A Modest Proposal

I was catching up on this blog just now, and noticed that the old question about the identity of X-Phi came up in several posts and comments. Perhaps this is enough of an excuse for me to made a modest proposal. There is, of course, plenty written about the relationship between X-Phi and more traditional philosophy. I think the simplest way to approach it is to ask whether it is psychology, and if not, why not? (After all, the disciplinary identity of psychology is a lot clearer than that of philosophy.) To lay my cards on the table, I think it is: both the methods it uses and the questions it addresses are those of empirical psychology. This is not, of course, to say that X-Phi studies are necessarily philosophically uninteresting – I agree with Tamler (sort of) that such issues of relevance must be settled in piecemeal fashion. But since the two disciplines have distinct questions and methods, it is important to avoid the sort of confusion that easily arises (and has indeed arisen) from using the same or similar terms for different types of inquiry. So I will advocate that people begin to describe themselves as doing psychology of philosophy or of intuitive judgments when that’s what they do. The label ‘experimental philosophy’ and talk of ‘philosophical experiments’ has generated a lot of heat and no light. The admirable modesty of many smart X-Phiers is incompatible with the alleged continuity with the philosophical tradition.

First of all, I hope it’s not controversial to define psychology as the science that studies how the mind actually works. I don’t know what else psychology could be. (Perhaps it also has some limited otherworldly ambitions.) To answer questions about how the mind actually works scientifically, what is needed is systematic observation and controlled experiments of all sorts, including surveys, lab experiments, studies of the brain, and so on. Now, according to Josh Knobe, X-Phi studies how the mind actually works. It does so by way of systematic observation and controlled experiments of all sorts, including surveys, lab experiments, and studies of the brain, among others. Ergo, X-Phi is a branch of psychology that specializes in how the mind actually works when we make judgments about philosophically interesting topics.

At this point Josh would likely say, as he has recently often done, that philosophers, especially prior to the 20^th century, did make claims about how the mind actually works, so this is actually just a return to the classical tradition. (There may also be an underlying narrative of progress: put in my terms, the suggestion is that philosophy has always been psychology, but only now it is done properly and scientifically.) Behind this response, there may be a causal hypothesis: certain misguided developments within philosophy (maybe Moore, Russell, Wittgenstein, Husserl) led to an aberration that has now finally run its course.

I think this line of response is deeply unconvincing. It is true, of course, that many philosophers from Aristotle to Nietzsche did make claims about how the mind actually works. (Though I’d be surprised to learn they had an interest in folk classification.) But then Aristotle made all sorts of biological claims, too, and pioneered just about any other empirical discipline as well, Descartes worked on optics and geometry, Kant wrote about astronomy, and so on. No one would now say that they were doing philosophy when they did so (though at the time they may have seen no significant difference between their various projects, and all their work would have been then labelled as philosophy). Why? Because we can see now, in retrospect, that these are issues that are best settled with empirical evidence. Psychology was among the last sciences to gain independence from philosophy in the 19^th century; that is the real reason why only 20^th century philosophers needed to and were able to distinguish their own project from psychology, not the emergence of some misguided doctrine. But that is not to say that we can’t trace the non-empirical and non-psychological questions and methods back to earlier work, and see that they form the core of what the people we now regard as great philosophers did.

This brings me to my second point, which picks up on comments made by Confused Traditionalist, Clayton Littlejohn, Dave Chalmers, and Joachim Horvath in the Sommers thread. That is, there are first-order, traditional philosophical questions, the questions that draw most people into the discipline, that cannot be answered by studying how the mind actually works. Here’s just a few examples: Is free will compatible with determinism? Which changes will I survive? Why is there something rather than nothing? Which patterns of inference are entitlement-preserving? Are there objective moral truths? Why and when should I obey the law? There are also second-order, methodological cum epistemological questions: How can we discover whether free will is compatible with determinism? What is evidence for or against the existence of objective moral truths or irreducible phenomenal states? These questions, too, are non-psychological. They’re not about how the folk categorize things and why. It is only if we give certain optional substantive answers to these second-order questions that such psychological facts (or other kinds of empirical facts) become philosophically relevant.

This is no place to argue for it, but I think some of the second-order views that X-Phiers explicitly or implicitly rely on can be seen as nonstarters when plainly stated. We can’t conclude from the fact that most people think or talk in a certain way that things are that way (with the exception of some response-dependent properties). For example, it is no strike against the existence of objective moral truths that people disagree or fail to make some distinctions, nor that some ordinary people may hold the metaethical view that there are no objective moral truths. Now, of course X-Phiers draw inspiration from appeals to intuition or ordinary language, which continue to be employed in philosophical debates. But as people like Ernie Sosa, John Bengson, Tim Williamson, and myself, among others, have argued in different ways, it is a big mistake to read such appeals as predictions about how people would respond to a survey, even a well-designed one. That’s just not the kind of evidence appealed to. And if it isn’t, that’s not just bad news for optimistic experimentalists, but also restrictionists who find order effects or other kinds of instability in survey responses.

