the double life of objects
Thomas Sattig interviewed by Richard Marshall.
Thomas Sattig is a chillin’ metaphysician who thinks there’s good sense in studying language and reality together. He thinks that foundational metaphysical analysis should preserve common sense conceptions, that temporal supervenience is an equilibrium problem, that tense doesn’t run metaphysically deep, that there are problems with a temporal-parts account, that a departure from the Aristotelians helps, that there are puzzles that pluralism about material objects raise, that objects have double lives, that just because physicists don’t include some objects in their theories doesn’t mean they don’t exist, that mereological indeterminacy is worldly but not fundamental, and that his interest is not in the clash between philosophy and physics but in philosophy and the ordinary conception of the world. After this, reboot yourself….
3:AM:What made you become a philosopher?
Thomas Sattig:It happened in high school. Exhausted by teenage vertigo, I longed for something stable to direct my energy towards. So I started developing a sense of how fundamental a question is compared to others and went after the deepest questions I could find. They were philosophical questions. From some of the answers I could understand I gathered that philosophy matched the creativity I had encountered in fine art, while having stricter rules. A very special and attractive mixture. Back then Nietzsche was my guide. I think that was a good start.
3:AM:In your book about language and realityyou study them together rather than separately. Why put them together?
TS:Metaphysicians often begin with prephilosophically accessible phenomena and then go deep by asking what the phenomena are like fundamentally. Given that the phenomena are familiar, we have common-sense beliefs and intuitions about them. What role does common sense play in the metaphysical enterprise? I believe that foundational metaphysical analysis should aim to preserve our common-sense conception. The task is a difficult one. Soon tensions between our metaphysical principles and our ordinary thought and talk start appearing. But we should resist giving up our prephilosophical beliefs too easily. For they are prima facie guides to how things are and to how they could be. So I recommend searching for an equilibrium between the metaphysical analysis of the deep structure of the world and our ordinary, linguistic and mental, representation of the latter. The way in which I recommend establishing such an equilibrium is by giving a semantical account of ordinary discourse, which links familiar linguistic behaviour with deep metaphysics. To be sure, this is a type of semantics geared to the demands of metaphysicians. Semantics as done in linguistics and philosophy of language doesn’t share the aim of uncovering the metaphysical basis of ordinary thought and talk.
3:AM:You begin with what you call temporal supervenience. Can you explain what you mean by this term?
TS:The problem of temporal supervenience is an equilibrium problem of the sort I just mentioned. There are different conceptions of time. While ordinary space is three-dimensional, ordinary time is one-dimensional — it can be represented by a line — and consists of past, present, and future. This is the ordinary conception of time, in virtue of being the conception to which we are committed by our ordinary temporal discourse. When we describe the world in ordinary time, we describe it from the perspective of the present time — we use tensed language.
Modern physics, by contrast, offers a different conception: there is only a four-dimensional spacetime of which timeis merely an aspect. In its most general form, the problem concerns the metaphysical status of our ordinary conception of time.
One instance of the problem concerns the status of tense. Is temporal perspective an aspect of the reality represented by ordinary thought and talk — are there fundamentally present-directed, past-directed and future-directed facts — or do we merely represent a fundamentally tenseless reality in a tensed way? Another instance concerns the relationship between ordinary time and spacetime. How is what goes on in ordinary time related to what goes on in spacetime? I find it quite plausible that all facts about ordinary time are grounded in facts about spacetime. But how are they grounded? (I now prefer Kit Fine’snotion of ground where I used to invoke the notion of supervenience.) As parts of the latter issue, there is the task of explaining the status of tensed language vis-à-vis the spacetime conception and the task of explaining how ordinary facts of material objects’ persisting and changing through time are grounded in spacetime facts.
3:AM:What’s the shape of ordinary time? This involves working out whether tense talk is irreducible or not doesn’t it? You say truths about spacetime can’t be tensed don’t you?
