Wednesday 25 January 2012

UCLan PhD Studentships, School of Health

Reference Number: RS/11/17-19

The School of Health wishes to appoint 3 full time PhD studentships.  Each studentship is tenable for up to 3 years for a PhD (via MPhil route) [subject to satisfactory progress].  The studentship will cover the cost of tuition fees for UK/EU residents plus a stipend (currently £13,590 per annum).  The successful applicant will start on 1 April 2012. International applicants may apply but will be expected to pay the difference between the UK/EU and International Fee Rate.
Applicants need to undertake research projects within an area of methodological and subject expertise within the School of Health details of which can be found on the website
The School of Health has a broad range of expertise and was ranked in the top 10 Nursing and Midwifery Schools in the UK in the Research Assessment Exercise in 2008. Our main subject areas are:-
Potential applicants are strongly encouraged to discuss their research proposal with an appropriate lead prior to application.
Applicants should have a first degree (Honours) in a relevant subject at 2.1 or above for standard entry onto MPhil/PhD or Master's level qualification in a relevant subject (Essential for ethics and philosophy applicants)
Requests for an application pack (quoting the reference number RS1117-19) should be directed to the Graduate Research School Office. Tel: 01772 895082 or
Closing Date: 10 February 2012

Monday 23 January 2012

CHSTM Mental Health Group Programme Jan - May 2012

Centre for the History of Science, Technology & Medicine & Wellcome Unit for the History of Medicine

Mental Health Group Programme Jan - May 2012

Thursdays,, Room 2.57,
2nd Floor, Simon Building, Brunswick St.

26th Jan Matthew Smith, University of Strathclyde
The uses and abuses of the history of hyperactivity

1st March Len Smith, University of Birmingham
Work as treatment in the home and colonial lunatic asylum; England and the West Indies, 1815 – 1890

29th March Sarah Collins, University of Manchester
Second stories: dementia, narrative and memory in conversation

3rd May Jen Wallis, Queen Mary, University of London
The male brain: pathology and gender in the nineteenth- century asylum'

31st May Ian Cummings, University of Salford & Martin King, MMU
Representations of post-traumatic stress in modern detective fiction

The mental health group is an informal, interdisciplinary forum for academics, practitioners and people with an interest in mental health to share and discuss their mutual interests. Everyone very welcome. For more information see

Friday 20 January 2012

Metaphor and anomalous self-experience

I have been having a look at Josef Parnas et al’s EASE: Examination of anomalous self-experience with the hope of adding discussion of its attempt to codify, or at least increase the degree of codification of, a complicated diagnosis based on anomalous self-experience to my chapter in the OUP Handbook of Philosophy and Psychiatry. But before turning to that I’m intrigued by an initial comment in the context of the difficulty of patients putting such abnormal experiences into words. The difficulty is this:

The experiences may be fleeting, perhaps even verging on something ineffable. They are not like material objects that one can ‘take out of one’s head’ and describe them as if they were things with certain properties, or redescribe the experience at different occasions in exactly the same terms. The patient may be short of words to express his own experiencing. [Parnas et al 2005: 237]

Given that, then patients employ metaphors. And that prompts the authors to say the following:

Use of Metaphor
The patients employ metaphors to describe what they experience; this is also the case with healthy people – it is a universal process. A metaphor is usually defined as a transfer of meaning from one conceptual domain to another, like in the expression: ‘life is a journey’ (the concept of life is made meaningful by an appeal to a journey, belonging to another domain). In the context of a psychiatric interview, a metaphor should not be seen as ‘just a metaphor’ or ‘just a manner of speaking’ that somehow, distortingly or conventionally, stands for an underlying (more) true or authentic anomalous experience, i.e., a metaphor is not only a signifier (sign), distinct from, and contingently attached to the signified content (‘signifié’ = the sign’s meaning). Rather the following is the case: an experience (non- or prelinguistic), especially of the prereflective type, becomes progressively conceptualized, i.e. transformed into a conceptual (linguistic) format, in order to be grasped by the reflecting subject, thematized and rendered communicable to others. The metaphor should be seen here as a basic functional aspect of this symbolization process, where it operates as a linguistic vehicle or medium through which the experience first articulates itself and so becomes reflectively accessible. The metaphor is therefore the first stage of making a prelinguistic or prereflective experience explicitly accessible to oneself and to the other. The choice of metaphor is linked to the nature of experience in a noncontingent way, i.e., experience and metaphor are not entirely independent. [ibid: 237-8]