This is a live battleground, and I think a piecemeal approach is best here, much as I love sweeping conclusions. For example, Eddy Nahmias just recently made an impressive case for the relevance of what he calls folk intuitions to the free will debate over at Flickers of Freedom. (Having read some epistemology lately, I hesitate to label folk judgments as ‘intuitions’, though I’ve used the term more liberally in the past.) I can’t properly respond to it here, but I’ll just note that it’s undeniable that philosophers appeal to the crowd to motivate a position such as indeterminism, or shift the burden of proof, I don’t think they appeal to the psychological fact that it seems to people (or to themselves) that things are so-and-so to support premises or thought experiment-based judgments. This distinction is important, because it means that psychological studies won’t get directly at the evidence for philosophical claims. If they did, they could indeed be used to support substantial views so intimately that they should be a part of philosophical methodology.

Now, I think many self-identified X-Phiers will agree with me that X-Phi can’t directly address either substantive or methodological questions. Perhaps, on reflection, they’ll agree with me that they presuppose certain (armchair) answers to second-order questions in order to merit philosophical attention. If they do, there may be no substantive dispute between them and me, depending on what their armchair answers are. But given the way the mind actually works, ‘mere’ labels matter, too. (Yes, that’s an empirical claim.)

To make the case clear, I’ll pose a dilemma for Josh Knobe in particular and for modest X-Phiers in general. Here are two theses that they seem to hold:

Continuity

Experimental philosophy is significantly continuous with traditional, pre-twentieth century philosophy.

Modesty

Experimental methods do not address or yield answers to the traditional first-order philosophical questions or methodological/epistemological questions. (Modesty+: Experimental methods presuppose non-experimental answers to some traditional first-order philosophical questions or methodological/epistemological questions.)

I believe Continuity and Modesty cannot be coherently held together, though they don’t directly contradict. Either one has to go. For both a priori reasons (at least most of traditional philosophy is concerned with other topics than how the mind actually works) and a posteriori reasons (X-Phi hasn’t made a significant contribution to first-order or second-order questions), there’s no denying Modesty. It’s Continuity that has to go. If both the core targets and methods of X-Phi are different from those of traditional philosophy, it cannot be significantly continuous with it. At the same time, it is significantly continuous with psychology and empirical study of cognition and judgment in general. Apart from slight variations in subject matter, it looks indistinguishable from where I sit. It talks like psychology, walks like psychology, even quacks like psychology. So why not call it psychology, maybe psychology of philosophy, to borrow a phrase from Jonathan Weinberg’s comment earlier?

Why does it matter? Well, no good will come out of calling surveys or lab experiments ‘philosophy’. It gives people (including, it seems, some philosophy grad students) the wrong idea of what philosophical questions are, and consequently makes philosophical methods look comically inadequate (how could you possibly settle from the armchair what the majority of people think about this or that, or which causal factors influence their judgments!). Moreover, if you grant Modesty+, your choices are a throughgoing, self-defeating scepticism (since you can’t justify the use of any empirical methods without non-psychological assumptions), or the accepting that there are distinctive, indispensable non-psychological philosophical methods that make it possible to arrive at more or less justified answers to (at least some of) the traditional philosophical questions. You better not saw off the branch you’re sitting on, so you better not attack or cause confusion about armchair philosophy.

Not only does talk of X-Phi give the wrong idea about philosophy, but it also misleads people about the psychology of philosophy. Given the name of the movement and the way some results are presented, some philosophers, particularly those of an older generation, tend to go apoplectic. (Is the idea to cause heart attacks and open some tenure-track lines?) It is easy to miss Modesty when the headlines scream about new discoveries concerning intentional action or free will. If the talk is of a psychological discovery about the causes of laypeople’s (or philosophers’) non-inferential judgments, well, nobody’s vessels will burst. The press coverage may be in smaller print, and the novelty less apparent, but sometimes that’s the price you pay for intellectual honesty and an increased light/heat ratio. If you make a philosophical case that this factual discovery undermines a consideration presented as evidence for a philosophical view, or even that it favours a first-order philosophical conclusion, well, bully for you – there’s no reason why the same person couldn’t be both a psychologist and a philosopher. (But you will have to make a real case, and not just cobble together a few incautious quotations about "what we would say" from a more innocent time.) Just don’t pretend or speak as if you’re doing the same thing when you’re not.

So try it with me: “I’m a philosopher, but I also do some psychology of philosophy.” There, that doesn’t sound so bad, does it? So why not give up the misleading label ‘experimental philosophy’ and start talking about the psychology of philosophy or psychology of intuition?

 

(This is already too long, so I’m going to address some specific points made in earlier discussions in a separate comment in a while.)

Comments

[NathanEmmerich]

Interesting take, the direction of which I appreciate. Have you seen Fiest's Psychology of Science. You probably have but I thought it might be a good way to think through X-phi in this way.

[Mark Phelan]

Interesting post, Antti...as usual. I’m not sure if people other than Josh are supposed to respond, but I will anyway. (I do take myself to be primarily a modest X-Phier.) Here are a few questions that occurred to me while reading through this:

1. What’s the relevance of the initial identification of experimental philosophy with experimental psychology? It seems that this is supposed to support the position that x-phi isn’t philosophy, but how? After all, as you point out, “the disciplinary identity of psychology is a lot clearer than that of philosophy.” So why should we assume at the outset that the purported fact that (some) x-phi is experimental psychology should lend any support to the idea that it isn’t philosophy? Are you assuming a tidy view of the academy without disciplinary overlap? Aren’t some of the questions addressed by art historians the same as those addressed by philosophers of art? Aren’t some addressed by biologists the same as those addressed by philosophers of biology? Don’t they use precisely the same methods? You start off with this idea that x-phi is experimental psychology, but as far as I can tell that is totally irrelevant. All the argumentative weight of your post rests on the well-trodden ground of defining philosophy and attempting to show that x-phi doesn’t fit the definition. So let’s turn to those projects.

