Leiter Reports: A Philosophy Blog

News and views about philosophy, the academic profession, academic freedom, intellectual culture, and other topics. The world’s most popular philosophy blog, since 2003.

  1. F.E. Guerra-Pujol's avatar

    Apropos of Sagar’s wish to foist the A.I. industry by its own petard, this article appeared in print in yesterday’s…

  2. Claudio's avatar

    I teach both large courses, like Jurisprudence and Critical Legal Thinking (a.k.a Legal Argumentation), and small seminar-based courses at Edinburgh…

  3. Charles Pigden's avatar

    Surely there is an answer to the problem of AI cheating which averts the existential threat. . It’s not great,…

  4. Mark's avatar

    I’d like to pose a question. Let’s be pessimistic for the moment, and assume AI *does* destroy the university, at…

  5. A in the UK's avatar
  6. Jonathan Turner's avatar

    I agree with all of this. The threat is really that stark. The only solution is indeed in-class essay exams,…

  7. Craig Duncan's avatar

Rosenberg on intractable moral disagreement

Brian Leiter

I usually ignore the generally awful Stone Blog, but Alex Rosenberg sent along his nice piece on a topic I'm quite interested in.  I particularly liked this succinct rejoinder to those who like to compare mathematical and moral knowledge:

A few philosophers claimed that we have a moral sense that perceives the moral rightness or wrongness of things directly and immediately. This theory might be worth taking seriously if morality were like mathematics. Mathematicians all agree that we know with certainty a large number of mathematical truths. Since experiment and observation could never be the source of such certainty, we (or at least mathematicians) must have some other way of knowing mathematical truths — a mathematical sense that directly perceives them. For this argument to work in ethics, there would have to be little or no ethical disagreement to begin with. Since many moral disagreements seem intractable even among experts, the hypothesis that we are equipped to know moral truths directly is very difficult to sustain.

Since the comments at the Stone Blog are generally worthless, I thought I would open comments here for those who want to discuss the philosophical issues raised by Rosenberg's piece.

Leave a Reply

Your email address will not be published. Required fields are marked *

50 responses to “Rosenberg on intractable moral disagreement”

  1. Daniel S.

    I don't really do moral philosophy, but I wonder if one *might* be able to defend something like that notion that we have a moral sense that approximately resembles mathematical sense, but the propositional content of moral claims leads to more widespread disagreement. Moral perception might be quite distinct from the ability to interpret our immediate moral intuitions correctly and effectively.

    Reply
  2. Marc Hamann

    The intent of this comparison betrays a lack of knowledge of the history and philosophy of mathematics.

    It is probably true that many mathematicians without philosophical sophistication might agree with that there is an objective "faculty of mathematical truth", in part because they have spent a lifetime in the culture of mathematics which often assumes such a faculty.

    But there have been major foundational and philosophical disputes about the nature of mathematical truth (some still ongoing), just as you get with any other type of judgmental reasoning.

    "Innate faculties of judgment" are very difficult to tell apart from unexamined deep cultural biases and beliefs.

    BL COMMENT: Can you or someone else give some examples from mathematics of this kind of disagreement?

    Reply
  3. Branden Fitelson

    The debate about the Continuum Hypothesis is about as intractable as debates get (in any domain).

    http://plato.stanford.edu/entries/continuum-hypothesis/

    Reply
  4. Branden Fitelson

    Perhaps a better article to look at this following (the SEP entry is rather inside baseball).

    https://www.ias.edu/about/publications/ias-letter/articles/2011-fall/continuum-hypothesis-kennedy

    Reply
  5. Robert Gressis

    Imagine if, in mathematical debates, it was in many people's direct self- or group-interest to deny or assert certain claims. You might see a level of disagreement akin to ethics.

    Reply
  6. Marc Hamann

    A major example is Constructivism vs Formalism.

    https://en.wikipedia.org/wiki/Constructivism_%28mathematics%29
    https://en.wikipedia.org/wiki/Formalism_%28mathematics%29

    Reply
  7. anon ex-pro-philosopher

    Comparing the two seems apt to me, or at least more apt than Alex and Brian suggest. We know with certainty a large number of moral truths, too (e.g., that killing an innocent person for no reason is wrong). Those who dislike the comparison tend, I think, to overlook the widespread moral *agreement* that underlies moral disputes.

    Reply
  8. Kenny

    Re: examples. Everything in here:

    http://www.amazon.com/Consequences-Choice-Mathematical-Surveys-Monographs/dp/0821809776/

    and every place where constructive analysis departs from classical analysis, e.g., the intermediate value theorem.

    Reply
  9. Gideon Rosen

    But the 'debate' over the continuum hypothesis isn't really a disagreement. As I understand the situation, everyone agrees that CH is an open question (which may or may not have a determinate answer), and everyone agrees about the theorems that are relevant to assessing it. Set theorists may lean one way or another given this inconclusive evidence. But you don't find one school of mathematicians asserting CH outright and relying on it as a premise while another school rejects it. So this is quite unlike disagreement in ethics.

    Reply
  10. Lexington

    It should be pretty obvious that there's a huge difference between the debates mathematicians have over CH and those moral philosophers have over most moral propositions. Namely, mathematicians agree on a enormous body of mathematical propositions, and may disagree on very, very few, and for moral philosophers it's the reverse. And of course the reason mathematicians even disagree on such things as the CH is the ever surprising result of Godel's demonstrating that mathematics is incomplete, combined with Cohen's astounding proof that CH in particular is independent. Not a lot of other propositions in mathematics with any intuitive appeal that are known to be independent.

    What moral propositions do all moral philosophers agree on? Maybe something as nearly vacuous as that it's wrong to inflict suffering on an innocent person for no other reason than one's own pleasure in doing so? Anything else of significance come to mind?

    Pretty big difference. Mathematics embodies already a vast system of knowledge about which mathematicians agree; moral philosophy at most a few almost degenerate propositions to which all might assent.

    Reply
  11. Daniel S.

    There's certainly a lot of interesting disagreement in work on the foundations of mathematics (the set theory vs. category theory debate, for instance). There have been interesting debates, historically, over whether mathematics has empirical consequences or not. Then, of course, there are the debates between constructivist mathematicians and mathematical platonists (which touches on debates between proof theorists and model theorists). Foundations work and metamathematics work are rife with controversy.

    Reply
  12. Formalist

    Whatever one refers to by "a mathethematical sense that directly perceives [mathematical truths]" is largely mysterious. I would think there is little to gain in comparing moral "perception" to mathematical "perception" until we understand and agree on exactly what we mean by the latter.

    Reply
  13. Michael B

    I'm not sure the rejoinder does as much work as first appears. There are lots of open problems in mathematics that are so complex that no one has any idea what the answer is or even what a solution should look like (this is something Richard Feynman talked about having to do in his day job – try to figure out even what might be an answer). There is no direct perception of truth in such cases.

    But there are other truths for which we have answers but they cannot be directly perceived by intuition e.g. the four-colour map problem and its associated proof – that not all mathematicians even accept as a proof.