The picture is this. In the context of EASE, the use of a metaphor to express an abnormal experience is not mere metaphor where this normally involves either a distorting or a conventional representation which stands in for a true or authentic experience. So the first point of note is the idea that experiences of this (normal) sort could be true. I suspect that is not what is meant at all. A normal perceptual experience could be true if were representational and represented things correctly. (Lots of philosophers, including, eg., McDowell have thought this. Others deny that experiences themselves ever have any content.) But in this passage, there is no suggestion that this is the relevant dimension of truth (between the experience and the world). I think they mean just between the metaphor and the experience. It is as though if a metaphor is distorting or false, there must be a standard of truth and the thought here is that it is the thing for which the metaphor is a metaphor which is itself true. (As though: if a sentence falsely describes something, there is a fact which is true. Or if the sentence is true then there are two true things: derivatively the sentence and originally the fact. But facts are not true; they just are.)

Rejecting that thought – but not rejecting it as senseless – Parnas et al suggest that in the case of EASE, the metaphor is a basic feature of the symbolization process. It serves as vehicle through which the experience articulates itself (another fishy phrase) and thus experience and metaphor are not independent.

That seems very odd to me. Surely if one selects a metaphor to express one’s normal experience, the metaphor and the experience will not be independent. The one will be selected to fit, in whatever way metaphors fit, the experience. If, for example, one thinks that metaphors have content then the content of the metaphor will (be selected to) fit the experience. So the idea of saying that in EASE the two relata are not independent, and that choice of metaphor is instead linked to the nature of experience in a noncontingent way, seems not to distinguish it from the everyday case properly or normally understood.

So my hunch is that this is an attempt to dig deeper and say something like this: in the case of normal experiences, one might put them into words in a non-metaphorical way as well as in metaphorical ways. But in the case of EASE, all there is, is a metaphorical expression. If so, though, I have two worries.

First, ‘metaphor’ is the wrong word because, I suspect, there will be no possibility of unpacking how the metaphor applies to the abnormal experience. There will be no unpacking because there will be no way to weigh up the content of the metaphor and the content of the experience as potentially distinct matters. In fact, that is part of what Parnas et al are trying to say when they contrast the EASE case with normal cases where there is, they claim, by contrast no connection of content. That very closeness (in the EASE case) suggests that this is not a matter for metaphor, however.

Second, the idea of anomalous experience becoming conceptualised having initially been unconceptualised seems odd. Again I say this in part because of something they say: when they say that abnormal experiences are not like material objects that one can ‘take out of one’s head’ and describe them as if they were things with certain properties. For that reason, such experiences seem the wrong sorts of things to be first independent of, but then clothed in, conceptual form.

PS: Look here for a video made by Josef Parnas.

Parnas J, Moeller P, Kircher T, Thalbitzer J, Jansson L, Handest P, Zahavi D. (2005) ‘Examination of Anomalous Self-experience’ Psychopathology, 38: 236-258

Philosophy for the theory of translation and interpretation

I’ve been invited to give a guest lecture on a post-graduate module on the theory of translation and interpretation here at UCLan. I learn that ‘part of the programme content involves studying post-structuralist approaches to fluid textual meaning and to the (im)possibility of equivalence and semantic fidelity within translating, especially of literary texts with fluid meaning, which are open to interpretation/deconstruction by individual readers and translators’. So that prompts the question of which bits of the philosophical larder to raid, of what is sufficiently-free standing to be teachable to non-philosophers and then, finally, what one should think about the matter.

Obvious first thoughts are:

1) Quine’s attack on analyticity based on the impossibility of a sufficiently independent characterisation of such a notion. The argument in a nutshell is that there is no distinction between the analytic and synthetic if we cannot explain what that distinction amounts to. But all attempts at explanation presuppose equally mysterious concepts.

Against this, Grice and Strawson memorably argue:
If we take these two conditions together, and generalize the result, it would seem that Quine requires of a satisfactory explanation of an expression that it should take the form of a pretty strict definition but should not make use of any member of a group of inter-definable terms to which the expression belongs. We may well begin to feel that a satisfactory explanation is hard to come by. [Grice and Strawson 1956: 148]

2) Quine’s radical translation and Davidson’s radical interpretation. Under slightly different ground- rules, both offer accounts of what underpins facts about meaning by examining its ascription by field linguists starting from scratch to interpret native utterances. They go on to suggest that our own access to language is constrained by the same epistemic predicament.

