:: Article

epistemic consciousness

Daniel Stoljar interviewed by Richard Marshall.

Daniel Stoljar thinks all the time about what we can and can’t learn from introspection, about ignorance and the imagination, the epistemic view of consciousness, the ignorance hypothesis, slugs and tiles, distinctions between empirical and philosophical questions, physicalism as weltanschauung, whether materialism is part of a scientific world view, on materialism and physics and on whether metaphysics harmonising with science is any different from tourism doing so also. This one keeps hooking to the body. Brawlin’.

3:AM: What made you become a philosopher?

Daniel Stoljar: It’s not too far wrong to say I was persuaded to get into the subject by David Armstrong; that is, D.M.Armstrong, then Professor of Philosophy at Sydney. I originally went to university intending to become a historian, though I took philosophy too in part because my older brother and sister had done it and liked it. (My sister stayed in the profession and is currently on the faculty at McGill.) After one of the courses—a second-year course on philosophy of mind, as a matter of fact—Armstrong took me aside and told me in no uncertain terms where he thought my future lay.

I was also attracted to philosophy for a different reason, not a very academic one I’m afraid. I could see a mile off that it was an international discipline—very different from the way history seemed to me for example. At the time I was desperate to ‘see the world’, which more or less meant going to the US. That turned out to be true when I went to MIT for graduate school—a fantastic experience though part of a different story. It’s ironic that I now live in Canberra, just across the lake from where I grew up.

3:AM: Many philosophers of mind think that introspection can’t tell us anything about the world in its most fundamental nature. This is a big problem for explaining consciousness isn’t it because it claims there isn’t anything in fundamental reality to explain. So why do some contemporary philosophers find the idea of no experiences so compelling?

DS: Well, I certainly don’t think that introspection can tell us much about the fundamental nature of the world, not anyway if ‘fundamental’ carries the meaning it often does in contemporary philosophy, i.e. something like a basic element of nature.

When people talk about learning such and such ‘by introspection’, they usually have a particular type of fact in mind, the type reported by first person present tense psychological reports. So you can learn by introspection that you have a particular sensation or that you feel a particular way. But it is implausible that you can learn by introspection a very different type of fact, for example, that you have or that you instantiate some fundamental property. Maybe if you knew by introspection that you had a particular sensation and if you knew independently that having that sensation was a fundamental property, you could infer that you instantiated a fundamental property—but it is stretch to say that you learnt that further thing by introspection.

So I don’t think we learn about fundamental properties by introspection. On the other hand, I don’t really see this as presenting a major obstacle to explaining consciousness, which isn’t to say there are no obstacles. After all, here are phenomena which are the proper objects of scientific study and which are not fundamental. Take chemical properties—chemistry studies them, and yet they are not fundamental. In principle, properties like having a sensation could be of that sort too.

3:AM: Why are you not convinced that any of the leading alternatives to the epistemic view of consciousness – eliminitivism, primitivism, a priori entailment and a posteriori entailment are plausible enough?

DS: These are the labels I gave to the standard positions in philosophy of mind in my 2006 book Ignorance and Imagination. (The labels are in all but one case non-standard because of issues internal to the book, which don’t really matter here.)

As regards eliminativism and primitivism (aka metaphysical dualism) my views are fairly conventional. Eliminativists deny that people are in pain or believe things. I regard that as denying obvious facts – something you should never do. Dualists say that various psychological properties are fundamental—but saying that makes it hard to see how such properties are integrated into the rest of nature in the way they clearly are.

As regards a priori entailment (aka a priori or type A materialism) and a posteriori entailment (aka a posteriori or type B materialism) my approach is perhaps a little less mainstream. Both these views say that psychological properties are necessitated by something else, e.g. non-psychological or physical properties. The difference is that the necessitation is a priori in one case and a posteriori in the other. One point I am keen to stress about these views is this. While the difference between them is important and interesting, it makes no difference to the assessment of the key bits of reasoning in philosophy of mind against necessitation, by which I mean arguments like the knowledge argument or the conceivability argument. Most people I think assume the opposite, which is why my position on this issue is a bit non-standard. (Though I don’t think it is that non-standard. I would argue that Kripke holds a very similar position in the last lecture of Naming and Necessity.)

Of course, it is still true that if you hold either of these views you need to respond somehow to the knowledge argument and the conceivability argument. And philosophy of mind has no shortage of suggestions about how to do so: the ability hypothesis (in the case of the knowledge argument), the phenomenal concept strategy (in the case the conceivability argument), various acquaintance views, and many others. Unfortunately, I think all of these are subject to fatal objections. That is what motivates the search for a more radical approach, which I offer in the form of the epistemic view.

