I’ve just finished a (long) draft of an entry on Intuitions for the SEP. It is focused on the nature of intuitions and their actual and normatively appropriate role in philosophical inquiry. It also includes some discussion of experimental philosophy. Any comments or suggestions (including regarding what to cut) would be quite appreciated.
|
|
Here’s an abstract for a paper I’m planning to write soon. Comments very welcome.
Experimental Philosophy and Apriority
Abstract. One of the more visible recent developments in philosophical methodology is the experimental philosophy movement. In some of its more negative paradigmatic instances, experimental philosophy collects some empirical data—perhaps You can Buy Clomid Online here and also Buy Nolvadex Online in our drugstore
some survey data from folk judgments about philosophical cases, or perhaps some neurological data about what happens in philosophers’ brains when they make certain kinds of judgments—and uses it to cast doubt on traditional armchair philosophical methodology. If these experimental philosophers are right, then philosophical methodology, as traditionally practiced, is importantly misguided, and stands in need of substantial revision, with a much greater emphasis on empirical investigation.On its surface, the experimentalist challenge looks like a dramatic threat to the apriority of philosophy; ‘experimentalist’ is nearly antonymic with ‘aprioristic’. This appearance, I suggest, is misleading; the experimentalist critique is entirely unrelated to questions about the apriority of philosophical investigation. There are many reasons to resist the sceptical conclusions of negative experimental philosophers; but even if they are granted—even if the experimentalists are right to claim that we must do much more careful laboratory work in order legitimately to be confident in our philosophical judgments—the apriority of philosophy is unimpugned. The kinds of scientific investigation that experimental philosophers argue to be necessary involve merely enabling sensory experiences.
A priori knowledge is knowledge for which sensory experience need play no warranting role. As has long been recognized, however, experience can play many important roles in the genesis of knowledge beyond warranting ones; the merely enabling role of experience in concept-formation has been widely recognized as non-warranting, and the necessity of such experience for knowledge is consistent with apriority. Of course, the experience involved in psychological experiments does not play this role in concept-formation—we needn’t perform experiments, for example, to acquire the concept knows—so such experiments cannot play merely enabling roles in our epistemological judgments in the same way that, for example, experience of red things may be necessary for possession of the concept red and therefore for a priori knowledge about redness. But being necessary for concept acquisition is not the only way in which experience can play merely enabling roles. Experience can also play enabling roles by being necessary for properly cautious and reflective thought. The visual experience I have when double-checking a mathematical proof, for example, plays important roles in my eventual knowledge of the conclusion—but not a warranting one. They are no part of my evidence; they are part of the process by which I acquire the evidence. The experiences that experimental philosophers claim are necessary for philosophical knowledge ought, I suggest, to be understood on this model. If the experimentalist critics are right, then we must perform more experiments in order to have philosophical knowledge. This is consistent with the traditional idea—once it is properly spelled out—that philosophical knowledge, when attained, is often a priori.
We need a term for it, so let’s use “cathedrism” for the thesis that core analytic philosophy can legitimately by and large be practiced as it is today, as basically an armchair discipline. A full statement would need appropriate exceptions for philosophy of physics, etc., and clearly different ways of configuring cathedrality would lead to different strains of cathedrism. Indeed, if one’s conception of the armchair is as something decidedly a priori, then most of the rest of this post will not be relevant to you — I’m interested in the increasingly popular line that defends an a posteriori-involving armchair (or an a priori/a posteriori distinction-jamming one, such as Williamson’s).
Now, I am not looking to contest here that some sorts of a posteriori knowledge is fair game for some construal of cathedrality. There is, surely, lots of knowledge about the everyday world around us to be had just by unaided human observation. But I cannot see why, if one is pursuing a project of inquiry that is also not something like a foundational project, there would be any reason not to avail oneself of any good methodological resources that might be available to one. And I particularly cannot see why one would rule out using better resources where they might be available. And for nearly any substantial theory about the world, surely having the results of a proper scientific investigation of its empirical components would be epistemically better than our not having such. So there’s no room for an in-principle objection to be had against using extracathedric empirical methods in philosophy.
