Thursday, August 06, 2015

Baffling Philosophy?

What positions (or whole debates) in philosophy do you find most baffling?  For each such case, how confident are you that your bafflement is warranted -- that the view or debate in question just doesn't make any damned sense -- as opposed to just being due to a lack of understanding on your part?

I'll offer a few examples that spring to my mind, and encourage others to comment with either (i) more examples of philosophy that you find baffling (and feel free to pick views of mine -- I promise not to take it personally!), or else (ii) defenses of any identified examples, to help those of us who initially find them "baffling" to better understand why they are (in your view) actually well-motivated after all.

So, to get the ball rolling...

(1) Moral views on which "the numbers don't matter", such that you should flip a coin to decide whether (in a forced choice situation) to save a group of five people or a distinct (non-overlapping) group of just one person.

There are plenty of non-utilitarian views that I "get".  I'm positively sympathetic to the view that we should be partial towards our loved ones. I think that traditional deontological constraints (and "rights") ultimately reflect an unjustiable status-quo bias, but I at least understand why some people find them plausible.  But the idea that, even when there are no potential rights violations at stake, nor any "special obligations" (e.g. to loved ones), so that the only reasons left to consider are reasons of beneficence... that still we shouldn't do what's best, but rather ensure that the few are just as likely to get helped as the many... that just seems nuts to me. [99% confidence that the view is indeed objectively nuts.]

(2) Welfare hedonism. It's not as nuts as the above, but I don't really get the motivation for it, in light of the obvious psychological fact that we care about all sorts of stuff besides just our own (or even others') happiness.  I would have thought that the "default" view of wellbeing would be some kind of preferentism, and if you think (as I do) that some ultimate preferences are more reasonable than others, then you should shift to some kind of pluralist objectivism.  I can get thinking that there are certain defective preferences (to do objectively pointless things like count blades of grass) that don't contribute to your wellbeing.  I can't understand why anyone would think that every non-hedonistic preference (including for love, friendship, etc.) is defective in this way.

[98% confident that hedonism is false; 80% confident that it is nuts, in the technical sense of being "significantly less respectable -- or worthy of being taken seriously -- than it is currently generally regarded as being."]

(3) Candidate Pseudo-debates: I'm sure every philosopher has some suspicions about debates that seem to them merely terminological, and not really about anything of substance.  Some I'm dubious about include:
- (i) debates about mereology and composition -- whether there really are tables, or just atoms-arranged-tablewise, that kind of thing.  [80% confident there's no "there" there.]
- (ii) representationalism vs direct realism about perception [75% confident it's all terminological]
- (iii) epistemology focused on "knowledge" (as opposed to, say, rational credences) [70% confidence that knowledge is not an important concept, and epistemology would do better to focus elsewhere.]

I'm sure there are more, but those are the examples that initially spring to mind. (But I'm open to having my mind changed, if you disagree with any of my assessments...) Your turn!


  1. A certain response to the xPhi: Call it the "Look, philosophers argue for their points, they don't merely intuit them!" objection to xPhi. Sure, philosophers might provide arguments to defend their intuitions, but that is consistent with philosophers merely rationalizing their intuitions — and would anyone be surprised if that is what philosophers are doing? So if some intuitions are faulty in certain situations and we build arguments to support such intuitions, then does the existence of the argument overcome the problem with the inauguration of the argument (i.e., the intuition)? It seems that it doesn't. After all — as this post highlights — we can argue for all sorts of stuff and accomplish little more than baffling our colleagues. [To be fair, I found the response intuitively appealing at first. Also, I don't think xPhi's greatest virtue is the negative program...or the positive program for that matter].

    1. Interesting, not sure I've come across that one before. A better response in the vicinity, I think, is the "intuitions are just premises" one (I really like Liz Harman's paper here, for example). There's nothing wrong with appealing to unargued-for premises in an argument (especially if at least some of your interlocutors are likely to accept the premises in question). And while you should of course be willing to reconsider any such premises in light of further arguments against them, such arguments can only get traction against you if they start from (other) premises that you accept.

