Skip to content

David Lewis: Expressivism is Fictionalism

October 19, 2010

In his essay “Quasi-Realism is Fictionalism” (in Kalderon 2005), David Lewis argues that… well, that quasi-realism, Simon Blackburn’s particular flavor of moral expressivism, is in fact a version of moral fictionalism.

Lewis grants, for the sake of argument, that Blackburn’s quasi-realism succeeds on its own terms–that is, that the quasi-realist succeeds in earning back the right to say everything that the realist says (that action X is wrong, that it is true that action X is wrong, that if action X is wrong, action Y is wrong, and so on–our entire realist-sounding moral discourse). On that assumption, Lewis asks, what distinguishes quasi-realism as a distinct position from realism at all? By stipulation, there is nothing that the realist wants to say that the quasi-realist will not. So Lewis goes looking for something that the quasi-realist wants to say that the realist will not. And he claims to find such a thing–namely, a fictionalist preface.

Lewis distinguishes between fictionalist prefixes and fictionalist prefaces by listing some examples–“According to the Sherlock Holmes stories…” is a paradigmatic prefix, while “Let’s make believe the Holmes stories are true, though they aren’t,” is a paradigmatic preface. Lewis articulates the distinction this way: a prefix cancels the assertoric force of what follows it, but replaces that force with the force of another kind of assertion. The prefix “Acccording to the Sherlock Holmes stories…” cancels the assertoric force of what follows and replaces it with the force of, as it happens, another assertion–but an assertion about the content of the Holmes stories. A preface, by contrast, cancels the assertoric force of whatever follows, but doesn’t suggest what kind of force it will have. If this sounds familiar, it should. I think it is not uncharitable to take Lewis to be making the same distinction I laid out (via Joyce) in the previous post, between talking about a story and telling it–in Selim Berker’s vocabulary, between tacit operator fictionalism and force modifier fictionalism. Lewisian prefixes are ways of talking about the contents of a story. Lewisian prefaces are introductions to telling one.*

In any case, Lewis thinks that Blackburn’s endorsement of projectivism–his first step in dealing with moral nihilism, which is to say that rather than being sensitive to moral properties in the world, we project our moral attitudes onto the world–serves as a fictionalist preface, a “Once upon a time…” for the quasi-realist reconstruction of moral discourse that follows. Projectivism is the thing that quasi-realists will say and that realists will not. His very quick formal argument is as follows:

  • Blackburn doesn’t just happen to earn the right to echo everything that the moral realist says–it is in fact one of his stated goals.
  • That means he either wants to be, or to pretend to be, a realist.
  • Blackburn also says his goal is to avoid the errors of realism.
  • That means he does not want to be a realist.
  • So Blackburn wants to pretend to be a realist.

On Lewis’ view, Blackburn’s articulation of projectivism–that is, his initial inversion of the basic tenets of moral realism, related to the third point above–is what robs otherwise straightforward moral claims of their assertoric force and turns them into quasi-assertions.

Later–perhaps tomorrow morning–I’ll go through Blackburn’s response, and also C.S. Jenkins’ refereeing of the dispute. But for now I just want to clarify that while Lewis has singled out Blackburn’s brand of expressivism for attention, in principle his argument should apply to any expressivist project, as long as the expressivist (1) is motivated by a desire to avoid realism’s errors and (2) attempts to earn back the right to use all of our everyday moral discourse. For tomorrow–who’s right? And can you run the argument the other way–that is, is fictionalism reducible to expressivism?

* I think it’s possible that Lewis is ignoring a point that Joyce makes–Joyce explicitly says that one shouldn’t interpret non-cognitivists (read: expressivists) to be arguing that moral sentences include a tacit “Let’s pretend that…” clause (2005, p. 296). I’m not convinced that this applies to Lewisian prefaces, but it might. I’ll take this up later.

Introducing Fictionalism (Take Two)

October 19, 2010

After sleeping on yesterday’s post, I’ve decided that it wasn’t just putting the cart before the horse, but in fact was pretty pointless. After all, nothing about my topic depends on moral fictionalism being true–the issue at hand is simply whether fictionalism and expressivism are viable, coherent and distinct metaethical options. So I’m going to start over, and instead of running the argument for fictionalism, simply try to pin down exactly what fictionalism is claiming about moral beliefs and discourse. (As before, I’ll be following Richard Joyce’s 2005 essay “Moral Fictionalism,” with some help from Selim Berker’s excellent lecture handouts.)