And here’s the kicker: in every discipline of inquiry out there that trafficks in any sort of knowledge of the world around us, it has become accepted as necessary to deploy empirically rigorous methods. Not just in the parts of the sciences in which instrument-aided observations are really the stock-in-trade, but also in such fields as history, anthropology, or economics. This is so even in such near-to-philosophy fields as linguistics: people have surely been wrangling about language from the armchair for as long as there have been armchairs, but the explosive growth of modern linguistics in the last half century is due in no small part, I would argue, to its combining both formal and empirical rigor in an unprecendented way. Ordinary, unaided human cognition of the world around us has proved, across a vast range of domains, simply not up to the tasks that serious inquirers have demanded of it. It seems to follow that, if we are pursuing world-involving philosophical projects, then we should follow the lead of, well, everybody else, and look to make the world-involving aspects of our pursuit of those projects at least somewhat rigorous. Figuring out exactly how & where such methodological rigor should be developed would be an excellent methodological question. Alas, the discourse so far still seems somewhat bogged down on the question of whether such rigor should even be developed at all, which strikes me as unfortunate.
The oddness of our current methodological situation manifests rather tellingly in a substantial & unresolved tension at the heart of Timothy Williamson’s overall metaphilosophical package. On the one hand, he argues against a range of exceptionalist theses, endorsing instead a robust realism about philosophical truths; denying that philosophy, unlike the sciences, must take psychological states as its basic evidence; and contending that philosophy need not deploy any special bit of human cognition as part of its psychological underpinnings, asserting rather that human cognition of counterfactuals is all basically of a piece. All of that sounds pretty good to me. But on the other hand, he claims that philosophy is at the same time & in the same places, a properly armchair discipline. Yet if my prior paragraph is correct, then in making this latter claim, Williamson is in fact committing himself to a rather extreme form of philosophical exceptionalism: unlike all the other empirically-informed modes of inquiry, philosophy can safely rely on merely casual, unsystematic observation of whatever bits of the world happen to float by whichever windows our armchairs happen to face out of.
And it does seem to me that TW’s cathedrism is very much not of a piece with the rest of his articulated metaphilosophy. One could buy pretty much everything else he says in PoP, including all the (rather substantial) bets on the particulars of human psychology that he wants to make , and then say, “…and that’s why it’s terribly important that we make philosophy as thoroughgoingly experimental as we can, including doing lots of experimental philosophy, so as to better understand our most central instrument of philosophical inquiry, i.e., human judgment.” Suppose one were to swap out TW’s cathedrism for claims like that — would doing so be at all in conflict with the rest of views in PoP?
And if one resists that swap, then we have to ask: what is it that the world-involving parts of philosophy are doing, such that unlike all other sustained and deliberate forms of inquiry that significantly involve the empirical world, philosophy can get away with a practice of trafficking in its observations and generalizing inferences in a merely informal way? In particular, if one is interested in answering the sorts of questions that TW wants to answer, and if further one wishes to avoid metaphysical and epistemological and even psychological exceptionalism about such questions, and take them to be entirely of a piece with the sorts of questions that everybody else is interested in — then how is one to preserve this robust methodological exceptionalism as well? And, to return to a question I asked in an earlier post, are there any good theoretical reasons to even want to do so in the first place? (I take it as obvious that there might be entirely practical reasons to want to do so. But I would think that those practical considerations cannot ultimately trump the epistemic ones.)
One can also see this tension manifest in a different form in the Papineau paper that was recently linked to here. On the one hand, Papineau emphasizes that he’s not calling for any change from philosophy’s current armchair methods. But on the other, “philosophical intuitions do not qualify as knowledge until they have been subject to serious a posteriori assessment”, and this assessment will frequently (on his account) involve specific & substantial results from the sciences. So this seems to be an odd picture, in which the question of whether philosophers actually have the evidence they take themselves to have, is a question that itself lies outside of the purview of philosophy. I would think we’d want to have a bit more command of our own basic disciplinary inputs than that!