    2. That seems sensible. I'll take a look at Harman's paper ASAP. Thanks for the reference.

      I had only heard the "Philosophers argue for their points..." response conversationally until recently. It seems that Max Duetsch is making this kind of move in his new book The Myth of the Intuitive. (I am inferring this from an interview about the book, not the book itself [the interview]).

    3. I find the "intuitions are just premisses" reply baffling. ;) I really don't see how it is supposed to work. And skimming the linked Harman piece, I think it must rest on a confusion about exactly how typical negative critiques work. As I read them, most negative critiques are not launched from the mere fact of disagreement (though I think that's actually an interesting line of attack). Rather, they are launched from the alleged fact that our intuitive judgments or our inclinations to assert various premisses (depending on the interlocutor's view of "intuitions") are affected by or sensitive to epistemically irrelevant factors, like culture, socioeconomic status, or local environmental features. The thought is that if we are not warranted in asserting our premisses, then neither are we warranted in asserting our conclusions. Maybe we could try to be content with the sort of conditional, intellectual territory-mapping project that Harman mentions in the beginning part of her essay. But it sure seems like philosophers actually endorse the conclusions of their arguments! (Maybe that isn't right -- I haven't done any counting or constructed any way of measuring whether a premiss or conclusion is asserted.)

      Anyway, the force of the experiments is to show that we are not warranted in asserting certain specified premisses. (And then people generalize to greater or lesser degrees.) So, the "intuitions are just premisses" line looks a lot like people failing to reconsider premisses in the light of arguments against them.

      Harman comes really close to engaging with this in the revised version of the second epistemological objection that she considers on page 7. But I think she misses the most critical element: the experimentally-derived explanation as to why one finds a given premiss plausible in the first place. If finding a premiss plausible has nothing to do with how likely it is to be true and everything to do with irrelevant features of my local environment, then I have reason to refrain from asserting the premiss until I have further evidence, yes?

      I often feel like I'm missing something here because when I try to reconstruct things, the dialectic looks all wrong ... Meh. Like Nick, I'm not a huge booster for the negative program, but I really don't get this line of response.

      This is take-two, my first version, which was *much* better, I assure you, was eaten by the comment system when I logged in to preview it. I mean, really, that other version was a masterpiece of online commenting. This one is kind of crap. ;)

    4. Hi Jonathan,

      I wonder if your complaint is actually consistent with Richard's brief comment above (since he says that premises are only good if (a) most of your audience buys it and (b) you remain open to arguments against the premise. So if we find a robust experimental finding suggesting that the intuitive appeal of a crucial premise in an argument is suspect/flawed/etc., then -- on Richard's account -- that is a reason to reject the premise. Does that seem right Richard? Jonathan?

      (I haven't mentioned Harman because I haven't had a chance to read it yet. I look forward to reading it ASAP).

    5. "If finding a premiss plausible has nothing to do with how likely it is to be true and everything to do with irrelevant features of my local environment, then I have reason to refrain from asserting the premiss until I have further evidence, yes?"

      Not necessarily. That presupposes that the evidential basis for the premise P is the psychological state of "finding P plausible". But I don't think that my finding P plausible is necessarily evidence for P. It may instead merely be a cause of my believing P. But I don't accept the epistemic principle that we always ought to suspend any beliefs that are shown to have non-truth-indicative causes (or for which we cannot point to "neutral" evidence in their support). That way lies widespread skepticism, as I point out in my 'Knowing What Matters' paper.

      More generally, evidence that we are unreliable in our P-judgments is different from evidence that P is false (or likely to be false). As suggested in my old post, "Why Suspend Judgment?", I think the right response to general evidence of our unreliability is not to suspend judgment so much as to hold our judgments more tentatively. To reduce my credence in P in particular, I'd want to see some evidence that spoke to the truth or falsity of P, not evidence that spoke about me.

