real hallucinations

Here are some merits of Matthew Ratcliffe's new book: 

• He listens to what patients say and so develops a discussion which (like Merleau-Ponty's)
  
  turns away from philosophers' idealised hallucinations to look at the symptoms of actual hallucinators.
• He attends to disturbances to the modal structure of intentionality itself. He considers that psychotic experience, in particular, may involve breakdowns in the individuating polarities of different modes of experience (imagination, perception, etc.). So we don't have a business-as-usual set of psychological categories (faculties) in which to place hallucination.
• He draws on the promising phenomenological idea of a form of anticipation which is constitutive in different ways of different kinds of experience. Different types of experience, he says, incorporate their own characteristic patterns of anticipation and fulfilment. 
• He inscribes sociality at the heart of selfhood, rather than preserving some notion of a 'minimal self' which precedes or underpins such self-experience as is constituted by or in relationship.
• He draws on this relational conception of selfhood to make really important links between form of social engagement, phenomenological matters of form, and emotional and psychodynamic matters of content. This, as I see it, really is the oft-unacknowledged holy grail of psychopathology: to grasp the relationship between the emotionally salient meanings of a person's life and the form taken by their psychopathology.

Here are what I take to be the central elements of his theory:

• Hallucinations obtain on continua - for example they are experienced as arising both within (as more thought-like - in which case they are also on a continuum with thought insertion) and without (more sensory).
• Those experienced as arising without are due in part to misperception. Shame - a pervasive social emotion in those who suffer psychosis - may prevent using other's minds to reality test the experiences (p. 98).
• Hallucinatory voices may constitute what R. E. Hoffman calls the 'repopulating' of a 'barren interpersonal world' - one might say that they are not just hallucinations of voices but of otherwise absent relationships (p. 99).  
• Those experienced as arising within are due in part to (p. 82) anxiously anticipating an increasingly determinate thought content.
• Just as (p.85) 'fear of something that is taken to be past could disrupt the sense of that event as firmly anchored in the past', so might 'anxious anticipation induces verbal hallucinations... by shaping the sense of which intentional state one is in. ... Anxious anticipation of p ... contributes to the experience of relating to it in a perceptual or perception-like way. Anxiety is not ordinarily associated with our own thought contents and, when it is associated with them, they are experienced as the contents of a perception-like intentional state that also retains some of the features of thought.'
• Against the objection that we often anxiously anticipate what we are thinking about - without thereby hallucinating it (for example we might anxiously anticipate an intruder hiding in the cupboard) Matthew replies that he is not talking about our anticipation of what we are thinking about (the robber) but about the thinking itself (we are anxiously anticipating our thoughts about the robber).
• Consider abusive voices (p.89): 'in the case of an abusive "voice", there is an unpleasant emotional content p, which provokes anxious anticipation of a more determinate linguistic content q, one that is elicited by p and also consistent with p. Anxiety is intrinsically alienating and so its object, the thought that q, is experienced as alien, as something unpleasant that one faces and is unable to avoid. Whatever forms of anticipation our thinking more usually involves, anxious anticipation of thought content is not one of them. That style of anticipation is more typical of certain affectively charged perceptual experiences. So an unfamiliar, perception-like experience of thought content arises.'
• Such hallucinations are experienced as not being self-caused, even if experienced within, because (92) 'the sense of being immersed in a shared world is already altered and diminished. None of the person's perceptions and thoughts are embedded in a public world in the way they once were, and so the ordinarily taken-for-granted distinction between a consensus reality and his own experience of it is eroded.'
• Verbal hallucinatory 'contents ...are crystallisations of negative, self-directed emotions that reflect ... estrangement from the social world.'
• Why are hallucinated voices are attributed by 'voice hearers' to particular subjects? Matthew suggests (95) that the thematic consistency of what the voice 'says', the 'voice-hearer's' imaginative elaborations, and confusions between imagining and perceiving all feed into what he calls the 'personification' of the hallucinated voice.

Here are some issue-takings:

