Friday, March 23, 2018

Ideal Generation of Philosophical Theses

Round here I do a fair bit of meta-philosophy. Sometimes I am opine about how we ought  decide upon our research questions, sometimes about how we ought evaluate our answers. Today I am going to pontificate about both at once. My aim in this blog post is just to write down in one place what seems to be the consequences of the picture I have drawn in a scattered fashion. Not only do I not claim any originality for this, my self-impression is that I am just making explicit a fairly widely held and probably communally standard picture in Anglo-American philosophy. I would be very unsurprised if somebody has published this before, please link me in the comments! Finally, not only does not my work not live up to the ideal to be expressed, but this blog post was in very large part inspired by recent self-critique, so I most certainly do not offer my own work as exemplifying this process. What follows is an idealised processes by which one might hit upon a position in philosophy.

--------------------------------------------------------------------------------------------------------------------------

Inquiry begins in medias res, so the budding philosopher encounters a tradition or body of work leading up to now which includes some explicit pointers towards projects still to be undertaken. By familiarising oneself with that tradition, and perhaps bringing to bear an idiosyncratic sensibility or life history or knowledge of some external source, one identifies (i) a problem the tradition explicitly or implicitly points towards as open and significant (ii) and what would constitute an advance upon the presently available options.  That is to say, one formulates a position that, if adequate, really would constitute a response to a problem situation that tradition has identified as important. The work of (i) is coming to understand what problem situations the tradition identifies as important and why, the work of (ii) is coming to check that whatever position one formulates really does actually address that very problem.

None of this is any good reason to believe the position one thereby identifies. This is all the realm of Peircean abduction.  Rather, the role of the tradition is meant to solve the vexed problem of ensuring that one is not just pursuing higher order truths about chmess, that the position one seeks to defend represents something which, if correct, is worth knowing to be as such. The relationship to the tradition (as discussed here) in part achieves this in a quasi-Condorcet esque manner, just by appealing to the common consent and pooled wisdom of those who went before. If one has done one's job correctly in grounding the idea in what went previously, you have some kinda argument that this is the sort of thing a whole lot of folk past would say is worth doing. But, as was pointed out to me by Peli Grietzer in conversation about this, may also be thought to achieve this in a more constitutive fashion: the fact that it represents the culmination of a fine tradition may itself be thought to confer value upon considering an idea.

The position thus formulated, one submits it to test. This permits being broke into three substages. One (iii) identifies domains of application for one's position, (iv) deduces consequences of one's position in those domains, and (v) evaluates those consequences. Typically the processes involved in (i) and (ii) will have done at least some of the work of (iii) for you - previous folk were not discussing the idea in isolation - but it is worth keeping (iii) in mind as a separate stage. If only because one should be on the look out for novel or at least different places of application than that which one started with. It is a poor idea that only does what it was supposed to do.

People are also, to my mind, somewhat slapdash about (iv) - sometimes it is trivial to see what the consequences of one's ideas are, but things can be rather subtle and it is worth dotting one's ayes and crossing one's tees. (Here I have in mind Horsten's introductory book on axiomatic theory of truth, which I recently read and which amounts to an extended argument for the claim that in reasoning about truth the devil is in the details.) As far as possible I think one should approximate to actual deduction, genuinely valid argumentation from premises one can reflectively endorse. The further one departs from this, the less confident that one can be that what is being evaluated is the position one formulated rather than features of your argumentation. Lip service is often paid to this, but I think it is rarely actually done.