    6. [Jonathan Livengood replies...]

      Richard, thanks for your remarks; they're helpful. What follows is way too long, I'm sure ... on preview, looks like I have to break it up into parts! And I wish I could have gotten it down sooner. I've been mulling things over in between other work.

      I want to take up the discussion here:

      That presupposes that the evidential basis for the premise P is the psychological state of "finding P plausible".

      I don't think the criticism requires that assumption. At least, I hope it doesn't since I agree that the assumption is a bad one! I'm happy to grant that (at least for the most part) the fact that one finds a premiss plausible is not first-personal evidence for the premiss. (Possibly there are edge cases, like where a person has her brain manipulated such that she forgets all of her evidence for some collection of propositions, she still has feelings regarding their plausibilities, and she knows that she has been manipulated in this way. But I'm not really interested in weird edge cases. I'm interested in perfectly ordinary, typical cases.)

      I also agree with you in rejecting the epistemic principle that one always ought to suspend any beliefs that are shown to have non-truth-indicative causes. Probably all of our beliefs have some non-truth-indicative causes. The real question is how strong those causes are relative to the truth-indicative causes (if there are such) or whether the non-truth-indicative causes explain important disagreements. If I'm only a *little* bit sensitive to some irrelevant factor -- e.g. I give a 63 rather than a 60 on a 100-point scale, where tolerance for error is, say, 20 points -- well, then no problem. But if I'm *really* sensitive to irrelevant factors, and especially if what explains disagreement on an important question is an irrelevant factor, then there is a problem.

      I'm not so sure about the (distinct) epistemic principle that one always ought to suspend any beliefs for which one cannot point to "neutral" supporting evidence. I'm not sure what you mean by "cannot" here, but I'm worried that the principle is too dogmatic. If you mean only that one cannot immediately produce supporting evidence, then I'm probably on board with rejecting the second principle as well. But if you mean something more like, "If one thinks hard for a while but fails to be able to come up with even a hint of justification," then the principle seems more plausible to me.

      So, how does the criticism go if it doesn't rely on those principles or principles suitably like them? A lot here depends on the dialectic. As I see it, what happens is that the speculative philosopher makes a categorization judgment in response to some case, say, "That's an instance of knowledge." The speculative philosopher then draws some implications: knowledge has such and so properties, etc. The experimental philosopher steps in and worries, "Yes, but we've found that categorization judgments about cases very much like this one are affected by whether or not the person doing the categorizing had a sugary beverage in the previous hour." (Totally made up example.) "On the basis of our experiments, we think there is a 60% chance that had you recently consumed a sugary beverage, you would have said that this is NOT a case of knowledge."

      [to be continued...]

    7. [Jonathan Livengood continues...]

      If I were presented with such a finding for my own assertions, I would think it was troubling. Such a finding would give me reason to think that I was mistaken about why I actually hold the beliefs that I hold or make the assertions that I make. I *think* that I hold various beliefs and make various assertions on the basis of relevant evidence (although whether this is the case for intuitions treated simply as premisses, I'm not sure). What the experimental results do is to challenge the truth of the claim that my beliefs or assertions are based on relevant evidence by showing how and why confabulation is likely. Take for example a pretty famous experiment by Nisbett and Wilson (pdf) in which they offered participants a choice between two identical garments -- one placed on the right and one placed on the left. The participants made a selection and were then asked to explain their choice. The right-hand garment tended to be preffered. In some experiments, it was preferred four-to-one. But in explaining their choices, people never referred to the position of the garment. Nisbett and Wilson go on to say, "When asked directly about a possible effect of the position of the article, virtually all subjects denied it, usually with a worried glance at the interviewer suggesting that they felt either that they had misunderstood the question or were dealing with a madman" (244).

