Sharon Street

Sharon Street (New York University), “Evolution and the Schizophrenia of Quasi-Realism About Normativity,” with commentary by David Enoch (Hebrew University).  Both the paper and commentary can be found here.

Comments

[David White]

I enjoyed the paper a lot and am very broadly sympathetic to the arguments presented. I thought the challenge that David E presented at the end of his commentary was a good one and I had hoped someone smarter than me might try to take a stab at answering it. But absent that, I'll offer my best reply.

The potential problem goes like this: If the Darwinian Dilemma (DD) can be run successfully at our normative ethical beliefs then it also seems it should be just as successful when run at our meta-ethical beliefs. So one cannot use the DD to support anti-realism at the normative level and still be a "meta-realist". But let's look at just how the DD is supposed to work.

For some cluster of beliefs we can ask whether or not there is a good evolutionary explanation for us having them. If yes, then we can ask if there is any evolutionary reason that we should expect those beliefs to track truth, or whether there is an evolutionary reason to have them even if they are not true. For beliefs about mid-sized physical objects there is an evolutionary explanation for why we have the beliefs about them that we do, but that story requires that they roughly track the truth to make sense. So it seems we have some reason to think that they do track the (mind independent) truth. With normative ethical beliefs we can tell an evolutionary story for why we have them, but that story works regardless of whether or not those beliefs track truth, so the DD kicks in. It would seem massively coincidental if our normative moral beliefs did track the truth given that we would have them whether or not they did so.

Now what of meta-ethical beliefs? Is there an evolutionary story for why we have them as there is for mid-sized physical objects or normative ethical beliefs? Well, certainly not a direct story. Through most of human evolutionary history it is hard to identify any meta-ethical beliefs we might have had and it does seem that we would have had just as strong a tendency to survive with no meta-ethical beliefs at all, so it looks like the DD does not get off the ground at the first step. But David E writes, "our metaethical judgments – much like our moral and other normative ones – have been shaped by evolutionary pressures". So perhaps there is still some work left for the meta-realist to do.

The next reply would be to ask just what meta-ethical beliefs we have tended to form based on evolutionary pressures? Well, they seem to be, by and large *realist* beliefs about morality. This, too, makes evolutionary sense because people who believe that we are all bound, whether we like it or not, to a set of moral values that require us to care about the well being of others and act on those values, are going to be a species that has a better likelihood of survival than one that believes that morality is rooted in our individual, subjective sentiments. So, in fact, the DD *does* work against *traditional* meta-ethical beliefs, the realist ones. There is a reason for us to tend to be realists whether or not realism is true - an evolutionary advantage it provides.

So how, one might ask, could anti-realist meta-ethical beliefs ever arise? There is no evolutionary advantage or explanation for them, so why do we have them? The answer, I think, is this: There is no *direct* evolutionary explanation for them, but there is an indirect one. Our more sophisticated metaethical beliefs (ones that transcend knee-jerk realist assumptions) are ones based on beliefs we have about human psychology, language usage, and conceptual analysis. Beliefs in each of these domains, when applied to meta-ethical considerations, seem to indicate that the realist story does not work, but an anti-realist one does. And just as with mid-sized physical objects, there is good evolutionary reason to think that our beliefs in these realms should track the truth. Beliefs that do not would put us at a survival disadvantage.

Now if you have not noticed this already, let me make the craziest part of this answer explicit: I am essentially arguing that we have come to a point where our ability to identify meta-ethical truths transcends that which is to our best evolutionary advantage. Furthermore we are (from the point of view of species survival) better off believing that realism is true even though it is false. So, yes, meta-realism is true, normative anti-realism is true, but SHHHH!!! Don't tell anyone!

[David Enoch]

Thanks for your question, David. Let’s see.
First, Sharon argues – rightly, I think – that pretty much all our judgments were shaped subject to evolutionary pressures. So this seems to apply in particular to our metaethical ones.
Second, I don’t think that whether there are evolutionary advantages in believing realism is relevant here. So long as there is no reason to think that evolutionary pressures make our metaethical judgments track the (metaethical) truth, Sharon is still committed to meta-anti-realism, that is, to anti-realism about the realism-anti-realism debate. But we’ll see what Sharon thinks about this in a little while.
Third, I agree, of course, that evolutionary stories can be indirect, and in particular, that truth-tracking can sometimes be the byproduct of something that was selected for rather than the selected-for feature itself. And if the details can be filled in, perhaps it can be shown that such an indirect story can work for metaethics but not for ethics. Perhaps. But the details need to be filled in. And until they are, the analogy between the first- and second-order problem stands, I think.
As for not telling anyone: Well, I am not a fan of a publicity or transparency requirement even with regard to normative ethics. So violating it with regard to metaethics is certainly not something I have a problem with (not that the masses are waiting for an update on the latest developments in the metaethical literature…).