• Matthew accepts the notion that to perceive, think, remember etc is to be in an intentional state. I take no issue with the philosophical idea of intentionality; it's the relevance of the category of state I object to here. We (and other things) are in states and do not have them; states progress from one form to another; we are not in more than one at a time (if we are in a state of confusion and tiredness, we are in one not two states); states have no composition or location (Roger Squires 1970). Perhaps this seems phenomenologically fussy of me; well, perhaps I am fussy. But I think this kind of talk has risks - that it encourages a tacitly objectified conception of the exercise of perceptual and intellectual powers which illegitimately turns such exercises themselves into possible objects of transitive consciousness - objects of a transitive consciousness which may then go wrong (resulting in psychopathology as conceived by the philosophical psychopathologist). That it does encourage a reified conception of intentionality is, I think born out by the rest of my concerns which now follow.   
• Matthew's approach seems to me to reify or psychologise the notions of 'thought' or 'content'. As I see it, thought qua content, rather than qua thinking, is a purely formal, logical notion. What it is for my hearing and seeing and thinking and imagining a cat to share the same content (a cat) is that I would use the same words to express what they were of ('a cat'). This content is not any kind of thought in the mind and is not something to which we can turn our attention (er unless by 'turn our attention to a content' we simply mean 'answer the question as to what we think or hear or...'). (I will revise this if I've got it wrong!)
• Matthew appears to accept the notion, popular amongst certain phenomenologists who sit close to the philosophy of mind (e.g. Dan Zahavi), that our exercises of our perceptual and agential and rational powers come along with a a form of automatic 'self-consciousness'. Where by 'self-consciousness' here is not meant the ordinary idea of awkwardly feeling under inspection but rather a form of awareness of our own exercise of such powers which awareness is of both the type of the power (seeing, remembering, etc.) and the content of the intentional act (that cat again). He tells us that 'if I look at something and have a visual experience of it, I appreciate that I am perceiving it (and, more specifically, perceiving it visually), rather than imagining or remembering it. ... Put crudely, it is like something to remember, which differs from what it is like to perceive.' I think this approach is wrongheaded (i.e. not right or wrong) and will now try to say why (see also Peter Hacker 2006 and Joseph Schear 2009).
• So, yes, I'm not going to claim instead that there isn't anything it is like to remember or perceive or... My claim instead is that talk of there being something it is like to remember or perceive is misconceived, as is talk of appreciating that we are exercising this or that mental power (hearing, remembering, etc); the idea that seeing and remembering strike us similarly or differently itself strikes me as peculiar. 
• What is it that motivates the idea that there is something it is like to see a black cat? I suspect that one motivation is the notion that knowledge requires reasons. So let's accept (which I don't really) that it definitely makes good sense to say (in whatever context - that of us looking at a cat, say) that you know that you see a cat - both that it's a cat and that you see it. Then we might think to ask 'How do you know that you are experiencing it visually?' And then we might think we need an answer like 'Seeing strikes me a particular way', or 'Part of my visual experience is a pre-reflective auto-affection in which the modality is 'given' to me in self-consciousness. It is because of the deliverances of this inner sense that I know what sensory modality I'm in.' But, ok then, what is it like to see as opposed to smell a cat? Please tell me! It's no use waving around vague words like 'something', 'it', etc. - for, surely, if someone asks what it was like to go on the big bouncy castle then there being something that could form the content of an answer itself supplies the question with a purpose and a sense. 
• So too I may sensibly ask my neighbour how she knows that her cat - this cat here, wrapping itself around our legs, of visually indeterminate sex (the cat not the legs) - is a boy and not a girl. There is something about the cat (the just about locatable genitals) or the shared testimony (the certificate from the cat breeder) which justifies her claim to knowledge. But I may not sensibly ask her, in that same vein, how she knows that the colour sample in front of her which she is looking at, in normal daylight and with no perceptual barriers or other peculiarities in play, is pink rather than blue. An answer would be possible - for example she could tell me that she learned the names of the colours at school. But note the difference between the two situations: in the one we justify empirical knowledge, in the other we acknowledge the ancestry of conceptual know-how. I suggest that we know that we are seeing rather than hearing only in the same kind of sense that we know that the sample in front of us is blue not pink, or e.g. know what our name is. There is nothing about pink that tells us that it is pink, and there is nothing about seeing that tells us that we are seeing. We don't need telling when it comes to know-how! There isn't something about seeing which tells us that we are seeing. It's not as if we're in the predicament of needing to sort out whether our knowledge is visual or auditory. Similarly with remembering: remembering is not a nothing or a something and there isn't something it is like or unlike to do it. What it is to remember - to exercise this capacity - is to retain knowledge. (Perhaps we could say: what it is like to remember is to not be forgetful!)
• Finally, there are clear uses of the 'what's it like to...?' question. We use it to compare one thing to another. And we also use it to describe an experience where by 'experience' we don't now mean a perceptual act but rather a multifaceted event like going on a bouncy castle or taking a long walk in the rain. Talk of 'what's it like' and talk of 'appreciating' has its place here. But Matthew is not envisaging such a use for 'what's it like', and so I want to ask 'what use did you have in mind?' (I think there isn't one - i.e. that phenomenologists are here suffering from what Rupert Read calls a 'delusion of sense'. However it occurs to me I may be wrong - and that perhaps the idea of different senses being constituted by differently articulable anticipations gives us the content we need. I will revise this if it turns out I've been confused.)
• To take it back to hallucination: Matthew theorises hallucination as content apprehended in a modality which modality is appreciated as other than the one it is. I am claiming that content (being merely formal) is not apprehended and that modalities are not appreciated.
• Matthew's theory (along with various approaches taken by other philosophical psychopathologists) seems to me to sublime the logic of the notion of inner speech. (This is somewhat akin to the psychologising of thought qua content.) What I have in mind is that such theories of hallucination tend to imagine that it is both straightforwardly coherent and informative to say of a hallucinator that they are talking to themselves without realising it - that they have inner speech but mistake it for something more like hearing (or something which in some respects is between thinking and hearing). The idea is coherent yet uninformative if taken as simply definitional of hearing voices. But if it is supposed to do some explanatory work then we will need a criterion for talking to oneself other than one's sincere say-so. But none is forthcoming. Well, I don't think one is... Is there?
  