    Yes, there are lots of open problems in ethics that are so complex that there is no agreement about their solution either. But, like with mathematics, I think there is a whole raft of ethical issues about which there is vast agreement, but we don't notice because they are not interesting. The human race continues to survive, and in such numbers, because of such cooperation and agreement.

    A more interesting distinction between morality and mathematics, I think, is this: one cannot be considered an expert in mathematics without also being good at mathematics. Whereas the 'experts' cited in the piece above might be considered an expert in morality or ethics and quite plausibly be terrible at living well or acting morally.

    Reply
  14. Lexington

    It's frankly hard to see the dispute between constructivists and platonists as truly a mathematical issue, rather than a philosophical issue. Constructivists want to do their mathematics only using "constructions", because they don't believe in those platonistic thingies. Certainly both sides see that each side is right in their proofs, given what they start with as basics.

    But who can claim that that's true of the disagreements among moral philosophers? They disagree on very basic propositions regarding what's right and wrong, typically not for grand metaethical reasons, but because their "intuitions" differ on the specific propositions.

    Reply
  15. Daniel S.

    Constructivism vs. Platonism may be more properly philosophical, but the debate plausibly has consequences with respect to talk about whether or not mathematical truths are "perceived directly." Constructivists and Platonists would, at a minimum, disagree over what, precisely, is being perceived.

    Reply
  16. Aaron Lercher

    Rosenberg deploys a Darwinian skeptical argument. I have trouble understanding such arguments, but roughly:
    Natural selection chooses only traits that favor more offspring or better survival chances. Even if there are moral facts, the ability to recognize moral facts neither favors nor is favored by a tendency to have more offspring or improved survival. (And ability to recognize moral facts might even be disfavored by natural selection.) So there cannot be natural selection in favor of an ability to recognize moral facts.
    That's a sweeping argument. It seems to me that it might be equally destructive of claims for the existence of other cognitive abilities that are not direct results of evolution.
    (Please correct my Darwinian skeptical argument! I want to know whether this kind of argument can be made in a less sweeping way.)
    A naturalist might argue instead that (1) there are naturally selected abilities in humans (or other primates) to interact with other humans (or other primates) and to perceive their needs and pains, and that (2) when these naturally selected abilities are subjected to further *domestic selection* (education, social experience, historical precedent), these result in an ability to recognize moral facts, assuming that there are (at least some) such things as moral facts.
    But mostly I'm curious what folks think of Darwinian skeptical arguments, and what is the best way to state such an argument.
    Also, by the way, Rosenberg never spells out what honor killing is. Is it a punishment directed exclusively at women? So any claim is that it is or isn't a moral fact that honor killing is wrong is rather unclear. But I wanted to focus on the Darwinian skeptical argument.

    Reply
  17. Jimmy Doyle

    What about Cantor vs Kronecker & Poincaré on the uncountable transfinite? That looks like a very serious disagreement, in a way that the CH issue is not.

    Reply
  18. Lexington

    I don't think it's particularly helpful in general to focus on such notions as "perceived directly" when comparing mathematicians with moral philosophers. It's better to step back a bit from them and their mental processes, and think about just what both sides tend to agree on, or not.

    What seems obvious in so doing is that virtually all mathematicians agree about specific proofs, unless of course the proofs are themselves of extraoardinary complexity. They agree on the enormous mathematical systems they have developed using these proofs, even if they may, for example, disagree on whether assuming the excluded middle is reasonable (again, for philosophical reasons). Both the constructivist and the platonist can lie down together and agree that the proofs of each are correct, given the (philosophical) assumptions.

    But where are the big systems that all moral philosophers agree on, given any similar set of assumptions? Where are even the tiny systems for which such agreement can be found? Why do so many "moral systems" developed by philosophers seem to be all dressed up in formalisms with no place to go?

    Reply
  19. hopelessmisanthrope

    The mathematics and moral disagreement comparison is a familiar one. Rosenberg is by no means the first to bring it up. But that argument has never been convincing to me because it seems to me that there is huge disagreement on the most fundamental things within the foundations of mathematics.

    Even simple statements in math such as 1+1=2. I think most mathematicians would agree that for this statement to be true, there needs to be objects with certain properties which the terms refer to. But if there are no such objects such as numbers, that statement would not be true at least in a strict sense of true. Mathematicians would need to assume that they exist for any math to get done.

    Many would probably even say they believe them to be true. But Rosenberg's claim that all of them are certain of statements based on these foundational assumptions doesn't seem to be accurate. When pressed, I think many people would be surprised at how much doubt exists in the philosophical nether regions of most mathematicians' brains. They certainly exists among philosophers of mathematics. There may be more wiggle room in moral disagreements but the argument that there's a vast gulf of difference isn't convincing at all.

    BL COMMENT: Disagreements among philosophers of mathematics are not the same as disagreements among mathematicians. No mathematicians doubt that 1+1=2.

    Reply
  20. TJ

    Michael B is mischaracterizing the four color theorem proof. The qualms about that proof was that it was done by a computer. This meant that not only the mechanics of the proof were long and subject to the usual issues, but the software that did brute force analyses to resolve cases and put it all together also had to be reviewed for correctness.

    All this was daunting but it is not a fair characterization to say that there is a debate about the mathematics. What's unnerving is that it wasn't done by a human and can't easily be checked by a human.

    The fact that there are open problem in mathematics is not something that's to mathematics' discredit. What's wonderful is that all mathematicians agree on the open problems and don't split up in various camps who each think they've got their own incompatible solutions for several hundred years. Regardless of what might be the underlying nature of mathematics, if that means anything, it's objectively successful in reaching agreeing results in a way that is not endlessly disputable.

    Reply
  21. Justin Clarke-Doane

    I think that Gideon Rosen is correct about the attitude that most set theorists take toward CH. But I think that it is hard to interpret apparent disagreement over paradigmatic axioms – e.g., Choice, Foundation or Replacement – that way. Of course, if an appropriate form of pluralism is assumed, then these “debates” are like a “debate” over whether the Parallel Postulate is true (as a pure mathematical conjecture) – they are really just about which parts of the pluriverse to talk about. But not all participants assume pluralism. And with good reason — the debates look prima facie misguided in the context of pluralism.

    Reply
  22. Brad Cokelet

    The original passage runs: "Mathematicians all agree that we know with certainty a large number of mathematical truths. Since experiment and observation could never be the source of such certainty, we (or at least mathematicians) must have some other way of knowing mathematical truths — a mathematical sense that directly perceives them. For this argument to work in ethics, there would have to be little or no ethical disagreement to begin with."

    This looks like an inference to the best explanation that starts from widespread agreement that all informed parties know with certainty a *large number* of truths in the relevant domain. I am not an intuitionist myself, but I can't think of a moral intuitionist (or philosophic champion of the moral sense) who argues in this way.

