Saturday, 4 March 2017

understanding psychoanalytic understanding


Consider a psychoanalytic theory such as Freud offers for paranoia. In nuce this has it that paranoid beliefs - that others are harbouring hostile intent towards one - are caused by projection onto them of one's own hostile feelings toward oneself. I find my anger with myself utterly intolerable, so to keep from being overwhelmed by shame or incongruence or terror I instead attribute the anger to you. Now I no longer have the intolerable tear-myself-apart anger at myself - great! What was inside has been relocated outside, in you. But, unfortunately, now that I've populated my world with my hostile projections, my world becomes frightening. And it also makes me angrily resentful. Because, damn it, why are you being so hostile to me?!

That, at least, is a simple projection-based understanding of paranoia - but we might also offer a further or alternative dynamic to explicate at least some of it's forms. Namely that my live delusional sense of retaliatory anger from you arises because unconscious anger at you has awakened in me a particular complex, a particular transference dynamic, a particular unconscious phantasy, in which two people (you and I) are locked in an intractable angry retaliatory battle with one another, and yet of this I am only left in conscious touch with your feelings towards me. The thought being that for me to be able to acknowledge/own my anger, it is not so much that it would have to be reclaimed from you, but rather that it would not now be simply fuelling the phantasy in which your retaliatory hostility is provoked. (This is projection in a very different - non-dynamic - sense than in the first theory: projection type one involves a motivated misattribution; projection type two involves but a misreading of you along the lines of my fearful phantasy - combined with repression of my own hostility.)

My question is: what is the 'logical status' of such theories? Are they empirical theories of a sort which have a truth value which could, in principle at least, meaningfully be ascertained through experiment? Another way to put this is: is their objectivity of a sort evincible through testing? Or are they redefinitions of paranoia - now anything worthy of being called 'paranoia' necessarily is to be understood as a function of projection or what-have-you? Or are they let's-call-it phenomenological theories which aim to give us a deeper characterisation of the intelligible forms of paranoid thought?

One way to think about what it can mean to describe a cognitive (i.e. knowledge-providing) discipline as scientific, as opposed to non-scientific, is to focus in on its use of 'the' scientific method. Perhaps there are in fact many scientific methods, but the idea I'm here referencing with that phrase refers to a general methodological feature of good experimental science which always gives to nature alone the opportunity of answering the questions one puts to it. Imagine that one is investigating the causes of upper back ache. Various dietary and viral contributions are proposed. Naturally the form of our investigation will be to test the proposals - hypotheses are formed, and the diet and viral infections of people who have or who go on to develop back ache (and of those who don't) will be investigated. The hypotheses are clearly testable: if the achey people don't have more viral infections then the hypothesis would appear a dud. The results of the investigations may be predicted beforehand, but the fact is that back aches and viral infections are individuated utterly separately, and this external relation between them makes for a maximum ease of application for the scientific method. Nature can then answer quite independently of our intuitions, and this fact is valuable to us in establishing the objectivity of our thought.

Nothing in what I am saying is intended to promote the idea that the only method for establishing thought worthy of being called 'objective' is the scientific method. (To think in that way is to submit to a rather bizarre form of positivism.) Here, for example, is another way. I make a claim and we get clear on whether or not it has application by spelling out a range of alternatives and how to tell them apart. For example, at the beginning of this post I offered two psychoanalytic understandings of paranoia. The apt question here, I imagine, is not the general 'which is the right understanding?' since, unless one of the understandings makes less sense than I suppose, they surely may either of them find application in this or that instance - and will, when they do, reference two different forms that paranoia may take. Depending on how you are apt to individuate meanings we may even find ourselves wanting to say that here we have two different possible meanings that paranoia can possess. At any rate, so we know that we here genuinely do have two different explanations, and not simply two different sentences, what is required is that we can think how to tell them apart in practice. And this is fairly easy. The first question we would want to ask would, I imagine, be whether the feelings and thoughts which the paranoid person attributes to their feared other are feelings and thoughts which do in fact apply to themselves. Or do they simply characterise not what the observer might think of as aspects of the paranoiac's character, but the different matter of what the paranoiac expects to meet with in others? Does the patient appear more shamed by the idea of the content of the paranoid thoughts applying to them or by the content of the accusations that their feared object makes of them? And so on. The objectivity of my thought - the fact that I can't just whimsically say whatever I want, that I'm being properly accountable - is guaranteed here not by experimentally testing the ideas but by being very clear about what the differential criteria are for their instantiations. This provision of a more deeply specified description of a phenomenon is what I'm calling here a phenomenological understanding. (You will have noticed that I slyly answered my own question already: psychoanalytical understandings are often phenomenological rather than scientific.)

