on the reality of sherlock holmes etc
Amie Lynn Thomasson interviewed by Richard Marshall.
Amie Lynn Thomasson is a funky philosopher of fictionalism, phenomenology and ordinary things. She is always thinking about Moore’s arguments , about common sense, tables, when May was born, whether science and ordinary objects are rivals, the limits of metaphysics, what fictions are, why Sherlock Holmes is as real as a number, about links in phenomenology crossing the so-called Analytic-Continental divide, and what benefits this approach brings. All in all, boss.
3:AM: What made you become a philosopher? Were you always that way inclined or did something happen?
Amie L. Thomasson: I guess I was always that way inclined. I can remember driving my mother crazy with all of my ‘why’ questions when I was four. I didn’t know what philosophy was until I got to college, though. I signed up for an introductory course in ‘metaphysics and epistemology’ my first semester—I didn’t know what either of those words meant, but the description sounded intriguing. It was a small seminar, and several of us met after every session for lunch and continued discussion. There was no looking back from there.
3:AM: You wrote ‘Ordinary Objects’ because until then you didn’t think anyone had taken on all the arguments denying that ordinary things exist thoroughly enough. Why didn’t you think GE Moore’s response to these eliminativist skeptics – that we’re more certain about trees existing than any philosophical theory being true – was good enough?
AT: I guess there are two reasons. One is that, even if Moore’s argument gives us good reason prima facie to accept the existence of ordinary objects, we still might have an obligation to respond to challenges to that claim. A lot of new arguments against ordinary objects have been raised since Moore’s time, and many of those claim to show that the very concepts involved lead to contradiction—so to answer those challenges we need to do the detailed work of showing why the concept doesn’t really lead to contradiction, or where the eliminativist’s arguments go wrong. The other reason is that Moore’s response only gives us reason to suspect that something has gone wrong somewhere in the eliminativist’s arguments. It is more satisfying and far more informative if we can figure out exactly what has gone wrong. What was interesting to me as I did the work on Ordinary Objects was the way in which it turned out that many of the arguments against ordinary objects go wrong in a similar way. By working out exactly where those arguments go wrong I think we can learn a lot that we don’t get just from Moore’s argument: about modality, meaning, and the methods of metaphysics. Those are the real lessons I think we can take away from this work—even if we started out already believing that ordinary things exist (and finish there, too).
3:AM: It’s important to you to make philosophical sense of our common sense notions isn’t it?
AT: Let me put it this way: I think we should be suspicious of any philosopher who claims to have discovered that common sense is radically mistaken. (Things are quite different if there is empirical evidence against some common sense belief.) We can draw a distinction between adhering to common sense beliefs (or folk theories) and employing common sense concepts. I think we have to make use of the common sense concepts we have—unless we are explicitly and clearly engaged in suggesting new or revised concepts. Many of the ‘radical’ non-commonsensical claims we get from metaphysicians I think come only from misusing our common sense concepts, or tacitly revising them. But then—if you have implicitly changed the ordinary concept—you are not getting conclusions about tables and chairs (claims that would contradict something said using our ordinary concept). As a result, it is very misleading to characterize such revisionary work as overturning common sense views that are expressed using the ordinary concepts.
3:AM: Can you give us an overview of some of the best reasons why someone might conclude that ordinary objects exist?
AT: The reasons for thinking that ordinary objects exist are actually very simple: if we know how to use the concept table, then we can just take a look in a restaurant, see that the application conditions for the concept table are fulfilled, and conclude that there are tables. If an eliminativist denies that there are tables, yet accepts that in a certain situation there are ‘particles arranged tablewise’—where this is supposed to mean that there are particles arranged by an artisan or factory, so that they are in the shape of what we thought was a table, and so that the particles can ‘jointly’ do all the things we thought tables did—then again, those with conceptual competence are perfectly entitled to make a trivial inference from ‘there are particles arranged tablewise’ to ‘there is a table’, and conclude that there is a table.
Since the book came out, I’ve also been looking at other ‘easy’ arguments for entities of various sorts. I think, for example, that one may conclude that there are events by a trivial inference from “May was born on a Monday” to “May’s birth occurred on a Monday” to “There are births (and other events)”. In short, I think most disputed ontological questions can actually be answered by trivial inferences from undisputed truths.
So the arguments for the existence of ordinary objects (and many other things) are very easy. The hard part is defending that ‘easy’ methodology (which I am working on in my new book on existence questions) and showing where the eliminativists’ arguments go wrong (which I worked on in Ordinary Objects). For although the arguments are easy, they are open to challenge—and when they are challenged we have work to do in showing where those challenges go wrong.
