:: Article

The Impossible and The Real

Interview by Richard Marshall.

Mark Jago is currently working on what truth is; how we think about impossible things; what material objects are; how to think about essence; and what propositions are.

His major topic at the moment is about how we think meaningfully about impossible things. We might wonder whether some hypothesis is true or false, knowing full well that if it’s in fact false then it couldn’t possibly have been true. False mathematical claims represent impossibilities, for example, yet they are often meaningful. How come? Philosophical claims, too, are often held to be either necessarily true or necessariy false. So, quite probably, the majority of what most philosophers say (with their work hats on) is impossible—but it’s not all meaningless!

He’s also interested in philosophical curiosities such as holes, absences and negative facts. Most philosophers say that such things don’t literally exist.. He thinks they do. That’s why doughnuts aren’t just balls of dough, and that’s how claims like ‘there’s no milk in the fridge’ are grounded in reality. Such curiosities tell us something deep about our concept of existence.

He’s drawn to paradoxes—seemingly acceptable premises which lead to absurd conclusions. These lead to his interests in truth and vagueness. Far from being mere philosophers’ games, paradoxes are deep problems calling out for a philosophical solution. In the case of paradoxes about truth and vagueness, the paradoxes have had the best of it, so far. Here he discusses some of these things, including hypertensional contents, informativeness, how best to think about impossible worlds, why he rejects Meinongian beings that lack existence, how ertsatz worlds represent non-actual entities, bounded rationality and vagueness, grounding and fundamentality, essences, properties does the grounding relation have, what material objects are, what is fundamental to reality, and why heed the philosopher.

3:AM: What made you become a philosopher?

Mark Jago: I almost didn’t. I studied a very traditional analytic brand of philosophy at uni and it sometimes felt like cramming dry biscuits. I took a PhD in computer science, thrilled with the prospect of cutting-edge Artificial Intelligence research. One of the things I learned is that, if you pursue a question you find interesting and important, it doesn’t matter whether it’s labelled ‘philosophy’ or ‘science’ or whatever.

I ended up writing a PhD thesis on ‘Logics for Resource-Bounded Agents’, which turned out to be far less practical and far more conceptually involved that I’d thought. It got me thinking about questions like, how can we think meaningfully about something, if it turns out to have been impossible all along? Needless to say, these questions get a better hearing in the philosophy department! So that’s where I find myself.

3:AM: You’re an expert in impossible thoughts. Does your approach help answer the question whether there is a link between conceivability and possibility. On the face of it, it seems that if we can have impossible content then just because we can conceive of something doesn’t mean it is actually possible. Is that right?

MJ: We think about impossible things when we try to solve a sudoku, play chess, or discuss philosophy or religion. If we wonder whether it’s a ‘3’ or a ‘4’ in that box, or whether King to Q4 avoids the checkmate in three moves, we’re entertaining a thought that has a mathematically necessary answer. If the answer is ‘yes’, then ‘no’ is impossible. Yet we considered the impossible ‘no’ answer nevertheless. When a theist and an atheist disagree, they each express meaningful thoughts (at least some of the time) and they can’t both be right. But because of the subject matter, if they’re wrong, they’re necessarily wrong. That’s an impossible thought: a meaningful, contentful mental act that couldn’t possibly be true. Of course, we often don’t know that our impossible thoughts are impossible, otherwise we wouldn’t think them.
So we can think about something that isn’t a genuine possibility. And I think we can in principle imagine (or ‘conceive’) many of the things we think about (although perhaps many ideas are too complex). I can imagine being you, or being a disembodied spirit, but I don’t think that settles the issue. In general, imagining isn’t much help with possibility.

3:AM: What are hypertensional contents and what has this to do with informativeness?

