Error is Infinite and Chaotic

A conversation with a colleague the other day prompts me to explain in more detail why I think myself and most of those like me are wasting our lives. I have already explained that I think analytic philosophy is a degenerative research programme building shoddy structures from inadequate material. But the sunny optimism of Daniel Stoljar's engaging book has given me the tools to explain in somewhat more detail where exactly my worries lie.

Stoljar makes a rough and ready division of questions one may ask into three sorts: topic questions, big questions, and small questions. I will give a slightly wrong description of his division here to suit my purposes. Topic questions are agenda setting big picture concerns that get a person wondering in the first place, the driving curiosity that lie behind whole swathes of inquiry. They are or can be pretty vague and high level -- "What's everything made out of?" "How can I live a good life?" etc. Big questions are what developed (sub)fields cluster around in order to address their topic questions -- "Can string theory unify general relativity and quantum mechanics?" "Should ethical concern focus on the consequences of actions or the nature of the acts themselves?" etc. And small questions are aimed at addressing (elements of) particular answers to the big questions -- "Can one derive such and such claim from the particular formulation of string theory offered by So and So?" "If we characterise consequences in this way can we formulate a version of utilitarianism that does not entail one ought push people off bridges sometimes?" etc.  Stoljar wants to say that the question of progress in philosophy concerns whether or not we succeed in answering big questions with sufficient regularity.

After all, Stoljar notes, it's a bit trivial to say that we address the small questions successfully from time to time. We probably do know (to use his example) there is no intensional fallacy involved in Jackson's knowledge argument even though that was a worry people had upon its initial publication.  So think of the Topic Question as: what is the nature of mind? The big question as: are minds and all they contain entirely physical? (yeah yeah I know that sucks leave me alone) And the small question to be something like: is Jackson's knowledge argument to the effect that conscious experience involves non-physical properties guilty of the intensional fallacy? Then even though the topic question certainly remains open and and the big question arguably so as well, we have actually made progress on a small question here.

And in the other direction Stoljar thinks it is not informative to see that we haven't answered topic questions. There's a sense in which we still wonder "What's all this stuff made out of?" just as Thales once did. But modern chemistry is none the less some sorta advance on "Maybe it's all water". Or, at least, if philosophy hasn't made progress in the same way that modern chemistry is no advance on the pre-Socratics then that maybe isn't such a bad thing. So all the interesting debate concerns whether we make progress on the big questions.

I like this and I think Stoljar is right! Small questions are only derivatively interesting. We care about them only because we hope they shall advance our goal of learning about something worth knowing. Topic questions are cool but too loose to really admit of resolution. But it would still be nice is to learn things like "Are minds physical?" or "Does justice require equality before the law?" or "What elements are necessary for providing scientific explanations?" So I would agree on all this. But I would add just like to put emphasis on an element of this which I find optimistic philosophers often downplay. We want positive answers to these questions, not merely to rule things out.

Very often alleged examples of progress in philosophy (including some of those Stoljar gives, but explicitly discussed as such here) amount to reformulating the Big Questions we have because we realise a certain class of answers can't work. So one of Stoljar's examples of progress is that while Descartes and us may still be trying to work out the nature of mind - we know Descartes' particular version of the problem and the field of answers it suggests can't be where the answer lies, because the conception of matter he was working with is thoroughly refuted. The Big question Descartes was asking - are res extensa and res cogitans the same substance? - has actually been answered, basically in the negative by the denial that one of the terms refers. We now know not to look in that bit of the conceptual universe for our answer. In fact almost all of the examples given (and all the uncontroversial ones) are of this negative sort, not an answer we can affirm but a set of answers we can rule out.

And, well, I just don't think that really matters. If there were a finite set of possible answers to philosophical questions, or even some way of ordering the field allowing for a principled search procedure, then this would be a form of progress. Our inquiry would be an activity we could have some reason to believe will end in a desirable outcome. But it's not. 

Possibilities in their full array are an endless cacophony threatening to drown out all reason. But by the good fortune of our ancestors' suffering nature has rendered us unable to actually attend to the full symphony. So we focus on the isolated strands we are capable of working with, and at least here we can sift through what comes before us and impose some sense. Yet we have no reason to think that even this limited sphere of comprehension is such that we could exhaustively search through it to isolate any underlying melody. And it's clear this could only ever be a tiny snippet of the full soundscape. There are more things in logical space than can be dreamt of by any philosopher.*

So what does it matter if we have better refined the Big Questions we are asking? If we have ruled out whole swathes of answers? We would still have no reason to think that we are right, or close to right, or approaching the truth, or on a path which if followed for ten thousand years by dedicated workers would eventually start to illuminate how things actually are if followed for five thousand more. We are, for all we know, simply cycling through mistakes and no closer to the truth re the Big questions of philosophy than we would be if we had simply never asked them at all.

Of course there is lots of reasonable disagreement re my pessimism. And I think there is some cause for optimism about the teaching we do in philosophy in particular (for instance) that may save us from the charge of absurdity yet. I also don't think people like Stoljar are unaware of the above - he at least has his answers and I hope later to respond to his ideas. And it would be some sort of contradiction in any case for me to think I have established this metaphilosophical thesis with any certitude. Always the bit where I say it, always the bit where I take it back.

But I just want this point to get a bit more emphasis or consideration when philosophers talk about progress in our field, a bit more frank realism about our situation. There is a reason number theorists do not proceed by checking each number individually for whatever property they are interested in. As far as I can tell we are doing the equivalent of checking particular numbers individually for whether they are prime; but in a space far less ordered, with no search procedure and methods of reasoning far less secure, and with no reason to think we know something about our sphere of inquiry that makes all this ok. 

* The wonderful Cosma Shalizi reached out to me with some illustrations of the vast expanse of possibility space as applied to meta-ethics, from a blog post he wrote. I love it, he makes the point far better than I did here!