It turns out, they argue, that the underpinning account does not determine meaning sufficiently to rule out a lingering indeterminacy. Concretely, according to Quine, the native utterance of ‘gavagai’ could pick out rabbits, rabbit stages, or even rabbit flies etc. More abstractly, according to Davidson, there is holism that connects word meaning and belief content such that there can be no atomic equivalencies.

3) Kripke’s Wittgenstein-inspired (if hardly accurate) argument that there are no facts about meaning. In thinking about how I can justify that my use of a word today is in accord with the standards of correctness that governed my past practice, Kripke unleashes a form of scepticism which concludes from limits on what we can know to the claim that there are no facts about meaning. How can Kripke draw such a conclusion about the link between rules and applications from epistemological considerations?

The answer is that an important assumption is built into the sceptical approach. If there were some fact that constituted the relation between a rule and its applications, it would be independently identifiable by the idealised subject that Kripke postulates. Kripke supposes, for the purpose of argument, that one may have all possible information about one’s past experiences, mental states and inclinations. He then asks whether any of these would be sufficient to determine the rule that one were following. His conclusion, based on his interpretation of Wittgenstein’s arguments, is that none would be. Given the idealisations involved, and the assumption that had any fact constituted the rule one were following one would have known it, then there is no such fact of the matter.

The sceptical argument, then, remains unanswered. There can be no such thing as meaning anything by any word. Each new application we make is a leap in the dark; any present intention could be interpreted so as to accord with anything we may choose to do. So there can be neither accord, nor conflict. [Kripke 1982: 55]

What should we conclude from these arguments?

All three lead to revisionary consequences. They undermine our pre-philosophical commitment to facts about meaning, respectively: there are no facts about meaning equivalence; there are no determinate facts about meaning; there are no facts about meaning.

All three start by examining the nature of meaning and attempting to shed light on it (one might say, what we mean by ‘meaning’ but this is not an interesting reflexivity yet) from outside. Quine assumes a scientistic naturalism about the kind of facts there are. Davidson at least wishes to shed light on meaning without circularity (though he says:
Quine describes the events or situations in terms of patterns of stimulation, while I prefer a description in terms more like those of the sentence being studied; Quine would give more weight to a grading of sentences in terms of observationality than I would; and where he likes assent and dissent because they suggest a behaviouristic test, I despair of behaviourism and accept frankly intensional attitudes toward sentences, such as holding true. [Davidson 1984: 230]).
Kripke appears to place no particular limits on the reduction base (though it turns out he does).

All three, therefore, can be undermined by those with sufficient confidence to reject explicit and implicit attempts at reductionist naturalism. The arguments share the same defect: they infer, from the failure of a particular kind of reduction of facts about meaning, the conclusion that there are no, or at least limits on, those facts about meaning. But the cost of this diagnosis is that one has to grant that facts about meaning are sui-generis. And that in turn suggests that the ontology and epistemology of meaning is mysterious (or rather, will seem mysterious to anyone who feels any temptation at all towards physicalist reduction): the world contains facts about meaning which are not reducible to more basic facts and those facts are apparent to those, at least, with eyes and ears to see and hear them, to those initiated into this tract of the space of reasons.

Davidson, D. (1984) Inquiries into Truth and Interpretation, Oxford: Oxford University Press
Grice, H.P. and Strawson, P.F. (1956) ‘In Defence of a Dogma’ Philosophical Review 66: 141-159
Kripke, S. (1982) Wittgenstein on rules and private language, Oxford: Blackwell

Thursday 12 January 2012

Language and cold symptoms

I’ve spent the last three days not just working from home but from bed, laptop nestling in a growing mountain of tissues, mug of lemsip in hand, and angry and apparently underfed cats shouting from the doorway. A cold, in other words. I wonder whether there’s any truth in the idea that if one can be ill for a few days, after a period when it would have been very difficult to be, then one is more likely to be so. The only upside of still feeling lousy on, unusually, a third day is that – irrationally perhaps – this suggests to me a thorough biological underpinning (even if psychological factors helped let it in).