[Pablo Stafforini]

The reason why we should expect our beliefs to track meta-ethical truth is simply because this area of inquiry requires the exercise of faculties that are themselves evolutionary advantageous. We do not need to postulate that knowledge of differential calculus confers organisms a reproductive advantage in order to be entitled to claim that our mathematical beliefs are generally reliable. Similarly, we can be generally confident about the results of discussions about normative realism without committing ourselves to the claim that that our ability to identify meta-ethical truths gives us a competitive edge in the struggle for survival. This is because both mathematical and meta-ethical thinking are species of the broader capacity for rational thought. We shouldn't be surprised to see that some of the beliefs formed by the exercise of reason have no biological function at all. The point to bear in mind is that as a whole rationality, like sensory perception, and unlike moral “intuition”, is biologically profitable.

[David Enoch]

Again, I agree with the general point. But the important one is the one at the very end of Pablo's remark: Is it so clear that this kind of story can be told in a much more plausible way about metaethics than about ethics?
It seems to me that the answer is no. But I don't claim conclusiveness here, of course.

[Andrew Lee]

Both David W and Pablo make the same substantial point -- there is no reason to suspect that the Darwinian Dilemma applies to high-level theoretical thought about things like maths and metaphysics. The reason why the DD doesn't apply is just the same reason why Plantinga's forehead-slappingly bad arguments agains evolutionary naturalism fail. One can plausibly construct a consistent "lie, cheat, and steal" set of evolutionarily generated normative preferences that is quite conducive to reproductive success. But just try to imagine what a consistent "false, contradictory, and impractical" high-order belief forming system would look like that would routinely generate false beliefs but be of survival value anywhere outside of a brain in a vat. Once evolution coughs up a species capable of rational thought that does not consist in a set of rote stimulus-response algorithms, then any such belief-forming system will by definition have structural error-correcting mechanisms, and so we will more or less by definition "have reason" to trust the results of beliefs arrived at by reason.

But when one looks at the moral epistemology of cases like the Trolley Problem, it's not at all clear to me that what's going on is an abstract cognitive process akin to differential calculus or debates on coherence vs. correspondence. Rather, one is supposed to simply "intuit" or otherwise directly perceive that a certain event is wrong. The DD gives us good reason to be suspicious of judgments of this sort, in a way that it should fail to raise suspicions about higher-order reasoning.

So not only would I not say the DD applies to meta-realist beliefs, I would want to say that what Sharon Street describes as a "liberal quasirealist" shouldn't have any problem requiring "practical faith" as a solution. As I understand QR, it doesn't require faith so much as a basic confidence in our general reasoning capacities on the one hand and the notion of maintaining a consistently realizable set of norms on the other in order to say one thing with the metaethical hat on and a different thing without.

[Andrew Lee]

Of course, as David Enoch's commentary pointed out, the argument seems to work plausibly well against what Gibbard actually says in the more recent book, which I can't claim deep acquaintance with. The only reason I can think of for why he would want to be able to make a claim so strong about mind-independence as part of his imitative project is nostalgia; it doesn't seem to be required to make his QR work.

So may be a better title for the paper should be “Evolution and the Schizophrenia of Quasi-Real*ists* About Normativity,”

[Sharon Street]

Many thanks to David W, Pablo, Andrew, and of course David E, for weighing in with their comments. At the end of the conference, I’ll post a full reply to David E’s formal commentary, and there I’ll explain my own reply to the worry under discussion. In the meantime, though, I wanted to weigh in on a few points.

First, David E says above that I argue (rightly, in his view) “that pretty much all our judgments were shaped subject to evolutionary pressures.” While there’s a sense in which I do think this, I wouldn’t want to put it that way -- at least not without important clarification. In my view, all our judgments “were shaped subject to evolutionary pressures” in the following, limited sense: in rendering any judgment whatsoever, we can’t help but employ general capacities whose origins trace back in one way or another, at least in part, to evolutionary processes (which aren't limited to natural selection but could also include processes such as genetic drift, etc.). This by itself is a fairly weak claim, however, and (as implicitly noted in the discussion above) *this* kind of “evolutionary influence” on our judgments is by no means sufficient for the Darwinian Dilemma to kick in. For example, our judgments about the presence of midsized objects in our immediate environment are rendered using general capacities whose origins trace back, at least in part, to evolutionary processes, but that point alone doesn’t undermine the supposition that we’re tracking, with those judgments, independent facts about the presence of midsized objects. The same thing holds in the metaethical case: our metaethical judgments too are rendered using general capacities whose origins trace back, at least in part, to evolutionary processes, but that point alone doesn’t undermine the supposition that we’re tracking, with these judgments, independent facts about metaethics.

