Print This Post Print This Post

Chapter One is essentially a ground clearing exercise. Appiah’s aim is to argue that experimental philosophy is not the innovative and threatening enterprise that it might seem: instead, it is a return to philosophy’s roots. Philosophy has traditionally been closely informed by scientific work, and the best philosophers have often engaged in science themselves. It is the era of conceptual analysis divorced from mere empirical engagement that is the aberration, not the turn to the empirical.

I’m not sure how seriously to take Appiah’s genealogy. He begins with a gesture that undercuts his history, by noting how enterprises like the one in which he is engaging are motivated exercises in forgetting as much as recall: we cut the history of the enterprise to suit our current conception of it. Appiah cites Renan on history as a kind of motivated forgetting. He could have equally cited Kuhn on the Whiggish histories of science. In any case, some of what he says is uncontroversial: for much of its history, philosophy was deeply engaged with the empirical. I think, however, that Appiah overlooks the extent to which recent philosophy – and recent philosophy alone – has adopted formal structures that mirror science: the move from big ideas to far narrower and specialized questions and a corresponding move to the journal article as the unit of publication, rather than the book. This allows work to be far more rigorous and detailed, and debates to advance much more rapidly. In Kuhnian terms, philosophy – in the analytic tradition – has become puzzle-solving. And that’s a radical departure from the past. So perhaps we should come to a mixed verdict on the question to which philosophy turned its back on the science: it adopted the form of the sciences at the same time as it rejected the content. This might make reincorporating the content that much easier.

Of course, this is not inconsistent with Appiah’s claim that the neglect of the empirical is a recent phenomenon. I do wonder whether the connection between philosophy and science in the past wasn’t more focused on mathematics than on the more contingent sciences; even Aristotle, for all his own work on the natural world, suggested that knowledge concerning the contingent was of lesser value than knowledge concerning the necessary.

A final aspect of the ground clearing enterprise: Appiah is concerned with experiments in ethics, and therefore needs to dispense with the supposed naturalistic fallacy. His strategy here is to attack the ‘fallacy’ on its own grounds: He constructs a deductive argument moving from nonmoral premises to moral conclusions. Once again, he doesn’t mean us to take the argument all that seriously: as he says, it ‘has more than a whiff of sophistry about it’. His point is this: though the argument may be sophistical, it does demonstrate that the claim that moral and nonmoral propositions are categorically different, and the further claim that one cannot be deduced from the other, are from clear, never mind clearly true. We should not be sure that we even know what these claims mean. Moreover, as Appiah goes on to show, there are certainly and uncontroversially connections between the normative and the empirical.

All of this is well-taken. However, in a book that takes science seriously, I would have expected a discussion of identity claims. The naturalistic fallacy is often taken to follow from the open question argument: for any naturalistic property or set of properties, it is supposed to be an open question whether that property or properties is identical to some normative property. A competent speaker can, without making a mistake, wonder whether the two are identical. But the open question argument is sound only if identities have to be a priori, and the identities established by science are a posteriori. A competent speaker may well wonder whether water is H2O; nevertheless, water is H2O. A posteriori identities are frequently surprising, so the intuitions of competent speakers are neither here nor there.


  1. 1. Posted by Jeff Huggins | March 3, 2008 8:35 pm

    Thank you Ethics-Etc for hosting these groups. They are very informative and helpful. Thanks also to Kwame Anthony Appiah for such an informative and helpful (and enjoyable!) book. And thank you Neil for your helpful summary.

    I have a few comments regarding the Prologue and The Waterless Moat, all made in the context of the entire book, which I’ve read. My comments are mainly supportive, and I agree with Appiah’s arguments, at least as far as I understand them (with perhaps a couple exceptions later in the book that I’ll mention as we get there.)

    In my view, briefly, I think it helps to consider the following:

    “Science”, in a narrow sense, taken alone, can describe, understand, and explain, of course, but can’t justify. If we limit the term ‘science’ to observed data, pattern recognition, interpolation, and so forth, without including (in the term) much reason, imagination, justification, and so forth, then science is, of course, limited in those respects.

    On the other hand (it seems to me), “philosophy”, IF seen only as disembodied “pure thinking”, and IF not connected in any way to the warp and woof of the universe or to the warp and woof of human life, can’t get very far. Indeed, as Appiah seems to argue, key philosophies have always been tied, one way or another, to at least one aspect of terra firma. Aristotle observed. Many philosophers these days intuit (but they, and their minds, are human, and science thinks it knows, with a reasonably high degree of probability, roughly how humans developed). In any case, after all, what is disembodied pure thinking?

