Hi all –
Very happy to have Sharon Street joining us at PEA Soup for a stint in the Featured Philosophers chair! Sharon is Associate Professor and Associate Chair of the Department of Philosophy at NYU. She provides a very helpful introduction to her work below the fold. Please join me in welcoming Sharon!
-dd
I am grateful
for the invitation to appear here on PEA Soup, and would like to thank Dale
Dorsey, David Shoemaker, and David Sobel for organizing this series. I would be glad to discuss any aspect
of my work over the next few days.
Roughly
speaking, the goal of my work so far has been to develop an account of
normativity that coheres with a broadly naturalistic conception of the
world. While most of my writings
focus on the case of practical reasons, the view I develop is intended to apply
to epistemic reasons as well.
My work on
the topic may be divided into three parts. The first part is
the negative part, in which I argue against various “realist” accounts of
normativity (including non-naturalist realist, naturalist realist, and
“quasi-realist” accounts), according to which there are robustly
attitude-independent normative reasons for belief and action. The second part is the positive part,
in which I develop what I call a “Humean constructivist” account of
normativity. The third part consists
in an effort to show how even if we construe debates about the
attitude-independence of normative reasons as “substantive normative debates,”
as theorists such as Simon Blackburn, Ronald Dworkin, and Allan Gibbard
recommend that we do, we still have
to give up the view that there are robustly attitude-independent normative
reasons. The epistemological
problems that plague realism, I try to show, cannot be sidestepped by
interpreting claims about attitude-independent reasons as first-order normative
claims; rather, I argue, traditional metaethical problems end up reasserting
themselves in a substantive normative guise.
My published
and forthcoming papers, as well as a few unpublished ones, are available at https://files.nyu.edu/ss194/public/sharonstreet/Writing.html. Here is a brief summary of what topics
I address where:
I argue against non-naturalist versions of
normative realism in “A Darwinian Dilemma for Realist Theories of Value,” “Objectivity
and Truth: You’d Better Rethink It,” and a draft on Parfit’s metaethics
entitled “Nothing ‘Really’ Matters, But That’s Not What Matters.”
I argue against naturalist versions of normative
realism in “A Darwinian Dilemma for Realist Theories of Value” (especially
section 7), “Reply to Copp: Naturalism, Normativity, and the Varieties of
Realism Worth Worrying About,” and a rough draft entitled “Normativity and
Water: The Analogy and Its Limits.”
I argue against quasi-realism in
“Mind-Independence Without the Mystery: Why Quasi-Realists Can’t Have It Both
Ways,” and section 7 of “What is Constructivism in Ethics and Metaethics?”.
I argue against realism about epistemic reasons in
“Evolution and the Normativity of Epistemic Reasons.”
I argue against theism, exploring some parallels
with secular normative realism, in “If Everything Happens for a Reason, Then We
Don’t Know What Reasons Are: Why the Price of Theism is Normative Skepticism.”
I develop my
positive “Humean constructivist” account
of normativity in “Constructivism about Reasons,” “What is Constructivism
in Ethics and Metaethics?”, and “Coming to Terms With Contingency: Humean
Constructivism about Practical Reason.”
I address objections to the positive account
in “In Defense of Future Tuesday Indifference: Ideally Coherent Eccentrics and
the Contingency of What Matters” (addressing the allegedly implausible
consequences of subjectivist views like my own) and the draft “How to Be a
Relativist About Normativity” (addressing the question of how one may
consistently be a constructivist about normativity “all the way down,”
including with respect to the attitude-dependence thesis itself).
I argue that
epistemological problems cannot be
sidestepped by interpreting claims about attitude-independent reasons as
substantive normative claims in “Objectivity and Truth: You’d Better
Rethink It,” “Mind-Independence Without the Mystery: Why Quasi-Realists Can’t
Have It Both Ways,” and “Nothing ‘Really’ Matters, but That’s Not What
Matters.” I discuss the sense in
which theism is a substantive normative claim (one that, like realism, leads to
epistemological problems) in “If Everything Happens for a Reason, Then We Don’t
Know What Reasons Are.”