So to recap: moral fictionalism begins from the premise that moral nihilism is true–that is, that there are no objective moral facts, and nothing that we are required, morally speaking, to do or refrain from doing. To the question “What now?”, the moral fictionalist says that we ought to hang on to our moral judgments, because they’re practically useful, but that we should give them the status of fiction. When we talk about what is right and wrong, the fictionalist says, we should think of ourselves as engaging in an extended session of make-believe. The question, then, is how we should understand our moral beliefs and moral discourse once we see them in this new light. The fictionalist has to make some basic choices about his position, and running through them will help us pin down exactly what it is to have a fictive belief, or to make a fictive assertion.

First, the fictionalist has to decide whether or not his view is descriptive or prescriptive–that is, whether fictionalism is an attempt to describe how we have been thinking and talking about morality all along, or whether it is arguing that we should revise our understanding of our moral beliefs and discourse and move forward with a newly fictionalized understanding of morality. The former option is usually called hermeneutic fictionalism, and the latter is revolutionary fictionalism. Joyce is a revolutionary fictionalist, which gets him out of having to argue that even the most committed moral realists (religious fundamentalists, say) have only fictive moral beliefs, even if they insist wholeheartedly that their beliefs are sincere. But arguments for both strands exist, and the difference between them might have some bearing on the comparison with moral expressivism. (Expressivists, for instance, face roughly the same choice, but tend to come down on the hermeneutic side–that is, they tend to see themselves as giving an alternate account of what we should be taking ourselves to have been doing all along.)

Another choice concerns exactly how to understand what is going on, in the fictionalist picture, when someone makes an assertion about a putative moral fact. One option is to hold that any fictionalist sentence includes an unspoken clause that changes the content of the assertion–something along the lines of “According to the moral fiction…”. On this view, someone who says that wanton cruelty is wrong is in fact sincerely asserting the proposition that according to the moral fiction, wanton cruelty is wrong.

As Joyce points out, there are real problems with this characterization. For one thing, it elides the meaningful distinction between “describing the story and telling it” (291). Describing a story involves making sincere assertions about the contents of the fiction. Telling a story involves pretending “that we are a person who has access to a realm of facts that we are reporting,” that is, it is a different speech act entirely (291). The distinction can be illustrated thus: someone telling a story about Sherlocke Holmes, when pressed, will “emerge” from the fiction and admit that it is not literally true that Holmes lived at 221B Baker Street. But if what she was asserting was “According to the story, Holmes lived at 221B Baker Street,” it should always be true, even from a perspective that is external to the fiction.

(There’s another worry, which is that adding a tacit clause to every moral sentence means that they don’t embed correctly in complex contexts, and perform strangely in logical arguments. This is complicated and I won’t get into it now, but it’s interesting in that it’s a version of the Frege-Geach problem, the main argument against moral expressivism.)

Joyce’s account takes the distinction between describing and telling a story seriously, and so he argues that when the fictionalist uses moral discourse, he’s not talking about the contents of the moral fiction but rather actively engaging in it. That is, he is pretending to assert moral truths. In everyday life, in the grip of this fiction, he puts forward moral propositions as something he believes, and as something his audience should believe, but in more “critical contexts” (including the philosophy classroom), he is able to assume an external perspective and admit that he’s not actually serious.

In my last post I summarized Joyce’s argument for fictionalism. The idea is that sincere moral beliefs are useful in that they help us coordinate our actions and emotions–that is, they help us live in the world together, and give us psychological incentives (such as guilt) to cooperate with others. Fictive moral beliefs, he argues, can give us the same practical benefits, as bulwarks against weakness of will and temptation to lie, steal, and cheat. Joyce sees moral beliefs as precommitments, ways of “binding ourselves to the mast,” as Jon Elster puts it in his book Ulysses and the Sirens. This practical value, on Joyce’s account, is reason enough to maintain moral discourse as a fiction, rather than jettisoning it entirely in the face of moral nihilism.

There’s one other element of Joyce’s fictionalism that’s worth mentioning. Joyce thinks fictionalism is most fruitfully thought of as being a strategy that a group of people use to deal with a problematic discourse. If one imagines a single fictionalist in a world of sincere moral realists, Joyce acknowledges that there’s an open question about whether the fictionalist and the realists can really communicate at all, or only talk past each other. Even if you think they can (since, as a practical matter, everyone would understand each other, and the realists might never even realist that there is a fictionalist in their midst), you might wonder whether the fictionalist, in pretending to have moral beliefs and pretending to assert them to an audience of realists, is being disingenuous, or even whether his fictionalism can get off the ground at all. Joyce’s response is that it’s better to think of fictionalism as something adopted by a community of speakers. He compares fictionalism to sarcasm, another instance in which we withdraw or warp the apparent assertoric nature of our claims. A single sarcastic person in an otherwise irony-free world runs into the same set of problems as the lone fictionalist. But that doesn’t mean that a community can’t successfully establish conventions and norms of sarcasm.

