:: Article

Without mirrors

Huw Price interviewed by Richard Marshall.

Huw Price is an ice cool pragmatist philosopher with global expressivist deflationary thoughts that he writes about in his many books. He thinks about time and causation and truth but isn’t a metaphysician. He doesn’t think there’s a time’s arrow. He thinks Stephen Hawking gets things wrong. He thinks Bertrand Russell is an armchair anarchist. He is indubitably a groovy jive.

3:AM: What made you become a philosopher? Was it always something you felt drawn to or was it a surprise?

Huw Price: A surprise. In my first year as an undergraduate at ANU I added philosophy after signing up for the maths and physics courses I thought I needed. Later, when I wanted to drop physics and concentrate on pure maths, I had to make up another major, so I did a couple more philosophy courses. That’s when I started to get drawn into it (though it was a couple more years before I made the switch – my first year at graduate school was in maths).

3:AM: You argue in ‘Facts and the Function of Truth’ that usual ways of making a distinction between factual truths and non-factual truths fail. Before looking at your own position, can you outline the main difficulties with the alternatives?

HP: At this distance it’s hard not to be anachronistic, but what I was criticising was the common view that there is a ‘bifurcation’ in language between those declarative utterances that are genuinely ‘descriptive’, or ‘fact-stating’, and those that have some other function. Traditional non-cognitivists employed this distinction, arguing say that moral claims lie on the latter side of the line, not the former (and hence that there is no metaphysical issue about the nature of moral facts). I was (and am) sympathetic to that anti-metaphysical move, but I think the bifurcation thesis turns out to be unnecessary to it, confusing, and ungrounded. In FFT I argued that the various ways in which people try to draw this distinction tend just to take in each other’s washing, and that attempts to find a firm place to stand don’t work.

For example, it was a common idea that the notion of direction of fit would do the job. Beliefs were said to have ‘mind-to-world’ direction of fit, meaning that we try to conform our beliefs to the way the world is; whereas desires were said to have ‘world-to-mind’ direction of fit, meaning that we try to conform the world to our desires. But apart from the fact that this distinction is no use at all for drawing the distinction needed in non-normative cases – that of non-cognitivism about probability claims, for example, to mention one case that I was interested in at an early stage – it turns out to be completely unhelpful even where it is supposed to work.

In order for it to work, we need to exclude the possibility that there could be a proposition Val(P) such that believing that Val(P) could be compatible, even identical, with desiring that P. And the direction of fit test doesn’t tell us that. Having world-to-mind direction of fit about Val(P) might be entirely compatible with having mind-to-world direction of fit about P, simply because P and Val(P) are not the same proposition. In order to apply the direction of fit test, in other words, you need to already have an answer to the question whether there is such a proposition Val(P) – and that’s just the original issue.

3:AM: You argue that we should understand truth in terms of a functional analysis. You argue that truth has a central normative function, and that is to get people to argue. You link this with arguing that there is no sharp boundary between factual and non-factual uses of language. Is this a version of deflationism?

HP: Yes. It shares several features with more familiar versions of deflationism, such as those of Quine, Horwich, Brandom and many others. Like those writers, I think that asking ‘What is truth?’ is asking the wrong question – a question that mistakenly assumes that truth is the kind of thing that has a nature, as it is sometimes put. Instead, we should be asking explanatory questions about the term, or concept: Why do we have a truth predicate? What is its function, and how would matters be different if we didn’t have it? My view diverges at this point, since I link truth to a central norm of our assertoric practice rather than to the kind of logical role (e.g., in generalisations such as ‘Everything that Fred says is true’) central to other deflationary views. But I agree with other deflationists completely about what questions we should be asking.

3:AM: Since you wrote that book, back in 1989 there have been other functionalist approaches to truth. So Beall argues that truth is merely a mechanism that allows us to generalize, a short cut mechanism required because of our medical limits (so god wouldn’t need truth in her vocabulary)? How has your view developed over the subsequent time? One thing I’m interested in is whether philosophers do actually change their minds much, so this is partly a question about that as well.

HP: Beall’s view is another version of the ‘transparent’, deflationary account of truth I just associated with Quine, Horwich, Brandom. I haven’t changed my mind about that kind of deflationism not being the whole story. On the contrary, I have defended my own version again more recently in a couple of papers, including ‘Truth as Convenient Friction’. Roughly, I think that my view trumps the other kind of deflationism in this sense: our conversational practice needs a norm of the kind I associated with truth; but once we have one, and a means of making it explicit, as Brandom would say, then we have something that will do the logical job, too. But the same doesn’t work in reverse, so the normative notion is the more basic one.

I don’t have much to say about philosophers changing their minds, but I do have an anecdote. When I was in Edinburgh in the early 2000s I once had the pleasant task of introducing Michael Dummett, who was there to give a named lecture. He told me that he had once done the same thing for Putnam. Putnam’s advertised title was something like ‘Theory Change in Science’, and Dummett said that in his introductory remarks, he commented on what a suitable title this was for Professor Putnam, who was famous for his readiness to change his own mind. He said that Putnam then got up and said, ‘It’s funny you should say that, Michael, because I’ve decided to give you a different paper!’