MJ: ‘Hyperintensionality’ is something of a buzzword right now. Any word over four syllables runs the risk of putting people off an otherwise fascinating area. But, like inflation, Trump, and people putting cycles on the walls of hipster cafes, we seem to be stuck with it. Hyperintensionality enters the scene when you can’t replace one thing with an equivalent one, without anything else changing. Belief is a classic example: I may believe that 1+1=2 without believing Fermat’s last theorem, even though they are mathematically equivalent. It’s an important concept because meaning, knowledge and belief, and information — as well as technical concepts like truthmaking and grounding — are all hyperintensional.
Information covers many subject areas, including logic and mathematics. Sometimes, mathematicians discover something exciting and new.

They gain some information they didn’t have before. But what can that information consist in, if it was in some sense guaranteed all along by the axioms of their theory? That’s a question I tried to address in my book, The Impossible. The idea is quite simple: you gain information by ruling out some scenario. If you look out the window and see that it’s raining, that for you rules out the situation in which it’s not raining. Logical and mathematical information likewise consists in ruling out scenarios. The difference is that those scenarios were never genuinely possible in the first place, even though they may have seemed possible to you.

3:AM:  Why can’t possible world semantics provide fine-grained hyperintensional contents, why do we want them to and if they really can’t then why not abandon possible worlds semantics completely?

MJ: To continue the story: you may gain mathematical information by doing some reasoning. That reasoning rules out some would-be possibilities. Perhaps you thought that there’s no way to colour this map using just four colours so that adjacent countries never have the same colour. But the Four Colour Theorem tells you it can be done. Those ruled-out scenarios — like the one in which you need more than four colours to colour the map — aren’t genuine possibilities. They don’t correspond to any possible world. So, if we have only possible worlds in our toolkit, we can’t make any sense of logical or mathematical information. For the same reason, we can’t explain our logical and mathematical knowledge. There’s a difference between knowing a simple mathematical truth, that 1+1=2, and a complex one, like the Four Colour Theorem. But they’re both true in exactly the same possible worlds. If all you have are the possible worlds, you can’t explain the difference.

3:AM: So what kind of metaphysics do we need for this move to work? Are we left being committed to the existence of abstract objects – in this case, ersatz worlds? Does this link to your idea of propositions as truthmaker conditions, where you make metaphysical sense of merely possible states of affairs?

MJ: Impossible worlds are probably best thought of as abstract entities, since nothing genuinely real is impossible. That’s probably the best thing to say about possible worlds too, since nothing genuinely real is merely possible. So it’s not a special move you need to made for impossible worlds. In fact, I argue that worlds of all kinds are abstract constructions, in contrast to the reality we inhabit. In terms of their metaphysical construction, impossible worlds aren’t so different from possible worlds.
A really interesting question is whether Truthmaker Semantics, which is a very new idea in town, can do as good a job as the impossible worlds approach. It’s hard to say, since Truthmaker Semantics is such a new idea and the metaphysical details haven’t yet been worked out fully. It’s an approach I’m very interested in — as you say, it provides a very good account of propositions. But I think it has much more trouble accounting for our knowledge and beliefs, particularly about logical, mathematical, and philosophical matters.

3:AM: Why do you reject Meinongian beings that lack existence? Hasn’t Graham Priest given a pretty good account of how non-existing being might work? Aren’t they a sort of negative fact, something you seem to endorse in allowing for truthmaker maximalism?

MJ: When I think about what it means for something to exist, I don’t see a gap between existence, reality, being, or whatever you want to call it. When it comes to Pegasus, Santa Claus, and the Tory party’s concern for the poor, it’s not that there’s something out there in reality that somehow lacks the property of existence. There’s simply no such thing.

That’s just what my concept of existence tells me. Other people claim to have a different concept: that to exist is to be physical, or causal, for example. But numbers aren’t physical or causal and yet they exist. (Are there any numbers between 2 and 5? Yes! So there are numbers.)

That aside, I don’t see much philosophical utility in non-existent entities. If Pegasus doesn’t exist, then it isn’t part of our physical world. So it’s hard to think specifically about Pegasus (as opposed to the infinitely many other non-existent winged horses). And if giving Pegasus some kind of being doesn’t help explain how we think Pegasus-thoughts, it’s hard to see what the point of it is.

