On Graber's, "Creating Truth By Winning Arguments..."

I read a really interesting paper today by Abraham Graber entitled, "Creating Truths By Winning Arguments: The Problem of Methodological Artifacts in Philosophy" (forthcoming in Synthese). The paper sort of crystalizes a worry that I've had about philosophical practice dating back at least to graduate school, and which I explored previously here and here: namely, that a whole lot of philosophy--just about anything having to do with conceptual analysis (including analytic metaphysics and meta-ethics)--is little more than philosophers (1) taking themselves to discovering objective philosophical facts (i.e. what material objects are, what morality is, etc.), where what they're really doing is (2) making more-or-less arbitrary semantic decisions to settle the semantics of fundamentally vague concepts that (3) objectively have no determinate satisfaction-conditions before those arbitrary semantic decisions are made.

There's another, simpler way to put this worry: namely, that philosophers engaged in conceptual analysis, far from discovering philosophical facts, are making them up arbitrarily out of whole cloth--in which case a lot of what appears to be "philosophical progress" or "good arguments" at any given point in time are more-or-less-arbitrary semantic decisions that have "won out" rhetorically in the literature and seminar room. This is more or less the worry Graber is pushing as well. So let me briefly explain why I've had the worry for so long, and then briefly comment on Graber's paper.

When I was in graduate school, I found that I often didn't share the intuitions I was "supposed" to. When doing philosophy of language, metaphysics, and mind, I found I didn't share Kripke's Godel/Schmidt intuitions, I didn't share Putnam's Water/XYZ intuitions, I didn't share Davidon's Swampman intuitions, and indeed, I didn't share any externalist intuitions about pretty much anything. But there I was, encountering a veritable tsunami of externalists trumpeting the arguments! But of course not everyone was an externalist. There were still some internalists like me. We just seemed to be getting crowded out by those with externalist intuitions.

And so I worried:was what was going on a kind of band-wagon, Hawthorne effect where once some influential people (Kripke, Putnam, Davidson, etc.) affirm an intuition, other people simply tend to follow suit--in which case it's not actual arguments that are winning the day so much as a sociological phenomenon of certain intuitions becoming popular (in turn rather arbitrarily redefining what counts as a "good argument" in that literature--namely, those that begin with the popular intuition!).;

I ended up taking my "messed up" intuitions home, as it were, leaving philosophy of language for other areas of philosophy where my intuitions weren't considered so "off": namely moral and political philosophy (my AOS now). And yet...the worry has crept up yet again! Indeed, Graber focuses on another case that has troubled me even more lately: the case of moral realism.

Anyone who does meta-ethics knows that there has been a moral realism tsunami, with realist after realist contending that, "only a realist account of moral semantics can account for the face value of moral language [see, for e.g., Shafer-Landau (2003) and Brink (1989)]." (Graber, p. 5) Except here's the problem: many people--people like me, any expressivist on the planet; my students, if my class discussions are any indication, etc.--just don't think the face value of moral language is anything like moral realists claim it to be. Most of my students, for instance, explicitly say that they think moral statements (e.g. "Murder is wrong") merely state matters of emotion, opinion, or otherwise mind-dependent phenomena (such as desires)--all of which flies in the face of moral realists' claims.

And so I have worried: how has moral realism achieved such a preeminent place in philosophical discussion? Is it the arguments that have done it, or is it more of a sociological phenomenon where the philosophers with realist intuitions have the upper-hand not in terms of arguments but rather mere number and prominence of people with their intuitions? 

Now, of course, as I have expressed them, these are just worries--nothing more! However, in his paper, Graber argues that on any plausible theory of meaning (internalist, externalist, whatever), philosophical inquiry changes the meaning of words and concepts, altering their satisfaction-conditions so that (A) at one point in time, and for one set of speaker, one answer to philosophical questions (e.g. moral realism) is correct, whereas (B) at another point in time, and for other speaking, a different answer is correct (e.g. moral antirealism).

Graber's final point then is that if he is right, then philosophical projects that depend on "meaning analysis" (i.e. analytic metaphysics, meta-ethics, etc.) all have to lapse into incoherence. Why? Because, very roughly, if we take ourselves to be (1) discovering answers to philosophical questions, but (2) our inquiries change semantic meaning of contested terms, it follows that (3) we're not discovering answers to philosophical questions. We're making up "answers" on the fly...which is to say, we're not discovering answers at all--we're playing rhetorical philosophical games.

Anyway, I don't know if the worries are good ones, or if Graber is right. These are just some issues that have bounced around in the back of my head for a while--and, while I've never thought about them very carefully, Graber seems to have: and his analysis of them seems to me worth taking seriously. But what do I know? I'll be curious to hear what you all think.

Comments

[Eugene]

Hi Marcus,

A few super quick thoughts without having read Graber.

(1) Graber's view is vulnerable to his own argument, no?

(2) You seem to suggest that moral realists are the only metaethicists that would be vulnerable to Graber's argument, but expressivists would be, too.

(3) If Graber was right, then those who hold minority views (say, about morality) would be using moral terms incompetently, right? But no one thinks that (say) expressivists are incompetent moral language users. They just think the view is wrong.

[Marcus Arvan]

Hi Eugene: Thanks for your comment!

On (1), Graber considers and rejects that objection in section 7 of his paper. His response is that his argument only calls into question philosophical inquiry that involves conceptual or ontological analysis--which he claims his argument isn't doing.