      I think that such studies do and should reduce one's warrant for assertion, and in the case we're considering -- where a proposition is simply being put forward without any support -- it seems that even a little bit of a challenge requires an answer. I'm *not* saying that no adequate answer could be given. One might make a further argument that rests on premisses that are not of a sort that has come under threat from experimental critique. Or one might offer arguments against the experimental work itself. Maybe the studies were badly done or unlucky. But I don't see how the dialectic leaves room for speculative philosophers to gives "intuitions are premisses" as a *defense* against experimental critiques.

      Your second paragraph raises some interesting points. I agree that evidence that an agent is unreliable in P-judgments is not evidence that P is false, provided that "unreliable" here means merely "not reliable" and not "anti-reliable." If the P-judgments of some agent S are unreliable in that they are sensitive to the wrong things and insensitive to the right things, then knowing that S judges that P is no reason to think that P is true. But it also isn't any reason to think that P is false. However, if being unreliable means being systematically biased against the truth, then knowing that S judges that P is reason to think that P is false. I think that this should be true even first-personally, though I admit that things are trickier when we go from third-person to first-person.

      I wonder what you think about the following, though. I have been trying throughout to stick to belief or warranted assertion, rather than truth. Do you think that in light of an experimental critique, the intuitions-as-premisses response is enough to preserve warrant for asserting the premiss? Two sub-questions: Do you think the response preserves warrant in the case where the experiments show that people are sensitive to irrelevancies and insensitive to the truth? Do you think the response preserves warrant in the case where experiments show that people are systematically biased away from the truth?

      Finally, I'm not sure exactly what you have in mind with the distinction between suspending belief and holding a belief more tentatively. Is one supposed to be about credence and the other about acceptable update rules?

      - Jonathan

    8. Yeah, 'holding a belief more tentatively' here was meant to mean something like expecting you'll end up revising it (in either direction) in light of future evidence / further reflection. (Importantly, you've no basis for thinking that future revisions are more likely to involve lowering rather than increasing your current credence levels, which is why I think there's no current basis for changing your credence.)

      If we're shown to be anti-reliable, or "biased away from the truth", then we can -- and should -- recalibrate onto a more warranted level of credence. So I agree with you that the "intuitions are premises" response is irrelevant in that case. But if all that's shown is that our starting points are non-reliable, but without any further guidance as to what alternative starting points would be more reliable, then I think it's fine to stick to one's (admittedly fallible) guns.

      I guess this stance is really a mix of "intuitions are premises" + "epistemic conservatism", with the latter doing most of the work. I don't think we have to convince critics that we're right (or even reasonable). I think it's up to the critic to convince us that some alternative starting point is better warranted than our own. And a purely negative skeptic can't do that.

      A complication for the experimentalist critiques (and analogies to debunkings of empirical beliefs): Fundamental philosophical questions concern a priori, necessary, causally inefficacious truths. So we already know that our philosophical inquiry cannot be causally sensitive to the truth of the matter. So we'd better have an epistemology (like the a priori reliablism set out in my 'Knowing What Matters' that allows us to have warranted a priori beliefs nonetheless. I don't see that the experimentalist critiques really add anything to the challenge we could have identified a priori; and whatever solution we came up with to the a priori challenge should naturally carry over to the experimentalist critiques as well.

  2. I'm in favor of the position "that knowledge is not an important concept, and epistemology would do better to focus elsewhere," but it does have a heavy lift in that "knowledge" is just about the only epistemological term that's also widely used in common parlance. So those of us who think knowledge isn't a very useful concept for epistemology have to explain why it seems to be a useful concept for The Folk.

    Of course I think I have an answer for this; I think saying we know something is a way of wrapping up many other intrinsically useful epistemological concepts in one short word. "Knowledge" is something like a prescientific concept--unsuitable for heavy-duty epistemological inquiry and even inconsistent but good enough for everyday use. But even if I don't think the knowledge debate itself is productive, we need a debate about whether to have the debate, as it were.