I would note that this seems to me to apply to even Daniel Nolan’s very liberal form of cathedrism. Let me briefly consider two of the “a posteriori armchair” projects that Daniel proposed in his talk, if I remember them correctly. One project that we can conduct (he argues) is that of “gathering the platitudes”; but discerning what’s really platitudinous across the linguistic community at large, and not just a quirk of one’s local subcommunity, will likely require deliberate investigation from outside the linguistic inclinations of one’s fellows. Armchair investigation is, practically by definition, limited to rather local observations. It will also be very hard to avoid general human cognitive foibles like the various sources of confirmation bias, without using some appropriately extracathedric methods. Another project Daniel suggests may be pursued from the armchair is that of taking in competitor theories (which will have an a posteriori component), and then evaluating from the armchair their comparative theoretical virtues: but seeing how one ought apply a virtue like simplicity in a given theoretical situation will require fairly particular sorts of information about one’s sources of evidence of a sort one is just unlikely to be able to gather from armchair methods, as already discussed a bit here. Both of these projects, then, are ones that will be better pursued in a more empirically engaged and systematic manner than Daniel’s invocation of the armchair would suggest.
In short, it sure looks like liberal, aposteriori-involving notions of the armchair are generally going to have empirical methodological commitments and presuppositions of a sort that the rest of the world of inquiry have learned cannot in fact be handled sufficiently from the armchair. If philosophers are to remain methodologically exceptional in this way, they really ought to pony up an account of how some difference — of their methods, or their subject matter, or of the inquirers themselves — can license that exception.
I have just posted a draft of a paper at the x-phi blog that may be of interest to the methodologists in residence here.
The third Arché/CSMN graduate conference will be held November 7-8 2009, at the University of St Andrews. The call for papers invites high quality papers in the areas of Philosophy of Language, Philosophical Methodology, Philosophy of Logic, Epistemology, Metaphysics, Philosophy of Mind, Philosophy of Action/Rationality and Moral Philosophy.
Accommodation and travel expenses for all graduate speakers will be covered. The Keynote speakers are Ernest Lepore (Rutgers) and Susanna Siegel (Harvard).
Ernest Sosa and I are preparing an annotated bibliography on intuitions for a new online resources from OUP. Here is a draft; corrections and suggestions — either for things to alter, or things to add — are welcome.
At the Joint Session a couple of weeks ago, David Papineau’s Presidential Address was of considerable methodological interest. I’m wrestling with it now, and hope to have a few things to say about it in the near future, but in the meantime, interested readers might want to check it out. Here’s the abstract:
The Poverty of Analysis. I argue that philosophy is like science in three interesting and non-obvious ways. First, the claims made by philosophy are synthetic, not analytic: philosophical claims, just like scientific claims, are not guaranteed by the structure of the concepts they involve. Second, philosophical knowledge is a posteriori, not a priori: the claims established by philosophers depend on the same kind of empirical support as scientific theories. And finally, the central questions of philosophy concern actuality rather than necessity: philosophy is primarily aimed at understanding the actual world studied by science, not some further realm of metaphysical modality.
Here’s the official version from the Aristotelian Society. Here’s a pdf I found on the internet; a very cursory glance suggests it’s pretty much the same.
Here is an inconsistent set. I think its members are all pretty widely accepted by philosophers.
I. We do not take the belief attitude towards the contents of fictions; indeed, when we know we’re presented with mere fiction, we’re not even inclined to believe its content. Instead, we do something like imagine its content.
II. Thought experiments are fictions, and judgments about thought experiments are intuitions about their contents.
III. Intuitions are beliefs or inclinations to believe.
We could reject (1); that would be an interesting result, and require some interesting treatment of the content, or a shifting of the world of evaluation, or something. Or maybe we need to generalize our account of intuitions, denying (3). I don’t see that denying (2) is going to be the way out. Any thoughts?
A view that Williamson is at pains to argue against (and I am in overall agreement with him on this matter) is the view that our evidence in philosophy should be viewed as only the psychological. I take the phrase “psychologization of evidence” to indicate a kind of shrinking of one’s entire evidential set to things that count as psychological, and indeed, not just psychological but the particular thin slice of the psychological that have privileged introspective access to, of a sort not easily challenged by the maneauvers of the traditional skeptics. It is of a piece with the kind of thinking that has on occasion led philosophers to think that there is a deep, perhaps intractable “problem of the external world”. The psychologization of evidence here is a kind of retreat, a place to hold a last (probably unsuccessful) epistemic stand.
The way in which the experimental restrictionist challenge to the armchair brings in psychology has basically nothing to do with any of that.
