Thursday 15 September 2016

(Ab)normal psychology

In a presidential blogpost at the BPS last month, Peter Kinderman reiterated an argument from his book that there is no such thing as abnormal psychology. I have spoken to this debate once before, when I reviewed his book on this blog. Here is what I said then:
A superficially appealing argument raised here is that "abnormal psychology" is an unreasonable field of study; after all, we don't speak of "abnormal physics". There is an important idea here with which I find myself aligned. Using the word "abnormal" is indeed a needlessly unpleasant way of speaking about people, but the physics analogy doesn't fly. All physical phenomena are subject to the same basic laws (as far as we know), but that hasn't prevented the fruitful subdivision of their study into solid state physics, condensed matter physics, and so forth. When people have experiences of psychological distress, these tend to manifest in a propensity toward particular states of mind. Is it really so unreasonable to study these states in their specificity, cautiously categorising them until some better framework is offered?
I still stand by that more or less, but when I re-read Kinderman's argument this time around I felt more disposed to agree with something in the point he makes. What is he driving at here? It's a fun idea to probe.

Psychiatric diagnoses like schizophrenia can be said to be hypothetical constructs. That is to say they are theories about the nature of entities (what type of entities is controversial) that are held to exist. Because it is still hard to find solid external criteria by which to independently validate their existence, they are sometimes said to fail as valid constructs. This is not a fringe argument. It is acknowledged far and wide within academic psychiatry. That is why the validity/utility debate has such traction within the discipline. I have pointed out before that psychiatric diagnoses survive because they come to act as a sort of stand in term for quite real seeming experiences. Where they aspire (and fail) as hypothetical constructs, they succeed as intervening variables.

What does that mean? In the 1948 paper that introduced and distinguished intervening variables and hypothetical constructs, the former are simply a convenient shorthand for some collection of already observed (but potentially unexplained) empirical facts. The latter are supposed to be things, that have some "explanatory surplus"; if you can propose a successful hypothetical construct, you will be able to make new (and accurate) predictions about reality.

That term explanatory surplus is key. Although there is a way of reading Kinderman's argument that is unfavourable to him (namely that one can of course plausibly divide the study of psychology into common and and relatively uncommon processes), he is certainly on to something. Here are two reasons why:

1. In any given case in which an individual has a psychiatric diagnosis, I can make some rough empirical predictions based on aggregated statistical facts about that diagnosis. But because knowledge about psychiatric entities is generally obscured by how poorly defined they are (for now), I am largely at a loss to make tightly-specified predictions about individuals based on a decent theory. "Abnormal psychology" is a collection of useful observations about how certain people behave and what processes are present in particular groups. 

2. Even if I did have a well specified set of facts about such and such a psychiatric entity, the majority of any given person's behaviour will still be best explained by facts that are common to all people. Thus I can usually understand instances of aggression in terms of things like a person's likely fears and wishes, combined with the situational context they found themselves in. I can then add on some nice sounding post-hockery to the effect that they have "poor impulse control" (a variation on a capacity we all have more or less of) or something similar. In the absence of a well set out aetiological theory of any disorder (giving it "explanatory surplus"), I don't really have an explanation yet. Most people's behaviour (even psychiatric patients) can be mainly explained by principles that have been derived from general psychology. Only a little is added by factoring in the useful observations of psychopathogical research. 

I have to be careful. I am not downplaying the value of clinical psychological research. Nor am I one of those people who wants to deny that there could be something like illness processes present in many cases of DSM diagnosis. I think it is unambiguously clear that abnormal psychology exists in the context of neuropsychology and neurology. But I suppose I agree with Kinderman insofar as I think that most of the behaviour of most people can (and should, as far as possible) be understood in terms of the things that are common to everyone.