    1. Yes, absolutely -- I think many such "meta" debates (meta-ontology, etc.) are very interesting and important.

    2. It seems rather strange to say that "knowledge" isn't a useful concept for epistemology. Without that concept, how is an epistemologist going to say what business he's in? I would distinguish between whether a concept is useful and whether clarification of that concept is useful. Most of the time we use our concepts perfectly well without any need to clarify them. So, when I say that "knowledge" is a useful concept, I don't mean to say that it's useful to clarify it. Since I favour a pragmatic approach to philosophy, I would encourage epistemologists to emphasise thinking about how we acquire knowledge, and how we can do that better, over thinking about the meanings of words.

    3. If we reject the significance of knowledge, we might instead understand epistemology as in the business of addressing questions about rational credences (degrees of belief), for example.

    4. "If we reject the significance of knowledge, we might instead understand epistemology as in the business of addressing questions about rational credences (degrees of belief), for example."

      Hi Richard. Thanks for replying. Let me clarify my last comment. I'm quite happy to continue describing epistemology (very roughly) as the study of knowledge. In the course of that study, we might employ various concepts, including both "knowledge" and "rational credence". My main concern is not with whether we focus on the concept of "knowledge" or some other related concept like "rational credence". My concern is with what kind of questions we ask in relation to those concepts. The traditional approach is to ask what X is, where X is a concept of interest. I'm suggesting that more attention be paid to pragmatic questions, like "how do we (best) acquire knowledge?" or "how can we improve our chances of having appropriate rational credence in a belief?". That said, I don't entirely reject the value of asking what knowledge or rational credence are, but I want to de-emphasise such questions, and, if we do pursue them, I would do it in a rather different way than the traditional way. (To explain my way would take a while. Suffice to say FWIW that my approach to philosophy combines an emphasis on naturalised thinking with a Wittgensteinian view of language.)

      In my last comment, I referred to "the meanings of words". I was assuming that philosophers who address the question "what is X?" are mostly addressing the question "what does 'X' mean?". But that was too simplistic, which is why I've made the switch in this comment to the "is" question.

    5. P.S. I didn't appreciate that you were probably just replying to my question: "Without [the concept of "knowledge"], how is an epistemologist going to say what business he's in?" OK, I guess he could use some other related term, like "rational credence". But this is just a matter of terminology, as far as I'm concerned. To study rational credence is effectively to study knowledge, even if you aren't using the word "knowledge". In my view what matters most is what kind of question you address, and what kind of reply you try to give, not which of two related concepts you choose to focus on.

  3. 1. I'm totally with you in thinking that ethical view that "numbers don't matter" is utterly insane. To anyone who thinks that it is only consequentialists or those who have been corrupted by technocratic utilitarians who think that the numbers matter, I like to quote Aristotle's line "To secure the good of one person only is better than nothing; but to secure the good of a nation or a state is a nobler and more divine achievement" (Nicomachean Ethics 1094b10).

    2. I also think that the idea that Larry Temkin has defended, that the predicate 'x is better than y' does not stand for a transitive relation, is totally crazy. 'Better' is comparative adjective, for goodness' sake... Whatever next? Will someone start arguing that "being longer than" or "being hotter than" do not stand for transitive relations?!

    There are plenty of other examples, but those are just the first two that came up mind...

    1. Yeah, rejecting the transitivity of 'better than' is a striking example.

  4. [Tim Scanlon writes in:]

    "Just for the record, I do not believe that numbers never count. What I am unsure about is the best way to explain when and how they do count without admitting what are (to me at least) implausible aggregation in which tiny considerations on one side, if sufficiently numerous, can always outweigh more significant considerations on the other side. Perhaps Parfit's modification of my contractualist view is the best way to do this, but I am unsure."

    [RYC comment: I just had in mind Taurek and others who adopt a much more extreme view here!]

