Pascal Engel on Experimental Philosophy

The distinguished French philosopher Pascal Engel has an interesting new piece exploring the broader implications of recent work in experimental philosophy. 

At one point, he suggests that experimental results might be philosophically relevant if one has already accepted an 'anti-foundationalist' conception of the relevant domain -- but that the experimental results do not themselves provide any evidence for such a conception. In other words, the suggestion is that these results should only be considered relevant by philosophers who think they have some kind of independent (non-experimental) reason for adopting an anti-foundationalist view.

I'm certainly no great translator, but I was curious to hear what people here might think of Engel's argument, so I did my best with the relevant passage:

Am I not being too severe? Are there not domains in which experimental philosophy can have a positive impact, in which it can establish or reinforce the truth of certain philosophical theses? I believe so, but the consequences are doubtless more limited than the promoters of these empirical methods would suggest.

In epistemology, recent work has discussed knowledge attributions and the question as to whether, in cases of so-called ‘pragmatic encroachment,’ knowledge is not fundamentally relative to practical stakes... Empirical studies can help us to better understand these effects. Can they also allow us to derive consequences with regard to the validity of evidentialism in the theory of knowledge? It is doubtful.

In metaethics, experimental philosophy can play an important role if one is an expressivist, i.e., if one maintains that our moral ideas and principles do not rest on a moral reality but rather are ‘response dependent.’ It then becomes important to know what our responses are, and moral psychology can therefore play an essential role. But does any of this give us the slightest argument in favor of expressivism in metaethics? In my opinion, no.

[…] I understand perfectly well that philosophers who have deflationist, relativist or skeptical positions in epistemology (relativism, contextualism), in metaethics (expressivism, minimalism) or in metaphysics (positivism, eliminativism, quietism) would be tempted by these developments. But I do not see why philosophers who are not tempted by these developments, who still have ‘foundationalist’ positions, should adopt these methods.

Any thoughts?

Comments

[John Turri]

Thanks for this, Josh.

"Can [empirical studies] also allow us to derive consequences with regard to the validity of evidentialism in the theory of knowledge? It is doubtful."

To the extent that it is doubtful, I think that's probably because evidentialism hasn't been explained well enough to have clear and testable empirical implications.

Anyway, I disagree that it is doubtful. Once we're told what evidence is, and how it supposedly must be related to belief so that belief is justified/counts as knowledge, we can look and see whether: (1) within an overall pattern of competent epistemological assessment, people judge beliefs to be justified/knowledgeable only when related to evidence in the specified way; (2) whether and how the concept of evidence or justification that evidentialists have articulated relates to ordinary people's ways of evaluating beliefs and intellectual performances; (3) whether there is any evidence from the cognitive sciences that justified believers "have" the relevant sort of evidence (e.g., conceptually structured perceptual states, beliefs standing in the relevant causal-inferential relations, or whatever).

[Mark Alfano]

I take on some of these questions, though only with respect to ethics, here: http://alfanos.org/Blog/?cat=10

[Florian Cova]

Note that Engel speaks here only about "positive impact". If we distinguish the "negative impact" of experimental philosophy (i.e. assessing the truth of certain premises used in philosophical arguments) from its "positive impact" (i.e. proving a certain philosophical thesis to be true), I think this thesis seems more reasonable.

Clearly, it would be wishful thinking for a foundationalist to think that his position are immune to certain empirical challenges (at least if he appeals to intuitions). However, I think that it is not what is claimed here. I interpret Engel as saying that the foundationalist is not immune to such challenges, but has no "positive" gain to earn from experimental philosophy (for him, nothing can be proved true by experimental evidence).

[Jonathan Weinberg]

I consider most of my own arguments to be driven by (though they need not be committed to) a pretty straightforward realism about epistemogy, ethics, and so on.

Are we correctly understanding what Engel is meaning by what's being translated as 'foundationalist' here? Is it the same as the position in epistemology that we use that term for in Anglophone philosophy?

[Joshua Knobe]

Jonathan,

The phrase I translated as 'foundationalist positions' is, in the original French: des positions « fondationnelles »

It definitely looks as though Engel is understanding this notion in a slightly different way from the way in which you are probably thinking of it, since he contrasts foundationalist positions with deflationist, relativist and skeptical positions (apparently implying that these latter positions involve a rejection of foundationalism).