    I would think that the relevant analogy with math would start from the claim that mathematicians all agree that we know with certainty a relatively small number of basic axioms which are then said to be self-evident. Then moral intuitionists make an analogous argument about basic moral truths. If I remember correctly this is how Roger Crisp develops his view in his fine book "Reasons and the Good". He admits that it will he tough going to build anything like a complete ethical theory from truths he takes to be self-evident, and he is admirably humble and sensible about the challenge that disagreement poses.

    Reply
  23. Christopher Morris

    I informally poll my students about different claims, and in my intro ethics course, which focuses on the ethics (and politics) of killing and letting die, I invariably get 100% of the students to assent to the statement that "killing people is generally wrong". ('Generally' is inserted to allow for the principle to be overridden. I ask this question at the start of the course.) The 100% here is not hyperbolic; I get NO dissent on this proposition. On virtually other topic there is considerable disagreement. And, of course, as we learn in the class, there is considerable disagreement about the scope of the principle (e.g., whether it applies to fetuses).

    This comment does not directly address the debate above, but it is important to remember the agreement that does exist in ethics.

    Reply
  24. Michael B

    Just wanted to add, as a minor point, the disagreement among people per se is larger than among moral or ethical experts (I presume we mean here mostly moral philosophers?). But since the comparison was with mathematicians, I think my point here is valid to compare like for like.

    Just think: how much must two people agree on, in terms of values and behaviours, to both get and keep university jobs as academics? Probably a whole bunch of stuff to do with killing, violence, theft, work, trust, reliability, effort, contracts, sociability, collegiality, even down to trivial matters like punctuality.

    There is a LOT – that comes way before whether or not you think intentions or consequences or character is the main focus of ethics – that is agreed upon implicitly through behaviour outside of class, by two moral experts with opposing views, just to get into a job where people call you a 'moral expert'. You're probably not off doing anything that any other moral expert considers incomprehensible, like trying to exterminate endangered species for a laugh, trying to bring back slavery, or trying to open a restaurant serving human. Rather, you probably differ in your view of what mode of transport is OK to use frequently, what non-human animals are ok to eat, or how much people should be taxed and on what it should be spent. And no one is suggesting the ethical thing to do is to defeat colleagues with opposing views by violence, or bribery or coercion. There is a ton of agreement in place for such disagreement to take place.

    Are any of these enormous disagreements coming from outside the academy? Would such a person with, say, a blog but not PhD even be considered a 'moral expert'?

    So I'm not sure the contrast, between mathematical experts who all agree on very much about mathematics and moral experts whose moral disagreements are cavernous and many, is very good.

    Reply
  25. Daniel S.

    I see your point. I think there's something instructive about constructivism/anti-realism vs. platonism/realism, though. Even if we set aside the question of "direct perception" and just focus on what mathematicians and moral theorists tend to agree upon (or not), we still end up with interesting questions. If constructivism proves convincing, for instance, and we can assume that moral sentences have the same ontological status as mathematical sentences, then we'll naturally wonder why sentences about mathematics enjoy basically widespread agreement and ones about morality do not. In both cases we grant the constructionist intuition that the sentences in question are not about the world but rather reflect our own concepts/cognitive structures, but the behavior of each set of sentences is quite different in the world. Why? It doesn't seem like adding more restrictive formalisms to moral analysis that would roughly parallel mathematical formalism would help.

    Reply
  26. Alex Rosenberg

    I should not be surprised at how much more interesting this discussion is than the one in the NY Times comment space, where many readers think I am in favor of honor killing, or at least not against it. There are of course huge disagreements about the foundations of mathematics, but the key difference between ethics and math is the proportion of disagreements about relatively derivative truths, such as computational ones, which are quite different from CH. By contrast it's hard to state interesting ethical truths on which there is equivalently widespread agreement. A couple of commentators have tried, but they run up against Hume's dictum that 'Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger.' What we need is whole sets of easy to state, unqualified, universal claims that every one endorses in ethics, not just one or two, or even ten…Then we can begin to do their epistemology.

    As for my few paragraphs in the Stone piece on the Darwinian pedigree of our core morality. they are expanded up on at length in my contribution to The Cambridge Companion to Darwin, both editions, and in Tamler Sommer's and my "Darwin's Nihilistic Idea: Evolution and the Meaningless of Life" Bio and Philosophy 2002 or 2003 I think, and chapter six and seven of "The Atheist's Guide to Reality" where I address some of the things said above explicitly. I am traveling and regret I dont have access to the bibliographic details.

    Reply
  27. Eddy Nahmias

    Picking up on a point made by others above (and by philosophers of the past), it's much less clear that moral disagreement is widespread, heated, or intractable once we align other beliefs (and perhaps practices) among the relevant individuals or cultures. Most of my students are ready to embrace cannibalism as morally permissible, perhaps obligated, when they pretend to sincerely believe that eating the dead bodies of one's enemies is the only way to prevent their spirits from harming one's family. (This suggests another example of a presumably nearly universal moral belief, something like: one is morally required to prevent harm to (and perhaps to help) one's family, unless another moral obligation interferes. Yes, I know utilitarianism might suggest that a whole lot interferes!)

    The example Rosenberg uses may work this way. If we could convince people in honor cultures that they and their family would not, in fact, be dishonored if they allowed their daughter to survive after she had sex or was raped, perhaps by convincing them that their community members (or religion or ancestors) do not require such killing, then it's hard to see why those people would continue to think honor killing was obligated or permissible. And a good bit of the amazing moral progress–as I would call it, but Rosenberg can only 'emote it'–on gay rights seems to be driven by converging beliefs about the fact that homosexuals do not cause harm (even when they marry!), that they do not chose their sexual orientation, that they are our friends, etc. Once we correct the mistaken beliefs about how God feels about homosexuals, it's not clear there will be much left to drive disagreement.

    I don't think I accept intuitionism (or a 'full-blooded' moral realism), but there are fewer messy facts to align in mathematics before we test people's convergence on the mathematical facts. And yes, there's some tricky slippage between non-moral facts, moral facts, and 'thick' moral facts in these here parts.

    [Brian, I know you ran a poll on the Stone's editorship years ago, but have you run one on whether your readers think that, all things considered, it's a good series and not "generally awful"–e.g., it's better that it exists, even in its current form, than if it didn't (or more benefit than harm for academic philosophy). I'd be curious if readers (or you) disagree with me that the majority of Stone columns are pretty good (some excellent, some lousy). If we could just align all the non-normative facts before we judged…]

    Reply
  28. Scott Hagaman

    The putative "massive, widespread ethical disagreement" is vastly overestimated by philosophers. I want to echo those who have pointed this out above because it cannot be repeated enough. There is widespread ethical agreement across cultures. Many philosophers seem to be deeply confused about the scope of ethical disagreement. This probably has something do with being blinded by their perfectly reasonable focus on the interesting cases over which there is disagreement and their pretense at a rationality unfettered by deep cognitive bias.