2. It seems that according to your post, philosophy is the study of traditional first-order philosophical questions and certain second-order methological/epistemological questions, all of which are non-psychological. Assume I’m willing to concede that the questions that you list cannot be answered by studying how the mind works. So what? As you admit, that’s only a partial list. Lots of other first-order philosophical questions seem as though they could be addressed by studying how the mind works, questions such as: Are the objects of perception external to the mind? Are concepts all copies of percepts or are some innate? Is all human action self-interested? Could those questions be addressed by studying the mind independently of more traditional philosophical methods? I doubt it. But it’s a caricature of x-phi to suppose it attempts that. Are these also questions addressed by experimental psychology? Perhaps, but as we’ve already seen it’s a red-herring to give that argumentative weight. What I sense is supposed to be really pulling the weight here is a subtle imperialism about philosophical questions. The questions you innumerate are the more important, traditional questions—“the questions that draw most people into the discipline”. But I was drawn in (at least in part) by the pidgin questions I enumerate as well. And the philosophy courses I took that addressed these questions—before I’d ever heard of experimental philosophy—took research into the mind seriously. Lots of (misguided?) people wrote doctoral dissertations on these questions before experimental philosophy came along.

3. I think your characterization of experimental philosophy is also too narrow, though this mistaken conception has been aided and abetted by some characterizations of experimental philosophy due to experimental philosophers themselves. It is no longer correct to say (if ever it was) that all experimental philosophical projects involve the scientific study of pre-theoretic, intuitive assessments of particular cases (the “thought experiments” or “intuition pumps” of traditional philosophy). Just as philosophers have traditionally appealed to a disparate assortment of evidence in favor of the positions they have supported, so experimentalists have been systematically assessing not only intuitive judgments, but other sources of evidence as well. For example, a team led by Eric Schwitzgebel has attempted to gain insight into the ancient and oft asserted claim that philosophical reflection on ethical issues can improve one’s moral behavior by examining how professional ethicists actually behave in a variety of circumstances. In other work, Joshua Knobe and Jesse Prinz have examined how frequently people use different kinds of mental state sentences on the internet, and argued from their findings that people are ordinarily willing to ascribe beliefs and desires, but not experiences or emotions, to disembodied or distributed entities like corporations. (Adam Arico, Shaun Nichols, and I have done follow up work that uses a similar, non-intuition-soliciting approach, but argues contra Knobe and Prinz.) In my own experimental work, I have argued against the claim that it is distinctively difficult to express what a metaphor actually means by systematically comparing how people paraphrase metaphorical and literal utterances. This is a straightforwardly empirical claim and one that I argued in that paper bears much of the brunt in arguments for non-cognitivist theories of figurative language. (Incidentally, in making these arguments I don’t merely, “cobble together a few incautious quotations about "what we would say" from a more innocent time.” I think that’s an unfair characterization of the field.) I don’t deny that these other projects attempt a systematic study of the mind—or at least of human behavior. But intuition isn’t the target, so the “psychology of intuition” is an ill-fit. Nor are all these projects concerned with the psychology of philosophers, or the psychology of philosophy in any obvious sense, so the “psychology of philosophy” won’t do either. And, more substantively, your narrow conception also obviously raises trouble for your attempt to shade x-phi off from traditional philosophy. (You also seem to assume the oft repeated mischaracterization that x-phi adopts a first-past-the-post approach to philosophical questions, as though experimental philosophy papers said nothing more than: “Survey says...incompatabilism!” Experimental work has always been supported—at least in the best examples—by careful philosophical argument. As was heatedly discussed here a while back, it’s ad hominem to suggest otherwise without offering a careful analysis of experimental philosophical papers.)

4. So, as for your dilemma, I reject Modesty. (I suppose I’m an immodest philosopher after all!) And I also deny both the reasons you suggest prohibit me from denying it. You claim that “most traditional philosophy is concerned with other topics than how the mind actually works”. And I say, perhaps, but so what, some traditional philosophy is concerned with how the mind works (nor do I think it’s clear that how the mind works is irrelevant to most traditional questions, which I see as the key issue). This reason for rejecting modesty is akin to claiming that epistemology isn’t philosophy because most traditional philosophy is concerned with other than epistemological questions; or metaphysics isn’t philosophy for the same reason. Any sub-genre of philosophy will entertain a minority of the traditional questions. So what? And I deny, too, that “X-Phi hasn’t made a significant contribution to first-order or second-order questions”. Though I think there’s no point in going over the details here. (And even if it hadn’t, why isn’t the relevant question that it couldn’t? How long does a new discipline get before the bills come due in your estimation?)

I recently wrote a general introduction piece about x-phi and in thinking through that I came to believe, too, that the name isn’t entirely satisfactory. But obviously not for your reasons. And I certainly don’t think psychology of philosophy or psychology of intuition are improvements.