Again however, it’s the relationship of such an ordinary experience to something quite bizarre which strikes me and yet, even so close up, it seems to resist capture. What I mean is the way the dysfunctioning of a cold upsets the relationship of language and experience.

So I am not particularly surprised by the ‘content’ of what is always for me the primary symptom: a feeling that I need to be in horizontal, to be in bed. What might have been surprising is that that seems to be a primitive content. It’s not, eg., that a feeling of dizziness (no doubt the result of blocked sinuses) leads to an inference that it would be wise to lie down. Rather, there’s a kind of direct normative demand in the feeling itself. But that seems to be language clothing experience in the way we expect.

What is odd is the way that lying down and trying to sleep I have experiences which seem to last, stably, for some considerable time and yet be a bit bizarre. Last night, for example, I became convinced that getting to sleep was a kind of practical project, an action which involved subsidiary elements. Further, I interpreted these through the metaphor of the need to shut down complex machinery. Whilst at the same time realising that this was nonsense, the thought that levers had to be moved and valves shut floated around my mind, stopping, so it also seemed, me actually falling asleep.

But I didn’t really think that to sleep required turning something off. It wasn’t that false belief. Nor was it the mad parade of imagery of a dream. Nor a fleeting fancy. For what seemed an hour I had an experience I want (and wanted, I think) to put in those words whilst also thinking them unsatisfactory. That was just one bit of the typical experience of a cold. (And why one can spend three days in bed and still feel unrested.) Given my wakeful thoughts about the relationship of experience and language, I’m not sure what to make of this.

Tuesday 3 January 2012

Charles Travis ‘A sense of occasion’

I am going to summarise just what I take to be the main argument of Travis’ 2005 paper ‘A sense of occasion’. Travis starts with two claims from John Cook Wilson, one of which, the ‘core conception’, he supports and one, the ‘accretion’, he rejects.

The core conception: knowledge is unmistakeable
The positive claim is also familiar from McDowell’s ‘Knowledge and the internal’. It runs:

[I]f N knows that p, then what he sees as to whether p leaves no possibility (for him) that p is not so. To see enough of how things are to qualify as knowing that p, one must see no less than p itself. He says, for example (p.100),
In knowing, we can have nothing to do with the so-called ‘greater strength’ of the evidence on which the opinion is grounded; simply because we know that this ‘greater strength’ of evidence of A’s being B is compatible with A’s not being B after all.
The point for the moment is this: if N knows that p, then he could not have the grounds he does for taking it that p while p was not so. There is simply no such possibility, not even a very remote or outlandish one. We can see this as rejecting a Lockean suggestion: that knowledge may be merely ‘certainty as great as our frame can attain to, and as our condition needs’. [Travis 2005: 288]

Travis illustrates this first with a mathematical case which both admits of a proof and for which there is no possibility of a sceptical ringer. But the more interesting application of the claim is to perception.

Now consider a perceptual case. N faces a pig, and, eyes open, sees it. Suppose N grasps that he is doing that – sees what he is seeing (in this respect) for what it is. N thus sees there to be a pig before him. To grasp, as he does, what he thus sees is to grasp himself to see what excludes all possibility of there being no pig (for all he can see) – to grasp himself as seeing what excludes a ringer case. [ibid: 288]

Now in one sense this is not very radical. Seeing is factive so if N does see a pig that does exclude all possibility of there being no pig there. But the point seems to be that seeing the pig involves taking oneself to be so seeing and thus is unmistakeable. That seems more contentious. Seeing requires taking onself to be seeing. (Throughout the paper, Travis rejects what he calls externalism which does not seem simply to mean something like reliabilism.)

Knowledge that one faces a pig, if available, is as unmistakable, secure, as arithmetical knowledge, if that is attainable. There is a describable position which, if attainable, would be knowing one faces a pig. Nothing less than this would do. There is (pace Locke) no second-class variety of knowledge. [ibid: 289]

But Cook Wilson adds to this a further thought.