    Where there is ethical disagreement, it is important to try to identify the source of that disagreement. Not all ethical disagreements are, at bottom, ethical disagreements. Many of them are factual disagreements. For example, perhaps I believe of something that it is a poison (i.e., a vaccination) and you know that it isn't. If this is the source of our ethical disagreement over the permissibility of vaccination, our ethical disagreement is not at bottom an ethical disagreement. Similarly, I might believe that it morally wrong to sacrifice virgins on altars, as the Aztecs, did, and for the same reasons. This disagreement is not at bottom an ethical disagreement either. At bottom, it's a factual disagreement over a religious belief concerning whether the sun will continue to shine if virgins are not sacrificed. And we could go on and on…

    Moral matters are also not the only ones over which there is widespread, persistent disagreement. We could cite plenty of factual matters where there is widespread, and apparently (at least for now) unresolvable, factual disagreement. Are human persons purely physical things? Does some kind of powerful creator being exist? Is Obamacare going to worsen the economic situation of the poor? Is communism a better system than unregulated free market capitalism? These are not moral questions, but they clearly concern objective matters of fact and they are all answered correctly with either a "Yes" or a "No". They are all widely debated by both the religious and non-religious alike, and persistent widespread disagreement concerning these questions exists. Yet nobody draws the lesson that there aren't right or wrong answers to these questions or finds themselves tempted by something akin to non-cognitivism in these domains.

    Finally, widespread ethical disagreement is to be expected. First, many ethical disagreements are really factual or religious disagreements, etc… Second, human beings are incredibly biased creatures who are often poor at reasoning, and especially so when it comes to difficult issues. Our moral beliefs are often the result of self-interest, self-image and group membership. Swapping moral beliefs can often be incredibly costly. It can break group and family ties, for example, which might result in a direct and difficult to overcome monetary loss. Huemer's discussion of disagreement in _Ethical Intuitionism_ is stellar, and I recommend it for anyone who is inclined to vastly overestimate the extent of real ethical disagreement. I've only scratched the surface of his discussion in what I've said here.

    Reply
  29. Brian Leiter

    There is massive and intractable disagreement about all the foudational questions in ethics, as the entire history of moral philosophy shows. And it's not obvious that this intractable disagreement can be explained in an non-question-begging way by some cognitive defect or impairment of some parties to the disagreement. I discuss this in some detail at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1315061

    Reply
  30. Yet Another Anon Grad Student

    I have three points.

    Lexington @ 10 says: "What moral propositions do all moral philosophers agree on? Maybe something as nearly vacuous as that it's wrong to inflict suffering on an innocent person for no other reason than one's own pleasure in doing so? Anything else of significance come to mind?"

    Point 1: There is not even this much agreement. Rather, at best people agree that you should not inflict suffering on an innocent person for no other reason than one's own pleasure *when the person's (expected/actual?) suffering is greater than your (expected/actual?) pleasure*. Hedonistic act utilitarianism may have some faults, but it is still very much an open position with much to say in its favor. The construction of utility monsters is counterbalanced to a large degree by our capacity to construct cases in which people have seemingly perverse preferences (e.g., to be tortured in the near future). Thus both preference-satisfaction utilitarianism and hedonistic utilitarianism seem to break down in psychologically odd cases. Add to this the possibility that people's preferences are often irrational or based on faulty information and the gap between the two positions is narrowed considerably.

    Point 2: The fact that there may be widespread social agreement on the acceptability of certain actions does not imply that there is widespread moral agreement. For there to be widespread moral agreement there must also be widespread agreement as to *why* these actions are acceptable/unacceptable. Consider an analogy. There was widespread agreement among the common people in the ancient world that objects fall toward the earth when dropped. But this did not entail any sort of widespread agreement on the underlying physics. Most people probably didn't even think about it. So the mere fact that more people today agree that slavery is bad in no way implies that we've made progress in our moral theories. This brings me to my third point…

    Point 3: Perhaps the most important disanalogy between moral theory and mathematics is the simple fact that people actually listen to mathematicians. That is, whenever people use mathematics they defer to mathematicians. Platonists like to talk about the abstract virtues of mathematical theory, but the fact of the matter is that we would not study mathematics if it were not incredibly useful in describing the physical world. Yes, many mathematical truths strike us as obviously true from an a priori standpoint, but there is still no way that mathematics would have the authority that it does if it were not useful. In contrast, the applications of moral theories almost always lose out to our pre-theoretic moral judgments. In mathematics the theory always comes first and always overrides our particular judgments about cases when the two disagree, whereas the exact opposite seems to be true in moral theorizing. Let us put this in the context of point 2. Calculus had to be developed as a mathematical theory before physicists started using it. In contrast, it is far less clear that theoretical moral objections to slavery were a necessary prerequisite historical development before abolitionism. More likely, the nascent attitudes against slavery developed first and were then justified in a post-hoc manner philosophically.

    Reply
  31. s. wallerstein

    Eddy Nahmias

    Even if you could convince people in certain cultures that their family honor is not at stake if their daughter is raped and thus, avoid honor killings, what seems more difficult to do is to convince those who do the honor killings, the males of the tribe, that they don't have certain rights over the women of the tribe and that the women don't "belong to" them in some sense.

    Why is it more difficult to convince them that the women don't "belong to" them?

    Simply because they obtain benefits from the oppression of women, not to mention the pleasure and ego inflation that coming from lording it over other human beings. At least in this case, it seems that in order to understand honor killings we need to pay a bit more attention to Nietzsche's theory of the origins of morality as well as the Marxist or semi-Marxist concept of ideology, that ethical systems often are rationalizations or justifications of power/economic interests.

    Reply
  32. Daniel Muñoz

    I'm not sure I trust people who say that agony isn't bad while taking such extreme measures to avoid it!

    A less cheeky point: the realist could say (following Ross, Moore, Ewing, Russell, and many others) that moral disagreements are way more tractable when they're about the basics—Is pain ever bad? Is freedom from coercion ever good?—and only really trenchant when it comes to weighing things up. But that's to be expected: weighing is hard. And before you think that the hardness of weighing cuts against moral knowledge in general, consider how hard it is to weigh up the various respects in which an experience can be painful. It's often difficult even when the pains are of the same type, as Sidgwick admits, but that doesn't show that there aren't any determinate facts about comparative painfulness.

    And even if we grant that there's deep, widespread, expert disagreement about moral questions, the realist can always deny the premise that this is a serious threat to knowledge. One way to do this is to deny that anti-realists are the realist's epistemic peers—which is quite rude, but not all that persuasive. A better move would be to deny that disagreement is a problem even among peers. The nice thing about this: if the anti-realist objects, she may have to admit that there's deep, widespread, expert disagreement about whether such disagreement undermines knowledge. But now the view that disagreement undermines knowledge is unstable, because it's itself is undermined by disagreement! The argument against moral realism turns out to be self-defeating, in that it attacks its own assumptions about what you need to have knowledge.

    BL COMMENT: For those interested, I address the self-refutation argument in the paper above, the final version of which is in the 2014 Oxford Studies in Metaethics.