[Antti Kauppinen]

As I said at the end of the post, I’m going to take the liberty of responding to some earlier comments that bear on what I said. In the main post, I in effect endorsed a distinction between substantive, methodological, and factual questions, which corresponds to a distinction made by Dave Chalmers and others between first-order and second-order questions in an earlier thread on Tamler Sommers’s work (http://experimentalphilosophy.typepad.com/experimental_philosophy/2011/05/tamler-sommers-on-the-death-of-the-x-phi-debate.html). In that thread, several people sought to challenge Dave’s distinction or show that X-Phi has contributed more directly to traditional questions (which Chalmers doesn’t want to rule out a priori). There were really good comebacks by Joachim Horvath and Clayton Littlejohn, which I pretty much fully endorse. (In fact, reading the thread more carefully made me realize how very little new there was in my post.) But I want to address a few things that went unchallenged there. (When I started writing this comment, I thought there were many more, but the less enthusiastic people really did a very good job.)

In that discussion, Chandra Sripada argued that X-Phi can after all provide first-order philosophical insights. He said:

"For example, with manipulation cases in the free will debate, everyone agrees there is a strong intuition that a globally manipulated agent is unfree, but the debate is over why. Incompatibilists say the intuition arises in virtue of the agent’s lack of ultimate control over his character, while compatibilists say the intuition is in virtue of the agent’s being psychologically damaged."

This may look innocent at first sight, but there’s a crucial slide from first-order to second-order questions. What incompatibilists and compatibilists disagree about is why a globally manipulated agent is unfree, not why the intuition that the agent is unfree (which both indeed share) arises. (This confusion may be just due to an ambiguity – we use ‘intuition’ for a) a type of psychological state, b) a non-inferential, non-perceptual belief, c) the intuited proposition, and d) a putative psychological faculty, and this is not an exhaustive list. In the sense c, we could (clumsily) say that incompatibilists think the intuition is made true by lack of control. This, of course, is no psychological claim, unlike a claim about why the intuition in the sense of a or b arises.) Neither side needs to take a position on the psychological sources of the intuition – though as usual, there may be psychological explanations that call into question the epistemic credentials of the intuition (or, as some would describe it, the genuineness of the intuition). So even if sophisticated X-Phi methods can probe into which factors influence the intuition (in the psychological sense), we haven’t got an example of a first-order philosophical insight here.

Jonathan Weinberg, in turn, challenged Dave’s distinction between first- and second-order questions:

"Just think of how much totally mainstream, paradigmatic philosophical work nonetheless has a significant higher-order component. E.g., huge parts of the consciousness literature is about the conditions under which we should credit an intuition that p is metaphysically possible. In some sense, it's not a first-order debate, at least, not a first-order debate about consciousness. But in another, and I think more important sense, it just doesn't matter: that epistemic meta-level is just part of the relevant terrain to be fought over. To fight there _just is_ a way to be participating in the battles about zombies, dualism, and so on."

I would say that the debates are certainly intertwined, but there’s still a distinction to be made. They’re about different premises, after all:

1) Conceivability is a guide to possibility.
2) Zombies are conceivable.
3) Hence, zombies are possible (and if zombies are possible, blah blah…)

The second-order debate is about 1, of course. If all sides grant 2 (do they?), it will be key to the first-order debate about 3, so it’s unsurprising that it gets a lot of attention. But it’s still meta, though quite immediately relevant. And of course 1 isn’t about how the mind actually works. (I do actually think that discoveries about the limitations of our imagination are potentially relevant to evaluating its truth, though.) Now, perhaps there are debates in which it really is impossible to separate the substantive and the methodological; even so, the gap between the substantive and the psychological is even bigger, so I’m sceptical of the chances of finding an inseparable empirical component.

Finally, a minor historical point. Dave Chalmers wrote:

"Re the Knobe case, it's also not entirely obvious that the argument from x-phi to a first-order conclusion about intentional action should count as a primary insight as defined. After all, it's basically a quantitative version of a thought-experimental intuition that might have put forward by a single philosopher first and can be construed as supporting that intuition. But as a matter of fact the x-phi came first in this case, so if there is a first-order conclusion here, x-phi should get some credit."

Since a lot of people share this historical belief, it’s surely worth pointing out that it’s actually not the case that X-Phi got there first. Here’s a footnote on the so-called Knobe effect that somehow never made it to the final version of my Rise and Fall paper: “In fairness to traditional philosophers, something very similar about the influence of moral considerations on attributions of intentionality was noted a long time ago by Georg Henrik von Wright, who presents a case in which a person opens a window and foresees that another person will shiver as a result of his action, and argues that whether producing the side effect would be considered intentional depends on its moral status: "If the agent can be blamed for what he foresaw without intending to bring it about, then the foreseen consequence is something which he did intentionally and for which we hold him responsible." (von Wright 1971, 90)” So I’d say von Wright pretty clearly felt the asymmetry, and as an ordinary language philosopher of sorts drew the first-order conclusion, as the quote shows. (He doesn’t claim any originality, so I wouldn’t be surprised if this was a widely held view at the time.)

I'll respond to Nathan's and Mark's comments soon.