3:AM: What is the epistemic approach to the problem of consciousness – and why do slugs and tiles help explain it?

DS: There are two parts to the epistemic approach. The first, which I called the ignorance hypothesis, says that we are currently ignorant of a type of non-experiential or physical fact that is relevant to the nature of consciousness. The second part is that the truth of the ignorance hypothesis makes the key arguments in philosophy of mind unpersuasive.

Here’s a good way to see the second part of the view. Take the main claim of the conceivability argument, namely, that it is conceivable (and so possible) there is someone identical to me in respect of all physical facts but different in respect of some psychological fact. Now focus on the phrase ‘all physical facts’. How is that to be understood? If the ignorance hypothesis is true we have two choices: either it includes the facts of which we are ignorant or it does not. Either way the argument is unpersuasive. On the first option, the thing that is claimed to be conceivable will not be, at least not in the relevant sense. On the second option, certain facts will be left out of the story which means that even if successful the argument shows only that psychological facts come apart from some physical facts not that they come apart from all.

The slugs and the tiles are just a device to show how a story like this might be true. (The slugs are in fact borrowed from Frank Jackson’s famous article ‘Epiphenomenal Qualia.’) Imagine a mosaic constructed about two sorts of tiles, triangular tiles and pie-shaped tiles or sectors. And imagine that there are some intelligent slugs who live on the mosaic. The slugs don’t know about the pie-shaped tiles, perhaps because their sensory systems don’t detect them. They might wonder about how circles are formulated on the mosaic, since no arrangement of triangles is going to add up to a circle. You can formulate counterparts of the main positions in philosophy of mind here, but it is also the case that the epistemic view is correct.

3:AM: So ignorance is the key to the analogy isn’t it, coupling ignorance with the problem of experience?

DS: Yes ignorance is the key to the analogy. But notice that I am not recommending that you say “I don’t know” or “We don’t know yet” to any philosophical question of about consciousness that might come along Rather I have in mind a quite specific question: viz., are the conceivability and the knowledge arguments persuasive? The epistemic view gives an unequivocal response to that question, viz., no.

3:AM: Do you think this is true or just plausible? I think you have found Russell’s metaphysical speculations helpful don’t you, as well as some of the linguistic understanding characteristic of the seventeenth century and the discussion of chemistry later?

DS: I think it is true and plausible, or better, true because plausible.

It is a bit hard to talk about ignorance of the sort I have in mind in the first person. You can’t say ‘the fact that I don’t know or don’t understand is such and such.’ Still it is fairly clear that the ignorance hypothesis describes a situation we could be in—others (and ourselves at other times) have been in such situations.

Why think we are in that situation? I have three main grounds. General grounds—we should never assume that we know everything. Explanatory grounds—if it is true it explains a lot about our situation. Historical grounds—there are historical precedents for our current situation. That is why I am interested in Descartes on linguistic understanding, C.D.Broad on chemistry, and Russell on the abstractness of physics.

The case of Russell is in some ways particularly interesting. Russell’s discussion in The Analysis of Matter and related works presents as I read it a version of the epistemic view but one which is placed in a philosophical context I don’t share—Russell defended a kind of empiricism which entailed that in physics, and in empirical inquiry generally, we learn only about what he called ‘the causal skeleton of the world’. Something a bit like Russell’s position has in recent years come to be called ‘Russellian Monism’ and has been the subject of a lot of debate—for example there is an excellent forthcoming collection with that title from OUP, edited by Torin Alter and Yujin Nagasawa. I am interested in that view and have written papers on it but in the end I take it to be one form of the general view I am interested in, and not the most plausible form.

3:AM: Why aren’t you convinced by the idea that we can explain consciousness by making a distinction between objective and internal standpoints, or truths that concern structure and function, and those that don’t?

DS: My engagement with these ideas comes from what I think is probably the most interesting objection to the epistemic view. The objection is basically that the ignorance hypothesis is false. Nobody denies we are ignorant in some sense or other; the suggestion is rather that what is at issue in (e.g.) the conceivability argument is a certain type of truth, for example, an objective truth or a structural truth. Since, the objection says, we know what this type of truth is, we are not ignorant of any relevant type of truth.

I have gone into battle over these ideas on a few occasions now. The report from the front is this. I keep giving completely clear and devastating reasons why the objection fails, but my opponents don’t change their minds. I can only conclude that they are irrational. (Though I admit they probably think the same about me.)