Florian,

Reading Engel's discussion, it looks as though he is continuing an earlier debate with you or in some other way responding to your work on this topic. Do you think you might be able to give us some further background about the earlier discussion here?

[Florian Cova]

Hi,

I cannot help with disambiguating what Engel means by "positions fondationelles", I am not familiar enough with his work. However, I suspect he means something like positions for which philosophical knowledge is reached through a priori reliable evidence. He somewhat opposes that to positions for which nothing is enough it itself to provide philosophical evidence, such as coherentism (no single evidence is enough in itself) or various positions that make philosophical truths dependent from our responses. I grant this is not very clear. My overall feeling is that his claim is that there might only be a positive import of experimental philosophy (as opposed to critical)for those who consider that philosophical truths are determined by our responses. Though, I don't think he claims that other positions are totally immune to experimental challenges. If he does, then it clearly becomes very weird.

As for Joshua's question, this is not a continuation of an earlier debate with me. To give you the context (and explain why my name appears) : the author of this blog (not Engel) has recently written a review of Engel's book "Epistémologie pour une Marquise". This book is intended for a wide audience, but strangely featured a very short (one paragraph) and cryptic dismissal of experimental philosophy as "blind". The author of the blog was legitimately puzzled and expressed his puzzlement. In a comment, I gave my interpretation of this puzzling criticism. Engel's current post is supposed to be an explanation of his previous criticism (though I still don't get it) and he gently signals that my former interpretation was wrong (hence my name).

[Florian Cova]

However, on a related note, Engel's worst grief about experimental philosophy is that experimental philosophy (specially the so-called negative program) is tied to eliminativism about concepts (http://www.unige.ch/lettres/philo/enseignants/pe/Engel%202009%20Philosophical%20thought%20experiments.pdf).

This seems a very strange and inaccurate claim to me, but I would not want to rely on my intuitions. Apart from Edouard, who here is a concept eliminativist?

[Joshua Knobe]

Thanks to Florian for this background information. The review of Engel's book explains that it is presented as a dialogue between an epistemologist and a marquise and that, quite apart from anything in the philosophical content, the dialogue itself is a pleasure to read.

In any case, the relevant paragraph from Engel's book is (again, in my possibly flawed translation):

If philosophy often appeals to thought experiments, that is not because it is purely conceptual and removed from all control of experience, nor because it is an empirical discipline like psychology or anthropology. It is for this reason that what is today known as ‘experimental philosophy,’ an attempt to test our philosophical ‘intuitions’ (particularly in ethics) by means of psychological experiments, is blind. Conversely, a purely speculative ‘armchair’ philosophy is empty.

[This passage is presumably an allusion to Kant's famous claim: 'Thoughts without content are empty, intuitions without concepts are blind.' But I'm not sure quite what it would mean to say that experimental philosophy is blind. Is the idea that our work involves intuitions without concepts?]

Florian's (very helpful) thoughts on these metaphilosophical issues appear in his first comment at: http://www.philalethe.net/post/2012/03/10/Engel-sur-la-philosophie-exp%C3%A9rimentale.

[pascal engel]