    So, I think, neither “half” of the picture can accomplish the whole task, the wisdom-quest, on its own. They are both necessary to bring the whole picture into substantially better focus, (in this case) a picture of morality that is both explanatory and justificatory.

    These two “halves” of the picture are not precisely the same, of course, in terms of detail or in terms of several other important considerations. They are different but not unbridgeable, and they “live” in the same universe, of course. So, one must ask, if they aren’t the same, what do we mean by considering both of them, bringing them together, or bridging them? How do the explanatory and justificatory relate, and what are their common elements?

    In my view, the essence of the answer is in finding, and understanding, what’s common between them, i.e., between the two halves. It’s in finding what they both “point to” and “indicate”, in finding what “bridges” them. On page 2, in his Prologue, Appiah uses the very helpful word “concordances” as part of his phrase “the concordances between our traditions of explicit moral reflection and these posited primatives.” Without going into the matter here and now, a central and foundational “concordance” involves the notion of ongoing human survival.

    Finally, I also agree with this quote, from page 1 of his Prologue, which captures much of the essence of “Experiments in Ethics”:

    “Yet in making our choices we must sometimes start with a vision, however inchoate, of what it is for a human life to go well. That was one of Aristotle’s central insights. It is my argument that we should be free to avail ourselves of the resources of many disciplines to define that vision; and that in bringing them together we are being faithful to a long tradition. In the humanities, I think, we are always engaged in illuminating the present by drawing on the past; it is the only way to make a future worth hoping for.”

    I began my own book (aside from Acknowledgments and Preface) with a Prologue in the form of a short dramatization, titled “The Reconciliation of Science and Moral Philosophy.”

    I hope these thoughts are helpful in some way. And, I repeat the “thanks” I mention above.

  2. 2. Posted by Roman Altshuler | March 4, 2008 1:43 am

    Hi Neil, thanks for the helpful summary, and for taking note of some of the quirks of Appiah’s exercise in selective historical forgetting. There are a couple of points that, I think, are only implicit in what Appiah says, but that help to situate this chapter:

    First, his emphasis on the experimental and empirical concerns of pre-20th century philosophy seems aimed not simply at opposition to those who would purify philosophy, but also at some overly exuberant proponents of x-phi, who sometimes present it as a radical break with the history of philosophy.

    Second, Appiah’s major target here–or, at least, the target that his arguments really address–is not just those who want to preserve a role for philosophy separate from experimentation, but those who attempt to justify this separation through a very specific and idiosyncratic (though, for a while, dominant) mode of conceptual analysis, i.e., followers of the legacy of the linguistic turn. The kind of experimentation that involves asking for people’s intuitions often seems to me (though I could be very wrong) as an extension of the linguistic turn, which may conflict with the turn to empirical data in psychology, which is motivated precisely by a desire to get away from the idea of pure linguistic analysis. But the idea that pure philosophical analysis is identical to pure linguistic analysis is a very recent one, and rebutting the latter is not the same as rebutting the former. (A specific point on this: questioning the naturalistic fallacy as a thesis about the logical relations between intrinsically moral and non-moral propositions does not cast doubt on the naturalistic fallacy under non-linguistic interpretations.)

    I worry that the emphasis on the experimental tendencies of many historical figures really might be a red herring in some ways (as in Neil’s point about Aristotle’s interest in necessary truths). An example: Kant did write about the volcanoes on the moon, but he didn’t consider this work to be philosophy. And he did write philosophically about ether, but this is because he thought its existence could be proven a priori and was not an empirical matter. In fact, Kant was very strict about maintaining disciplinary boundaries, both within and without philosophy. Given this, how is the fact that he had experimental interests relevant to the claim that philosophy has historically been experimental? Similarly, in the Hegel example Appiah cites, it is true that Hegel criticizes Newton for experimental failings; but of course his point is that good metaphysics is a precondition of good experimental work–that we have to get the role of reason straight before we can apply it competently to the study of nature.

    So there seem to be two currents involved: on the one hand, a current of drawing philosophical conclusions from experimental evidence. On the other hand, a current of taking philosophy as methodologically independent of experimentation and involved in providing interpretative schemas for the sciences. These currents don’t always fit well together, and represent very different relations of philosophy and experimentation. If we follow the first current in moral philosophy, it turns out that we can get normative conclusions from empirical evidence. If we follow the second current, we might see normative philosophy as providing us with a schema for what normative conclusions to make from experimental findings.