[jonathan weinberg]

I find myself having trouble making sense of a lot of this. Much of the problem is that Antti is committing what is a common error in people writing about x-phi, namely, treating it as much more uniform and monolithic than it actually is. I mean, (Modesty) seems like it describes at most one practitioner, Joshua Knobe, and maybe Shaun Nichols in some of his moods. (Though I would add that as written it seems to me to be something that not even Joshua would agree with, since Joshua has in various places speculated on how x-phi results could have _some_ bearing on traditional debates.) (Modesty+), on the other hand, is so obviously true that I can't see it as the, er, modest addition to (Modesty) that Antti seems to intend it to be. For example, I think everyone in x-phi takes for granted, non-experimentally, the validity of modus ponens; the statistical tests that we use presuppose, non-experimentally, a great deal of mathematics; and so on. I have no idea whose ox is supposed to be gored by this sort of thing. It seems that Antti is assuming that x-phi folks hold to a kind of super-duper-hyper-scientism, such that only experimentally-gained knowledge is knowledge. But that's just crazy.

* * * * *

Regarding the continuity issue, the examples are chosen somewhat poorly, I think. Sure, let's grant _arguendo_ that Aristotle's _Parts of Animals_ can be treated as non-philosophical, for contemporary purposes. Descartes's _Meteorology_, too. Berkeley on tar water, and so on. But it's utterly obvious that most philosophers' views about the _the nature of the mind_ are still treated as part & parcel of their philosophies. It would be bizarre to teach an undergraduate philosophy course on Descartes without his dualism; Hume without his impressions & ideas & principles of association; or Kant without his faculties. I suppose one could teach Plato without going into innateness, but obviously the Meno is not treated as an aberrant, non-philosophical part of Plato's oeuvre. So there's really a very strong case to be made for the continuity of the sort Joshua claims, connecting what he's interested in with these elements of the patently philosophical inquiries and theories of the greats of the tradition.

Of course there are many other lines of continuity as well, that different experimental philosophers might want to view their work through. One obvious candidate would be ordinary langauge philosophy (which, I suppose, is a misnomer on Antti's account, since apparently it was never _really_ philosophy). For my own work, in restrictionist/"negative program" x-phi, I see a continuity not so much with the "philosophers who were philosophically interested in the nature of the mind" line, but rather with the "philosophers who were interested in ways that philosophy was, by their lights, doing things wrong": Descartes against the Scholastics; Hume against "sophistry and illusion"; Kant against what he saw as the errors of rationalists and empiricists alike; the positivists against "metaphysics"; Wittgenstein against, well, everyone, I suppose. I'm personally eager to claim James as an intellectual forebear, in this regard.

* * * * *

Fwiw, I don't think I was challenging Dave's distinction itself, so much as challenging the idea that it can do much interesting work for us. I still have yet to see why anyone should care even a teensy bit about whether an insight is "first-order" or "second-order".

"If all sides grant 2 (do they?)" Not that it much matters here, but: no, they don't; those that don't are type-A materialists, in Chalmers' taxonomy. It's not the modal view in the literature, but it has more than a few adherents. I don't think that anything in x-phi is particularly relevant to that debate though. The example was just meant to illustrate, again, that the first-order/second-order distinction is not one that is of metaphilosophical interest beyond the merely taxonomic. I think it would be best to keep the first-order/second-order issue raised there cleanly distinct from the psychological/philosophical issue that you're trying to raise here.

[jonathan weinberg]

I also wanted to magnify Mark's first comment, concerning Antti's mistaken background assumption that psychology and philosophy are contraries, so that if he can show that x-phi is the former, it would follow that it can't be the latter. But counterexamples to such an assumption of disciplinary partition are plentiful -- _lots_ of philosophy is also, unproblematically and at the same time, something else. Much work in logic and formal epistemology is also mathematics; much philosophy of language is also linguistics; much history of philosophy is also history. Why think that any of this is any sort of problem? Of particular relevance here is the way in which much philosophy of cognitive science is part of cognitive science itself. Fodor & Pylyshyn's attacks on connectionism, for example, are obviously to be categorized as part of both the philosophical and scientific literatures, as evidenced also by the diverse sorts of engagements there were with those papers across a number of different disciplinary boundaries.

So, suppose that you're right, and x-phi is psychology (at least when the science part of it goes well!). Na und?

[Antti Kauppinen]

W, you sound like somebody's told you a couple of times too many that what you do isn't philosophy! Note that I'm not saying that. I'm saying you're a renaissance man of sorts - you do a bit of psychology and then philosophy on top of it. (I happen to disagree with some of the philosophical commitments, but that's a different issue.)

I'm a little surprised that you don't subscribe to Modesty - I included you among 'the many X-Phiers' who did. Perhaps we're reading the thesis differently. I certainly meant it to be compatible with what I said earlier, namely that with suitable bridging assumptions, psychological discoveries do potentially bear on many trad questions. Modesty just says that by themselves, experiments don't address or answer them, as they're not about how the mind actually works. I also didn't mean to suggest that anyone denies, on reflection, Modesty+ (though Jesse Prinz says some wild things about all questions being empirical). The point is rather that if you do accept it, you're pro-armchair, and you're better not igniting it or, erm, creating a smoke screen that prevents grad students and the public from finding it.