Many thanks for your interest in my grumblings and mumblings about experimental philosophy, especially since they are in French. Too often French speaking philosophers complain that because they write in a minor language they are not read by their English speaking colleagues (this is bad particularly in Quebec , where two important linguistic philosophical communities coexist). So I am particularly happy that the language issue is less of a problem now.
I am afraid, however, that the contents of my remarks is much less interesting than what the commentaries which Joshua Knobe's translation of part of my text seem to anticipate. My critiques of X phi are disappointingly similar to those which have appeared in the literature, and will very likely seem boringly repetitive to those who know the literature much more than I do. But since I am invited to react to these comments, I am happy to do so. My French post being rather long, I am not going to repeat it, and shall just focus on some points (I hope, although do not assume, that your readers will have a look at my French post).
The context is the one reminded by Florian Cova. In a small book meant to introduce playfully French readers to some issues in the philosophy of science, I have a chapter on thought experiments at the end of which I say that thought experiments are neither based on intuitions, either rational or empirical, nor on reasoning, but are part of conceptual analyses and philosophical arguments of a broader kind. In this sense I said that experimental philosophy , if it meant to test our intuitions only is blind, and armchair philosophy if it is supposed to go completely without intuitions, is empty. This rather banal remark has seemed to some an excessively severe judgment about experimental philosophy, and I wonder why !
My - very humdrum - point was that there is a tendency in-X phi to take thought experiments as being only ways of testing our ordinary intuitions. Although there has been a tendency in current analytic philosophy to overemphasize the use of intuitions within thought experiments in order to give plausibility to some philosophical views or to attempt to refute them, it seems to me that, both on the part of “armchair” philosophers and on the part of their experimentalist critics, one forgets that thought experiments and intuition pumps are part of more extensive argumentative strategies backed by conceptual analyses and sustained arguments. In this sense, for instance, reliabilism in epistemology does not stand or falls only on the basis of Bonjour’s clairvoyant case or on the basis of Lehrer’s Truetemp case, whatever weight the proponents of these thought experiments may want to give to these. If one considers, for another example, Strawson’s famous chapter II of Individuals, it’s not clear to me that it can simply be compressed into form of a thought experiment ( although there is undoubtedly one in it) and separated from the rest of Strawson’s strategy in the book. As everyone knows there is a transcendental argument there, and it’s not clear that the thought experiment can be separated from it. To take another much discussed example, it’s not clear to me that Kripke’s case for direct reference puts so much weight on the Schmidt / Gödel example
Setting aside all the issues having to do with what is exactly tested in empirical studies and how, there is also an issue about the nature of the project of testing intuitions many critics of X phi have also expressed the concern that intuitions do not really test philosophical analyses and concepts and are not really evidence for these, but that the relation between them is much more of a constitutive type. Much of X phi is written on the assumption that if the intuitions which are evidence for these analyses fail or are show hazy, then the analyses are wrong. But that is not clear to me. Neither is it clear to me that philosophy is only or mostly conceptual analysis: much of meta-ethics , for instance, is not meant to analyze, for instance, the concept of norm or the concept of moral property, but normativity and moral properties themselves. Perhaps there used to be a style of analytic philosophy as what Ryle called a “second order” kind of subject, dealing with concepts and meanings instead of things and the world, but it seems to me that rather few philosophers today think that philosophy is confined to conceptual analysis (perhaps some Wittgensteinians like Peter Hacker do, but they seem not to be the majority, and possibly some philosopher favoring contextualist approaches in epistemology where knowledge ascriptions are the main focus of interest, and not the nature of knowledge).
My French post was mostly concerned with the impact and philosophical significance of experimental philosophy. I discerned three “metaphilosophical “ strands which seem to coexist with what is today called experimental philosophy : (i) a very sensible claim that philosophy ought not to be divorced from sciences, in particular cognitive science, and that it should not be confined to its own games, so to say, (ii) what I called a Menchevik or moderate project consisting in doing psychological experiments upon folk intuitions about various subjects (mostly in ethics, but also in epistemology and in some issues which respect to causality), (iii) what I call a Bolchevik project calling for a radical reform of the methodology of philosophy, the rejection of conceptual analysis and all form of a priori knowledge, and a radically naturalist project aiming at reconstructing the field, often with eliminativist echoes. I have no objection to either (i) or (ii) and welcome such studies, which are very often interesting and informative, and they are, at their best interesting contributions to social and cognitive psychology. But I have serious doubts about the Bolchevik project. It’s not clear to me that empirical studies about folk judgments affect, and ought to affect, traditional problems in epistemology like skepticism, the existence and nature of a priori knowledge, or issues in metaphysics and metaethics. It may be that many philosophers who recognize themselves as belonging to the “experimentalist” camp do not actually subscribe to the Bolchevik project, and conceive of their studies as having more limited and local impact. Many recent declarations in that respect seem to me to contrast with the enthusiastic manifestos launched ten years ago. But it would be nice, given that there are manifestos, hence pronouncements and hopes, to know more about what is expected. I suspect that the returns are more modest than what some advertisings seem to make us expect. But once again, I clam no originality for this!