So there’s my not-so-brief introduction to fictionalism. I still haven’t said much about how it relates to expressivism, but that’s to come. In my next posts I will briefly (briefly, I promise!) summarize two arguments: that expressivism is reducible to fictionalism (David Lewis), and that fictionalism is reducible to expressivism (Matthew Chrisman). Until next time.

Fictionalism – A Brief Introduction

October 18, 2010

As a reminder, my first paper topic is this: is moral fictionalism in some way reducible to moral expressivism? or vice versa?

To start, I’m going to grind through some exposition, largely because I’m having quite a bit of trouble articulating my own views on the subject, and this lets me put off having to stake out a position and simultaneously will serve as some good groundwork for when I do have to take a side. (Productive procrastination is the best kind…) So, what is moral fictionalism?

To start more broadly: to be a fictionalist about some domain is to believe that we are deeply mistaken about it (all our sentences about it, and maybe even all our beliefs about it, are false), but that we should continue talking about it as we do because it is a useful fiction. You can be a fictionalist about anything, really, and in fact various philosophers have argued for fictionalism about mathematics, possible worlds, events, colors, and religion, to name a few. I’m going to be focusing on what it is to be a fictionalist about morality, and so when I say “fictionalist” in what follows, that’s what I mean.

As introduced by Richard Joyce (2005), moral fictionalism starts from the same premise as expressivism. (This is, as we’ll see, just one of the many formal parallels between the positions.) Fictionalists assume that moral nihilism is true, and thus that there is no such thing as objective, out-there-in-the-world right and wrong. In Kantian terms, there are no categorical imperatives, nothing that we must do irrespective of our beliefs, desires, or inclinations. The question that fictionalists (and some expressivists) are interested in is: what now? What do we do with the countless everyday ways in which we think and speak as though there were moral facts? Joyce lays out four options:

  • Conservativism: keep going just as before–hang on to all our moral beliefs and discourse.
  • Eliminativism: get rid of our moral beliefs and stop talking as though there are moral facts.
  • Propagandism: admit in private that moral nihilism is true, but keep it a secret so that the general public continues on unaware.
  • Fictionalism: give up the sincerity of our moral beliefs and discourse, but continue to have fictive beliefs and make fictive assertions about moral facts, because participating in such a fiction is useful somehow.

Joyce dismisses conservativism and propagandism without giving them much attention. Even assuming we could hold onto our moral beliefs, Joyce thinks that hanging on to so many false beliefs would come with a prohibitive cost. (Would we have to somehow forget the argument for nihilism? Would we have to simply mark off whole areas of life as unfit for critical inquiry?) Promoting a conspiracy in which the philosophical elite manipulates and lies to the rest of the general public would be, for obvious reasons, even worse.

The main competition, then, is between fictionalism and eliminativism. Why keep moral beliefs and discourse around at all, if they are fundementally in error? Joyce’s strategy is to try to identify the value of our moral beliefs and discourse, in order to figure out what would be lost by eliminating them, and to see if maintaining a moral fiction would preserve some practical benefits, if not completely then at least in part.

In a move that should seem familiar to readers of Gibbard and other expressivists, Joyce argues that we have strong pragmatic reasons to act cooperatively, i.e. to coordinate our actions and emotions in order to live together in the world. One benefit of moral beliefs, he argues, is that they act as bulwarks against weakness of will.

When a person believes that the valued action is morally required—that it must be performed whether he likes it or not—then the possibilities for rationalization diminish. … The distinctive value of categorical imperatives is that they silence calculation, which is a valuable thing when interfering forces can so easily hijack our prudential calculations. In this manner, moral beliefs function to bolster self-control against practical irrationality.” (301)

Joyce argues that this is a benefit that is preserved, at least to some extent, when one pretends that one has moral obligations. His example is of doing sit-ups. It is likely true that doing roughly fifty sit-ups almost every day would have essentially the same benefit to my health as doing exactly fifty sit-ups every day. “But by allowing myself the occasional lapse,” Joyce writes, “by giving myself permission sometimes to stray from the routine, I pave the way for akratic sabotage of my calculations–I threaten even my doing more-or-less fifty sit-ups on most days.” Joyce argues that it is better, as a practical matter, to convince himself that he must, in a non-negotiable sense, do exactly fifty sit-ups every day. There’s no need to be wedded to these beliefs (or as he puts it, to “believe in these thoughts”). When not doing sit-ups, Joyce is happy to acknowledge that doing forty-nine sit-ups would get him into just as good shape. But while exercising, saying to himself “must… do… fifty!” is at least a partial defense against weakness of will.

