explanation and understanding

Richard Bentall and David Bell have rather different psychodynamic theories of paranoia. Not just different theories of what causes paranoia, but different forms of theory - different ways of relating causes to effects - differences which, perhaps, could not unnaturally be said to spread into what is meant here by talk of a 'causal explanation'. I mark these differences with the terms 'empirical' and 'phenomenological'. In calling them that I'm not trying to categorise them in already understood categories, but just to advertise (prior to explicating) a conceptual difference that needs after all to be marked out by using some or other terms.

Thus Bentall the scientific psychologist wants to develop psychological hypotheses and test them. He wants to show us that paranoid people really do process information in the way that his theory suggests. By contrast it never seems to occur to Bell the psychoanalyst to derive general testable hypotheses about paranoia from his Kleinian theory. He proceeds instead by giving us formulations and examples. That is 'all' we get, and it seems reasonable to assume that it is all he takes himself to be required to provide. Why - the psychologist asks - is this? Is Bell a scientific failure - is he not schooled in actually substantiating his claims with empirical evidence? Not schooled in putting a question to nature so that she can as it were now answer all by herself? ... Well, I think not! Below I explain why.

Here is the general empirical method by which, as far as I understand it, Bentall the psychologist proceeds. Take a state: paranoia. Develop a measure of it. This state is our explanandum: it is what we want to explain. The kind of explanation we seek is: what in the individual's psychology makes her likely to experience paranoia? Next identify some external triggers, internal states and internal traits which may conceivably give rise to the paranoia. The latter two - the inner states and traits - are our psychological explanantia. Develop measures of these inner states (degree of implicit low self-esteem - how the person deep-down feels about herself; the quality of her underlying 'self-representations'; degree of explicit self-esteem - how the paranoid person consciously and explicitly represents herself to herself) and inner traits (habits of information processing such as having a bias toward making external and personal attributions for why the triggers obtained). Finally correlate the measures. If there is a positive relation between the measures of the explanantia (the degree of low self-esteem, the attribution bias) and the measures of the explanandum (the paranoia) then this constitutes evidence for the truth of the psychological model. The character of the theory might be summed up like this: paranoid people are people like this; it is in part because they are like this that they are now paranoid; the data we collect are empirical evidence for the truth of the theory.

By contrast with Bentall, Bell the psychoanalyst proceeds according to what I am calling a 'phenomenological' method. He too has an explanans (A = projection) for the explanandum ( B = paranoia), but he doesn't try to collect evidence of an increased level of projection leading to an increased amount of paranoia. A is not by him conceived of as a psychological trait; it isn't an independent phenomenon which throws up paranoia when triggered. It is rather a psychological process - a defence mechanism. Bell isn't saying that the paranoid person always deals with their distress through projection. He is saying that projection characterises the paranoid reaction to experienced threats to selfhood. (If we wanted we could say that the reference to projection is a way of understanding, rather than explaining, paranoia. What would be important, in saying this, is that we don't take ourselves to have done more than index the phenomenon - we haven't, simply by using this terminology, thereby either explained or understood it better.) What Bell offers us is a way of seeing paranoia: paranoia is, he suggests, the relocating of disturbing feelings from oneself into others-as-one-sees-them. Now, I'm not suggesting that it would be wrong to say that he sees paranoia as caused by projection, but it would be wrong to think of 'caused by' here as meaning 'precipitated by', and wrong to contrast it with 'characterised by'. Yet we might here still describe B as 'a function of' A. We could also, if we wished, describe the differences in terms such as: Bentall is on the whole trying to tell us more about what makes paranoid people vulnerable to paranoia; Bell is trying to deepen our understanding of what it means to be paranoid.