    Reply
  33. Yet Another Anon Grad Student

    Scott, you might have a point if moral theory were simply a matter of listing acceptable and unacceptable actions. But it isn't. A moral theory must actually include some systematic explanation of why certain actions are acceptable or unacceptable. See my "Point 2" in comment 30 above. In trying to explain many ethical disagreements in terms of religious disagreements you are blatantly overlooking why these religious disagreements are important in the first place; i.e., because a huge number of people subscribe to some form of theological volunteerism. There might be widespread agreement as to whether torture is wrong, but this likely belies the fact that people think it is wrong for very different reasons (e.g., it goes against God's commands, it is ineffective and causes more suffering than good, it violates fundamental rights and is wrong no matter how many lives it might save, it's a dishonorable form of cowardice to treat prisoners that way, etc.). To reuse the analogy I gave above, superficial agreement among the folk as to whether certain actions are or are not acceptable has no greater implications for ethical theory than superficial agreement among the folk as to how physical objects behave has for theoretical physics.

    Reply
  34. Scott Hagaman

    @BL It seems likely that your latest post was in response to mine. But whether or not that's the case, we should distinguish between various kinds of ethical disagreement. The sort of disagreement you appear to be appealing to in your comment and linked article is a kind of metaethical disagreement. It's not something like disagreement about first-order evaluative claims (e.g., abortion is morally permissible/impermissible). When it comes to first-order moral claims, there is widespread, cross-cultural agreement (and when there isn't, the disagreement is often not ethical at bottom.) That there is widespread metaethical disagreement is not, I think, particularly interesting or surprising. While there is no first-order disagreement amongst mathematicians (as you yourself note) about the claim that 1 + 1 = 2, there is widespread and perhaps intractable disagreement about what makes it the case that 1 + 1 = 2 and whether or not that mathematical proposition is objectively true or not. So just as we have philosophical (or metamathematical) disagreement, we have metaethical disagreement. But ethical disagreement (of the first-order type) is, again, not nearly as widespread as philosophers like to pretend.

    I also note that in the article you linked you take Nietzsche to hold that moral facts are not objective. And you seem to think that if this is correct, it follows that there is no moral knowledge. But given your characterization of "objective" (p. 4, fn. 4), it looks to me as if no such thing follows. First, on your characterization of "objectivity", it may well be a perfectly objective matter of fact that some act is subjectively (or non-objectively) morally wrong. Second, it might well be that it's a subjective (or non-objective) matter of fact that some joke is funny, say. But from that, it's hard to see why one should think that one can't know that the joke is funny. So as an epistemologist, I find it difficult to see (on a quick first glance) how the type of foundational (better: metaethical) disagreement you're appealing to in this article provides interesting support for moral skepticism.

    BL COMMENT: it's not metaethical disagreement, it's the kind of fundamental disagreement between say, utilitarians and deontologist. It is first-order, substantive normative disagreement, and it is intractable.

    Reply
  35. Scott Hagaman

    @33 I am happy to grant that there is widespread (and potentially intractable) metaethical disagreement. But there is also widespread (and potentially intractable) metamathematical disagreement. Such is the nature of fundamental or foundational philosophy. What I reject is the claim that there is either widespread first-order ethical or mathematical disagreement that is ethical or mathematical at bottom. And, of course, Rosenberg's primary example concerned a first-order ethical claim, not a metaethical claim. Generally speaking, this is typical of how arguments from putative widespread ethical disagreement go. Perhaps you think that a lack of agreement (or the existence of widespread disagreement) on the what the correct metaethical theory consists in casts first-order ethical positions into doubt? I see no reason to believe that, but perhaps you can provide one? Remember, one can know that oil and water don't mix well without knowing why they don't. Knowing the higher-order chemical or physical explanation for why they don't mix well isn't a requirement for knowing that they do not.

    Reply
  36. Andrew Sepielli

    Brian — Does it matter for your purposes that the attribution of a disagreement to a cognitive defect is grounded solely in claims external to the ethical domain, or just that the attribution is not question-begging?

    And a question for Professor Rosenberg, if he's still reading these comments: You write that "These conclusions encourage tolerance of ethical differences and an appropriate diffidence about our own moral judgments. But they also make it harder to condemn honor killing or even more extreme or violent actions. If “honor killing is wrong” reports our emotional horror at the practice, and not its objective wrongness, then even worse moral catastrophes will be hard to condemn."

    Do you mean these claims about "encouragement" and the "hard"ness of condemning as merely psychological, or also as claims about what's warranted? If the latter, why do you think this? I don't ask this in an entirely skeptical spirit. Just curious.

    Okay, okay, and another question for Prof. Rosenberg, if he's reading. You write: "Many people will not find this a satisfactory outcome. They will hope to show that even if moral judgments are expressions of our emotions, nevertheless at least some among these attitudes are objective, right, correct, well justified."

    Agreed. I'm interested in why people have this hope. Any ideas? Does anyone else have any ideas?

    BL COMMENT: Thanks, Andrew, I will reply, can't right now, but wanted to get your question out there.

    Reply
  37. Yet Another Anon Grad Student

    Scott, you seem to be characterizing ethical views in an exceptionally narrow manner such that they consist only of ascriptions of rightness/wrongness to specific actions. Generally speaking, there is a distinction to be drawn between applied ethics, normative ethics, and metaethics. The debate between Kantians, consequentialists, virtue ethicists, etc. belongs to normative ethics, while the debate between realists, non-congnitivists, error theorists, etc. belongs to metaethics. Granted, the distinction is not a sharp one. But I think most people would clearly count the disputes in normative ethics as ethical disputes. At any rate, there's no particular need to engage in a terminological dispute. One of the consequences of your narrow definition is that widespread "ethical agreement" does not actually count as evidence of progress in ethical theorizing, and it's fairly clear that progress in ethical theorizing is precisely what people in this thread are interested in.

    "Perhaps you think that a lack of agreement (or the existence of widespread disagreement) on the what the correct metaethical theory consists in casts first-order ethical positions into doubt? I see no reason to believe that, but perhaps you can provide one? Remember, one can know that oil and water don't mix well without knowing why they don't. Knowing the higher-order chemical or physical explanation for why they don't mix well isn't a requirement for knowing that they do not."

    You are right insofar as widespread disagreement regarding normative ethics and metaethics does not cast first-order ethical positions into doubt when simply taken by itself. But no one has really made this claim in the first place. Rather, it is this widespread disagreement alongside other considerations (e.g. the lack of a naturalistic explanation for ethical claims, etc.) that form the case that the first-order ethical claims are mistaken. (Recall also that many of our commonplace intuitions about how physical objects behave are woefully mistaken and had to be corrected by systematic physical theory; e.g., the intuition that one should drop the bomb when one is over the target rather than before, or the intuition that heavy objects fall more quickly than lighter ones.) If you couple the widespread disagreement about normative ethics/metaethics with an undercutting evolutionary/sociological explanation for why the first-order ethical beliefs are prevalent, then this gives one evidence that ethical theorizing is a sort of post-hoc rationalization and that such beliefs do not really track a realm of universal sui generis moral truths that exist independently of social context. At any rate, my aim was not to defend such a theory but rather to point out what sort of evidence folk belief actually gives us.