As regards objectivity, I roughly follow Thomas Nagel in thinking that an objective truth is one that can be understood in many ways, while a subjective truth is one that can be understood in only one way. Suppose now that P is some truth that is objective and Q is some truth that is subjective. It is immediately clear that understanding P does not entail or necessitate understanding Q—for the way you understand P may not be the way required to understand Q. But it does not follow at all that P does not entail or necessitate Q. Since the conceivability argument is concerned with that second issue, objectivity has nothing to do with it.

As regards structure, the main issue here is that ‘structure’ is a word with umpteen meanings. If you doggedly work through all the relevant meanings it can have, it turns out that none of them permit a workable version of the objection summarized above. I won’t go through all of these here. For those interested look at my paper in the OUP volume on Russellian Monism I mentioned above, which discusses three interpretations of ‘structure’ in some detail.

3:AM: You bring three new elements do you bring to the epistemic view – the difference between empirical and philosophical questions, and the particularity of the problem of experience, and the way you frame ignorance. Can you say something about these?

DS: On the distinction between empirical and philosophical questions (I wouldn’t quite put the point this way now), the main point here was to emphasize what question precisely the epistemic view could answer. As I have said, for me the question is: are the key bits of reasoning in philosophy of mind persuasive? For that question the epistemic view gives a straightforward answer. But there are loads of interesting questions about consciousness and related things that are quite different. For example, do conscious states represent the world, and if so in what way? (That is the sort of question at issue in works by many philosophers: Susanna Siegel and Charles Siewert to take two excellent examples.) Those questions are super interesting but the epistemic view has little to do with them.

On the second, I wanted to emphasize that the ignorance hypothesis bears on the conceivability argument because of a particular fact about conceivability, viz., that you can conceive that p in the relevant sense only if you know what it takes for p to be true. If you want to bring to bear the ignorance hypothesis on other questions, you would need to formulate those questions clearly, and show that it is relevant. You can’t just assume it is relevant.

On the third, I wanted to decouple the epistemic view from positions often associated with it. One example is the view associated with McGinn, which says that we are cognitively closed with respect to the solutions to philosophical problems, and the mind-body problem in particular. I am sympathetic to some elements of McGinn’s position, but I make two points: first, the argument he offers for his view is unpersuasive; second the issue of cognitive closure is irrelevant to the assessment of arguments like the conceivability argument.

3:AM: You’ve also thought and written extensively about materialism. This has become the dominant contemporary philosophical position hasn’t it, the Weltanschauung of modern analytic philosophy, as Carl Gillett and Barry Loewer say?

DS: Yes it is really a striking fact that materialism (aka physicalism) has become such a dominant view in philosophy since about the early sixties. It is sometimes thought that materialism is less popular than it used to be. Maybe, but it is still extremely popular, and the views that replace it often agree with it to a considerable extent. But it wasn’t always popular. In his (1925) preface to Lange’s classic The History of Materialism, Russell spoke of the 1860s as the ‘the materialist sixties’, and singled it out a unique decade in the history of philosophy in which materialism became popular. For us the sixties means of course the 1960s, but that could easily described as another materialist 60s—maybe the fifty years sense is the long materialist sixties.

3:AM: What do you say it means to say that everything is physical?

DS: The project of transforming the vague statement ‘everything is physical’ into something more explicit and tractable is something I am very interested in. A plausible suggestion here is to say something like this: materialism is true just in case every instantiated property is necessitated by (and not metaphysical distinct from) some physical property. That gives you an outline of a definition. Then to fill in the outline, you need to say what it is to be a physical property, and what it is to be necessitated by and not metaphysically distinct from a property. I have some things to say about the second question but mostly I am interested in the first.

To put it roughly, a physical property is one that is either a distinctive property of a physical object or is a property that explains the distinctive properties of a physical object. This account isn’t circular (or at least not too circular) because ‘physical’ as it attaches to objects means one thing, and ‘physical’ as it attaches to properties means something else. The second disjunct is required to accommodate the fact that physics discusses properties that are not distinctive properties of physical objects ordinarily conceived.

In my 2010 book Physicalism I developed an argument that if a physical property is understood this way, there is no version of materialism that is true and deserves the name. The basic thought can be put like this: in the definition of a physical property (and so in the definition of physicalism) does the second disjunct mean ‘which actually explains’ or ‘which possibly explains’? If it means the first, materialism is bound to be false in possible worlds at which intuitively it is not. If it means the second, it is bound to be true at possible worlds at which intuitively it is not. Either way, there is no version of thesis here that is both true and deserves the name.