TS: In my monograph The Language and Reality of TimeI argued for the view that tense doesn’t run metaphysically deep — that it is merely an aspect of our representation of reality, not of reality itself. One of my worries concerned how dynamic, irreducibly tensed conceptions of time are to be formulated in spacetime terms. If spacetime has a “static” temporal dimension built into it, how do we make sense of the “flow of time” in spacetime? Do we need an extra temporal dimension, a “hypertime” to explain how things undergo dynamic change in spacetime? That would be extravagant. Perhaps, one might respond, we should stop thinking of spacetime as being temporal in the first place. There is a dynamic time with its primitively tensed facts on the one hand, and there is that four-dimensional manifold described in physics on the other. But then all hope of grounding ordinary time in that manifold seems lost. Reductionists about tense are in a much better position to offer a unified picture of time and spacetime.
3:AM:So how do you say objects get located in spacetime?
TS:The question I started with in The Language and Reality of Timewas how ordinary, material objects persist through time. According to one of the classical accounts, they do so by having different temporal parts at different times. According to another classical approach, they persist by being wholly present at different times. A common objection to the second account used to be that it is unclear what being wholly present amounts to. This and other considerations led me to reformulate the persistence debate in terms of spatiotemporal location. In this alternative conceptual framework I distinguished a range of views, including the view according to which a material object persists by being exactly located at a temporally extended spacetime region and the view according to which a material object persists by being exactly located at multiple temporally unextended, or instantaneous, spacetime regions. The latter picture now goes by the label “multi-locationism”. It is the picture I defended in the book.
3:AM:You criticise a temporal-parts account of temporal supervenience. Can you explain what that is and why you don’t think it works?
TS:The temporal-parts account, as I see it, grounds a material object’s ordinary temporal profile in facts about the object’s temporal parts. First of all, a material object that is exactly located at a temporally extended spacetime region has a temporal part corresponding to each instantaneous part of its region. Moreover, an object exists at a time in virtue of having a temporal part at that time, and an object has a property at a time in virtue of having a temporal part at that time that has the property. I’m no friend of this account, because it doesn’t do a very good job in capturing our common-sense conception of ordinary objects. There are a number of objections. I just want to mention one that took me a while to appreciate and that has not received the attention it deserves.
The temporal-parts picture of material objects is extremely liberal about how these objects are composed of parts. Given that an object exactly occupies a certain four-dimensional spacetime region, the temporal-parts view has it that any way of slicing the object’s location is exactly occupied by a part of the object. So the object has arbitrary parts along the spatial and temporal dimensions. This is a very popular idea. By the lights of common sense, however, this picture is too liberal. Here is a simple example. Suppose a statue was made from a block of marble. The statue has two arms. Are these arms also parts of the block of marble that constitutes the statue? They don’t seem to be. The block is made up of the same physical particles as the statue, but the block is not a kind of thing that has arms. There are more extreme examples. Without going into detail, the standard temporal-parts picture of composition has the exotic consequence, as Kit Fine has shown, that I currently have a part that itself has Socrates as a part. Surely, though, I don’t have parts of that sort. What these types of example bring out is that how ordinary objects are composed seems to depend on the kind of the composite and the kinds of the parts; an object of a certain kind only has parts of certain kinds. So it does not seem to be the case that any way of diving the spatiotemporal location of an object has a corresponding part of the object.
3:AM:So what’s your favoured alternative?
TS:Focusing on the temporal profile of ordinary objects, I used to think that ordinary objects are multi-located material objects without temporal parts. Now I deny that ordinary objects are straightforwardly material objects in the first place. In my forthcoming book The Double Lives of Objects(to appear with OUP in spring 2015) I develop a picture of ordinary objects, according to which they divide into different metaphysical components with different “lives”. Following Aristotelians, I distinguish between an object’s matter and its form.
Departing from Aristotelians, I view an object’s matter and form as having different qualitative profiles. Accordingly, ordinary objects can be described differently from different perspectives, where a description from one perspective is made true by one component of the object and a different description from another perspective is made true by the other component of the object. This position occupies a middle ground between the positions that have dominated traditional metaphysics of material objects, classical mereology and Aristotelian hylomorphism. It combines some of their individual strengths. For instance, it is able to deal with the intuitions about parthood mentioned earlier without incurring commitment to metaphysically suspect Aristotelian forms. And it has problem-solving powers that its competitors lack.