    Reply
  38. Lexington

    I find myself mostly agreeing with Yet Another Anon Grad Student's analysis.

    It may be — and it would probably be surprising if it weren't so — that we might mostly agree on any number of judgments of whether a particular action is right or wrong. The real problem is when we try to construct broader theories that subsume the actions upon which we agree: they always seem to break down, and pretty fast. We try to come up with basic principles that might subsume these judgments, but those principles always seem to justify things that are wrong, or fail to justify things that are right, and are intended to be handled by those principles.

    This contrasts, again, with the experience in mathematics. Mathematics too begins, certainly historically and pedagogically, with strong, well, "intuitions" about how particular geometric objects and numbers behave. But, over time, those "intuitions" can be generalized and generalized again, and sharpened and sharpened again into a set of basic propositions each of which is itself strongly "intuitive", which are all together consistent, and which explain all the more limited, more concrete "intuitions" from which we started. The activity of theorizing reinforces, rather than undermines, our original sense of what's going on the mathematical domain.

    It's really at this systematic, theoretical level that moral systems always fail.

    And I think it's likely that it fails, not because doing moral theory is so hard, or because those who work at it somehow aren't as smart as mathematicians, but because there just is no coherent theory subsuming all of our judgments about concrete cases. Indeed, our intuitions about many cases seem often to be incoherent, as some of the X-phi experiments seem to show. Both Rosenberg and Yet Another Anon Grad Student, I think, are onto the source of the problem: the connection between moral judgments and evolution (and culture).

    My own point here would be: if indeed our moral intuitions regarding particular concrete actions are derived from evolutionary demands, then the situation we are in is exactly what we would expect. Evolution demands that we (or our ancestors) experience the strong need/desire to avoid and/or punish certain actions and engage in and/or reward other actions. Evolution just doesn't care whether the "intuitions" we experience regarding the rightness or wrongness of these actions can be assembled into a larger, coherent theory. Its interest is in the actions themselves, and, really, only in the statistical tendency to perform them or avoid them, and in the statistical tendency of those actions to promote or degrade one's ability to have more surviving offspring. If, indeed, there were actually a "correct" morality above and beyond what we find "intuitive" on a concrete level, why would it ever seem intuitive to us, and how would we ever know it, given that evolution has the whip hand here? There are very few broad moral principles we might all agree on in any case, and there's really no hope of constructing a theory of any really important application from them. How, from such a flimsy reed, might we get to a point that we could make out the difference between those "intuitions" that are "objective", and those that were dictated to us by evolution?

    I don't think this leads us to moral anarchy. For better and for worse, we can't just stop being who we are because we can regard all this sub specie aeternitatis, and so cease being concerned about morality. But it does, I think, give a reasonable person some pause when it comes to how we regard alternative moral principles in other cultures.

    Reply
  39. David Wallace

    Replying to Scott Hagaman at 34:

    "While there is no first-order disagreement amongst mathematicians (as you yourself note) about the claim that 1 + 1 = 2, there is widespread and perhaps intractable disagreement about what makes it the case that 1 + 1 = 2 and whether or not that mathematical proposition is objectively true or not."

    I don't think there's any disagreement at all among mathematicians about what makes it the case that 1+1=2. It's this:
    1) 1 is definitionally the successor of 0, 1=S(0); 2 is definitionally the successor of 1, 2=S(1).
    2) The "+" function is definionally given by n plus zero is 0, n plus the successor of m is the successor of (n+m) – n+0=n, n+S(m)=S(m+n)
    3) 1+1=1+S(0)=S(1+0)=S(1)=2

    Philosophers of maths are – rightly! – interested in just what kind of thing just happened there- is it a piece of pure logic, does it tell us something substantive and if so in virtue of what. Mathematicians aren't. (These are external questions, in Carnap's sense; practitioners aren't interested in external questions.)

    In ethics we might find (some) statements that all ethicists agree are true. "Other things being equal, don't torture children", maybe, or perhaps (given background context) more specific things like "Bob acted wrongly in cheating on Alice". But it's not only the meta-level understanding of what makes that claim true that's contested; it's the first order argument advanced as to why it's true. There's no analogy in ethics, so far as I can see, to that completely uncontroversial proof I sketched above.

    Also – and I assume this is Brian's point – uncontroversially-agreed-upon ethical claims are a lot thinner on the ground. Are there circumstances where torture is permissible? Is there a bright line between infanticide and abortion? To what extent can property rights be set aside in pursuit of equality? Should parents be permitted to advantage their children, and if so, when and how? In what circumstances, if any, are we permitted to kill innocents in war? Or, indeed, kill the guilty in war? When is sex selection of children permissible? When is incest permissible? Outside set theory – itself a very esoteric corner of modern mathematics – there don't seem to be any analogous intractable questions in maths. Even in set theory, there's no controversy about how to handle unprovable claims: just go to the meta level, treat "Is X provable?" as itself a mathematical question, and explore what kind of additional axioms can be introduced in order to render X or its negation provable.

    Reply
  40. Tamler Sommers

    Alex Rosenberg (an objectively awesome dissertation supervisor) writes:

    "These conclusions encourage tolerance of ethical differences and an appropriate diffidence about our own moral judgments. But they also make it harder to condemn honor killing or even more extreme or violent actions. If “honor killing is wrong” reports our emotional horror at the practice, and not its objective wrongness, then even worse moral catastrophes will be hard to condemn."

    I'd like to see some support for this claim. For one thing, the people you disagree with might have false beliefs about the relevant empirical facts. Second, if the parties agree about the relevant empirical facts there is still the epistemological problem: Ok, there’s a fact of the matter, but how do I know I’m right? Why is it any easier to condemn someone without epistemological justification? It could be (as Blackburn argues) that the first order moral debates would proceed in virtually identical ways for expressivists as they do for objectivists. And of course, at least one prominent ethical anti-realist, Brian Leiter, doesn’t seem to have much difficulty condemning practices that are far less horrific than honor killings…

    Reply
  41. Beckett Sterner

    For anyone interested in how mathematical disputes actually get carried out and settled (at least sometimes), I highly recommend "Weaving Self-Evidence: A Sociology of Logic" by Claude Rosental. Google Books Preview here: https://books.google.com/books?id=LqTVFkh5tbsC&lpg=PP1&dq=weaving%20self%20evidence

    Reply
  42. TJ

    I want to echo David Wallace @ 39. If you ask "does my math add up to launch a rocket to Pluto" all mathematicians applying mathematics correctly get the same answer. If you ask any moral philosopher "can I pull this lever?", good luck getting all of them to reach a consensus. https://en.wikipedia.org/wiki/Trolley_problem#Views_of_professional_philosophers

    This isn't to say that this is a failure of philosophy, it's a different domain and it's interesting that there is this divergence. And much we need to learn from it. But any attempt to somehow equate their epistemology boggles my mind — is there some sort of misguided defensiveness at play here, the older sister thinking it should be able to do everything its younger but now grown up sister can do because it used to be so?