The accretion: no ringers
[T]here is no such thing as a ringer for knowing that p – a condition which, if N were in it, would be indistinguishable to him from the condition in which he is in knowing that p. N would not know that p if not-p. So this means that N knows that p – is in the right condition – only if there is no such thing as a ringer for his situation with respect to p – a situation in which not-p, but which, were he in it, he could not distinguish from his actual one in knowing that p. Conversely, in the knowing state, N can distinguish his condition from all conceivable states in which not-p. I will call this Cook Wilson’s distinguishability principle, or, for short, the accretion. [ibid: 290]

The problem with this is that it seems to rule out knowledge of the external world (though not knowledge of maths) because it is always possible to construct a ringer. Cook Wilson puts the accretion forward because of a version of the argument from illusion which starts and finishes thus:

Premise: there is a ringer for N’s situation with respect to p (o)
Penultimate conclusion: N might be (for all he can tell/see) in a ringer situation with respect to p (o)
Conclusion: N does not V that p (or does not V o) [where the values of V are such things as know (that p), see (that p), see o (e.g., a pig)]. [ibid: 291-2]

According to Travis, Cook Wilson thinks that for knowledge of the external world the premiss is clearly true and leads to the penultimate conclusion and hence the conclusion. The challenge Travis takes on is conceding the premiss whilst denying the penultimate conclusion and hence the conclusion.

Rejecting the accretion
The suggestion for how to do this is first made in this passage. The first stage is to link the objectivity of judgement to the standing possibility of error:

I take something to be so only where there is that which I thus take to be so – something so, or not, independent of my so thinking. Where what I take to be so need not have been so (if not elsewhere), there is at least that much room for making sense of the idea of my being mistaken. Environmental judgements, in the nature of the case, always make for that much room... [ibid: 296]

But despite that, although the standing possibility implies the logical possibility of a ringer, it does not follow that I may actually be subject to such a possibility:

It is always possible, in this sense, that I may be wrong: where I take p to be so, that fact, so far, always leaves it open that I might be wrong. But, Austin reminds us, for it to be possible that I may (might) be wrong is not yet for it to be so that I may be. That depends, he tells us, on circumstances (in ways not yet spelt out). If it does, that makes the situation this: there are, or may be, circumstances in which, though there is, recognizably, such a thing as a ringer for my situation with respect to p, it is not so that I might be in a ringer situation (for all I can see or tell). If that is right, then the argument from illusion is invalid. The accretion accordingly drops out. [ibid: 296]

In other words, Travis concedes the first premiss of the earlier argument but not its penultimate conclusion. Why not?

The answer, predictably, is occasionalism. There may be a multitude of standards by which to judge whether a feature of the world is the way someone has said that it is. The concepts expressed in the words used do not determine a single view. What does narrow down the possibilities (when all goes well and thus there is a determinate answer) is context: the occasion of the utterance. This applies to familiar cases such as what counts as toast being blackened or Lake Leman being blue (blackened by burning or Marmite; blue from the sun and sky or dye?) but also to knowledge ascriptions themselves.

The main suggestion for how this is the case comes from a discussion of something else Travis talks about elsewhere (and in truth I’d not before seen its point in his picture): factive meaning. Because Travis supports Cook Wilson’s core conception of knowledge, knowledge cannot be generated merely from strong evidence. It needs something stronger such as factive meaning. He takes a typically Travisian example: the signs of Sid’s infidelity.

[C]onsider the tell-tale signs – that strange perfume, the receipt for lunch at a secluded spot, Sid’s suddenly depressed libido. There are at least two statuses such things might have. That scent may be evidence that Sid is seeing someone. But it may also mean that he is. Here ‘mean’ is factive: if a means that b, then given a, b. If, despite a, b does not obtain, or happen, then a did not mean b after all. Suppose Zoë sees that a, and recognizes it as meaning b. (Thus, ‘recognize’ being factive, a does mean b.) Then Zoë has, not mere evidence that b, but proof in the strictest sense: it is not just so that, for all the grounds that Zoë has, perhaps not-b. Zoë sees, unmistakably, that b. She thus knows, on Cook Wilson’s conception of knowledge. Evidence is beside the point. Strange scents can be acquired in many ways – crowded elevators, over-zealous department store personnel. But suppose Sid manifestly had no such opportunities. Then in his case that strange scent may mean that he is seeing someone. If Zoë is au fait enough with his wonts, she may recognize the scent to mean that. [ibid: 301]

So scent can be either mere evidence or factively mean Sid’s guilt. If the latter, Zoë can acquire knowledge even according to the core conception. But still why does this connect to occasionalism? Because factive meaning depends on what might have been.