3:AM: You’re a metaphysician and you’ve taken a look at all the big themes. One of them is times arrow. When discussing this you suggest that we step back and look at the idea from a point of view outside of time. But before you tell us your thoughts about time’s arrow, could you say something about the block universe theory of time because it is very peculiar isn’t it. It conceives of time as being tenseless, which to many folk will seem paradoxical. How could there be time without past, present and future. It sounds like a redefining of the concept to avoid agreeing with McTaggart that time is unreal. How can time be tenseless? Is this just another case of folk talk being totally cut off from theories of physics?

HP: Well, just to disagree with a couple of bits of the first claim, I’m not a metaphysician (at least not in what’s now the familiar sense – more on this below), and I’m sure I haven’t looked at all the big themes! But about time, if someone wants to insist that time can’t be tenseless I’d be very happy to give them the term, leave them to the mercy of McTaggart, and use a different term for the notion associated with the block view.

I’d expect to be able to reclaim the term, eventually, when the lessons of McTaggart, Mellor and many others finally sink in on the other side of the fence – or, more likely, when that degenerating research program finally fizzles out. And even if that never happens, no matter. One of the advantages of not being a metaphysician is that I’m not in the grip of idea that there’s a thing, Time, whose nature it is our job as metaphysicians to figure out. So it’s easy for me to walk away … though that’s not incompatible with nipping over the fence occasionally to join McTaggart, Mellor, et al, in exposing the internal difficulties in that program. (Nor would it be incompatible with talking to physicists about whether something was missing from the block view for the purposes of physics, if someone came up with a plausible argument of that kind.)

3:AM: You argue that there is an asymmetry in how we think about the world that contradicts the block universe idea, or at least looks strange from that view. This is how you began to think about those in-built temporal asymmetries in the way we think and develop your theory about times arrow wasn’t it? Can you tell us about this and your theory of time’s arrow?

HP: Time itself doesn’t have an arrow, in my view. Someone creating a world like ours from scratch (and from outside, of course!) wouldn’t have to make a choice between two versions, one the exact temporal mirror-image of the other. Those are the same world, differently described. (This is the temporal equivalent of disagreeing with what Kant says about a world containing only one hand: he thinks there is a fact of the matter about whether it is a left hand or a right hand.) Indeed, I think it is very difficult to make sense of what it would be for these to be different worlds, in the intended respect.

There’s lots of temporal asymmetry with the world as we know it, of course, but I see no reason to think of any of that as an indication that time itself has a direction. (As I said, I find it hard to see what that would mean.) But there are fascinating questions about the nature and source of these various asymmetries – which of them are fundamental, and in what sense, for example. Another fascinating kind of question, in my view, is which of these apparent asymmetries are entirely objective, and which might in some sense be merely projections of our own temporally-asymmetric viewpoint.

But to get these questions into focus, you need a clear sense of the background, the base from which the project starts. One reason for starting with the undirected block universe – apart from the fact that it’s what physics gives us! – is that if you’re sure that you’ve stripped all the asymmetry out, it’s not so hard to keep a close eye on how it gets back in. That way, we can hope to avoid what I call double standard fallacies – illicitly assuming a time-asymmetry in some subtle form, in the course of trying to explain another one.

3:AM: Philosophers have had a pretty torrid relationship with physicists recently, but you were hammering away at Stephen Hawking’s views about the direction of time in public when his best seller came out in the early 1990s. What did Hawking get wrong, and more generally, why should philosophers and scientists listen to each other?

HP: Hawking was interested in the question whether cosmology could explain why entropy was so low around the time of the Big Bang. By that stage, thanks especially to the work of Roger Penrose, that question was clearly in view as the one to which the task of explaining the time-asymmetry of the second law of thermodynamics was pointing. The second law says that entropy goes up asymmetrically towards the future, and as Penrose saw perhaps more clearly than anyone at that time, the real puzzle is why it goes down towards the past. (In a sense this is the same thing, but focussing on the past directs our attention at the anomaly: statistically speaking, as had been clear since the late nineteenth century, one would expect entropy to be high in both directions. The puzzle is why the statistical inferences don’t work towards the past.)

Penrose also saw that there’s a problem with explaining why the Big Bang has low entropy in terms of a time-symmetric fundamental physics (which is essentially what we have). In such a physics, any argument applicable to the Big Bang ought to be equally applicable at the other end of the universe, in a ‘Big Crunch’ – or even in the kind of mini big crunches that produce black holes. But that would mean that the second law would reverse direction, as one approached such a final singularity. Penrose thought that that was unacceptable, and proposed instead that cosmology must invoke some time-asymmetric law (what he calls his ‘Weyl curvature hypothesis’).