Finally, I
offer a summary of the phenomenon of
undermining explanations in “Does Anything Really Matter or Did We Just
Evolve to Think So?”. (This paper
was written primarily for an undergraduate audience, but it contains some
material on undermining versus vindicating explanations that I don’t discuss
elsewhere.)
First, thank you for your work. Is your contribution to Does Anything Really Matter: Parfit on Objectivity essentially your response to Parfit’s critique of your view in On What Matters?
Hi Professor Street! We briefly met when you came to Austin (I suggested to you a paper by Gideon Rosen, if that jogs your memory). Thanks for being so gracious back then, and thanks for being here now!
To be honest, I don’t know if I understand the modestly attitude-independent concept of a normative reason that you discuss in “Nothing ‘Really’ Matters.” I see how it figures into the dialectic of the paper, but I don’t intuitively get it it the same way I get externalist reasons.
I’m initially inclined to take this as intuitive evidence for the externalist story. Situations make claims on us, but only if we have the opportunity to respond in the appropriate way—what could be simpler? That said, the more of your work I read, the more I worry that you’re on to something and that I should be less closed-minded.
So my question is this: Can you explain to a non-naturalist how your modestly attitude-independent normativity is still normativity? This isn’t a challenge—I just want to understand your position better!
And if you have time to answer a bonus question, do you know of any historical predecessors to your view? I know Ross and Kant assumed robust attitude-independence, but I’m curious to know how long it took before people starting getting skeptical.
Thanks!
Hi Sharon,
my question is about a point you bring up in some of the papers, but which comes out especially nicely in the new paper for the Parfit volume. Namely, that – as you see things – we can point to a certain experience, and thereby get people to understand what we mean when we use the concept of a normative reason. The experience is that of seeing something as a reason “counting in favor” of some action or other response.
You write, for example, that “we have a conscious experience of certain features of the world … as what we can only describe as calling for, or counting in favor of, or demanding, or requiring… certain responses on our part.” (“Nothing ‘really’ matters… p. 4)
My question is this: do you think of these experiences as all having a very distinctive element in common to them? I am asking since I am thinking of, for example, the difference between spontaneously thinking that going to a particular restaurant might be a good idea, on the one hand, and feeling a sense of obligation or responsibility to keep a promise, on the other. Perhaps the thought of a certain dish that is available at the restaurant makes going to the restaurant seem appealing, thereby making it feel as if we have a reason in favor of going there. And the feeling of being “bound” by a promise (which might be a promise to do something one does not particularly feel like doing) can be said to feel like having a moral reason one cannot ignore.
Is it, as you see things, the case that these two kinds of experience of things as giving us reasons have a distinctive character in common? An alternative way of thinking about these different kinds of examples might be that the experience of something as appealing (as an “apparent good”) is rather different from the experience of something as a “duty” or as a responsibility we “have to” discharge. Both kinds of experiences can surely be thought of as experiences which can be said to make us feel like we have a reason to act in some way, but the experience of apparently having the former kind of reason seems rather different (does it not?) from the experience of apparently having the latter kind of reason.
Since these different kinds of experiences of having different types of reasons for action are different in their character, one might wonder why to think of them as experiences of the same thing (namely, of seemingly irreducible reasons). Is the idea that both types of experiences (finding something appealing, and feeling as if one has a responsibility to do something) both have to do with the choice of action that you think that it is enough to talk broadly about experiencing things as “counting in favor” of certain responses? Or do you think of there as being a very distinctive element that is part of both kinds of experiences (and also of any other experiences agents might have relating to their choices of actions)?
Lastly, thanks for making available some of the as yet unpublished materials!