The argument is similar to a famous dilemma in this area of philosophy called Hempel’s dilemma, which has been discussed interestingly by a lot of people including Barbara Montero, Jessica Wilson and Andrew Melnyk. One problem with Hempel’s dilemma is that it gets bogged down in an apparently fruitless debate about how optimistic we should be about present science. I was trying to find an argument that was similar but avoided any such problems. One piece of advice David Lewis offers philosophers (I am paraphrasing) is that when you are faced with a bit of philosophy involving time, translate into a bit of philosophy involving modality, and see what happens. I was following Lewis’s advice here. I think that if you take Hempel’s dilemma and transpose it into a modal framework, the argument you end up with is very different and much more plausible.

3:AM: Is it part of the scientific world view?

DS: It is certainly often thought so. But in fact the whole question of whether materialism is part of the scientific world-view is much harder than it appears.

For one thing, there are many different things people have in mind by ‘materialism’, whether legitimately or not. We might have in mind the materialism of the ancient Greeks. That is definitely not part of our world-view. Or we might have in mind the materialism of Smart and Lewis. But I doubt that is part of our world-view either, mainly because it is incredibly optimistic. It says first that we know the basic facts of nature and second that we are in a position to derive every other fact from these basic facts. But if we have a world-view at all it will presumably have to say something about the epistemic position that we human agents find ourselves in; I doubt it will describe our position in such optimistic terms.

For another thing, it is far from clear that there is any such thing as ‘the scientific world view’. Freud said that a world-view is a “single overarching hypothesis” that answered “all the questions for our existence”. I doubt contemporary science (or even future science) provides any such thing. Maybe we have fragments of a world-view. We know that the fundamental properties, whatever they are, are unlikely to be the properties we know about in ordinary experience. We know that we are made of matter (though we don’t know what the properties of matter are). We also know of course lots of issues of detail. But these facts don’t really add up to a world-view.

That is not to deny there is a determinate way the world is, or that we can find out important things about the world. I am a very old fashioned realist about this sort of thing. But it is one thing to say that there is a determinate way the world is and it is quite another to offer some pithy summary (in ordinary English, here and now, in a philosophy department!) of what it is. That philosophers of mind sometimes think it is their business to do that has always struck me as faintly ridiculous.

3:AM: Is one of the reasons for investigating materialism motivated by the thought that so much else in philosophy hinges on it being assumed to be true – are you out to destablise that assumption?

DS: Yes one of the big reasons materialism is interesting for me is its role in meta-philosophy. To put it incredibly generally and roughly: you might say that materialist philosophy of philosophy shares with positivist philosophy of philosophy, which it replaced, a sort of rationalism about philosophy, the idea that philosophical problems are a matter of explaining things in terms of some basis which is assumed itself to raise no further questions.

It’s pretty curious: the topics I have worked on—what if anything is wrong with the conceivability argument and similar arguments; how exactly to formulate materialism—look at first glance like standard questions in philosophy of mind, bits of ‘normal science’ as Kuhn might have said. But it turns out that to answer them properly you need to reject some very big pictures about what philosophy is and how to do it.

3:AM: How far does the philosophy of materialism depend on understanding contemporary theories of physics?

DS: Not that much I think, which is lucky for me because I am certainly not in any position to contribute either to physics or the philosophy of physics. The basic reason is that materialism is supposed to be an abstract doctrine, which is not supposed to depend on the details of any particular physical theory. And in any case, some people, like say Van Fraassen, who know an enormous amount about physics, are as puzzled about materialism as I am, and for recognizably similar reasons. (Though the point is not decisive, since other people who know a huge amount about physics are not so puzzled.)

3:AM: Many metaphysicians want to ensure that their metaphysics is in harmony with the best science – do you agree, or are you arguing that the best science needs to understand that there are philosophical issues that science can’t investigate?

DS: Metaphysics needs to be in harmony with the best science. But that shows nothing special about metaphysics. The same applies to tourism studies. If the best science says p, that means we either know p or else have extremely good evidence for p. Neither metaphysics nor tourism studies should contradict what we know or have very good evidence for.

But I also don’t see any tension with this point of view and the idea that there are philosophical issues that science can’t investigate. There are loads of issues that science can’t investigate. Take the Spanish Civil War. That happened and we can read fantastic books about it, like Orwell’s Homage to Catalonia. But I don’t think there is a science of the Spanish Civil War, not at any rate if by science you mean something like what happens in physics—a highly mathematized investigation of empirical laws. It seems to me quite likely that a lot research in philosophy of mind is going to look more like history than physics; it will take the form of detailed (truth evaluable) descriptions of consciousness and other interesting phenomena.

3:AM: And do you think philosophy makes progress? How can we decide this? Won’t any sort of evidence for advance become fodder for philosophical counter-argument?