• Finally, on 'personification'. Matthew asks why it is that someone experiences hallucinated voices as coming from particular people. The question is made room for, within the theory, by the fact that it is what I call a 'psychological' theory: namely, it attempts something of a psychological reduction of voice hearing - seeing it as an experience which is made sense of in certain ways. (Unlike certain psychological theorists, Matthew at least allows the sense-making to go on within the experiencing itself. And that surely takes us closer to the phenomenological facts.) I suggest that this gets the phenomenology wrong: hallucinations of voices present themselves ab initio in a range of distinct voices, voices which have tones (high female, low male, etc) utterly different from the speaker's.

Here are some alternative suggestions:

• Sensori-motor anticipation is an important notion to draw on in a theory of hallucination, and, yes, this anticipation is not to be hived off from the hallucinatory experience itself. However what is anticipated is not that a particular thought qua content will have a particular form. Rather what is anticipated is, after all, the experience of something happening (not 'in the mind' but) in the world (that one will hear someone being mean to you by saying 'you stink', for example). 
• Such anticipations are forms of readiness or preparedness essential for negotiating social reality.
• Normally the subject who is in contact with reality 'cancels' or 'relinquishes' such sensori-motor anticipations when they are unfulfilled. This happens automatically. The cancelling of sensori-motor anticipations enables the experience of changelessness.
• The psychotic subject is not well connected with reality. Some of his or her sensori-motor anticipations are therefore not cancelled despite the absence of the anticipated stimuli.
• A sensori-motor anticipation which is both unfulfilled and uncancelled simply is an hallucination. A hallucination stands to a perception like a photographic negative stands to a positive (the photo): it is an anti-experience if you like.
• We can grasp this most easily by thinking on the experience of getting on a broken escalator. We can see that it's broken, and so in one sense do not anticipate that it will move. However in another sense our body still readies itself for getting onto a moving staircase. This is the sense in which we do still anticipate that it will move. But then, of course, the broken escalator does not move. What we then tend to experience, unless we've really gotten used to it, is a lurch. This may seem daft because, after all, you can see that the thing is not moving. However that lurch, which is a feeling as of the escalator moving backwards, is itself the form taken by an unrelinquished anticipation of forward movement in a situation in which no actual movement is detected. It is, if you like, a haptic hallucination; it is a kind of shadow thrown by the sensori-motor anticipation.
• Other aspects of Matthew's theory I should very much want to maintain. I'm thinking here, in particular, of the focus on loneliness, terror, and especially shame and social anxiety - not just as the psychological context which happens to inspire (be the efficient cause of) the hallucination but as the existential matrix which in-forms or de-forms the modal structure of intentionality. Anxiety doesn't relate to what for some silly reason we call reality-testing ('maximal grip' gets it better, but really we're talking about the ability to instantiate a distinction between perceiving and imagining, not some merely epistemic capacity to tell the two apart - i.e. the problem is that there aren't two modes here any more!) by way of one thing (anxiety) making something else (reality testing) hard to do. Anxiety is the shaking apart of reality-contact itself. When we are badly anxious we find it hard both i) to imagine anything much and ii) to achieve a good experiential understanding of our situation. Finally I should like to retain the idea of crystallisation (pp. 88, 94). Here we have the idea that hallucination involves the condensation of quasi-sensory solid out of an initially diffuse existential atmosphere. I would relate this to the human disposition to 'symbolism' (Langer) more generally - i.e. to what psychoanalysts also call 'dreaming' or 'thinking' - the giving distinct shape to as-yet-unthinkable-because-too-diffuse experiences. However if I read him right I think that Matthew might be thinking that the crystallisations into hallucinations involves the quasi-realisation of what is anticipated, whereas I think they are a function of a failure of the relinquishing of anxiolytically crystallised anticipations despite the absence of their manifest non-realisation.