  3. 3. Posted by S. Matthew Liao | March 4, 2008 7:00 pm

    Thanks for the very nice summary, Neil! As you know, I’m very sympathetic to the experimental philosophy project, but (like Roman) I too wonder whether the genealogy that Appiah has presented will persuade those who aren’t already sympathetic to this project. In particular, they might accept the ‘descriptive’ genealogy but argue that nevertheless one *ought* not to do philosophy this way. In that case, it’s hard to see how the ‘descriptive’ genealogy would challenge the normative claim.

    This said (as Jenn Neilson has helpfully pointed out in our offline reading group), Appiah could be trying to persuade those who haven’t made up their minds. In that case, his strategy may be effective.

    A related point: As Neil has pointed out, Appiah tries to attack the argument from naturalistic fallacy head-on. I was puzzled though by the point that Appiah tries to make from his arguments. On p. 28, he says,

    What [his arguments] suggest, at the very least, is that it isn’t entirely clear what we mean when we say that you can’t derive a moral claim from nonmoral ones. And until we understand the claim, I think we are justified in not allowing anyone to rely on it too heavily.

    I found this ‘lack-of-clarity’ argument odd, because plenty of things have been said regarding the naturalistic fallacy. So, it seems that the issue can’t just be the lack of clarity regarding this argument.

  4. 4. Posted by Roman Altshuler | March 5, 2008 6:33 am

    My worry is also that if the goal is to persuade those who haven’t made up their minds, this chapter isn’t likely to persuade in the right way. There is a world of difference between showing that experimentation has historically been crucial for drawing philosophical conclusions, and showing that the historical figures we consider philosophers were often also experimenters. The first point is the important one; the second, on the other hand, provides no justification for any claim about the relation between philosophy and experimentation. There is a bit of both in this chapter, but they’re lumped together as part of making the same point. It would have been nice to see more of the way philosophy has traditionally depended on experimentation. (That said, I’ve spent the past few days trying to peel raspberries, and am starting to wonder if there’s some trick to it.)

  5. 5. Posted by Neil Levy | March 5, 2008 2:24 pm

    For what it’s worth, I think the way to show that the empirical is relevant to the philosophical is to go straight to the evidence. Once one has delved into, say, the heuristics and biases tradition, it is difficult to be quite as comfortable with appeals to intuitions as one was before.

    Appiah’s strategy might be to appeal to people by saying ‘see, this isn’t such a radical break at all, so don’t worry’. Some people may be attracted by this. But some others would be more attracted by the claim ‘this is a new departure from philosophy as it has traditionally been practiced’. Given that such different approaches can be expected to work with different people, perhaps we shouldn’t read him as attempting to persaude the uncommitted at all.

  6. 6. Posted by Jeff Huggins | March 5, 2008 6:29 pm

    I agree with Roman’s point when he says:

    “There is a world of difference between showing that experimentation has historically been crucial for drawing philosophical conclusions, and showing that the historical figures we consider philosophers were often also experimenters.”

    And, I would imagine or guess that Appiah would also agree with that point.

    That said, it seems to me that it’s rather easy to demonstrate that observation, experimentation, and/or so forth (i.e., some “anchor” back to the warp and woof of the universe and/or to the warp and woof of human life) are central, and have been central, to nearly any school of philosophy, past or present. A book that does so (in detail, and beyond the degree that Appiah’s book does so) could well be very helpful to certain audiences, if it doesn’t exist already.

    On the other hand, covering the matter from the standpoints of history and less detailed examples (than those that would be systematically involved in the approach mentioned above) is, of course, very helpful, insight-giving, thought-provoking, and ground-clearing. Indeed, some audiences approach these matters mainly with history and self-identity and precedence in mind. For example, some people may hold on to one interpretation of one sentence that Hume (for example) wrote at one time, without carefully considering his context, his other sentences, and the fact that he died before Darwin was even born, to mention just a few considerations. For them, it might be more powerful to consider some of the history that Appiah points out than to read an entire detailed analysis of all the ways in which nearly every school of philosophy anchors itself (to varying degrees) to the physical universe and/or to aspects of observed human behavior.

    Besides, it’s great to read a well-written, accessible, insightful, 204-page book (without the notes), yes?

  7. 7. Posted by Constantine Sandis | March 6, 2008 12:22 pm

    Thanks for the great critical summary Neil. Regarding the Hume quotation that Jeff Huggins mentions, he is right that it’s taken (seriously) out of context. Annette Baier has located the passage, it’s from “Of the objects of allegiance”, Treatise Book 3 part II. sec 10.15 . Hume writes:

    “In this particular, the study of history confirms the reasonings of true philosophy”.

    But what of other particulars (e.g. past regulaties and ‘true philosophy’ regarding inductive reasoning)? This is hardly a sweeping general statement about the study of history. Also, the relation between Hume’s histories and his philosophy is a vexed question. For example the explanations he gives of the actions of Kings and Princes do not easily fit the ‘Humean’ theory of motivation. And we all know how unkeen he sometimes was to apply ‘true philosophy’ to his day to day life.