More rarely, but sometimes, people are resentful of the idea that they should offer clear and valid arguments linking their position to its consequences, and throw that Aristotle quote at me about only seeking as much precision as due. In such cases I am often reminded of Russell's quip about those who emulate the ancients in all but their virtues. But I will say this in concession to such people, the request that one be as precise as one can be on this front does need to come with a warning. Stage (iv) will often require precisifying one's initial looser or more broad formulation to such a point that it admits of being the basis of an argument that has consequences for some specific domain of application. A perennially tempting error of analytic philosophy is then to refute the precisification or show it has ungainly consequences in the particular domain in question, and consider this by itself a refutation of the original position that motivated the precisification. But it is always possible that there was a slip here, and the real spirit of the position was not adequately captured by the proposed precisification. Hermeneutic charity is very important at this stage, and it really would be inappropriate (a debators trick) to insist on the philosophical significance of a particular precisification just because it is dialectically useful to do as much.

Finally one evaluates the consequences of one's position in the domains of application one has identified. This is... difficult. Yet some such evaluative process must be carried out if one wishes to show that having outlined one's position one can now defend things that are true, useful, good, beautiful, edifying, or whatever positive value judgement one wishes to secure, since none of the preceding has yet done anything towards that. This list is deliberately varied and open ended to indicate that - while stages (iii)-(v) are consciously modelled on hypothetico-deductive models of confirmation in the sciences, I do not intend this analogy to be taken too seriously, and recognise that there are all sorts of values one might hope a philosophical position instantiates. The difficulty chiefly arises because rarely in philosophy is it the case that one's position will have clear consequences in a domain wherein we know what ought be said. Philosophical puzzle cases are sometimes constructed for this purpose, but by this point notoriously fail to induce uniform judgements about what is the normatively preferable response. I do not know of any general advice to give here, beyond that it should be reasonably clear why it is that the answer your position generates within the domain of application is an attractive answer to have generated.

To give a somewhat - only somewhat! - more concrete example, let me take a made up process in social epistemology. There is by this point a mass of work arguing for and against the claim that democratic judgement aggregation procedures are especially good means of discovering the truth about whatever the demos are reasoning about. Reading through this, one may be able to identify as worthy of consideration the claim that under some specially salient set of conditions democratic judgement aggregation will perform especially well by some measure of epistemic success. (Perhaps that had not yet been fully appreciated due to the cultural biases of those participating in the discourse hitherto.)  Voting theory being what it is, it may be possible to mathematically demonstrate the kind of if-then relation. This means wherever those conditions are met one's position commits one to thinking that there will be success in the relevant sense.

One may find that the relevant conditions have recently been met in deliberative panels concerning the likely effect of urban housing policy in Sydney, with some feasible modifications these conditions could be met in medical consensus conferences, and while these conditions are met by some powerful social institution, which claims to be epistemically conscientious, in fact a non-democratic judgement aggregation procedure is used therein. One can then test the truth of one's position by seeing how those panels in Sydney actually went, show that one can offer fruitful normative advice to those organising medical consensus conferences, and issue stern but righteous remonstration to the powerful social institution. From immersion in a tradition of research, bringing to bear one's own idiosyncrasies, one travels gradually from the formulation of a position to various grounds of test or evaluation.

--------------------------------------------------------------------------------------------------------------------------

This, then, is a picture of a full process of philosophical theses generation in its ideal form. Before closing, two notes on some ways I think we typically deviate from this ideal, and three on what I take to be its personal implications. One especially salient deviation is that the kind of person who is typically sufficiently aware of the tradition leading up to now and the kind of people who are especially skilled at the hypothetico-deductive testing stage are not typically the same people. (I know of many exceptions to this - but I do think it is a frequent enough deviation from the ideal to be worth commenting on.) In theory a communal division of labour could make up for this kind of problem, but at least in philosophy - and I suspect many other fields - there is also often mutual hostility between the relevant folk, so they won't even read each other's work to offer guidance or take instruction. At worst, the practical consequence of this can be a community beset by valorised antiquarian irrelevance on the one hand and fad driven displays of virtuoso chmess performance on the other. Such a community achieves less than the sum of its parts. A more pleasant communal atmosphere of trying to recognise and respond to each other's varied strengths seems to me like it would be a real epistemic good.