Negative facts are a tricky business. The few who believe in them — me included — say they exist, in just the same way regular things exist. Some are physical, occupying space: the wine-rack holes would be useless if they didn’t. Others have causal powers. What could explain current government policies other than an absence of empathy? Meinongian objects, by contrast, don’t exist, don’t take up physical space, and don’t cause anything. They may both be weird metaphysical categories, but they’re not the same kind of weirdness.

3:AM: Sider and Melia have accounts of how ertsatz worlds represent non-actual entities but you reject them. What’s your account and why is it better than theirs?

MJ: Their accounts are good at representing the ways things could have been. They’re not so good at representing how things couldn’t have been. One problem with representing a particular impossibility is that the impossibility might not respect logic. It might represent that p but not that q, even though p logically implies that q. Impossible worlds need to represent facts explicitly, by listing every little fact within the impossibility, one by one. And that’s hard to do on Sider’s approach, which tries to list all the possibilities in one go. We might then try to list all the impossibilities too, all in one go. But, it turns out, for technical reasons, that no sentence is big enough to do that. (The theory’s representations need to be possible, even if what they represent is impossible!)

3:AM:  What is bounded rationality, how does it connect with vagueness and why is it a problem for getting the fine-grained contents from ersatz worlds? Does your solution to the sorites solve the bounded rationality problem?

MJ: Bounded rationality is the state we find human reasoning in. We’re rational (or at least, we’re capable of rationality), but not in the way mathematical theories are. Some questions we find easy to think through, some very hard. Often when we’re problem solving, we run out of time or memory. Or we lose focus, get distracted, or just can’t be bothered anymore. Even the most powerful computers have their limits. That’s bounded rationality: rational problem solving, within physical constraints.

One problem associated with bounded rationality concerns knowledge. If you know anything at all, then you know some further things by inference. But you don’t know all the things that follow from what you know, because your rationality is bounded by physical constraints. So, given that you know something, what else must you know? Anything? It seems absurd to say that Anna knows that it’s cold and grey outside, but doesn’t know that it’s cold outside. Yet if knowledge obeys every trivial bit of reasoning like this, we’d know every consequence of what we know; and we clearly don’t know all of that. That’s what I call the problem of rational knowledge.

My response is this. We can’t say where knowledge by inference gives out, because it’s metaphysically indeterminate. And we mistake not being able to put our finger on a precise point with there not being any such point. For every trivial little inference, it seems absurd to say that someone knows the premises without the conclusion, even though there must be some such point where one’s knowledge-by-inference gives out.

The problem and this solution are parallel to the sorites paradox, the problem of the heap. If you take one grain from a heap of sand, is it still a heap? Of course! But then, it follows from logic alone that a single grain of sand is a heap, which is isn’t. So it’s false that taking a grain from a heap always results in a heap. The problem is to square this with the seeming absurdity of saying, for example, that 1,234 grains makes a heap whereas 1,233 doesn’t.

It’s indeterminate how many grains you need for a heap. Because it’s indeterminate, it’s unknowable. So we can’t ever rationally assert some instance of, ’n grains makes a heap, but n–1 doesn’t’ (even if it’s true). Because no instance can be rationally asserted, the general idea seems absurd. Incidentally, this line is compatible with many of the specific theories of vagueness. What goes for ‘heap’ also goes for ‘knows’, ‘believes’, ‘has the information that’, and the like.

3:AM:  This work links with your interest in grounding and fundamentality, in what grounds what. Can you sketch for us the contemporary landscape of this area by telling us about the big metaphysical questions that are central to the philosophy of grounding?

MJ: Philosophers have always been interesting in grounding, although not always under that name. They talk about the ways in which one thing or phenomenon depends on another; some talk about one thing being derivative upon another; some have even talked about one thing being more or less real than another. These are (probably) all ways of talking about the most important metaphysical connections we find between parts of reality. Some examples (which may or may not be true!): wholes depend on their parts; sets depend on their members; the mind depends on the brain; meaning depends on use; conjunctions depend on their conjuncts.