Does the scent mean that Sid is seeing someone? Or is it merely evidence? What does that depend on? Suppose that, though Sid might have got the scent through seeing someone, he also might have got it in other ways – elevators, for example. Then the scent is at best evidence. (If seeing someone is a likely way for Sid to have got the scent, then it is good evidence. If it is an unlikely enough way, then perhaps the scent is no evidence at all.) But if Sid would not have come by the scent in any other way – if no such scenario is actually a way things might be – then the scent means that he is seeing someone. It is then up to Zoë to appreciate what is there to be appreciated... What might be thus depends, for one thing, on how things are. To that extent, it is a matter for discovery. But it also depends on what is (needs) to be treated as fixed in how things are, and what is allowed to vary. That is a matter liable to vary from occasion to occasion for saying what might be. [ibid: 301-2]

In a different example Travis considers whether bleating means sheep. Well it does if there are no goats but not if there are. It can in Umbria, eg., if there are no goats there. But what if there might have been? What if it is just luck that there are none? Travis suggests that fixing this is occasion-sensitive.

Cook Wilson’s core conception could be right only if his accretion is not. Austin’s view (once established) earns us our right to jettison the accretion. It earns us a right to reject the argument from illusion. For, on some occasions for stating the argument (of particular cases), its premise may be true while its penultimate conclusion is false. Sid faces a pig – in plain daylight, and he knows as well as most of us what a pig looks like. There is plainly such a thing as a ringer for his situation. There is such a thing as facing a robotic pig with artificial flesh. Might Sid, for all he can see or tell, be in a ringer situation? Might he be facing a robotic pig? Perhaps yes, perhaps no, depending on the occasion for discussing Sid. Where not, the argument from illusion, as then stated, is invalid. It is not generally reliable. [ibid: 304]

So the shape of the position seems to be this. Travis holds something like an aspect of McDowell’s picture in Knowledge and the internal’: that knowledge does not stop short of the fact known. That is the core conception. Also like McDowell’s paper, he rejects the idea that there can be ringer-free exercises of knowledge of the external world (McDowell calls that a fantasy). But, unlike McDowell, he argues that one needs occasionalism to block the inference from the mere possibility of a ringer to the undermining of knowledge via the idea that, as far as the subject is concerned, the obtaining of the veridical, non-ringer case, is external to him/her: that as far as he/she is concerned it might equally be the case.

There is a further clue as to why this is so. Travis argues that it is a mistake to say, as McDowell does say, that a subject needs luck, a favour from the world.

The favour the world must do me if, as I find myself at time t, I am to know that p, is that ‘it actually is what it appears to be’ – that p actually is so, rather than (for me, then, undetectably) merely appearing to be so. But, the idea is, once that favour is done, that it has been may be part of my standing in the space of reasons: part, that is, of what it is for me to know that p. In which case, my enjoying the standing means that I need no favours. [ibid: 313]

Now if I understand him, McDowell’s point is that one cannot have justification on one side of a divide (between internalism versus externalism) and truth on the other. It cannot be that justification requires a further favour to yield truth and thus knowledge. If so, even when combined fortuitously with truth, such ‘justfication’ would not amount to knowledge. So instead, McDowell suggests, given that we must also reject the fantasy that ‘reason must be credited with a province within which it has absolute control over the acceptability of positions achievable by its exercise, without laying itself open to risk from an unkind world’ then getting oneself into a state of justification already assumes some luck but no more is needed to get from that to knowledge.

But Travis objects to this view saying:

[T]here is no room, I suggest, for the sort of favour McDowell here envisages. I face a pig, just on the other side of the railing from me. What I see (whether I realize this or not) is a pig before me. If I take in (register, am aware of) my doing that, then I have that – that I see a pig before me – as my reason for taking there to be a pig before me; in which case I have proof, in the strictest sense of proof, that there is one before me. To take this in – to register it, and not merely, say, to surmise it – I must have a suitable ability. I must be able to tell when I see a pig before me. If I have such an ability, then one favour I do not need from the world is that things ‘be what they appear’. My ability is precisely one to tell how (in the relevant respect) things are. If I lack such an ability, then I cannot be said to know there is a pig before me – in the crucial instance, to register, rather than surmise. In that case, favours from the world cannot help me. So there is no room for McDowell’s envisaged favours. [ibid: 313]

