The latest issue of The Psychologist contains a letter - from Christina Richards - which takes issue with a previous writer therein - Miles Thomas - for his claim that Rudolf Steiner's agricultural belief in 'Preparation 500' - 'that filling a horn with cow's manure and burying it until the spring solstice (sic) will ensure good vitriculture' [wine production] - would be found 'contrary to scientific understanding' by 'many psychologists'. Here is her central argument:
Presumably the people who undertook the burying of manure-filled cow's horns had some, at least anecdotal, evidence of the practice's efficacy. As I am unaware of any double-blind scientific studies on the subject, I suggest that in asserting that the practice is unscientific a value judgement is being made, in that a scientist cannot see how it could work, ergo it does not. This is of course deeply ironic as value judgements by individual scientists are themselves unscientific.
The argument is not a good one, for several reasons which I will present in what are, I hope, increasing levels of philosophical interest. First, perhaps Thomas meant that the practice was unscientific because it wasn't supported by (e.g.) 'double-bling scientific studies'. It would be ironic, then, if the absence of such studies was given as a ground for rejecting his description of the practice as unscientific.
Second, it seems that we are being presented with an unargued-for Empiricist's conception of science - as a domain of truths which are either self-evident or which are evidenced by good (e.g. double-blind-trial-procured) empirical data. But perhaps what it takes to be scientific must be understood in far broader terms - as a matter of participating in a collective sensibility, training, authority, the community one works in, aptness for testing of the ideas, ready commensurability with what is known, and so on. To suggest that the adjective 'scientific' must be restricted to claims which have passed a double-blind test, and 'unscientific' be applied to claims which have failed such a test, would itself - ironically - seem to amount to a curious 'value [idiosyncratic, personal] judgement'. (I suspect that the understanding, articulated by Richards, of what counts as 'scientific' is one which psychologists 'just (believe they) know' - without having done the respectable scientific thing and going and looking at what the sociology and philosophy of science has to say about the matter...)
The issue of real interest however is one which crops up both later in Richards' letter, and also in disputes between Forteans and Sceptics. This is what Richards writes:
The last argument often made in this regard is that 'exceptional claims require exceptional proofs' and that it is not down to the observer to prove that the thing is not true, rather it is down to the person who makes such a claim to prove its veracity; as with the flying spaghetti monster and Bertrand Russell's teapot orbiting the earth. In general I agree [but why?], but I wonder who decides that these claims are exceptional? (And on what scientific basis?)
And this is the beef between Forteans and Sceptics: Both take a sceptical look at superstitions, but Forteans take a similar approach to supposedly scientific pronouncements. Both are held up to the same standards of evidence. It is supposed that, if I am to be counted as epistemically respectable in my holding to some claim, then I must be able to furnish the evidence for it.
This, I want to suggest, is just wrong. Whilst it is very instructive to realise that I have no better reasons for believing that speciation is a function of natural selection than my neighbour does for believing he was abducted by aliens, this does not entail that I am no more rational in my belief than my neighbour. Yet many, especially those of an empiricist temperament, will think it evident that this is not true. For them the entailment is obvious. That they are not right is what I want to argue here.
Anyone who takes reasonableness to amount to the having of reasons must surely be given pause for thought by the possibility of a vicious regress (which amounts to an instance of Wittgenstein's rule-following paradox). Take Richards' implicit claim that any judgement as to the exceptionality (bizzareness) or otherwise of Russell's teapot or flying spaghetti as explanations needs to be made on a 'scientific basis'. By this she presumably means that there must be some inductive or deductive or theory-driven procedure used for determining whether or not some explanation counts as reasonable or not.
Or take a more abstract example: We are trying to decide whether some decision was or was not reasonable. Now sometimes we can, perhaps, invoke various scientific theories, bits of knowledge, axioms, etc. But it could always be asked: How do we know that these are reasonable? And at some point our answer would just give out. We would just have to say: That is what counts as reasonable. That itself is a constitutive example of what it is to be reasonable.
Now I take it that a key insight of Wittgensteinian, Marxist, Heideggarian, and pragmatist philosophy has been that rationality does not reside in a decontextualised sempiternal set of self-evident axioms, but rather resides in distributed historical sets of practices. Similarly, I want to urge, with what counts for 'scientific' and for whether something is 'in accord with scientific understanding'. There are certain beliefs that strike us as 'far out' whilst their 'far-out-ness' cannot be explicated other than through their lack of concord with beliefs which in turn strike us as 'definitely sound' and the soundness of which cannot be explicated by reference to further such beliefs.
After all, we have to stop somewhere. And where we ought to stop, so the Wittgensteinian et al argument goes, is at various distributed, historically and culturally contextualised, different points in our lives. As with Neurath's boat, we may sometimes come to challenge the trustworthiness of some part of this structure of cases what otherwise passes for exemplary science. But when we do this we don't move forward by deploying some set of supposedly atemporal axioms against which the particular cases are held up. Rather we rely on the integrity of the rest of the distributed contextualised hull of the boat whilst changing any particular piece.