    As for the attack on so-called ordinary language philosophers it’s far from clear why we should derive an ‘ought’ concerning language use from an ‘is’ about language use. I see no contradiction in the claim that the majority of the population may use a term of phrase incorrectly (‘begging the question’ comes to mind). No doubt constant misuse over long periods of time could lead to a changed or added meaning, but there seems to be more to meaning than current usage. This is not an objection to ‘experimental philosophy’, indeed what Austin was doing was arguably in keep with it (philsophical ‘experiments’ should not ignore grammatical rules and etymological data and the expense of finding out more about what the person on the street would say. If nanny’s grammer is better we should not let her be outvoted.
    Is Joshua Knobe on this list? It would be great to have his input on the ‘nany’ argument in Appiah.

  8. 8. Posted by Jussi Suikkanen | March 7, 2008 9:51 am


    You should have made it to the live session on Monday. Big thank you for everyone for the illuminating discussion. I’m sure Joshua will check these discussions. But, here is a paper where he discusses the worry you have in mind (I could be wrong about this):

    This is a response to Antti’s Rise and Fall of Experimental Philosophy. As far as I can see, the response is that X-Phi is not even in the business of revealing the meanings of our terms. Rather, it is in the business of investigating how our minds work and for this actual concept application is supposed to be relevant. Joshua isn’t quite clear how we get from the actual concept application to the functioning of the mind but in any case, I think this is the answer.

  9. 9. Posted by Constantine Sandis | March 7, 2008 1:20 pm

    Thanks Jussi,

    I normally teach on Mondays (which is a pain as I also miss all the A Soc meetings) but not on the 17th so I’ll try to make the next one.

    Joshua’s paper was great, and answered my question. But I take it that on his account experimental philosophy and Oxford linguistic philosophy are concerned with two completely different things (human nature and conceptual analysis; Hume and Aristotle being more interested in the former).

    Appiah, on the other hand, makes it sound as if the two methodologies are in direct competition, chiefly by appealing to Quine’s attack on analyticity (which not all of us are convinced by).

    As for why we should be interested in conceptual analysis at all I think the answer is as old as Plato. If you want to know about the nature of(for example)beauty, you would do better to recollect various correct usages and conceptions than to collect data on which kinds of things human beings happen to find beautiful. The latter is what you do if, like Hume, you’re primarily interested in human nature and thereby want to find out what kinds of things we (rightly or wrongly) consider to be beautiful (what Plato would call ‘mere opinion’, contrasting it with knowledge ).

    Arguably Hume (as opposed to, say, the later Wittgenstein) appealed to an empiricist theory of meaning to conclude that by ‘beautiful’ and ‘just’ etc. we just mean those things which we naturally have a certain sentiment towards (the latter being the impression which gives rise to the idea). I see Appiah as siding with (my) Hume on this, whereas Joshua Knobe resists the temptation to think that the study of human nature is the ONLY philosophical project that is not meaningless (even if he does maintain that it is the most important one).

  10. 10. Posted by Andreas Mogensen | March 26, 2008 10:37 pm

    “But the open question argument is sound only if identities have to be a priori, and the identities established by science are a posteriori. A competent speaker may well wonder whether water is H2O; nevertheless, water is H2O.”

    I have been sceptical of this approach to the Open Question argument for some time now. The kind of argument you here deploy against it is a very common tactic in philosophy, but I think it must be used with caution. Many philosophical arguments say things such as ‘X cannot be the case, because X is of the kind K and all K’s have a property Y that X does not/cannot have’ – thus in the philosophy of mind it is sometimes said that one cannot have knowledge of epiphenomenal qualia because knowledge requires direct causal connection to the object of (true) belief and this cannot obtain under epiphenomenalism. Philosophers who believe that X is, in fact, the case therefore proceed to attempt to knock out the claim that the kind to which their particular X belongs must be Y (Brown, for example, attacks the causal theory of knowledge in an article called ‘Pi in the Sky’).

    Now, obviously, this kind of strategy has pretty good intellectual credentials, but it can also be used in a way that is unfair to one’s opponents. If one finds an exception to the claim that all things of kind K must be Y, one had better be sure that the exceptional case looks somewhat like one’s X, otherwise one’s opponent may revise his thesis slightly, so as to take account of the exceptional case, whilst nonetheless maintaining a solid case against X being of kind K. In the debate surrounding knowledge of epiphenomenal qualia, for example, it is useless, I feel, to argue (for example) that mathematical knowledge does not require causal connection to its object, because mathematical knowledge is of such a radically different nature to introspective knowledge of qualia that defenders of the so-called ‘causal theory of knowledge’ may simply recast their thesis so that it no longer applies to all knowledge, but only to empirical knowledge or experiential knowledge or something of that kind.