    Reply
  43. Justin Clarke-Doane

    I’d just like to make the following observations.

    1. For any area, F, there are two kinds of F-disagreement. First, there is disagreement over what F-claims follow from others. Second, there is disagreement over what F-axioms are true. The first kind of disagreement is just logical disagreement. The second kind is peculiarly F-disagreement.

    2. It is obvious that there is “more” ethical disagreement than mathematical disagreement in the following sense. There are more pairs of people who advocate conflicting views with respect to ethical “axioms” than there are who advocate conflicting views with respect to mathematical ones. But this is not an important difference. The overwhelming majority of mathematicians, let alone lay people, seem to have no serious view as to what mathematical axioms are true (and I doubt that most could even list the axioms of a system of set theory). What matters is the proportion of those with views on F-axioms who disagree. Actually, it is natural to go further. All that matters is the proposition of those with “informed” views on F-axioms who disagree.

    3. There is a difference between agreement in practice and agreement in belief. It is clear that the natural science community has decided to use the Least Upper Bound axiom. I don’t think that it follows that most natural scientists thereby believe that Weyl’s arguments are fallacious.

    4. Moral disagreement evidently tracks with personal and religious investment in a way that mathematical disagreement does not. But it is unclear how this could show that the moral disagreement is better evidence for moral antirealism than mathematical disagreement is for mathematical antirealism. Unlike the former, mathematical disagreement cannot be “explained away” as reflecting the aforementioned distorting factors.

    Reply
  44. Yet Another Anon Grad Student

    "There are more pairs of people who advocate conflicting views with respect to ethical “axioms” than there are who advocate conflicting views with respect to mathematical ones. But this is not an important difference. The overwhelming majority of mathematicians, let alone lay people, seem to have no serious view as to what mathematical axioms are true (and I doubt that most could even list the axioms of a system of set theory). What matters is the proportion of those with views on F-axioms who disagree. Actually, it is natural to go further. All that matters is the proposition of those with “informed” views on F-axioms who disagree."

    This is a flawed analogy. People might not care about choices among equivalent axiom-systems, and it may be that specialists are the only ones in the position to have an inforned opinion on decisions between equivalent axiom systems. But in moral philosophy the different axiom systems in question are far from equivalent: they prove very different things. If people think that torture is okay in some circumstances then this has a direct bearing on "axiomatic" disputes (e.g., consequentialism vs. deontology). There are some mathematical results at issue when mathematicians debate axioms (e.g., the continuum hypothesis). However, these results are not the central focus of contemporary mathematics. Rather, most mathematicians are concerned with using existing axioms to establish theorems like the Riemann Hypothesis. In contrast, the disputed axioms in ethical theory are directly and intimately tied to the primary questions ethicists are concerned with. Ethicists really are concerned with whether torture is always wrong. Moreover, it may be that some mathematical shortcuts are used in the sciences, but these shortcuts can almost always be justified as simplifications that are "close enough for government work". That is, there is usually an explanation for why one can take those shortcuts in specific applications and have them yield relatively accurate results. There is no corresponding explanation for why one can treat torture as okay in some circumstances in spite of the fact that it is always wrong.

    Reply
  45. An Anon Undergrad

    It seems somewhat strange to me that so many realists/objectivists insist on the existence of considerable ethical agreement. So let me stress again what I think some commentators have already brought up before.

    Distinguish first between two kinds of disagreement:
    1) There can be disagreement about what propositions are true/false. Call this propositional disagreement.
    2) There can be disagreement of what makes it the case that some propositions are true/false. Call this explanatory disagreement.

    Explanatory disagreement as regards ethical questions itself comes in two kinds. In asking the generic 'What makes it the case that p?', where p is some purported normative truth, we can ask either (i) a meta-ethical question about the, say, metaphysical ground of p, or (ii) a normative question about the practical reasons that support p. Possible answers to (i) include: because p is an irreducibly normative truth, because p follows from a suitably idealized constructivist procedure, etc.; answers to (ii) include: because the act described by p maximizes happiness, because only the act described by p manifests a virtuous character, etc.
    As was pointed out before, the kind of disagreement relevant to Alex Rosenberg's objection to the mathematics/ethics analogy is not meta-ethical disagreement, but some kind of first-order normative disagreement – we are interested in moral knowledge, not meta-ethical knowledge. So we can disregard kind (i) of explanatory disagreement. Hence, what remains as possible construals of 'ethical disagreement' is explanatory disagreement (ii) and propositional disagreement.
    Now, it seems clear to me that only explanatory disagreement (ii) is a plausible construal of what is meant by 'ethical disagreement'. To think that moral philosophers agree, in any interesting sense of the word, merely because they all accept as a normative truth 'It is wrong to kill innocent babies for fun' is to hold on to a highly degenerate picture of moral philosophy. As Alison Hills argues in 'The Beloved Self' (if I remember correctly), moral philosophy does not aim at 'moral knowledge' (in the sense of propositional knowledge), but at 'moral understanding': We want to understand what practical reasons make a particular act wrong, how the different reasons that we cite in explaining this wrongness connect to each other, what even counts as an admissible reason in moral discourse, what more general normative conceptions imply the reason that we think there are, etc. And it should be clear that on this understanding of ethical agreement, there is none. There is not even agreement on the seemingly simple 'fact' that killing innocent babies for fun is wrong, because such ethical 'agreement' deserves its name only if there is agreement about whether killing innocent babies is wrong because in so acting we fail to will in accordance with the laws of pure practical reason, or because we reduce the net balance of pleasure over pain, or because we fail to make the world maximally perfect, or because we display a vicious character, etc. As long as no such explanatory agreement is forthcoming, it is obviously misguided to assimilate the level of ethical agreement to that of first-order mathematics.
    So on the analogy between mathematics and ethics, I think we can say this: There is a similar degree of meta-disagreement (explanatory disagreement (i)) in both mathematics and ethics; but this disagreement is philosophical and therefore not relevant to the question of first-order knowledge in these domains. There is huge propositional agreement in mathematics, probably non-negligible propositional agreement in ethics as well, but this is not the sense of ethical agreement that matters. In terms of explanatory disagreement (ii), the kind that matters, there is, again, very little disagreement in mathematics (see David Wallance's comment, who also gives an example of a mathematical 'explanation' that virtually all mathematicians agree on), but there is very much disagreement of this kind in ethics. This is at least one respect in which the two domains of mathematics and ethics are disanalogous.

    Reply
  46. Justin Clarke-Doane

    Hi, Yet Another Anon Grad Student.

    I was not talking about “disagreement” over equivalent axioms systems, as in the case of Euclidean geometry. For instance, ZF + Axiom of Choice is not equivalent to ZF + the Axiom of Determinacy. We can stipulate that A and B disagree in believing P and Q, respectively, only if P is inconsistent with Q.

    I agree that disagreement over mathematical axioms is not the central focus of contemporary mathematics. That was part of my justification for not counting raw numbers of mathematicians who disagree.