DS: Philosophical progress is a topic I have became interested in a few years ago and have been thinking a lot about recently. I’m an optimist though of a qualified or reasonable kind. I think the pattern of success and failure in philosophy is not that different to the pattern you see in many fields, especially fields outside of the hard sciences whose success is really quite spectacular. In that sense philosophy is not an ‘exceptional’ discipline, to use Timothy Williamson’s phrase. (Of course Williamson’s point is that it isn’t exceptional. I very much agree with him about that.)

But curiously optimism seems a pretty unpopular view; pessimism about philosophical progress is very common both inside and outside the discipline. What is lying behind it? I think there are probably a number of factors but one key assumption is the view that the big questions being debated today are the same as the big questions debated long ago. Once that assumption is granted, many aspects of pessimism make sense. For example, suppose you focus on some big question currently being debated; it doesn’t matter what it is is, call it Q. It is clear that Q isn’t solved, otherwise it would not currently be the subject of debate. Now suppose Q is identical to some question debated centuries ago, call it R. It is pretty clear that R isn’t solved either; for one thing it is identical to Q and Q is not solved! So the history of Q is the history is an unsolved problem, and if Q is typical, the history of philosophy is the history of people not solving the big questions they are interested in. That makes pessimism seem almost inevitable.

But when you look at concrete cases it is quite unclear that the big questions debated now are the same as those debated centuries ago. Take for example Descartes’ discussion in the Meditations and Jackson’s discussion in ‘Epiphenomenal Qualia’, the famous paper I mentioned before. Are they talking about the same question? Well it is true perhaps that they are talking about the same topic or subject matter, both are interested in some abstract sense about mind and matter and the relation between them. But it doesn’t follow that they are asking the same big question about that subject matter. After all, historians today might still discuss the fall of Rome (that is, they might still discuss that topic) just as their forebears did 300 years ago, but they are presumably asking very different big questions about that topic. Maybe you could argue that while it doesn’t follow, it is nevertheless true; that is Jackson and Descartes are discussing the same question in fact. But I think that’s quite implausible. Descartes was mostly interested in thought and intellectual capacities. Jackson is not interest in that at all, at least not in that paper. Descartes thought that matter was extension, Jackson certainly does not. In other words, since they have quite different views of mind and matter, it is hard to see them as focusing on the same big question. And once you realize that, a much more optimistic way of looking at philosophy comes into view. At any rate so I would want to argue.

Will any sort of evidence for progress become fodder for counterargument? Well, I think it is true that when one sort of question is solved, a successor question can often be formulated. But nobody should conclude from this that philosophy makes no progress. After all the same thing is true in many fields. You might I guess decide to reject progress in any field. I think that would be very implausible, but in any case the interesting question about progress in philosophy is not so much whether progress of any sort is possible but whether progress is possible here.

3:AM: And finally, are there five books you could recommend to the readers here at 3:AM to take us further into your philosophical world?

DS: I tend to have crazes and read a lot of papers by one author on stuff that has only a marginal bearing on what I am writing. But some perennials I keep coming back to are:

1. Chomsky’s New Horizons in the Study of Mind and Language. Like many classics, this book provokes some extreme reactions. On the blurb Gilbert Harman says “these essays represent the most significant work that’s been done in the general area of philosophy of language and philosophy of mind”. But other people often dismiss it because of Chomsky’s notorious ‘London’ passage in chapter 2 in which he seems to say that London does not exist. I don’t believe at all he is saying that (or at least is not saying it in the way that most philosophers interpret it), but I won’t try to explain that here.
2. Fodor’s Representations. I am a huge fan of Fodor and the way he writes though I think it is fatal mistake to try to write like him (people who make that mistake populate the philosophical literature). This is a terrific collection that presents, as Fodor says in the preface, a theory of mind rather like Descartes—but without the “ontological and epistemological baggage”.
3. Jackson’s Perception. This is a book that defends the sense-datum theory of perception. I think of that theory as false (so does Jackson now) but his discussion of it is exquisite – a fantastic example of philosophical writing.
4. Lewis’s Papers on Metaphysics and Epistemology. This contains most (not all) of Lewis’s papers on philosophy of mind, and so it is something I have wrestled with repeatedly. It is also dedicated to the philosophers of Sydney and Canberra—something that appeals to me most (but not all) of the time.
5. Chalmers’ The Conscious Mind. An unbelievably fluent book that is usually described as a defence of ‘naturalistic dualism’. I think that description forgets that its official conclusion is three-way disjunctive: either interactionist dualism, or epiphenomenal dualism, or the position that is now usually called ‘Russellian Monism’. But that third view is not a version of dualism (whatever else it is).

Richard Marshall is still biding his time.

First published in 3:AM Magazine: Friday, February 28th, 2014.