3:AM: You think that these arguments have much in common, which is a happy circumstance isn’t it because you’re then able to show that defeat of one will go some way to defeating them all. Is that right?
AT: Yes, in a sense, though I take a somewhat different rhetorical strategy. I try to show how the eliminativist arguments rely on certain common presuppositions about reference, modality, and the methods of metaphysics. Then, if we can avoid all of those eliminativist arguments by giving up those presuppositions, that gives some reason (for those who agree that eliminativist arguments must have gone wrong somewhere) to think we really should give up those presuppositions. So I don’t expect a knockdown defeat of a single argument, and then try to use that against the others. Rather, I offer a unified diagnosis of where all of the diverse arguments might be thought to go wrong, and the plausibility of the diagnosis is increased by its ability to apply to so many different arguments.
3:AM: So how do you defend the common sense notion of ordinary objects?
AT: There’s no substitute here for the detailed work done in defending the view that ordinary objects exist against each of the different arguments that have been raised against them. These include arguments that ordinary objects would be causally redundant, would violate metaphysical principles against co-location (for example), or would run into trouble with Sorites arguments (given the vagueness of the concepts involved). They also include arguments that accepting such objects would run afoul of the demand for parsimony, the need to accept a scientific ontology, or the demand to find a clear and consistent answer to the special composition question. I’ll have to leave the detailed work of showing how to defend ordinary objects against each of these arguments to the book, since it would take too long here.
3:AM: I guess much of this is an important consideration when discussing the implications of scientific discoveries where fundamental reality seems so strange and make common sense ontologies suspect. You don’t just want to argue that there is no conflict between common sense and science, you want to provide a constructive argument for ordinary objects don’t you? Would this be a stronger position because it would mean science had to make sure they didn’t stray too far from common sense that they denied ordinary objects?
AT: I am mostly providing a defense of ordinary objects against arguments to the contrary, but I do also offer a (very simple) argument for the existence of ordinary objects. I suppose this does mean that science mustn’t deny the existence of ordinary objects, but I don’t think of that as putting any kind of constraint on science (though it may perhaps put a constraint on what scientists or philosophers say follows from the scientific results). I don’t think that sciences like physics typically use ordinary terms and concepts at all—they work within their own conceptual system—and I argue in Ordinary Objects that classic arguments fail to show that science and common sense are rivals. In short, I don’t think the physical sciences typically concern whether or not there are ordinary objects at all.There will be no danger of constraining the physical sciences if this is a question they don’t address.
3:AM: All this is part of the ‘what exists’ philosophical question that is one of the fundamental questions for metaphysics throughout the history of philosophy. What do you think are the methods and limits of metaphysics?
AT: That’s the core question I have been working on since Ordinary Objects came out. Basically, I am defending a view that was popular in the early twentieth century, among philosophers as diverse as Husserl, Carnap, Ryle, and Wittgenstein: that philosophy and the sciences have distinctive and complementary roles: that the work of the sciences is empirical, while that of philosophy is conceptual. This is a point of view that has been largely lost in post-Quinean metaphysics, but which I think has far more going for it than the alternative. So I think of metaphysics as capable of doing conceptual work (and generally undertaking this in the object language): work in determining the relations among our concepts, what follows from them, detecting problems or inconsistencies, and perhaps also (as Carnap would have had it) in conceptual engineering—building new conceptual systems as needed to serve different purposes. But I don’t think of metaphysics as properly engaged in quasi-scientific work of discovering what ‘really’ exists or what the ‘real’ natures of things are.
3:AM: There’s been a recent dispute between philosophers and scientists regarding ‘nothingness’ which seems to highlight a big misunderstanding about what philosophy and metaphysics is. Do you agree with this diagnosis that maybe scientists and the broader culture should be more philosophically sophisticated, or was Feynman right when he said philosophy is as useful to science as ornithology is to a bird?
AT: I think of philosophers as something like conceptual specialists—we think more than most people about how concepts are and are not to be used, about what inferences are and are not appropriate, and the like. Sometimes natural scientists, psychologists, and others will draw misleading or inappropriate ‘philosophical’ conclusions from their data—and philosophy can sometimes be useful in exposing mistakes of that sort. So I don’t necessarily think that scientists need to know a lot of philosophy, but rather that anyone drawing grand conclusions needs to be careful about the inferences they are making and about the proper use of the concepts they employ.
3:AM: You’ve written a deal about the role of fiction in metaphysics. Your original move is to stop treating fictional creatures as ‘ odd freakish entities, quite unlike common or garden objects…’ and to recognize how similar they are to other entities. You treat them as a combination of abstract entities and creations don’t you? Why did you take this stance? And who doesn’t, and why?