  5. I am totally baffled by people who think we should do epistemology without talking about knowledge. ;)

    I am also pretty baffled—though MUCH less informed—about literature on consciousness and phenomenology. Sometimes I wonder whether I'm a philosophical zombie, and don't really feel like I have any clue how to go about investigating. Lots of smart people who know the field orders of magnitude better than I do think there's something real going on there, so I guess I have a reasonably high credence they're probably right. But I've never been able to feel like I have a good grip on it.

    I guess I also don't understand why some metaphysicians think there are important metaphysical differences between the past and the future. That kind of asymmetry seems totally unmotivated to me.

    1. Hmm, interesting: you don't get the "past is settled / future is open" intuitions at all? (I'm happy to override them for theoretical reasons -- Eternalism seems much more plausible all things considered -- but I definitely have the asymmetry intuitions.)

    2. No, I've never really felt those intuitions, once it's clarified that we're talking about any kind of metaphysical modality. I'm happy to admit that it's often easier to know things about the past than about the future, so there's a sense in which the future seems more 'open'. But that's only the sense in which it seems open who won yesterday's game if I haven't looked at the score yet.

  6. What baffles me is not so much a particular view (though belief in mereological simples comes close) but rather the relative importance intra-disciplinarily ascribed to certain views, as indicated by how frequently they are discussed in publications by philosophers who are not among their proponents. A case in point is the knowledge-first approach, whose apparent importance I cannot help suspecting is due partly to the job title of its central proponent...

    1. The knowledge first approach plausibly derives from Plato.

      But to speak frankly, I find such speculation into intellectual motives rather insulting.

  7. This comment has been removed by the author.

  8. I'm baffled by a family of views in different areas, which seem to have grown up fairly recently, which have in common that they (i) maintain that there is, contrary to what we might be inclined to think, a certain sort of fact about some matter, but that (ii) we cannot know it. All views of the sort I have in mind seem to face a problem along the lines of: what would make the fact the way it is, rather than different? They seem unacceptably brute or free-floating, and arbitrary.

    The most prominent is probably Williamson's epistemicism about vagueness.

    Also, Breckenridge and Magidor have a view about 'arbitrary reference' (which occurs when we make arguments like 'There are Fs. Let a be such an F. ....') according to which we in such cases actually refer to a particular thing, but cannot know which. (I've tried to argue against this (early version here, but my main argument doesn't generalize to all views of this kind, and so is perhaps a bit superficial.)

    Finally, Nick J.J. Smith has a view, put forward in 'Semantic Regularity and the Liar Paradox', according to which the derivation of the Liar paradox should be viewed as a reductio proof, and that Liar sentences refer, not to themselves, but other things, where we cannot know which.

    I find these all very implausible indeed, in what feels like a pretty fundamental and important way.

    I'm trying to make a collection of these sorts of views and would be keen to hear about others. Also, general arguments against them, and whether other people are baffled.

  9. [Andrew Sepielli writes in:]

    Why should we think the debate about the existence of moral properties is substantive if debates about mereology and composition are not?

    1. That depends which "debate about the existence of moral properties" you have in mind. I take it there's clearly a substantive dispute between moral realists and Mackie-style error theorists, just as there is between (say) theists and atheists. We want to know whether a familiar sphere is discourse is accurately describing the world or not. And the naturalism / non-naturalism debate seems substantive in just the way the physicalism/dualism debate about consciousness is. We want to know what phenomena are basic or irreducible features of the world. But there's a further debate between "Platonist" (e.g. Enoch) and "Quietist" (e.g. Parfit / Scanlon) non-naturalists, and it's not clear to me that there is really room in logical space to distinguish those views. Perhaps that debate is more akin to the mereology debates, where the participants (arguably) agree about the basic facts and just differ in how they describe them. I'm not sure.

  10. Ah, figured out how to comment!

    Why do you say there's "clearly" a substantive dispute between realists and error theorists? It's not clear to me.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)