Now part of McDowell’s account is that – when all goes well – perceptual experiences take in, or stop nowhere short of, the facts. Relative to such experiences, no more luck is needed in gaining knowledge of the world. Still, it is not wholly within the subject’s control that what they have is this sort of veridical experience. I think that McDowell thinks that he has done enough to make knowledge possible by ruling out a picture on which it was clearly impossible. (There is an appeal elsewhere to the idea that we have a concept of action – and responsibility for action – even though what we are able to do depends on external contingencies not wholly within our power.) But Travis suggests that this is not good enough. If I realise that I see a pig, I must have an ability to tell when I see a pig. But if I do, then I have proof ‘in the strictest sense of proof’ that there is a pig. Given that it is proof then no further favour is needed. That it is a pig is as within my control as a mathematical case.

Of course, there remains a disanalogy between the pig and the maths case: the former permits a ringer, the latter not. But the appeal to occasionalism is supposed to block the more general possibility of a ringer and its potential application in this particular case. In this case, there cannot be a ringer.

Well that is how it seems to go. Readers familiar with the paper will realise that, aside from a number of subtleties, I have also missed out an account of what is central: how occasionalism applies to the ascription of knowledge. I have simply hinted that this follows from the fact that what might have been is assessed in context.

The reason for the omission is that whilst I know how it applies to saying that Lake Leman is blue (where what one says or means can differ between thinking it reflects the blue sky or has been dyed) I am much less sure what the different meanings are of saying that N knows that there is a pig before him or knows that there is a sheep nearby. Even if knowledge is an ascriptive status (by which I mean that we should think when it is right to ascribe knowledge to others) and even if the grounds for the ascription might vary depending on what one thinks might have been, I do not see how this makes occasionalism apply to knowledge such that one can say different things of N when one says he knows that there is a pig before him. To pinpoint my blind spot, let me quote at some length. This is where I get lost:

Max speaks truth in saying there might be goats. The truth he speaks is that there might be, on a certain understanding of something’s being what might be: what one ought to understand by this on this occasion. Again, if there is occasion-sensitivity, then there are not, in addition to such facts as to what might be when one understands might be in this or that way, further facts as to what might be anyway, occasion-insensitively. It is facts of the first kind, and not such supposed further facts, that bear on the truth of knowledge ascriptions, different ones on different ascriptions. Where Sid does not know, he is not to be treated as authoritative; where he does, he is. That rule applies equally in Pia’s situation and in Max’s. There is no difficulty in the idea that some people, engaging with the world in given ways, ought to treat Sid as an authority while others, engaged in other ways, ought not – even if the latter cannot recognize what the former ought to do.
Max ought not to treat Sid as an authority. For he ought to treat goats behind the barn as a way things might be. If he does so treat it, he will see what Pia said as indifferent to a possibility. But what Pia’s statement is indifferent to is what might be on a certain understanding of what might be. It need not thereby be indifferent to any way things might be, on that understanding of might be which its occasion calls for. (Nor is that more than Max, on his occasion, can recognize consistently.) So it need not be understood as crediting Sid with any status he might enjoy despite the existence of possibilities that he is wrong. It may be crediting him with a status he can only enjoy in having proof he grasps as proof. What may vary from one occasion to another (from Max’s, say, to Pia’s) is what would count as enjoying that.
What, if right, would demonstrate occasion-sensitivity is this. For us, both Pia’s occasion and Max’s may be fully in view. We can see all that would make things count one way on the one occasion, another on the other – if the relevant notions are occasion-sensitive. If there is not occasionsensitivity, then at most one of these occasions exhibits the facts as they really are. For there are then only occasion-insensitive facts as to what (really) might be, no matter what else passes for that on one occasion or the other. So either it really might be that there are goats behind the barn, or, really, that is not a way things might be, punkt. So which is it? What Austin and I think is that this question has no motivated answer. Nothing in the way things are gives the one answer any better credentials than the other as an answer to the question what (really) might be. If we are right, and if the point holds, not just for goats behind the barn, but reasonably systematically, then there can be no facts about what might be (or surely not enough) if those facts are not occasion-sensitive ones. That is always the mainspring of occasion-sensitivity. I think it is easy to confirm in the case at hand. [ibid: 308-9]

See this entry on ‘A sense of occasion’, this on ‘Reason’s reach’, this on ‘The twilight of empiricism’, and this on the discussion of rule following in Thought’s Footing.