On (2), I didn't mean to imply that moral realism is the only view in meta-ethics vulnerable to Graber's argument. In point of fact, I think almost *all* of meta-ethics is in fact vulnerable to the charge. I think this is a *big* problem for meta-ethics, and that meta-ethics should no longer be done in the way it is done (i.e. on the basis of our intuitions about moral language, facts, etc.). I think it should be done instead on the basis of empirical facts about human behavior (this is something I'm arguing for in the book manuscript I'm working on).

On (3), I don't think he'd say they're using terms incompetently (since meaning can shift, and expressivism might be one way that meaning can shift). I think he'd say instead that when moral realists convince enough people to adopt a realist interpretation of moral language, it makes expressivist interpretations *wrong* (which is exactly what you imply is the right thing to say!).

[Eugene]

Thanks for the reply, Marcus.

Re: (1): If Graber can help himself out of his own argument by claiming that he isn't doing conceptual analysis, then I suspect a lot of other people can make this same move to avoid it, which brings me to my next point regarding (2).

Re: (2): It seems to me that even if Graber is right then his argument has very few targets, namely, analytic reductivists. And not even all of them, because some analytic reductivists (e.g. Finlay) take the analytic method to be a kind of empirical method.

More re: (2): Why do you think metaethicists should be carried out on the basis of empirical facts about human behavior, and not empirical facts about linguistic meaning? What makes those facts any more privileged than the other? Maybe things are different in your book, but some of your posts on this stuff have always struck me as targeting a caricature of metaethicists who take linguistic evidence seriously. I mean, there is no intuition pumping going on when metaethicists (again, Finlay, but also Ridge) take the fact that communication would be enormously difficulty to achieve if not impossible as suggesting, against nonnaturalism, that normative terms like 'good' couldn't be lexically ambiguous between natural and nonnatural entities, which suggests further, again against nonnaturalism, that our terms couldn't be about some distinct realm of nonnatural entities. The starting point here in this kind of argument is a linguistic fact and the argument doesn't seem any worse for it.

[Marcus Arvan]

Hi Eugene: Thanks for your reply.

On (1), I don't think that's being fair to graber. A lot of philosophy involves conceptual or ontological analysis--but not all of it does, and I don't think his argument does. His argument is that on *every* plausible theory of meaning, the problem of "philosophical artifacts" (i.e. Creating truths out of whole cloth) is a problem for conceptual/ontological analysis. Because he is arguing that all plausible theories of meaning have these implications, if his argument is sound, he's on firm ground: *every* serious theory of meaning entails that a specific methodology in philosophy--conceptual/ontological analysis--is problematic. That seems to me a very strong way to argue. How is it cheap, self-undermining, etc? And how could proponents of conceptual/ontological analysis make the same move? (I don't see how!).

On (2): I'm with you. I think meta-ethics *should* be based upon empirical observation of language and concept usage. I just think the *kinds* of facts that are often appealed to in meta-ethics are flimsy (I don't think there's clear evidence that the face value of moral language is realist), and that there are better places to look for more solid evidence (evidence that expressivists, etc. won't be apt to deny). But that's another can of worms. (Maybe some of my critiques have caricatured some views. If so, I'm sorry! My main contention in my book manuscript is a positive argument that there's more stable linguistic/conceptual evidence to appeal to...but I'll have to leave that for the book--hopefully it'll eventually come out!).

[Angra Mainyu]

Hi Marcus,

I think many philosophical arguments do not pay enough attention to the possibility that the words that they're using are not precise enough for the task at hand - so, there is no fact of the matter about at least some of the issues under consideration, even key ones -, so in a way I'm sympathetic to Graber's argument.
On the other hand, part of your description makes me think I would probably disagree with a number of his points, maybe the central one. Still, I would need to read Graber's argument to assess that better.

That aside, I have a couple of quick points (but I have not read the paper, so there is that):

1. With regard to water and H2O, the vast majority of people are not familiar with the relevant philosophical arguments, and people keep learning how to use "water" as they did before, so probably the way they use the words has not been affected.
That suggests one can still test the theory about the meaning of the word "water" - as the term is used colloquially, at least usually - that holds that the meaning is such that if water is actually composed of H2O here on Earth, it's impossible that water be composed of anything other than H2O.
For example, one might present those scenarios to people who haven't read the relevant arguments, but who know that water is composed of H2O. Perhaps, students of physics, chemistry, medicine, etc., would be good candidates. Maybe even philosophy students who haven't been exposed to those arguments earlier.
If possible, the general public too, as long as they roughly understand the meaning of "water is H2O".

2. On the issue of realism, I think asking students - or anyone else - whether they think moral assessments have realist semantics is a question very different from asking people whether XYZ would be H2O after explaining the scenarios, since in the water case, one is probing [some] people's dispositions to use the word - which, it seems to me, is a way of finding information about the meaning of a word -, whereas in the realism case, one is asking [some] people about their theories about the meaning of a word.

3. Also, on the issue of moral realism, it's intriguing that Graber considers this case as an example, given that he (recently) defends what he calls "mad dog realism", or non-naturalist moral realism. Has he changed his mind on the matter?

4. While I think sometimes there is no fact of the matter until some semantic decision is made, it seems to me that at least the issue of whether some (most, perhaps) metaethical theories are true is not among them.
In particular, it seems very different to describe some feature of the world - mind-dependent or not; "mind-dependent" seems both obscure and imprecise to me -, and to give a command, or express a feeling but without describing the world (as in "boo").
As I see it, if it is the case that some people are describing something (real or not) when they make moral assessments, and some people are expressing some emotion (for example), then they're talking past each other, but in that case, it seems to me that the matter can't be resolved by means of an arbitrary semantic decision, since following such decision would radically alter the use of words by one of the groups - in other words, in a case like that, the meaning of the terms as used by different people would not be close enough for any patching.