    A similar problem afflicts the attempt to get around the Open Question argument by relying on the idea of a posteriori necessary identity claims. Scientific identity claims such as ‘H20 = water’ may involve a posteriori identity, but one might question the appopriateness of comparing naturalistic definitions of the good with such identity claims. One of the fundamental aspects of water that makes it possible for the identity statement ‘H20 = water’ to be a posteriori necessary is that water has so-called ‘surface properties’ and an ‘underlying essence’ (these determine, respectively, its primary and secondary intensions). Unless it is meaningful to attribute ‘surface properties’ and an ‘underlying essence’ to a certain kind, such as moral goodness, then the analogy with scientific identity claims will be void.

    Does moral goodness or rightness have ‘surface properties’? Remember that it must be capable of sharing those ‘surface properties’ with another kind that is not moral goodness, because it lacks an identical ‘underlying essence’; but is it possible for something to have the surface properties of goodness and not be goodness? These are undoubtedly strange questions, but I suspect that the shere oddity of the question signals the absurdity of thinking that moral goodness has ‘surface properties’. Hence, the Open Question argument can remain sound, because naturalistic defintions of moral goodness are not analogous to scientific identity claims, and thus the fact that some identity statements are a posteriori is strictly irrelevant.

  11. 11. Posted by Neil Levy | March 27, 2008 12:27 pm

    Interesting argument, Andreas. Yes, I fully agree that simply saying that identities can be a posteriori is not sufficient to show that (something like) the open question argument is invalid. It shows that the argument is too hasty as it stands, and calls upon its defender to show why the identity cannot be a posteriori in this case. You have attempted to shoulder the burden, by pointing out that moral terms do not look like natural kind terms. But is this the only model for a posterior identities? I doubt it. Moral terms might be nomological cluster terms (for instance). I think I can respond to your argument in the same way you respond to mine: by recasting the thesis slightly (except that, unlike my response to the OQA, your argument is not a refutation of mine as it stands, just a call for further details).

  12. 12. Posted by Andreas Mogensen | April 9, 2008 8:08 pm

    Thank you for your response Neil; I think you rightfully point out that I was being rather over-hasty. I must admit that I am not sure what is meant by ‘a nomological cluster term’; if you could explain in more detail, that would be very helpful. However – and here I am perhaps being a little overly hasty once more – I suspect that there is, in fact, no way of defeating the OQA by appeal to the idea of a posteriori identity statements, because there are no possible a posteriori identity statements involving moral terms. I shall now attempt to substantiate this claim.

    I shall address the issue using Chalmers’s two-dimensional modal semantics. As such, we may say that an a posteriori identity can occur only if an identity is asserted between two terms that differ in primary intension but agree in secondary intension. What is more, for this to be possible, it must be such that the secondary intension of one term must be capable of being unknown to competent speakers of the language in which the term figures, even though they are entirely knowledgeable of its primary intension; the secondary intension must be discovered through empirical investigation, thereby making hte identity a posteriori. Now, I take it that for any attempt to provide an a posteriori identity claim involving moral concepts, it will be the moral concepts whose secondary intensions are unknown.

    Now, to put it rather simply, the primary intension of a term determines its reference/extension in the actual world, while the secondary intenions determines its reference/extension across counterfactual, possible worlds. From this it should follow that to fail to know the secondary intension of a moral concept is to be ignorant of its extension across counterfactual, possible worlds.

    However, once the necessary conditions for a posteriori identity claims involving moral concepts are spelled out in this manner, it seems to me that it is sorely difficult to believe that such identity claims are possible, because it is, to me at least, implausible to suggest that we could know the meaning of terms like ‘moral right’ and ‘moral wrong’ but fail to have any knowledge of the extension of these terms across counterfactual, possible worlds. It would appear to me that the primary and secondary intensions of moral terms must be the same, in that no one could fail to know what kind of acts were morally right or wrong in counterfactual, possible worlds whilst knowing what kind of acts are morally right or wrong in this world. If it is wrong to lie (for person x in situation y) on Earth, then it is wrong (for person x’on situation y’) on Twin Earth, and so on for any kind of act one might imagine.

    Arguably this is a very rough outline of an argument, and I welcome any criticism, and especially an explanation of how nomological cluster terms function (or a reference to some relevant literature).

Post a comment

Name: (required)

Email Address: (required) (will not be published)