Take by contrast the kind of explanation of paranoia offered us by Richard Bentall. Bentall's is a kind of psychodynamic explanation, but he aims to provide the kind of explanation which is, at least in part, scientifically testable. Bentall's hypothesis is that 'paranoid patients show an exaggeration of the self-serving bias (the normal tendency to attribute positive events to the self and negative events to external causes) and ... when they make external attributions for negative events, their explanations usually implicate the intentions of others (the make external-personal attributions) rather than circumstantial factors (external-situational attributions).' He also claims that delusions are understood by thoughtful psychologists to be formed by people trying to come up with explanations to make sense of those of their experiences which otherwise puzzle them. Finally he claims that the self-serving bias has a function and doesn't just lead by chance to a happy outcome - the function is the happy outcome itself (i.e. the external attributions serve to protect self-esteem). His explanation as to why paranoiacs make negative attributions of others rather than of situations has to do with hyper-vigilance concerning threat-related information and/or impairment in reading situations.

There are several things I take issue with in Bentall's account. First I'd say that it is mainly the not-so-thoughtful (cognitive) psychologists who've accepted this notion of delusion-as-explanation, whilst the phenomenologists and psychoanalysts are more thoughtful in instead seeing delusions as rationalisations (not explanations) which are motivationally (rather than rationally) intelligible (i.e. they bring relief not through rationally making sense of things - which often they strikingly fail to do - but through locating inconvenient intentions or experiences elsewhere). Second I'd suggest that paranoiacs find it easier to make negative attributions to others (rather than to situations) because one can only meaningfully project an intention into, or take part in a dispute with, an other (situations aren't agents so can't be said to have intentions or to dispute). But leaving such disagreements aside, what I really want to focus on is the logical form of his explanation. It is not at all obvious to me how to test the idea that the patients are motivated in their self-serving bias. But it appears much clearer that the claim that paranoid delusions are a product of a heightened general self-serving bias is an explanation of a testable sort. Here all one needs to do is measure the self-serving bias of a population and see if higher measures correlate with higher measures of paranoia. In a way it's not a very satisfying explanation, since it just shunts our question a step further back: why does the patient have a higher self-serving bias? And if here we appeal to their need to defend against painful emotions we will, I suspect, be back in the territory of phenomenological understanding rather than scientific explanation.

Bentall does try to give scientific data correlating low implicit self-esteem and paranoia, but it all becomes very convoluted and controversial at this point. In any case our experience of defences is that they quite often cause more problems than they solve (e.g. the pains of now living in a hostile-seeming world), and are often not that effective despite being strongly motivated - so just because someone may use a defence of external attribution / projection to manage their latent negative feelings about themselves doesn't mean that we should expect any very clear correlation between one's latent feeling about oneself and the degree of one's paranoia. But be that as if may, my point here has been to contrast the kinds of explanations one may meet with in cognitive science, in which testing may be the way to establish objectivity, and the kinds of explanations one may instead meet with in psychoanalysis, in which the making of clear phenomenological discriminations may instead be the way to establish objectivity.

At a talk at the UK Institute of Psychiatry about 15 years ago a young clinical psychologist - Paul Tabraham I think it was - presented his case formulation of an individual psychotic patient. Causal relations between his patient's self-worth, self-understanding, symptoms, relational experiences, mood etc were presented one by one in diagrammatic form (with, if I remember rightly, the boxes and arrows beloved of cognitivist psychologists). The then head of the Institute, Robin Murray, kept intervening to ask what scientific (i.e. general) evidence there was that such associations actually obtained; he was clearly concerned about the objectivity of this psychologist's claims. Naturally, the psychologist had no such general evidence and became increasingly exasperated. Eventually Anthony David intervened to say something - I forget what - about the epistemological difference between individual and general explanations in psychology, and the rather tense matter was, if not understood, at least passed over.

Murray's idea seemed to be that the legitimacy of individual claims ought to be a function of the extent to which they instantiate general truths which have been scientifically tested. Yet this not only ignores the ways we make apt causal claims every day without any kind of general back-up (he broke his hip because someone tripped him up with a stuffed albatross), but also asks the wrong question for the current context. The right question is: how can I tell whether or not this perfectly intelligible scheme from amongst many such possible schemes actually is instantiated in this instance? To answer that question it helps me not at all to discern whether that scheme is to some or other degree generally instantiated - since the question still remains as to whether it is instantiated here and now. Contrast that with the kind of research claim which interests Murray - e.g. 'does cannabis use cause non-transient paranoia?' In the case of cannabis use it's clearly important that we employ the scientific method in population studies that correlate cannabis use and paranoia over time. And then, when we encounter a particular paranoid individual, and suggest that he went paranoid because he smoked a lot of weed, our suggestion regarding the particular will be all the more reputable if we can back it up with the general findings.  Yet nothing like that makes sense for the kind of psychoanalytic claims regarding paranoia with which I began this post. For such claims their intelligibility as schemes of understanding is perhaps all that is required at the general level, and at the particular level what is required is evidence that they are here and now - and not evidence that they are to some or other degree quite generally - instantiated.