On Continuity: sure, views about the mind are as core as it gets. But surely the ones that survived as unchallenged parts of philosophy after the birth of empirical psychology are not nomological claims about how the mind *actually* works. I'd think that's pretty clear about Descartes and Kant; with Aristotle and Hume, it's a mixed bag.

We've probably got a pretty different view about Ordinary Language Philosophy. I think the key thesis is roughly that there's nothing more to the essence of something than the way we correctly use the corresponding term. Note that this is a normative claim about proper use, not about what most people say. (OLP'ers like PMS Hacker are scathing about reading it as an empirical thesis.) If this methodological thesis about the subject matter of (some) traditional philosophical questions is correct, those questions can be answered by studying language. I can't see why anything I say would disqualify that as philosophy.

On Mark's first point: I admit that it's not really clear to me what philosophy is - it's a really big tent. I don't think I have the authority to be the final arbiter. We certainly have paradigms and a narrative. As I've said, X-Phi studies are pretty far from each, and smack in the middle of those of psychology. So I'm inclined to say it's psychology. But my main point is actually independent of the metaphysics of disciplinary identification. (I'm sure disciplines shade off in all directions, so that there are cases that are both borderline mathematics and borderline philosophy, for example.) It is a pragmatic one: labeling these studies as 'experimental philosophy' or 'philosophical experiments' is bad both for philosophy and for X-Phi. I think that's a more interesting point to debate. In a way, X-Phiers should be happy with that, since it calls for empirical data on the effects of labeling!

Out of habit, I do indeed use X-Phi somewhat narrowly. For example, I don't think that what Josh Greene does is experimental philosophy (nor do I think he describes it that way). Surely he's not doing philosophy when he puts someone in an fMRI machine! (I find it a bit disingenuous when people list such things as falling under the scope of X-Phi.) The Schwitzgebel studies I find problematic for many reasons (that's another blog post in the waiting), but in any case the thesis they purport to address is a straightforwardly empirical one. I don't see the philosophical interest there. (It's interesting only in the same way as claims about reading novels making you a better person are.) And finally, I think corpus analysis is a cool way to find patterns in language use, but that's still study of folk classification, just as surveys are (though it may avoid some problems), so I don't see a difference there.

More later.

[Chandra Sripada]

Hi Antti,

Thanks for this excellent post. You are right that I did not speak carefully enough when I was talking about the manipulation case example in the Sommers/Chalmers thread. Even if I didn't say it quite right, I still believe that some kinds of XPhi studies are best construed as directly addressing first-order questions. I wrote up an argument for this claim where I speak (hopefully) more carefully about the issue. But after the second page, it was clearly just too long. So rather than hijack this thread, let me wait and post the argument later, and hopefully folks can tell me why I am wrong. I should add that even while I am belaboring the issue of whether some XPhi is first-order, I am still quite sympathetic to Mark’s and Jonathan’s point that for many purposes, we don’t need to fuss so much about the first-order/second-order difference. I do, however, think it’s important to get clear on the different patterns of inference employed in XPhi-based arguments (not all XPhi is designed to assess the warrant of intuitions). That is my real interest in arguing that some XPhi is indeed first-order – I want to say this represents a distinct kind of XPhi-based argumentative strategy. But you were right to call me out for my earlier comments about manipulation cases. Thanks for that.

[John Dell]

Antti,

It is not particularly clear to me what the motivation is for trying to limit the domain of description for philosophical activity. You describe philosophy as being a big tent, but when you go into the details of what you take philosophy to be, you sound like are talking more about a pup tent.

I would really appreciate hearing a little bit more from you on what you take philosophy to be. (I'm not asking for final arbitration here, just something more to go on).

Philosophy is a *product* of those who practice it, no?

Historically, this *product* has been developed as a result of human efforts to understand ourselves, the world we live in, and how those two things fit together.

Is this a very broad way to understand philosophy? Sure. But why not make philosophy a place that allows for academic freedom as apposed to academic restriction? Why not spend more time on the actual word 'philosophy'? What does it mean to love wisdom? By trying to narrow the playing field for philosophy it seems to me like you are narrowing the playing field of wisdom. Wisdom certainly involves a whole lot more than exploring the kinds of questions you listed above. It sounds to me like you are saying a dog lover is really just a lover of Labradors! And if my suggestion is correct, then your position requires you to redefine what you take wisdom to be, so I would like to hear from you: what is wisdom? (again, just some thoughts, no final arbitration needed). Otherwise, I think you are the one who needs a new title for what he does.

Of course you would need a better title, but you sound to me like an "Intuiter"; i.e. an "intuiter of questions that cannot be explored with anything other than intuitions (and some logic motivated by those intuitions)". Certainly an intuiter could be considered a *kind* of philosopher, but hardly represents *all* kinds. No?

As it is, I don't see any reason to *limit* philosophy. But I may well be missing something (probably a lot of things)!

Do you think there is some objective form of philosophy out there and until recent times humans have just been stumbling around trying to find it? And now, at long last, we have finally got it.

[Foppe De Haan]

For fear of seeming rude by not engaging with your post directly, I would like to ask a question: Have you, by any chance, read Nicholas Capaldi's Hume's Place in Moral Philosophy?

[Tamler Sommers]

Now I know how Sisyphus must have felt!