There does seem to me to be positive and not only negative returns to be expected from such studies, but here too probably less dramatic than expected .In the passage translated by Joshua Knobe from my text, I said that I would expect studies about our ordinary moral , epistemological, aesthetical and other kinds of philosophical relevant judgments to show us certain facts about the nature of our responses and reactions to ethical and epistemological claims. If so, these studies are potentially relevant for views about metaethics and metaepistemology which , so to say, are on the Euthyphronic side of the Euthyphro dilemma, ie which take morality and knowledge to depend essentially and constitutively upon our responses, judgments and reactions. The idea is that if our normative think in one domain or other is so response dependent, we ad better be informed about what our responses ar. Hence it would seem that anti-realist views of ethical or epistemological norms would be reinforced, or at least better documented, by a naturalistic study of our “intuitions” in the relevant domains. But here again, I wonder how far we can go.
It’s not clear to me that debates about metaethics , such as the opposition between expressivists and cognitivists , or debates about epistemology, such as whether evidentialism or reliabilism is the correct view of justification, can be affected, let alone settled by empirical investigations into our judgments, and as far as I can tell even expressivists such as Blackburn, Gibbard or Schroeder do not rely very much or at all upon such empirical studies ( although some are tempted by a broadly evolutionist view of ethics : if such studies are meant to reinforce evolutionary psychology , fine, but that is a moot issue too). In other words the truth of expressivism vs cognitivism , or of reliabilism vs evidentialism seems to me independent from empirical studies about our ethical or epistemological intuitive judgments . Which does not mean that experiments might not show that these judgments are massively “expressivist”, say. But if they did, would that convince an ethical realist ? I doubt it. Or suppose a study massively showed that people are Kantians in their ordinary judgments. Would that establish the truth of Kantian ethics? It seems to me no at all. In that respect too, I do not see how studies about our intuitive judgments can affect our ( armchair) theorizing in metaethics or in epistemology ( bt the list could include metaphysics,
John Turri seems to disagree with me on that. He seems to say that studies about how people attribute or register evidence for they beliefs can explain , in some sort of causal way, the concept of evidence ( thanks to Mark Alfano for the references to these studies). I agree that it can shed light on what we mean by “evidence” when we say “this is evidence for this or that”, but does that show something about the nature of the evidential relation? Unless one thinks that there is no more to this relation than what we usually think about it, I do not see how it can show anything of this sort. If our experimental studies are able to tell “whether there is any evidence from the cognitive sciences that justified believers "have" the relevant sort of evidence” , this seems to presuppose that the concept of evidence that we are testing is already understood. And how to understand it is what is in question. Of course I agree that we must start from our intuitive judgments about evidence, as fragile and biased as they might be, in order to determine what evidence might be, but I also hope that we can elaborate or it, correct it, refine it, in order to have a genuine philosophical conception of it. In this sense, it seems to me that philosophical analysis indeed evolves from expertise and skill on the basis on intuitions, but does not limit itself to these in their brute and uneducated form.
Of course, behind such studies I see that there is a Nietzschean, or quasi Nietzschean, project of doing a genealogy of ethics – a quite Bolchevik one, I take it. As the quote from Knobe and Alexander ‘s manifesto says, the hope is to find the causal patterns underpinning or ordinary judgments, with the hope of determining what these judgments are actually about. But, even putting aside the issue of whether this would be sufficient to establish this kind of anti-realist view of ethics the expressivist, or anti-realist stance seems to be here a presupposition or at least a premise of the whole enterprise , rather than an upshot of the project.
When I said that X phi seems to me to marry better with “anti-foundationalist” views, such as deflationism or minimalism in ethics o in other fields , I meant with views which have renounced the project of giving a foundation of ethics ( here was referring to a recent book by Ruwen Ogien, with the deliciously French title “De l’influence des croissants chauds sur la bonté humaine”, here Ogien just says that). Can experimental philosophy be friendly with realist views in ethics? Of course that depends. If one is convinced by Parfit’s arguments in On what matters, it would seem not, since he discards all kinds of naturalism. Perhaps less heavy duty kinds of cognitivism can be married to X phi. I do not here see why not, but again, the bulk of the argument for realism would not rest on our patterns of ethics intuitive judgments. Jonathan Weinberg notes that his own views, which I classed more in the Bolchevik camp, are actually quite realist. Having read a former paper by him with Stephen Stich on metaskepticism and other cross cultural studies where relativist views of ethical and epistemological norms are suggested, I would have thought that these studies weigh more in favor of relativism or expressivist views than in favor of realism, but I prepared to be refuted on this.
When a book tells me that we can do without concepts, at least I take it that it is meant to be a rather eliminativist view, although I know that there are many ways to eliminate.
So, it seems to me that there are many ways in which the kind of empirical studies done by X-phers can be instructive and useful for various branches in philosophy. But not all, and not in all sectors, and depending on the kinds of philosophical views that on has independently of the experimental results. It seems to me that X phi has in part grown out of serious flaws within the practice of analytic philosophy during the last decades, in particular overemphasis on the role of intuitions, but these defects, once understood, do not seem to me to call for a radical revision of philosophical methodology. It may well be that it’s what experimental philosophers have been saying all along. If so, my remarks are just trivial indeed.