Hawking thought he could do better than this, using an alternative hypothesis called the Hawking-Hartle ‘No Boundary Proposal’. As Hawking explained in A Brief History of Time, he believed initially that this hypothesis would imply that entropy would reduce towards a ‘Big Crunch’ – in other words, he accepted the consequences of symmetry that Penrose found unacceptable. But then he became convinced that he had made a mistake – he calls it his greatest mistake – and retracted this part of the view. That left him in the situation of claiming to get an asymmetrical result out of symmetrical theory – the very thing that Penrose, like many others, had decided was impossible – without telling us how the trick works.

A natural thought, at least for a suspicious philosopher like me, was that he hadn’t pulled off the trick at all. Instead, he’d slipped in an asymmetry at some point (e.g., by relying on statistical arguments in one direction but not the other), thereby committing a double standard fallacy (and ending up in the same position as Penrose, only without being aware of his asymmetric assumption).

That’s what I wrote about in my little commentary article in Nature in 1989. I haven’t changed my mind about that one, either. Indeed my confidence went up considerably when I sent a copy of it to Penrose (I think this was couple of years later), and he said that he had been saying that kind of thing to Hawking for years.

On the general issue – why should philosophers and scientists listen to each other? – I think there’s a sense in which they don’t have a choice. There are questions that arise in relation to most (probably all) of the special sciences that are by nature philosophical questions. (It’s hard to say what that means, but you know them when you see them.) Of course, no one in either field needs to engage with these questions, if they don’t want to. But the questions are there, and interesting from both sides, to thinkers with particular predilections. At that point, each side is simply shooting itself in the foot, if it refuses to listen to the other.

3:AM: Is asking whether the past exists, or any tensed portion of reality, a badly formed question because it presupposes a bad metaphysics? Is there a general lesson here about how ignoring the metaphysical foundations of any statement can lead to hot water? Does your idea of metaphysical and semantic quietism help here?

HP: The main problem with the debate about ontological issues between presentists and like, on the one side, and B-theorists, on the other, is that it looks like they are simply talking past one another, by meaning different things by ‘exists’: one side means it in a tensed way, the other in an untensed way. The participants on both sides try to meet this charge by agreeing with each other that that the issue turns on some ‘fundamental’ notion of existence, whose nature is in dispute between them. Metaphysical and semantic quietists like me can simply tune out at this point, as for ‘time’ above, because ‘fundamental’ is not one of our words (at least as it is being used here). The same goes for ‘reference’ and other semantic terms that tend to be pressed into service – interpreted in a ‘thick’, non-deflationary sense – in trying to identify some real point of disagreement in these debates.

3:AM: Causation is another big topic you’ve grappled with, and again it seems to connect up with time. We tend to think of causation in terms of time: first something happens and then something else follows, caused by the first thing. At the start of your book Causation, Physics and the Constitution of Reality you and co-editor Richard Corry cite Bertrand Russell saying that causation was like the monarchy, a relic of a bygone age erroneously thought of as being harmless. But just like the Queen, causation hasn’t been scrapped and also like the Queen seems to be thriving. Nevertheless its causation that again draws you to thinking about asymmetries: how come causation is only one way? Have you an answer? Or should we be republicans about truth as Russell suggested?

HP: Richard Corry and I develop Russell’s metaphor a bit. We point out that in the political case, there are two ways of disagreeing with the traditional view that political authority is vested in kings by god. You can be an anarchist, and reject the idea of political authority altogether, but there’s also a more moderate option: republicans don’t seek to reject political authority, but just to bring it down to earth – to think of it as constructed by us, rather than instituted by god. By analogy, we suggest, a ‘causal republican’ doesn’t seek to eliminate talk of causation, but regards it as partly anthropocentric, not simply an element of the pre-existing ‘furniture of reality’. (Causation is regarded as a secondary quality, in other words, as I put it in an early paper with Peter Menzies.)

It is not really clear whether Russell should count as a causal anarchist or a causal republican, by these criteria. I think a charitable way to read him is as a kind of armchair anarchist, much like Hume, who thinks that talk of causation is fine when we’re out in the real world. We just need to avoid the mistake of thinking that it is something we should be investigating when we looking for the fundamental constituents of reality. And that’s much the same as causal republicanism.

One of the great attractions of the republican view, I think, is that it makes it easy to explain the difference between cause and effect, and the fact that, at least most of the time, this causal ‘arrow’ lines up past-to-future. The former stems from the asymmetry between means and ends, from an agent’s point of view (I think Ramsey was the first to see this, incidentally). While the latter turns on the fact that all human agents happen to face the same way, as it were: we all deliberate past-to-future. Creatures elsewhere in the universe may do things in the opposite direction, and if so then they take the causal arrow to point in the other direction – and there’s no objective sense in which one of us gets it right, of course, any more than it is an objective fact that Australia is on the top half of the planet.

Pages: 1 2

First published in 3:AM Magazine: Friday, September 14th, 2012.