    I agree that mathematicians are overwhelmingly focused on proving things from standard axioms. I would put it this way: mathematicians are overwhelmingly concerned with questions of logic, rather than with questions of non-logical truth.

    I did not understand your last comment.

    Reply
  47. Yet Another Anon Grad Student

    Justin,

    I guess I'm not quite clear on the dialectical purpose of your second point then. I took you to be providing an argument by analogy between mathematics and ethics for the conclusion that what matters is the level of disagreement between professional ethicists/mathematicians whose research focuses on the axioms in question, and thus that we can largely ignore ethical disagreement among the folk. My point was that the analogy is weak because the folk do in fact have skin in the game when it comes to moral axioms. Disputes regarding what to add to ZFC in order to get CH won't have a large impact on how the majority of society uses mathematics. In contrast, the axiomatic disputes ethicists engage in will usually have consequences for everyday moral decisions. For instance, the fact that most people don't think they have an obligation to give 50% of their income to charity clearly conflicts with Peter Singer's moral system. This appears to significantly weaken the analogy between mathematics and ethics. It might still be the case that disagreement among professionals is what matters, but you need to add a bit more to the analogy you've provided in order to show that this is the case; e.g. an appeal to professional expertise, or an undercutting psychological account of why folk moral beliefs are often mistaken. Perhaps you were tacitly assuming something to that effect. Or perhaps I simply misread your intent.

    Reply
  48. Lexington

    I just wanted to make one of the points I argued in my previous post a bit clearer.

    I should distinguish my point from one that Alex Rosenberg alludes to in his own response here. He quotes Hume "Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger." Hume's idea is of course the general skeptical one, essentially pointing out that from no "is" comes an "ought". But a common answer to that is that we have something like intuitions that settle in our mind whether something is right or wrong, and that's where we get our "oughts". Indeed, it's mostly because this sort of intuition based approach resembles how we think that mathematicians proceed that we bring up the comparison to them here.

    But my point is that the problems appear to be far deeper than the logical difficulty of getting our "ought" statements. It's rather that, across and even within individuals, our intuitions about morality simply aren't coherent enough or anywhere near complete enough to put together something that resembles a satisfying moral system or theory. Supporting this conclusion would be the results of a number of X-phi experiments, as well as the manner in which any such system or theory fails, usually at a very basic level, to gain broad assent.

    The question is, why should we be in this fix?

    It seems strange, if our intuitions were really in some sense "objective", revealing a domain of "truths" (however that gets cast out), that they would be so very deficient in their coherence and completeness. Certainly on this score the model of mathematics couldn't be more different.

    But my claim is that the best explanation of those deficiencies is that these intuitions really don't reveal anything about some domain, but rather that they are in effect instilled in us by evolution (combined with and shaped by culture). I would claim that this assumption would explain, almost perfectly, why we experience such difficulty putting together overarching moral theories based on our intuitions. Evolution has no reason to keep our intuitions globally consistent — its interest is in motivating us to perform or avoid certain actions, depending on whether those actions are statistically correlated with good or bad reproductive outcomes. We need just enough of the right sort of intuitions to motivate us in the suitable direction most of the time. Moreover, evolution can do its work only through the very blunt instrument of genes. So this is a very sloppy process. The reasonable expectation is that those intuitions might readily conflict in certain cases, or simply fail to cover others; in the end, it's no skin off the nose of evolution.

    Between the two explanations of our intuitions, the evolutionary account seems a vastly better one, certainly at first blush. And, as I had argued, even if our intuitions were sometimes based on something objective, how would we ever reach a point that we might know which were objective, and which fed to us by evolution, given the inadequacies of our moral theories at this point? There appears to be no point of strong certainty regarding "objective" intuitions from which the two kinds of intuitions might be sorted out.

    Reply
  49. Mike

    I think S. Hagaman's point is not that the disagreement is metaethical, though he uses that term. His point is that the disagreement is first-order, but theoretical; it is largely disagreement concerning the basis or foundations of morality. The agreement he notes concerns particular moral judgments or something like Rawlsian considered moral judgments. So the disagreement is first-order, but it's at a level that concerns almost no one except moral philosophers. The widespread agreement is also first-order, and it concerns particular moral judgments. In any case, that sounds like the position he's defending.

    Reply
  50. Justin Clarke-Doane

    Hi, Yet Another Anon Grad Student. Thanks for the follow-up.

    First, let’s bracket disagreement over CH, for the reasons touched on in Gideon Rosen’s comment and my first one. We can talk about Induction, Infinity, or Least Upper Bound. These axioms have major ramifications for “everyday” mathematics. By “everyday mathematics”, I don’t mean the kind you learn in grade school, which can be carried out in Robinson Arithmetic (though I think that honest ultrafinitists must reject even this). I mean the kind you study in college and graduate school, if you’re not a logician.

    Vanishingly few people ever seriously ask themselves about the truth of mathematical axioms. If you were to object to a proof in Real Analysis on the basis of skepticism about the Least Upper Bound axiom, you’d be met with confusion – not because that axiom is “obviously” true, but because the question of its truth is virtually never broached in the context of ordinary mathematics. Ordinary mathematicians are almost exclusively interested in establishing that IF standard axioms are true, THEN so too are standard theorems. Ask them whether the standard axioms are true, and you are unlikely to get a stable answer.

    What kind of mathematical disagreement matters for mathematical realism depends on how the argument from moral disagreement to moral anti-realism is supposed to work. Leiter suggests that moral disagreement that cannot be explained away as reflecting some independently-specifiable cognitive shortcoming is the kind that matters (e.g., “John is insensitive to the moral facts” is not an explanation in this sense). This suggestion is in harmony with Mackie’s “Argument from Relativity” and Wright’s discussion of “Cognitive Command”. A good question is: why would this kind of disagreement matter? One answer might be that it is evidence of indeterminacy in our language. If that were so, then the disputed propositions would be relevantly like the Continuum Hypothesis is commonly supposed to be.

    But it does not matter whether Leiter is correct. Either we count disagreement among non-“experts” in the two areas, or we do not. Suppose that we do. Then in the mathematical case this will add a mass of agnostics. How would that help show that moral disagreement is better evidence for moral anti-realism?

    A final comment: some commentators dismiss mathematical disagreement as “merely philosophical”. I think that there are three problems with this. First, where is the line between mathematical and philosophical considerations? Is the Limitation of Size doctrine “merely philosophical”? Is the cumulative hierarchical conception of set? How about the ban on impredicative definitions? Second, even if one says that there are paradigmatic cases of philosophical and mathematical disagreements, even if there is no principled boundary between the two, why is non-philosophical disagreement better evidence for anti-realism than philosophical disagreement? Finally, how do these commentators suggest that we insulate our “first-order” mathematical beliefs from such “philosophical” attacks? If Weyl’s arguments against impredicative definitions are compelling, then how can we continue to accept the “first-order” proposition that every non-empty set of real numbers with an upper bound has a least upper bound?

    Reply

Designed with WordPress