trained in restraint. The seeming contradiction reflects, however, an underlying question that can usefully be construed to involve the ‘validity’ of physical restraint as an intervention. Validity as a concept has several dimensions, including those central to the positivist research tradition, particularly in the behavioural sciences, namely ‘content’, ‘construct’ and ‘predictive’ validity. This approach to validity reflects a modernist worldview in which the purpose of knowledge is to provide a map of an objective reality, and validity provides a measure of truth in terms of the correspondence of the map provided with that reality. However, in an alternative postmodern worldview, the concept of an objective reality against which knowledge can be validated is discarded and with it the ‘correspondence theory’ of truth as the basis for understanding validity. The quest for certainty in knowledge is abandoned, replaced instead by defensible knowledge claims. Validation becomes the issue of choosing among competing interpretations framed as potentially falsifiable and of critically exploring the relative credibility of differing claims regarding knowledge. [ibid: 536]

This is heady stuff. Validity is not to be understood on the model of aiming to get an objective reality right but rather as choosing among competing interpretations and exploring the relative credibility of differing claims, but not, presumably, as credible with reference to their descriptive accuracy since that notion goes with (and is rejected with the rejection of) objectivity and world-mapping. There are a couple of further positive clues about validity. It is, with shades of pragmatism, linked to what works (although that is scare quoted to sever, I suspect, an objectivist notion of what really works). And: In this postmodern discourse, validity is a ‘linguistically, politically, economically, socially, culturally and professionally relative’ concept whose meanings are multiple and flexible rather then fixed.

There are a couple of distractions. Aiming to get reality right need not involve a commitment to certainty (although it would be nice) so the latter’s rejection is common and is distinct from the main radical claim. And the mention that rival interpretations are ‘potentially falsifiable’ might suggest falsifiable against observations of an objective reality. (That is what most of the staid and steady modernist Popperian scientists I meet would mean when they too endorse this view.) But in this case, it must mean reject-able in PoMo terms.

What is surprising, though, is that this radical claim plays no role, as far as I can tell, in what follows. The flavour of that is indicated in how the main argument begins:

If restraint as an intervention is to be considered valid from a pragmatic perspective it must be able to demonstrate that it ‘works’. However, the present literature largely precludes any safe answer to this question because the evidence base is, with limited exceptions, of very poor quality. [ibid: 537]

What then follows is a very helpful and plausible investigation of the evidence about the benefits of restraint. Poor quality seems to mean: quite likely not to be true. A typically balanced passage runs:

There are potentially strong arguments in favour of training in restraint as part of wider training in the prevention of violence. Fisher has argued that restraint can prevent imminent harm to self or others, substantial damage to the physical environment and the serious disruption of treatment programmes, and can decrease stimulation. He has also raised the issue that it may be valuable when implemented in response to service user requests... Reductions in the use of restraint following training have also been reported. Unfortunately, as Allen observes, negative outcomes have also been found in respect of these measures in some studies. In the UK, one explanation offered for the negative results sometimes associated with training has been that the ‘importing’ of training models from non-health services (prison) led to a widespread and persistent overemphasis on physical intervention during training in the prevention and management of violence in some programmes. [ibid: 538]

What strikes me in this is that the claims made (I’ve removed the supporting footnotes) function in pretty much the way you’d expect a ‘modernist’ argument to function. There’s no reliance on arguing over the credibility, understood in novel non-world revealing terms, of competing interpretations. The interpretations are lined up against the evidence and given an old fashioned modernist biffing.

Now one might say that, in that passage, there’s no appeal to truth so that’s obviously already radical. But, no, that is just what we’d expect in a dull modernist conversation. One need not say: the sky is blue because it’s true that sunlight undergoes electron scattering. One can just say: the sky is blue because sunlight undergoes electron scattering. Still, in making that assertion, one undertakes to be disciplined in a particular way: aiming at truth. (To put this point the other way, one could take the ground level claim and add in the it is true that operator. But it would add nothing not already implicit in the ground level claim. That is why one could always add it. If one could not, one would have to withdraw the ground level claim as well. So the explicit avoidance of truth in the ground level claim tells us nothing.)

I am not at all disappointed that the philosophy drops out of this paper. If it didn’t, I wouldn’t recommend it to students. It would be a pity if the only way to advance the key claim were to water down the standards of truth and objectivity to which the authors aspire. “We don’t care about truth and objectivity so now we can say whatever we want!” No, that’s a shiny ball just to attract the reader before the sober analysis begins.

I comment on this minor aspect of a substantial paper because I have long been interested in the way philosophy becomes combined with empirical results in thinking about mental healthcare (we have a module on such philosophically based research methods in our masters programme on Philosophy and Mental Health). At its best it is a natural and organic development from reflexive thinking about the conceptual complications of mental healthcare. In this particular case there are obvious and key philosophical issues raised, and calling for analysis, without the need to propose a new theory of truth. One is surely this: even when we know, through the kind of analysis that Brodie Paterson and Joy Duxbury offer, the extent to which restraint and restraint-training have the effects they do on mental healthcare, we will still face the question of whether those effects are desirable and / or justifiable. That is not an empirical question.

For information on the philosophy and mental health look here.

For further thoughts on social science and correspondence theories of truth see here.

Paterson B & Duxbury J (2007) Restraint and the question of validity, Nursing Ethics, 14 (4), pp. 535-545