AT: I really begin from following out what our ordinary conception of fictional characters commits us to their being like. (So this is again a matter of starting by analyzing our ordinary concept.) We think of a character like Sherlock Holmes as having been created by an author at a certain time—in that respect, as requiring intentional creation, and only coming into existence at a certain time—fictional characters are like artifacts. But also, pretty clearly, Sherlock Holmes doesn’t have any spatial location—you can’t, and never could, find him around London or anywhere else. In that respect, he seems to be abstract. So I have characterized fictional characters as abstract artifacts.
But historically, since Plato, most philosophers had assumed a simple dichotomy: there are concrete, contingent, created entities in space and time (such as individual tables, statues, or rocks); and abstract eternal necessary entities that exist outside of space and time (such as numbers or properties—Platonistically conceived). So there just wasn’t a standard category available to accommodate fictional characters.
Yet the problem is quite general: the problems that arise for categorizing fictional characters are much the same as those that arise for categorizing novels, stories and poems, symphonies, laws of state, and many other abstract cultural objects. So I argued that we really need to recognize this additional category of abstract artifacts to make sense of them all.
Many people reject this stance for various reasons: the most popular view (held by people like Russell, Quine, Ryle, Walton and Sainsbury, among many others) has been to deny that there are any fictional characters at all, in which case we don’t need a category of abstract artifacts to accommodate them. Some, like Terence Parsons, defended the Meinongian view that fictional characters are nonexistent objects (which there are although they do not exist, and are not created). Others (such as Edward Zalta and Nicholas Wolterstorff) took fictional characters to be (non-created) abstract objects or person-kinds individuated (roughly) by the properties they are ascribed in the fiction. Those realist views were developed before mine, so they weren’t rejecting my proposal.
But some later writers have found views that take fictional characters to be platonistic (non-created) abstracta more attractive than an artifactualist view on ontological grounds—thinking that abstract entities ‘just can’t’ be created, or that it is more parsimonious to do without a new category of abstract artifacts. In addition to these ontological concerns, there is a lot of room for discussion about which view—an anti-realist view of some stripe, an abstract artifact view, or a Platonist view of some stripe—provides the best analysis of fictional discourse in its many forms. Some important criticisms of my view have been written by people like Anthony Everett, Mark Sainsbury, Takashi Yagisawa, and Stuart Brock. I have tried to address some of these in detail in articles that have come out after Fiction and Metaphysics.
3:AM: So does your approach put ‘Sherlock Holmes’ in the same category as numbers, laws, theories, songs and stories? Can you explain how this works? Is Sherlock Holmes as real as the number twelve?
AT: Sherlock Holmes is as real as the number twelve—I don’t distinguish levels or modes of reality, and I think we should be simple realists about all of these things. Nonetheless, I don’t think fictional characters are in the same category as numbers—though they should probably be placed in the same category as law (of state—not nature), theories, songs and stories.
3:AM: What kinds of puzzles and paradoxes about fictional characters and worlds do you address? And how does your approach deal with them?
AT: Let’s see, there are a lot—many of which I discuss in Fiction and Metaphysics and in several later papers on fictional discourse (in some cases revising or updating my original treatment). As I argued later (in “Speaking of Fictional Characters”) there are a lot of tensions in the ways we talk about fictional characters, so any view has to give up appearances somewhere. We want to say, for example, that Frankenstein’s monster was a creation of Dr. Frankenstein, but also that he was a character created by Mary Shelley; or that Sherlock Holmes is a detective, but also that he is a fictional character who cannot be called on to solve mysteries. These are relatively simple puzzles, on my view, which can be handled by treating the first clause in each as implicitly in the context of an ‘according to the story’ (or, if you prefer, ‘according to the pretense’) operator, and treating the second as literally true. Traditional eliminativist views have to treat both clauses as either to be paraphrased or as involving pretense.
Of course some puzzles are more difficult, e.g. how can a claim like ‘Sherlock Holmes doesn’t exist, he is just a fictional character’ be true? It sure seems like the sort of thing we would ordinarily say (to some confused person who aimed to hire Holmes to solve a mystery). Yet if he doesn’t exist, how can we go on to say he is a fictional character? I discuss this in my later paper “Fiction, Existence and Indeterminacy” (where I revise some of what I said earlier about nonexistence claims involving fictional names, which present some of the most difficult cases for a theory like mine to handle). I’ll leave the details to that paper, since it would take a lot of space to explicate and argue for here.
3:AM: You say that your approach takes fiction ‘out of the metaphysical backwaters’. But you might also be considered as someone taking fiction out of its own backwaters, where metafiction and metaphysical fiction have perhaps been disregarded or underappreciated by philosophers. Are ideas of fiction writers important to you, or has your work been done primarily from with the framework of philosophical metaphysics?