It’s one thing to say that the mind depends on the brain, or that brain activity grounds mental activity, but quite another to say precisely what that notion of grounding consists in. Is it a relation in the world, or merely a notion in our theories? What are its properties? What is its logic? Does the same notion of grounding link parts to wholes, members to sets, or conjuncts to conjunctions? Or are we talking about a family of distinct but similar grounding relations? If so, what (if anything) unites them?

3:AM:  Is there a fruitful link between the idea of essences, on the one hand, and fundamentality and grounding on the other? For you, is the essence of something a modal notion or is it tied strongly to its identity? Van Inwagen makes much of the difference between artifacts and living things in his ontology – do you think this distinction matters when we think about essences?

MJ: I hope there is! What makes things the things they are, and how things relate to and depend on others, are some of the most fundamental of all metaphysical questions. These are questions of essence, on the one hand, and grounding, on the other. If there’s a productive link between essence and grounding, then that’s good news for those who like tidy metaphysical theories.

3:AM:  What properties does the grounding relation have – is it an irreflexive, asymmetrical and transitive relation, as the orthodox view would claim or do you agree with Naomi Thompson that this view is false?

MJ: This is one of the hotly contested questions. There’s an orthodox view — which is that grounding is irreflexive, asymmetrical, and transitive — but I haven’t yet seen completely convincing arguments for (or against) that view. It’s up for grabs. There’s a great book coming out later this year, Reality and its Structure, edited by Rikki Bliss and Graham Priest, which discusses exactly this issue.

My view is that it depends on what you want grounding to do in your metaphysics. For me, it’s a key ingredient in a story in which reality is built in layers, from the fundamental building blocks of subatomic physics, through the chemical and the biological, to the experienced world in which we find the wonders of consciousness, moral agency, and terrible road movies. When you build, you take some stuff and construct something new. Nothing is built from itself; no amount of new building gets you back to the components you started with; and components of components are components. So my concept of grounding is irreflexive, asymmetric, and transitive. But others may legitimately have a different notion in mind which lacks these features. Rikki Bliss, Graham Priest, and Naomi Thompson all have interesting (and different) notions of grounding.

3:AM: What do you say material objects are, and how does your approach explain the modal and sortal properties of coincident objects. (I’m aware that there’s a bit of jargon in this question which readers will find helpful if you explain!)

MJ: Material objects — things in space and time — each have a nature, a way they are essentially. You and I are essentially human, essentially people, essentially conscious agents capable of rational thought and action in the causal realm. That’s what makes us the kind of thing we are. But we take different paths through the world, from different starting points, gaining different experiences along the way. That makes us different, but these differences aren’t essential to us. So I think of a material object as an essence — a bundle of properties defining what kind of thing the object is — taking a trip through time and space.

One consequence of this view is that I’m not the same thing as my body. We travel along together but, when I permanently cease to be conscious, I cease to be, whereas my body will continue for a bit. Nor is my mind identical to my brain, or an artwork identical to the physical stuff which constitutes it (although in both cases there are important grounding relations between them). I still think of minds and artworks as material objects, since they exist in space and time and are subject to cause and effect. It would be terrible if my decisions had no causal effect on my actions! But I needn’t be reductive or eliminativist about mental or moral agency.

One possibility I find very interesting here is that there may be no essential difference between any of us. Gender, race, and disability are very much social characteristics. They don’t define what a person essentially is. If we are essentially no different, that opens the possibility of a notion of contingent identity: I could have been you and vice versa. That’s not a popular view amongst logicians and metaphysicians and I’m not even sure we can make sense of it. But I’d much prefer to defend a theory on which we’re all fundamentally the same, rather than one on which we’re all isolated in our difference.

3:AM:  So what do you say is fundamental to reality?