Now, Bentall's method runs into various self-confessed difficulties around testability (p. 339) - perhaps because it is (I suggest) hard to convincingly operationalise, or because it is (he suggests) hard to accurately test for, underlying as opposed to explicitly expressed low self-esteem. But I don't want to go into this here; instead I want to focus on another feature of his theory. This is that whilst his hypothesis-testing is geared up to assess whether paranoia may be an upshot of making external personal attributions when something triggers painful low self-esteem, nothing in his method allows him to test whether paranoia is motivated by the avoidance of painful low self-esteem. (NB I'm not saying that Bentall even thinks he's testing this aspect of his theory.) The method of taking measurements and making correlations does nothing by itself to establish the psychodynamic aspect of either his or Bell's understanding of paranoia, which understanding is of the motivation for the attribution bias / projection. And this is my central point: that the psychoanalytic model helps us understand paranoia - or at least certain forms of it - by seeing how it is motivated.

To see human behaviour (including inner behaviour - i.e. thought) as motivated is to see it as expressing intelligible desire. When we see it as such we do not do so by separately identifying the behaviour and the desire and then correlating or otherwise conjoining the two in thought. Instead the desire has its life within the action; it is not somehow stored up behind it; it is there in the action that we encounter it. The desire characterises the action, we could say, rather than having the action as its upshot. Imagine: you see someone withdraw her burning hand from a hot stove. You don't here separately identify her action and her desire to relieve pain, and then bring them together in your thought.

Naturally we may imagine strange cases (someone wants to burn his hand to win a dare, but he mindlessly withdraws it from the flame to scratch his itchy nose) but these do nothing to remove the default presumption that a hand withdrawn from the flame is, absent requisite strange defeating conditions, a hand withdrawn because of the burning or pain. And note, too, that we say all of this even if it so happens (Rundle) that the pain and the hand withdrawal are both effects of a common physiological cause (the burning), rather than the latter the upshot of the former. Our understanding that we are motivated to avoid pain is, then, not the understanding that avoidance is caused by pain. That we avoid pain and seek pleasure, rather than vice versa, is one could say not a contingent fact about our lives, and masochism must remain a special case on pain of unintelligibility. We are not to answer why we are motivated to avoid pain! Whilst we must be careful to avoid over-theorising the fact (a la 'simulation theory' etc), we understand the withdrawal of a burning hand from a flame in and by relating to the predicament: it makes immediate sense to us as such, and we are not left trying to make sense of one thing in terms of some other thing already understood. (The concept of 'immediacy' here is not temporal but instead has to do with the non-mediated nature of the understanding: we have here to do with something intrinsically intelligible (because it itself defines a form of intelligibility) rather than to something intelligible in terms of something else, or something made intelligible by doing something else. If it be insisted that we do it 'by empathic projection' (putting oneself in the shoes of another) then all that can be said is that this is: 'ok so long as our immediate grasp and our empathic projection are not to be thought of as two separate things, one done by means of the other'. No: 'empathic projection' is at best the form taken of this immediate grasp of motivational meaning.)

To return to paranoia: Bell aims at what I am calling understanding, whereas Bentall aims at what I've indexed with 'explanation'. But given that Bentall too clearly trades on our understanding what it is he is proposing but not demonstrating - that paranoia is motivationally explicable - then we do better to note that both Bell and Bentall aim at understanding, whereas Bentall aims in addition at explanation. If scientificity comes along with explanation then we may say that Bentall's account is the scientific one. But we cannot judge on that basis that Bell's account is un-scientific. All we may infer is that it is, on this rather limited criterion, non-scientific. And it is, in its reliance on our grasp of the intrinsic intelligibility of motivation, no less or more so than Bentall's. Bell isn't interested in noting how often paranoia is motivated, or in independently identifying features of paranoid people which appear to increases their likelihood of becoming paranoid. That just isn't his project. His project is instead to make the motivational forms of paranoia intelligible to us by providing us with a rich exemplary phenomenology - and what better method do we have for that than the case study? Bell's theory is not a scientific failure, but (I would argue) a successful attempt at providing the pre-scientific foundations for any meaningful understanding of paranoia whatsoever.