Second deviation, I think a couple of mistakes are typical at the evaluation stage. First, people think the mere act of being a genuine response to the open problems of the tradition gives one reason to believe the claim. This is, more or less, what I attacked in the post on inference to the best explanation in philosophy. Second, the domains of application philosophers will look to in (iii) will be limited in a fairly arbitrary way by the kind of shibboleths that make for disciplinary boundaries. If there is any productive use for such disciplinary policing at all, and I have my doubts, I think that one's grounding in philosophy comes from the tradition one responds to, not the peculiar domains one applies one ideas in. I would be happy to give a philosophy PhD to somebody who, developing a position by working their way through Saint Augustine's ideas on time in some quirky fashion, formulated a position that they then showed had novel, plausible, and genuinely interesting consequences to open problems in cognitive psychology and fundamental physics. And, having hired such a person, I would encourage them to collaborate with people in a position to see if those consequences bore out in those domains. (I am not saying I think that particular dissertation is a good idea! I am also confident that this is the paragraph people are going to come up with tricksy counter examples to. Bring it, analytic philosophy.)

First personal note, I have not said anything about what particular tradition one must be responding to. This because, and I realise this creates problems with the paragraph above, I would prefer to be laissez faire here. If at all possible I should like it that turning to bodies of high theory from around the world, musical or artistic movements, ongoing political or social struggles, or the development of scientific and mathematical research programmes, can all allow for processes (i) and (ii) to fruitfully generate philosophical positions worthy of consideration in fashion (iii)-(v). If nothing else this, it seems to me, captures the brute fact that what is now recognised to be very good philosophy has in fact been done which was responsive to all such traditions. I am not sure what to say about selecting what tradition to respond to. Relatedly, and also troubling for this picture, are traditions which seem to come with explicit recommendations that one not engage in the processes (iii)-(v). For instance, but not the only instance, think of philosophy in the mould of the later Wittgenstein. I am more confident about the passage between (i) to (v) than I am about what should be done at either end. The as yet untheorised first stage of this process, and the consequences it has for the rest of the process, shall be the subject of my further thought.

(Note that this is also to explicitly acknowledge that deviations from the above picture are not simply errors, but very often represent the conscious adoption of a different meta-philosophy. I would certainly not want to give the impression that I think everyone is just trying and failing to live up to this ideal! There's a broad variety of actually practiced meta-philosophies just as there are a broad arrays of actually endorsed philosophies. The use of the indefinite article - "an idealised process" rather than "the idealised process" - was conscious and deliberate.)

Second personal note, the picture above does not really allow for much direct evaluation of, or comparison between, philosophical positions. One knows one's philosophical positions entirely by their fruits. Perhaps in domains of application one may be able to say that one prefers the results of one position to another (and in some cases one may even be able to construct dominance arguments, though I think this will in fact be very very rare) but it is not clear how this translates into an overall evaluation. For now I will simply say that I think this reflects the reality of philosophy as it is presently professionally carried out.  The only direct evaluation that really seems to me possible is aesthetic. The way folk decide how to favour broad philosophical positions, where any such decision occur at all, usually proceeds in a quite arbitrary fashion and probably depends a lot on the quirks of those involved, what they had for breakfast that morning, etc. Maybe this could be improved upon though by a better meta-philosophy or theory of our own method.

Third,  as it stands, I am not especially confident that this is a good idea. Perhaps I shall be dissuaded. Suppose, though, that I continue to believe that the above represents one ideal worth striving for in the generation of philosophical theses. I am genuinely torn as to whether I should therefore teach it to graduate students as something to be explicitly attempted. I am, in general, nervous about taking on the role of graduate student advisor (note to future students who may be reading this: I have, of course, totally overcome this by the time you are reading, and should be looked to as a rock of faith who is certainly not an insecure mess whose only distinguishing feature is that his name is on the office door). I worry that even if I am right and this is a good model, to promote its explicit use would be like the mistake of the New Maths proponents who mistook logical for pedagogical order.

No comments:

Post a Comment