That, then, is what Joyce is proposing we do with morality. In the philosophy classroom and other critical contexts, we should avow our sincere beliefs–that categorical imperatives do not exist. But in everyday life, we should pretend that they do, in order to reap the practical benefits they have for our ability to get along with others and resist the temptation to lie, cheat, and steal.

There’s much more to be said about what it’s like to be a fictionalist in practice. In fact as I write this I’m realizing that I’ve sort of put the cart before the horse. I’ve run through Joyce’s argument for choosing fictionalism from among the other alternatives without laying out his account of fictionalist psychology (what is it to have a fictive belief?) and fictionalist speech act theory (what does it mean to fictively assert something?). The answers to those questions will start to make clear why fictionalism and expressivism are so intimately related, and how we should understand that relationship. To be continued…

Hello Again

October 13, 2010


Where I'll be living for the next year.


So here we are, almost a year later. I’m back at a new school, working on a new degree, and still working on issues in metaethics and the philosophy of language, with a focus on speech acts, non-cognitivism, intersubjectivity, and truth. I’ll be writing three papers and a dissertation over the course of the year, and I’m going to try once again to blog as a way of forcing myself to write as I go along. First up: is moral expressivism a kind of moral fictionalism? Stay tuned.

Illocution-based Humor

October 14, 2009

It’s been a while. In fact who knows who’s still checking this thing. But I found this, and couldn’t resist. If only apologies and declarations of romantic intent were this easy.

The Beginning of the End

August 13, 2009

It is getting to the point where writing a thesis blog turns into a form of procrastination about writing a thesis. That–combined with the fact that I finally have a pretty detailed outline of the thing, and that there are some application deadlines rapidly approaching–means that I will probably be posting fewer substantive musings about illocutions, authority, and the like in the coming weeks and months.

But fear not! I still have a few musings left in the tank, and I’ll still be posting links when something I read elsewhere strikes me as particularly thesis-related. In addition, this will probably become my main outlet for thesis-writing status updates. So stay tuned if you’re particularly interested in how stressed I am at any give moment this fall.

From The Comments

August 10, 2009

Just thought I’d direct your collective attention to this quick back-and-forth I’ve had with Mark Lance (of Kukla and Lance fame) in the comments of a previous post. I am very surprised, and very grateful, to be getting a little bit of feedback from someone I’ve been writing about.

The New York Times on Illocutionary Uptake

August 9, 2009

Modern Love picks up where Sports Night left off:

“I don’t love you anymore. I’m not sure I ever did.”

His words came at me like a speeding fist, like a sucker punch, yet somehow in that moment I was able to duck. And once I recovered and composed myself, I managed to say, “I don’t buy it.” Because I didn’t.

(hat tip: Silpa)

The New Yorker on the Second-Personal Sharing of Moral Responsibility

August 7, 2009

If Jesus informs you that you will betray him, and tells you to hurry up and do it, are you really responsible for your act? Furthermore, if your act sets in motion the process—Christ’s Passion—whereby humankind is saved, shouldn’t somebody thank you?

Maybe the Nuremberg Defense should be called the Judas Defense.

In Which Vicki Hearne Changes Everything and Nothing

August 5, 2009

Just back from New York, where I picked up a copy of Adam’s Task, by Vicki Hearne. I’m almost all the way through it now, and, as Korsgaard would say, I cannot go on as before.

Hearne was an animal trainer, a poet, and a philosopher (though she had no philosophy degree). The title of the book comes from Genesis, in which God creates the world’s animals, and gives to Adam the task of naming them. It’s a book about talking to animals, but that description really doesn’t do it justice.

My interest in the book (or at least my thesis-related interest) is in Hearne’s account, heavily influenced by Wittgenstein and Cavell, of how a dog and its trainer create a language-game and, thereby, a moral world that the two participants inhabit. In doing so, Hearne makes a series of observations which are helping me to better understand almost every element of my thesis. This is all sort of half-baked right now, but I’ll do the whole quote-extensively-and-begin-to-analyze thing, and see if I can get across what I’m talking about.

Read more…