Second, we must proceed very slowly and cautiously in exploring whether there is any additional, more interesting sense in which evolutionary pressures might have “shaped” our metaethical judgments. There is a high risk here (as elsewhere) of lapsing into purely speculative, unsupported “just so stories,” and of making the “adaptationist” mistake of assuming that every observable trait is an adaptation resulting from natural selection. (I don’t mean to be accusing anyone above of this; it’s just one of those points that’s always worth emphasizing.) The claim that evolutionary pressures shaped our metaethical judgments is, after all, obviously false on some interpretations. For example, it’s clearly absurd (though fun) to imagine that in the environment of our ancestors, there were full-blown realists, antirealists, and quasi-realists running around, or that esoteric metaethical beliefs are genetically heritable traits. Might there instead have been *extremely primitive* versions of realists and antirealists running around and being selected for and against -- a possibility mentioned, though not necessarily endorsed, by David W? (Note: presumably there’s no such thing as a primitive quasi-realist!) I am extremely skeptical of this hypothesis too, for a variety of reasons that are too much to get into here (reasons having to do with the nature of the “traits” in question, among other things).

A third observation: To the extent one thinks that there is an important discontinuity between normative ethics, on the one hand, and metaethics (and in particular, the realism/antirealism debate), on the other, one will think that our ethical judgments and our metaethical judgments are judgments of very different types, and one can seek to exploit these differences to show why realism about ethical truths, but not realism about metaethical truths, falls prey to the Darwinian Dilemma. This is the direction that David W, Pablo, and Andrew, as I understand them, are tempted to go. On the other hand, to the extent one thinks that there is a deep continuity between normative ethics, on the one hand, and metaethics (and in particular, the realism/antirealism debate), on the other, one may harbor more of a concern that the Darwinian Dilemma will work just as well against realism about metaethical truths as it does against realism about ethical truths. I take it that one thought that may be driving David E’s worry is the thought that there *is* a deep continuity between normative ethics and metaethics, and hence if the Darwinian Dilemma works on the one level, it’s at least prima facie plausible that it will work on the other (and hence perhaps will be self-defeating). Since *Gibbard’s* view in Thinking How to Live is that there *is* a deep continuity between normative ethics and the realism/antirealism debate -- and since I am wishing to assume this along with him for the sake of argument -- the worry at hand is especially salient in this context. (Note that Gibbard thinks there is a major *discontinuity* between normative ethics and metaethics when metaethics is understood as the study of what kinds of attitudes we’re expressing when we make normative claims, how to solve the Frege-Geach problem, etc.) As I’ll indicate in my official reply to David E, I believe that the worry at hand can be answered *even if* we assume that there is a perfect continuity between normative ethics and the realism/antirealism debate -- in short, my view is that it follows from within every practical point of view that antirealism is right -- but I won’t say any more about that for the moment.

Finally, about not telling anyone that antirealism is true: Go ahead and tell people -- it might actually help the state of the world! The idea that antirealists will tend to be any less devoted to the highest moral ideals than realists, is, in my view, a falsehood on par with the similar idea that atheists will tend to be any less devoted to the highest moral ideals than theists. Though I can see where such worries come from, they’re based on a very simplistic understanding of antirealism (and atheism, respectively). In my view, the strongest kind of “bindingness” can and must come from ourselves, and can be understood as such with a gain rather than a loss of force. So who knows -- maybe humanity’s best hope for survival over the long term is antirealism!

[Patrick Fleming]

I think Gibbard may be less committed to the full quasi-realist package in “Thinking How to Live” than the paper suggests, but I am not sure about this. The core argument of the book is a possibility proof. Gibbard aims to show that it is possible for concepts to work in an expressivist manner. Moreover, he argues all thinkers and planners are committed to concepts that work this way. Proceeding from hyperstates is supposed to show this. In the first six chapters he is not making any claims about how our actual normative concepts work. Instead he is setting out the logical features of a discourse that planners need.

Later in the book he turns to our actual concepts and argues that the theory matches our actual concepts pretty well. But I am not sure if he thinks there is a complete match. At one point he explains his approach by saying, “do the attitudes we normally take toward life and figuring out how to live it, I ask, lie anywhere in the vicinity of a coherent set of views?” (261). This seems to allow for the possibility that our actual concepts are incoherent and need to be revised.

Because of this it is unclear to me to what extent Gibbard is offering his theory as a revision of our ordinary moral concepts. He sets out the commitments that we can not coherently reject. It seems that if the theory does not match our concepts, then it is our concepts that need to be changed. The theory is forced on us by the logic of planning. No agent can coherently deny it (or so Gibbard argues). Any set of concepts used for the normative work of planning must have these features if it is to be coherent. Yet, it is an open question whether the concepts we actually use are coherent.

I am not sure if this gives him a way out of the Darwinian Dilemma. But an expressivist might say that any aspect of normal moral discourse that is incompatible with an expressivist understanding should be jettisoned. If the argument from hyperstates works, Gibbard might offer them principled reasons to do so. A proper understanding of planning and value cures the schizophrenia