Rather, the restrictionist is looking, not to (artificially) restrict our evidence to the psychological, but rather to (properly) include the psychological in our evidence set, and importantly, let various psychological facts, where appropriate, contribute to determining both what is in that set, and to what other propositions we justifiably may infer from that set. In this way it is completely of a piece with the rest of our evidence, which of course impacts what else can be members of the set of evidence and what can be inferred from the set on the whole. It is just a special case of what is, moreover, the methodological content of our evidence set. That is, very typically (though not at all exceptionlessly) we don’t just know stuff, but we know that we know it, and we know how we know it, and we know why it is that the way that we know it is a good way to go about knowing things. In the most standard & general cases (again, not exceptionlessly), when p is in our evidence set, so too are things like the following:
p is among my/our evidence
q is a reason p is among my/our evidence
q is itself among my/our evidence
r is a reason that q is good evidence for p
Often this is psychological — q may be, e.g., some facts about what was going on with my visual system at some time — but it need not be. E.g., I might know p because a well-run study reported that p; and there’s all sorts of good reasons to trust well-run studies. And note that all of this is perfectly trivial on an E=K approach, since we do typically know all those sorts of things about our evidence. (I’m not sure about the E=K approach myself, but it does seem to be popular around these parts.)
Now, one might think that this is all correct as far as it goes — but just deny that it goes very far at all. Perhaps our evidence about our evidence has something kind of like a Markov property, and once we have secured p among our evidence, then the rest of the methodological backdrop for how p got to be there might just become at that point unnecessary scaffolding, so far as any further inferences from p would be concerned. And (connecting up to BW’s recent post) this might easily seem to be so on a deductive conception of philosophical inference: once p is in fact deductively proved, then, in terms of what I can then go on and do with p, it doesn’t really matter by what means I proved it. Unless one has a specifically psychological or methodological interest, then it is the p itself, not the epistemic backdrop to p, that is of interest. But I take it that this would not in fact be a very popular view of how philosophical inferences work in general, and that philosophers typically think that the way that we go about securing our more general theories is by something much more like inductive or abductive inference, often from facts about at least a handful of cases. And to the extent that that is so, it will turn out that one cannot simply jettison the methodological facts, as they will play a substantial and ineliminable role in one’s inferences.
Perhaps the easiest place to see this is in the comparative weighting of different streams of evidence from different sources; or in different ranges of the same source. Propositions gained from highly trustworthy sources will be granted greater weight than those gained from shadier testifiers; propositions about middle-sized dry goods yielded by vision from a few feet off and in good light will be granted greater weight than those about the far distant, microscopic, and/or ill-lit. And we see a proto-version of such a move, with regard to philosophical judgment in Philosophy of Philosophy, in the vicinity of p. 164.
Robin Hanson, in an intriguing blog post on moral intuitions and moral theory, makes a point that generalizes to other sorts of intuitions “In the ordinary practice of fitting a curve to a set of data points, the more noise one expects in the data, the simpler a curve one fits to that data. Similarly, when fitting moral principles to the data of our moral intuitions, the more noise we expect in those intuitions, the simpler a set of principles we should use to fit those intuitions.”
The noisier our methods, the less subtlety we should allow in our generalizations. This perhaps sets up an argument similar in some ways to Brian Weatherson’s terrific “What Good Are Counterexamples?” paper. If our sources of evidence for “knowledge” attributions are sufficiently noisy, then maybe we should take JTB to be a better analysis than JTB+W (for some anti-Gettier W), even if we don’t have any specific reason to doubt our judgment about Gettier cases in particular.
And note that no number of propositions of the form “a1 is F”, “a2 is F”, and so on can by themselves justify an inference to “All A are F” (or even “most A are F”). That’s the kind of inference we often want to make in philosophy, but in order to do so, one needs information about the sample of ai that one is basing one’s inference on. Typically (and pretty much any time that you’re examining substantially less than the whole population) that information is to be had by knowing what methods were used in gathering the observations in the first place, and what steps were taken to try to ensure that they are a random & representative sample, or to compensate (via matching and the like) for when they aren’t; that the sample size is sufficient large; and so on. The evidence about one’s methodology, often including one’s psychology, is going to play a real role in making such philosophical inferences work — or, as the case may sometimes be, in making them fail to work.