One must imagine him happy.

[Antti Kauppinen]

Tamler, perhaps you're condemned to roll the rock because you insist on mislabeling what you do? Just let go; life will be lighter. You'll only have to shoulder the burden of proof. You may want to save your strength for that.

(Will respond to others later.)

[Tamler Sommers]

Now I feel like Marshall Mcluhan in Annie Hall...

[harvey brockman]

Can i point out, in order to encourage more hand-to-hand combat and fewer -- to mix metaphors -- ships passing one another in the night, that Antti offers x, y, and z as examples of "first-order" philosophical questions that "cannot be answered by studying how the mind actually works", and Mark offers a, b, and c as examples of first-order philosophical questions that "seem as though they could be addressed by studying how the mind works." If Antti and Mark could agree on the wording of a single question, and then each give an explanation of why that question cannot/can be so answered, we might gain some insight into what each of them means by "studying how the mind works". (I suppose "What is the mind?" can still be asked as a pretty contentious "philosophical question".) With some traction on "studying how the mind works", maybe connections (or a lack of them) between "traditional philosophy" and "experimental philosophy" will become clearer….

[APinillos]

Hi Antti,

you said: "Now, I think many self-identified X-Phiers will agree with me that X-Phi can’t directly address either substantive or methodological questions."

I disagree with you. here's an argument with 3 simple premises for why x-phi methods are importantly relevant to substantive philosophical questions. What do you think is wrong with it?

(P1) Facts about the semantics of 'knows', 'good', 'cause', 'must' etc. are importantly relevant for settling substantive philosophical issues.

(P2) Facts about the ordinary usage of 'knows', 'good', 'cause', 'must' etc. are importantly relevant for settling their semantics.

(P3) x-phi methods are importantly relevant for settling facts about ordinary usage of those expressions.

(C) x-phi methods are importantly relevant for settling substantive philosophical issues.

I wonder which premise you reject, or whether you think the argument is not valid?

Personally, I doubt there is much promise in rejecting (P1-P3). One could try to say that the argument is not valid by claiming that the facts of ordinary usage relevant for settling semantics cannot be importantly informed by x-phi methods. That would be an interesting and powerful claim. I would like to see an argument for it.

[Antti Kauppinen]

Okay, it seems it's going to be hard for me to catch up with all the comments, so I won't even try to respond to everything. I appreciate all the feedback, and hope that the things I do say will be helpful to those I won't get to as well.

Chandra: thanks for being so courteous! I think the kind of more sophisticated empirical methods you're using are moving the discussion forward, and I by no means intend to claim that the results couldn't be philosophically significant. I suppose one thing I want to resist is that psychological results, including those that have concepts or intuitions in the psychological sense as their object, are automatically or by default philosophically relevant. (Again, those X-Phiers who are Modest won't claim so either, and I thought they already constituted a majority.) If so, it's better not to label them philosophy.

John: surely you wouldn't want to call just anything, or even just any form of intellectual inquiry, philosophy. The reason why we talk of different disciplines is that we take there to be different subject matters to be investigated and different methods for addressing them. I believe that there are questions that no empirical evidence will settle, such as those listed in the post. (That's why logical positivists thought they were meaningless.) What better term than 'philosophy' do we have for inquiry into them, and why should we use it for something that addresses a different subject matter, such as how things actually are or what is nomologically possible?

Angel: that's a nice way to put the structure of an argument for x-phi. But first of all, even if all the premises were true, it would precisely establish an indirect role to experimental evidence, via a number of non-experimental premises. After all, to be even more explicit, you could add (P0): Substantive philosophical issues concern knowledge, goodness, causation, necessity, etc. The methodological premises then link that premise to the conclusion, but they don't say that the experimentally discovered facts just are the facts at issue in philosophical questions.

Now, the substantive (not merely verbal) methodological debate concerns the truth of your premises. I'm agnostic on (P1), because I can see the arguments on both sides. (Notice, incidentally, the contrast with Ordinary Language Philosophy, which crudely speaking advocates a deflationist view of philosophical problems: philosophical problems just *are* problems about the meaning of expressions like 'knows'. If you're not a deflationist of this type, you have the further challenge of explaining why it is that facts about semantics are importantly relevant to settling substantive issues.) I think (P2) is true, if we read 'ordinary usage' along roughly Kripkensteinian normativist lines. But then (P3) is false (unless a sophisticated form of dispositionalism is true and X-Phi methods are correspondingly adjusted to get at just the right sort of dispositions). Alternatively, if we read ordinary usage in the statistical sense, (P3) is true but (P2) false. (I argue along these lines in my 'Rise and Fall' paper.)

I hope that wasn't too telegraphic; the paper I mentioned has a fuller discussion (though it is in some other ways already outdated).

[John Dell]

Antti, you have got a lot coming at you so I understand that you did not address my questions (though I would still appreciate it if you did).

Notice that I did not say that the kind of questions you are promoting (nomological or otherwise) are not in the extension of 'philosophy'. My claim is that the kinds of questions you are endorsing do not exhaust what it is to *do* philosophy, and if you think they do, then you need a more limited term to accurately describe what it is you do.

I think by definition (and historical precedent) philosophy is a discipline that ought to be understood very broadly. For many people I think it is. Otherwise we would not have the great number of "Philosophy ofs..." (the list is enormous).