MJ: I think we don’t know the specifics. We should believe what the best fundamental physics says, bearing in mind both that that’s a project that likely will never be complete and that, at any state of the investigation, the science is open to multiple philosophical interpretations.

In broad terms, there’s a category of property-like things (which I think of as universals); there’s some basic physical space of points (which might be space-time, or points of some other kind of ‘space’); and then there’s mathematical entities. Properties get together with these basic points to give us the fundamental physical facts of our world. Re-arrangements of them provide the basis for alternative possibilities.

Getting more specific requires more guesswork. Some properties (those like charge) might take a value at each of the points in a field-like way. Others might be geometric properties of the whole space of points. Some physical theories suggest that space and time emerge from more basic categories. Fine by me, although then I have even less of a clue how to characterise the fundamental physical stuff. Physics easily outdoes metaphysics for weirdness! Either way, I think that material objects aren’t fundamental entities. They emerge from other categories and probably at a higher level than that of spacetime.

3:AM:  Now some might turn round to philosophers like yourself and wonder why they need to heed you – after all, physicists and scientists of various stripes are all in the business of finding out what really exists. So why should we heed the philosophers?

MJ: Scientists might not need to, at least for the times they’re focusing on building scientific theories. But most of us aren’t scientists, and some of those who are are sometimes interested in other questions, which science doesn’t and shouldn’t try to answer.

Both general relativity and quantum theory rely on concepts like mass and time. Specifically, they rely on the functional properties of those concepts. Whether time is or isn’t mind-dependent, and whether there is a universal of mass over and above all the particular instances of it, doesn’t affect the science. They’re metaphysical questions.

Ultimately, metaphysical questions seek answers which satisfy our craving to understand the world around us, both as we experience it and as it appears in scientific theories. Of course, you can get on in life quite well without ever asking those questions. It might even make things simpler if you never do! You also don’t need good music or literature to get on in life. It’s just that life is so much richer with them there.

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

MJ:

One of the best I’ve read recently is Potentiality: From Dispositions to Modality by Barbera Vetter (OUP, 2015). She starts out with the simple idea that objects have dispositions or potentialities, like the way salt tends to dissolve in water. And from that, she builds an entire theory of metaphysical modality. It’s very clever and a very good read.


Another important recent contribution to metaphysics is Karen Bennett’s Making Things Up (OUP, 2017). It’s about how certain phenomena are arise, or are built, out of others. How do aesthetic, mental, or moral properties arise from basic physical ones (if they do at all)? How do we get from facts about how the world is to facts about how it must be, or could have been? The way Bennett conceptualises these questions, in terms of building relations, is very appealing. And although the topic can be technical, Bennett’s writing is good fun.

The Structure of Objects by Kathrin Koslicki (OUP, 2008) tells a story about what material objects are. It brings together ideas from Aristotle, modern mereology (the part-whole relation), and common sense ideas of what is and what isn’t an object. This commitment to a common sense ontology differentiates Koslicki’s approach from many others working in contemporary mereology (me included). I disagree with a lot of Koslicki’s conclusions but I still find the book very persuasive, partly because she uses such good examples to illustrate her arguments.


One of the most influential books on modality since David Lewis’s On the Plurality of Worlds is Penelope Mackie’s How Things Might Have Been: Individuals, Kinds, and Essential Properties (OUP, 2006). I’m also lucky to have Penelope as a colleague and sometime collaborator.

Sally Haslanger’s Resisting Reality: Social Construction and Social Critique (OUP, 2012) is an important study of social phenomena, like gender and race, as social constructions. This is an important area in which metaphysics can step beyond its traditional boundaries. The notion of construction in social construction is, after all, a metaphysical notion. Understanding what it is to be a disabled person, or a trans woman, are questions to which metaphysicians, with their concepts of grounding and building, can fruitfully contribute.

ABOUT THE INTERVIEWER

Richard Marshall is still biding his time.

Buy his new book here or his first book here to keep him biding!

End Times Series: the first 302

First published in 3:AM Magazine: Saturday, June 9th, 2018.