AT: As an undergraduate I did a second major in English, so I came to this topic when I was in graduate school partially motivated by the desire to understand what sorts of things literary critics, historians and interpreters were talking about—which I think philosophers often fail to have in view. But to motivate the project in the historical context really required situating the position in the philosophy of language and metaphysics, so I end up speaking far more to philosophers’ concerns than those of the literary scholar, and I’m afraid my work may be of little interest to fiction writers or literary scholars.
Nonetheless, there are some works of fiction that have been important in these discussions. For example Jorge Luis Borges’ story “Pierre Menard, Author of the Quixote” raises lots of interesting questions about the identity conditions for works of literature, and has generated a lot of discussion in the philosophical literature.
AT: You’ve taken an original stance in yet another field of philosophy when in your editorial for the book on phenomenology you argue, with your co-editor David Woodruff Smith, that despite common assumptions, there really is a continuous link between the tradition running from Brentano, Husserl, Heidegger, Sartre, Merleau-Ponty and the so-called analytic tradition’s work on the philosophy of mind. Can you say why it’s wrong to think of these as being radically separate? After all, it seems difficult to initially see Patricia Churchland’s eliminativism as speaking to the same issues as Jean Paul Sartre, or Jerry Fodor’s LOT speaking to Heideggerian ideas.
AT: There are plenty of differences among individual philosophers, of course. But it is far easier to see Gilbert Ryle as speaking to some of the same issues as Husserl and Heidegger, and indeed drawing on their work on category differences, and knowing how versus knowing that. (I have a paper on this.) And it is easy to see Wilfred Sellars as taking a similar approach to Husserl’s on knowledge of our own mental states, or to see Sartre as speaking to contemporary issues about whether consciousness involves self-consciousness, etc. And it is easy to see Carnap as taking something like the same position as Husserl’s on the problem with the traditional debate between realism and idealism. And so on. So I think we do better in both appreciating the history of the subject and understanding the logical space of views available if we give up the idea that there is a great divide, and see some of the influences and commonalities that reach across what came to be thought of as two separate traditions.
3:AM: What is gained (on top of historical accuracy) by seeing greater unity between the approaches of these diverse thinkers?
AT: I think we can often understand the different positions available better by seeing where they have come from. For example, Ryle is often taught and thought of as a behaviorist (in The Concept of Mind). But unearthing the roots of his approach in the phenomenological tradition makes it clear that that is a complete misreading: he aims to ‘rectify the logical geography’ of our concepts, not to offer a reductive understanding of the mental in terms of the behavioral. Often, too, by seeing different ways in which a position is developed and expressed in the history of philosophy, we can better assess its plausibility: for example, I think the approach to self-knowledge developed by Husserl (in developing phenomenological method) and Sartre resembles in various ways the approach developed by recent philosophers such as Dretske and Tye—but in seeing its historical roots we can also see that the view needn’t be motivated by, or tied to, reductive ambitions. (I draw this line out in my paper “Phenomenal Consciousness and the Phenomenal World”).
3:AM: And finally, if we at 3:AM were wanting to go further, which five books (not counting your own which of course we’ll be dashing away to read straight away) would you recommend?
AT: That’s a hard one, since it depends on the topic, and also on whether you hope to understand the history of a debate or look to where I think the most progress can be made on it. On fiction, an excellent overview that has come out since my book did (and reaches different conclusions) is Mark Sainsbury’s Fiction and Fictionalism. Anyone wanting a primer on phenomenology (which also connects it to issues in analytic philosophy) would do well to start from David Woodruff Smith’s excellent book Husserl. More broadly, I think that many of these debates—about whether we should accept fictional characters, ordinary objects, or other things—end up resolving into debates about the methods of answering existence questions, and more generally about the status and methods of ontology. So that is where I think things have to turn next (and what I am working on next, in a pair of book manuscripts: one on existence questions, the other on modal questions). The classic source for a kind of deflationary approach to ontological debates of course is Carnap’s paper “Empiricism, Semantics and Ontology” (so I don’t have to count it against my five books, right?). There’s no better introduction to the recent debates on metaphysical questions than the Metametaphysics anthology edited by David Chalmers, David Manley, and Ryan Wasserman. The approach I favor to ontological questions, which I call the ‘easy’ approach to ontology, is developed and defended in Stephen Schiffer’s The Things we Mean. Finally, those interested in seeing how one may adopt a quietest approach that enables us to eschew many traditional metaphysical debates should read Huw Price’s Naturalism without Mirrors.
ABOUT THE INTERVIEWER
Richard Marshall is still biding his time.
First published in 3:AM Magazine: Friday, June 28th, 2013.