I agree with those who say we should not psychologize our evidence. But it is crucial that we not lose sight of the very important parts of our evidence, even in philosophy, that are already psychological.
Call for papers
Thought Experiments and Computer Simulations: Same End, Different Means?
International Workshop, 11-13 March 2010
IHPST, Paris, FranceScientific committee:
Anouk Barberousse (IHPST/CNRS/ENS)
Rawad El Skaf (IHPST/Paris 1)
Paul Humphreys (University of Virginia, Department of Philosophy)
John Norton (Pittsburgh University, Center of philosophy of science)Whereas thought experiments have long been a common practice in science and an important topic of philosophical analysis, the recent development of computer simulation has not yet paid much philosophical attention to simulations in relation to thought experiments. Conceptual investigation has focused on the links between thought experiments and experiments on the one hand and computer simulation and experiments on the other. However, striking similarities between thought experiments and computer simulations can be found. They are both used when experiments cannot be made. More generally, both are instrumental in answering questions of the type: “What would happen if X or Y were the case?” and in exploring the explanatory potential of theories.
The workshop aims to further investigate the relationships between thought experiments and computer simulation. A major question to be addressed will be whether computer simulations can be viewed as implemented thought experiments, at least in certain cases. The workshop is open both to case studies and to general considerations comparing computer simulations and thought experiments in all disciplines. While physics is undoubtedly an ideal forum for these investigations, talks devoted to fields in which theories play a weaker or different role, like economics, biology or artificial life, are especially welcomed.
Abstracts (1000 words) should be sent to anouk.barberousse@gmail.com by October 1, 2009.
The workshop language is English.
Important dates
- Submission deadline 1st October 2009
- Notification about acceptance: 1st November 2010
- Workshop: 11-13 March 2010
There’s a rather odd dialectic in Knobe & Nichols’s “An Experimental Philosophy Manifesto“. In §2, they articulate why they think that experimental inquiry is relevant to philosophy; §2.1 develops the sensible suggestion that experimental data can play an important but indirect role in philosophical theorizing:
When experimental research is understood in this broader context, one can easily see how it might have important philosophical implications. It is not that the actual percentages themselves are supposed to directly impact our philosophical inquiries. Rather, the idea is that these experimental results can have a kind of indirect impact. First we use the experimental results to develop a theory about the underlying psychological processes that generate people’s intuitions; then we use our theory about the psychological processes to determine whether or not those intuitions are warranted.
I agree with Knobe and Nichols that the legitimacy of this kind of empirical role in philosophical theorizing is plausible—indeed, I agree that it is obvious. Then they go on, immediately following the presentation of this truism, to write this:
But, unfortunately, what seems obvious to one philosopher often seems obviously mistaken to another. Instead of greeting these methodological remarks as simple truisms (which, we continue to think, is what they really are), many philosophers have reacted by offering various sorts of objections.
But none of the four objections they consider even pretend to be objections to the modest truism quoted above. (Additionally, none of the four considered objections include reference to individuals defending them.) For example, they cite the suggestion that intuitions among philosophers are more likely to be correct than are intuitions among the folk. I’ve taken this line, but I haven’t used it against the obvious truism above; neither, as far as I know, has anybody else. This line is an objection to a much stronger, much more incendiary view that many philosophers attribute to the experimental philosophy movement: namely the view that there is empirical evidence that demonstrates (or at least, raises serious worries indicating it likely) that traditional armchair philosophical practices represent a seriously flawed methodology. Of course there are problems in attributing views to movements — I’m confident that some experimental philosophers don’t hold that more incendiary view — but I’m also confident that some rather vocal exemplars of the movement do. (There’s a reason x-phi has taken the icon of the “Armchair in Flames”.) It’s disingenuous, therefore, for Knobe and Nichols to characterize experimental philosophy more weakly, then dismiss objections to it as missing the point; the objections mentioned are objections to a much stronger characterization.
I don’t think I’ve ever seen any philosopher object to the truism cited in the first quoted paragraph.
I’ve been asked to pass along this announcement for what looks like a pretty interesting event:
Conditionals and Conditionalization conference
Formal Epistemology Project (FEP), Centre for Logic and Analytic Philosophy, Institute of Philosophy, University of Leuven, Leuven, Belgium, Sept 4-6, 2009