How does this picture relate to the traditional debate between friends and foes of temporal parts about persistence and change? Much of this debate is about saving the appearances. My view is that many of the familiar puzzles concerning our common-sense conception of objects that have driven this debate have a plausible solution that is independent of the foundational metaphysics of persistence and change, and that is compatible with several of the traditional positions.
3:AM:Another problem about duration and change has been the subject of a recent paper about pluralism and determinism. So to set up the problem, can you tell us what you mean by pluralism about material objects.
TS:According to pluralism, distinct material objects can exactly occupy the same spatial region and be constituted by the same matter at a time — for short, they can coincide at a time. Consider my favourite example. A paper airplaneis made from a piece of paper. Since the piece of paper existed before the paper plane came into existence, these artefacts seem to be distinct objects that coincide over a certain period of time. Pluralism is typically motivated by common-sense examples of this sort. It is fair to say that the most popular metaphysical accounts of material objects today are common-sense inspired versions of pluralism.
3:AM:You say that this might seem like a common-sense position but it makes indeterminism too easy, and that it shouldn’t be as easy as that because determinism only loses its grip via physics. So how does it make indeterminism easy?
TS:Suppose that we create a piece of paper in the shape of an airplane. When the piece of paper is flattened at a later time, the piece of paper survives, while the paper plane gets destroyed. On the assumption that the two coinding artefacts have all the same qualitative properties until the time of flattening, the qualitative profiles of the two objects match before the time of flattening but diverge afterwards. In other words, our laws of nature fail to determine whether an object with a certain past survives flattening or not. Therefore, our world is indeterministic. The problem with this result is that our simple case seems to show our world to be indeterministic on mundane, a priori grounds. But that makes indeterminism too easy.
3:AM:So how do you prevent this mundane proof of indeterminism?
TS:I find the following strategy the most promising one. In the relevant cases determinism is violated by distinct objects whose profiles match up to a certain time but diverge later. Calls this scenario “qualitative branching”. Now distinguish between genuine qualitative branching, which is, in some sense, metaphysically deep, and nongenuine branching, which is metaphysically shallow. Given these concepts, we can say that qualitative branching involving coinciding ordinary objects does not violate determinism, because it is nongenuine branching, while determinism is only sensitive to genuine branching. So far, this is only a recipe for a solution. I argue against some ways of implementing this strategy and propose my own.
If coinciding objects are analysed as double-layered, quasi-hylomorphic compounds, along the lines I propose in The Double Lives of Objects, then it is natural to understand the qualitative branching of coincidents as concerning their form, but not their underlying matter. That is, at the deeper level of material objects, the qualitative branching disappears. And since only branching in the profiles of the strict denizens of space and time matters for questions of determinism — that is, since determinism is insensitive to metaphysically superficial vagaries of ordinary objects — the problem of cheap indeterminism is avoided. This is one of many instances where a philosophical puzzle about objects dissolves once we realize that they lead double lives.
3:AM:Wouldn’t a physicistrespond to this by saying that its theories just don’t include things like pieces of paper being paper planes or lumps of clay being sculptures or hands being fists, so the examples used to motivate the problem of pluralism just don’t get traction? They don’t think there is a double life of objects. Why would that not be a good enough response for you?
TS:That physicists don’t include these objects in their theories doesn’t mean that they don’t exist. If common sense is a guide, as I think it is, they do exist. And since determinism prima facie concerns the evolution of a whole world over time, the behaviour of ordinary objects is a prima facie candidate for the violation of determinism. So appealing to the restricted application of physical theories doesn’t do much work by itself. More would need to be said about why the crazy behaviour of ordinary objects is irrelevant for questions of determinism. What are the philosophical reasons for restricting determinism? I just sketched my answer. There may well be plausible alternatives.
[Pic: Rob Gonsalves]
3:AM:You’ve also thought about vaguenessand ontology and related vaguenessto the problem of the many. Can you first set out the problem of explaining mereological indeterminancy of ordinary objects that preserves our familiar way of counting objects, giving us an example so we can get a grip on it?