Out of curiosity, let me jump on my radical horse for a minute and suggest *any* sufficiently intellectual pursuit motivated by the desire for wisdom counts as a form of philosophy. What would you say?

Would you appeal to *only* practical reasons to argue against this point (such as the needed structure of the academy, and so on)? I doubt you would think practical reasons alone are good enough to take a stand against my broad notion of philosophy. So what's your argument? Why can't philosophy encompass numerous subject matters and methods as long as practitioners are engaged in sufficiently intellectual pursuits motivated by the desire to gain wisdom? [I realize sufficiently intellectual is vague, but time is short...].

Why can't philosophy be a place to find intellectual freedom, and the rare opportunity to synthesize subject matters such as psychology, anthropology, biology, sociology, and physics (for example)? Doesn't the academy also need a discipline with such freedom to go along with its division of labor? And if it does, what better choice than philosophy?

[APinillos]

Hi Antti,

you said "I think (P2) is true, if we read 'ordinary usage' along roughly Kripkensteinian normativist lines. But then (P3) is false (unless a sophisticated form of dispositionalism is true and X-Phi methods are correspondingly adjusted to get at just the right sort of dispositions)."

I disagree. I think (P3) is true in the sense of
'ordinary usage' relevant for semantics. You don't have to be a dispositionalist to think that in certain cases, ordinary people's dispositions will reflect their semantic competence and the semantic facts (the connection between disposition and semantic facts needn't be necessary or conceptual, it just has to be there, for certain cases, in the actual world--even mere correlation will do. BTW: This is how a lot of science works, you get at something hard to reach by observing things on the surface that are contingently connected to it). The x-phi researcher can isolate these cases and as you say, can get "just at the right sort of dispositions" with clever experiments.

Your rebuttal of my argument then needs that ordinary people's dispositions with using the words 'knows', 'good', 'cause', 'must' will *never* reflect semantic competence and the semantic facts (or that x-phi methods cannot get at these intuitions). This is an extremely strong claim, but it is what you need to refute P3. Is there an argument for it?

Let me give you an example of how x-phi can be relevant to semantics and to substantive issues. A number of epistemologists have claimed that in a conversational context, when the salience of error is raised, the epistemic standards for "knowledge" are raised. Some knowledge attributions will be more difficult to count as true. This assumption is not a minor one. Leading philosophers have used something like this to solve the old skeptical puzzles. Now, you don't have to have some sophisticated view about the philosophy of semantics (meanings are essentially tied to dispositions etc) to see that this sort of claim is open to experimental investigation. If you take competent subjects and put them in carefully constructed conditions where salience of error is raised, then you should see a change in behavior with respect to knowledge claims. Of course, all types of confounds can come up. But you can try to control for them with more x-phi methods.

[Josh Weisberg]

Meta comment on a debate in meta-philosophy:

I don't think this debate is about 1st-order vs. 2nd-order what's-it's, substantive vs. linguistic whatever, boundaries of fields, etc.

Nor is it about the burning need to save misled graduate students and the lay public from getting the wrong idea about philosophy (is changing the name from x-phi to "psychology of whosits" really going to solve that alleged problem?).

It's about sexy fads pissing off traditionalists, young turks of the new movement with their revolutionary rhetoric (complete with manifesto and Russian-inspired theme song!) guillotining too many innocents, and the incessant navel-gazing that's marked philosophy since Thales (probably the one methodological constant in the field).

You may return to your meta-debate now.

(Full disclosure: I am as guilty as anyone here, having posted on this debate a number of times and having been to x-phi indoctrination camp in Utah. ;>) )

[harvey brockman]

About Antti's reply to Angel's argument for the relevance of X-Phi (to settling substantive philosophical issues), Angel writes: "Your rebuttal of my argument then needs that ordinary people's dispositions with using the words 'knows', 'good', 'cause', 'must' will *never* reflect semantic competence and the semantic facts … This is an extremely strong claim." -- When i read this, i think it's a claim that is very far from having a clear sense. And this is my reaction to the argument as a whole. Before we're "agnostics" or believers, we need a clearer understanding: Is (P1) a contingently 'true or false' claim? Contingent upon what? -- We should be able to say what would count against its being true (as well as what would count as making it true). And we should ask this about (P2) and (P3), too.
But maybe (P1), (P2), (P3) are more like a premise (e.g.) "All cats are felines", and NOT LIKE (e.g.) "All cats have long tails"? If the question "Are X-Phi methods importantly relevant for settling substantive philosophical issues?" is itself a "substantive philosophical issue", we won't be able to use such premises to get IT settled, will we?

[Eddy Nahmias]

Hi Antti, thanks for starting up an interesting discussion. I had more to say, but it seems like things have died down here. Perhaps I can provoke you to say more if I ask you to explain briefly what you think is wrong with my arguments for why ordinary intuitions matter to the philosophical debates about free will and why x-phi methods can help uncover the relevant intuitions (judgments, beliefs, practices, etc.)?

(as you know since you linked to it) the post is here: http://agencyandresponsibility.typepad.com/flickers-of-freedom/2011/06/why-folk-intuitions-matter-to-the-free-will-debate-and-why-we-should-study-them-empirically.html