[Joshua Knobe]

Pascal,

Thanks so much for this clarification. I think you are bringing up some very important issues here, but I'm not sure I agree with your claim that the use of experimental studies in any way presupposes some controversial metaphilosophical view involving anti-realism, response dependence or anything of the kind.

Consider an analogy. Clearly, the aim of research in astronomy is not to study the responses of telescopes but rather to figure out how things really are in our universe. Still, one might well find that the best way to make progress on certain questions in astronomy is to engage in a detailed study of telescopes. The reason for this is, quite simply, that by learning more about telescopes, we can get a better grasp of what really is going on in the universe.

Now consider the study of moral philosophy. Suppose, if only for the sake of argument, that the aim of moral philosophy is to understand an objective property of actions that does not depend in any direct way on the emotions or attitudes that these actions elicit in human beings. Even if we make this assumption, it seems perfectly plausible that the best way of making progress in moral philosophy might sometimes involve studying the emotions and attitudes of human beings. Here again, the idea would simply be that examining people's emotions and attitudes can sometimes be the best way of gaining an understanding of objective features of the world. (For example, if we find that certain aspects of our cognition are biasing our judgments, we can use this knowledge of our own biases to arrive at a more objective and impartial understanding of the external world.)

This basic point has been developed in rich detail in the recent work of Regina Rini, and in my view, she makes a very strong case for it -- the importance of experimentation, it seems, does not depend on any strong claims about the nature of moral inquiry more generally.

[pascal engel]

Dear Joshua

I have not said that the use of X phi *presupposes* controversial metaethical views, but only that it seems to me that it should be of more interest and more attractive for those who have anti-realist ( non factualist, expressivist, subjectivist, fictionalist, relativist ) views of ethics and epistemology. Why ? Because these views hold that morality or knowledge ( or aesthetic judgment) depend essentially upon our responses or attitudes. So it would seem to be important to have a good knowledge of what these attitudes are. I gather that Hume or Hutcheson, or our Blackburns and our Gibbards are to be more favorable to an experimental study of human sentiments and feelings than Kand, Sigdwick and our Parfits and Koosdgards.
But the relation is not analytical or implicational. I do not believe that arguments in meta-ethics presuppose empirical work, just that may seem more plausible with it. Nor that empirical work implies any view in meta-ethics in the strict sense. For instanceI am not sure what a "fitting attitude" or "buck passing" conception of value would make of these results.

As to the telescope, indeed if we accept the idea that truth in a given subject entails what Wright calls "cognitive command", and if we suppose that we have actual knowledge of the stars, tracked by our instruments, then some properties of the tracking are worth studying empirically. But again no strict implication. But it seems clearly more important for the anti-realist to have a good grasp of our responses, for his view is that our knowledge of "moral properties" depends upon those responses in a way a realist does not acknowledge.

So I think that no metaethical, metaepistemological or metaphysical view is presupposed by empirical work , which in turn does not lead by implication to any meta-view, although it seems to me that anti-realist views should be more bedfellows of X phi. But again there is a spectrum here. Eliminativism or if you prefer deconstructionism à la Stich lies at the extreme end of the spectrum. By the way, I have never said or implied in the quoted paper or elsewhere, contrary to what Florian Cova asserts above , that "experimental work leads to eliminativism about concepts". I am neither saying the experimental work presupposes such views nor that it leads to them, but that meta-ethical or or other meta-views which make judgmental responses essential to he nature of a phenomenon would seem to be more concerned to learn about the nature of these judgments. If some of the X phi's project are truly Nietzschean, do they fit well with kinds of realism in ethics ? So much has ben made of Nietzsche that I am prepared to anything these days, but if one interprets comes to me and says that Nietzsche is a moral realist I would be surprised.

In addition it turns out that many X-phers have expressed attractions for such anti-realist views ( about epistemic norms for instance), and have claimed that some metapysical issues are undercut by experimental work ( e.g. skepticism). So they do seem to be at least tempted to draw some "meta- " conclusion from their work,and not to take it as neutral with respect to such broader philosophical issues.

[pascal engel]

Ps Apologies to Kant, Sidgwick and Korsgaard!