TS:Macroscopic objects seem to have fuzzy boundaries. Mountains are a popular example. Kilimanjaro is a lonely mountain on an open plain, or so it seems. It is composed of rocks. But it is unclear where the mountain begins and where it ends. For many rocks on Kilimanjaro’s surface it is indeterminate, or vague, whether they are parts of the mountain. So there are many ways of drawing Kilimanjaro’s boundary. Given that each boundary we can draw has a corresponding aggregate of rocks, each of these aggregates is a candidate to be Kilimanjaro. Which one is Kilimanjaro, then? Each of the aggregates seems to be an equally good candidate to be Kilimanjaro. None of them is special. If each of the candidates is a mountain, then we have many mountains on the plain. If none of them is a mountain, then we have no mountain on the plain. Either way, contrary to what we expected, it is not the case that there is one mountain on the plain. This is the problem of the many.
3:AM:So how do you argue that mereological indeterminacy has its source in how ordinary objects are? And what is a quasi-hylomorphic ontology and why is it important to your approach?
TS:The standard account of indeterminacy of parts locates it in how we refer to objects. That is, the indeterminacy is viewed as having its source in our representation of the world, not in the world itself. The main idea, as applied to the case of the mountain, is that the name “Kilimanjaro” is imprecise, in virtue of having multiple overlapping candidate referents with slightly different boundaries. I have argued that this framework makes it hard to capture our intuition that there is exactly one mountain on the plain. A more promising approach is to locate the indeterminacy not in our description of the mountain, but in the mountain itself. On this alternative view, we get to say that there is a single, mereologically indeterminate mountain on the plain. But how should we understand indeterminacy in the world? Most philosophers, myself included, shy away from accepting this type of indeterminacy as fundamental. What to do?
The alternative I propose is to construe mereological indeterminacy as worldly but at the same time as nonfundamental. I develop this account in my quasi-hylomorphic framework mentioned earlier. The crux is that an ordinary object is more than a material object. It is a compound of a material object and multiple, “superimposed” forms. For Aristotelians, forms have the function of unifying a complex entity internally. Accordingly, they wouldn’t countenance entities with multiple forms. In my framework, forms don’t have that function, and so the way is clear for multiplicity. Now, each form of a mountain encodes its own mereological profile. And when the different forms of a mountain disagree on whether a given rock counts as a part, then the mountain is mereologically indeterminate. This is vagueness in the object itself. But it doesn’t run deep, because all facts about superimposed forms have a simpler, indeterminacy-free explanation. Fundamentally, I claim, objects don’t have any indeterminate parts.
3:AM:Does common sense and relativistic metaphysics fit together. And for you has metaphysics got to agree with science or common-sense for it to be valuable?
TS:Metaphysics should seek a match with both physics and common sense. This is a hard task, admittedly. I have raised some problems about the relationship between relativity theory and common sense. I worry about their compatibility, because I take them both seriously. No need to worry, one might say. Surely, in a fight between physics and common sense, physics wins. I agree. But the problems that interest me do not concern a fight quite of this sort. They rather mark an apparent clash between common sense and certain metaphysical interpretations of physical theories. So we are really dealing with philosophical attacks on our ordinary conception of the world. Time to rethink the philosophy. If done carefully, the tensions may go away. (I don’t promise.)
3:AM:And finally, for the readers here at 3:AM, are there five books you could recommend to take us further into your philosophical world?
TS:Saul Kripke’s Naming and Necessity, his Wittgenstein on Rules and Private Language, and David Lewis’ On the Plurality of Worldsopened my eyes when I was a student. David Chalmers’, David Manley’s, and Ryan Wasserman’s anthology Metametaphysicscollects a range of views on what metaphysics is and on where it’s going. Finally, I recommend Kit Fine’s “Towards a Theory of Part”. This is a recent game-changing article (that feels like a book) by today’s most influential metaphysician.
ABOUT THE INTERVIEWER
Richard Marshall is still biding his time.
Buy the book hereto keep him biding!