Hi Clare, Daniel, and Sven,
Thank you so much to all three of you for your questions! I really appreciate them, and wanted to let you know I’ll be writing back about each of them shortly, at latest tomorrow. (I was in meetings the whole day today, but my schedule is very clear for tomorrow.) In the meantime, thanks again for writing — looking forward to being in touch!
Best regards,
Sharon
Hi Professor Street,
First let me say that I’m a big fan! I wish I’d have come across your work before writing my dissertation, since as far as I can tell your views are as close or closer to mine as anyone other contemporary philosopher. However, there are important differences, so I’ll focus on these, starting with what I think is the most important.
In Constructivism About Reasons (CAR), you attempt to answer the question of which competing sources of reasons ‘survive scrutiny’ by reference to the question of which sources are more deeply a person’s own. You then seem to conceive of that question as being answerable by the person’s own assessment. But it seems to me that people can easily be wrong about which of their values are ‘deeper’ than others, and even wrong about whether they value something at all. As just one example of how this could happen, there are strong socio-moral norms about which values are praiseworthy and blameworthy. Such norms motivate people to regard their deepest values as being among the praiseworthy, and not the blameworthy. In short, the truism that people can mis- or fail to perceive their own motives carries over to values, though this is somehow very widely unrecognized.
Relatedly, you seem to roughly equate the depth of a person’s values with the fundamentality of the principles they endorse. But again, a person might endorse a (fundamental) principle for reasons other than that it actually reflects her (deepest) values. In my view, this is the norm, but surely they can sometimes come apart. Finally, in CAR you seem to equate valuing with conscious valuing, but to me this seems not much more plausible than equating motivations with conscious motivations (in a footnote to Objectivity and Truth, it seems you might have changed your mind. You also seem to have moved from regarding normative reasons as being constructed from normative judgments to the much more inclusive category of evaluative attitudes, which seems much better to me. Is this right?).
The larger philosophical upshot of these observations is that in my view, normative antirealists—which I’ll define here as those who think that our reasons and/or what we should care about somehow depend(s) on our actual values and other evaluative attitudes—have failed to draw a very important inference from their own views. And that is that discovering what our values are should be an important, even central part of normative inquiry. For if our reasons depend on what we actually value, but we don’t well know what we value, or in what ways or how much, then we won’t be in a good position to know what our reasons are.
Of course there’s a lot more to be said, but at this point I’d like to know whether you think that your metaethical views have this sort of upshot for how we ought to conduct normative inquiry. I haven’t seen anything in your (or any other antirealist’s) work about the normative importance of discovering what we care about, despite the fact that it seems a fairly straightforward consequence of antirealism. Unless of course it’s trivially easy to know what we care about, which I think is quite mistaken. More generally, I wonder if you agree that your metaethical views have any important consequences for how we ought to conduct normative inquiry.
Finally, could you direct me to where you best describe what ‘surviving scrutiny’ amounts to? There is another point of departure between us, but I’ll let that wait.
By the way, I’m working on a book and since I hadn’t written about your work in my dissertation, all the material discussing your work will be new. So thanks in advance since this is a big help for me!
Hello Professor Street,
I quite enjoy your work and have found everything that I have read of yours intellectually stimulating. I have questions about almost all of your papers, but here I will just ask about the latest draft of your paper Nothing “Really” Matters, but That’s Not What Matters. I had the pleasure of seeing you present an earlier version of this paper at our SPAWN conference last year and wanted to (sort of) follow up to one of the questions I asked then.
In section 4 of the paper, you write
“Talk of Externalism simpliciter is ambiguous between externalism conceived of as a substantive normative position and what we can call Analytical Externalism, the view that it is a (presumably non-obvious) conceptual truth that the concept normative reasonNEUT and the concept normative reasonEXT are one and the same. Parfit’s way of dividing up the territory obscures a possibility to which we should be highly attuned, namely the possibility that both Analytical Internalism and Analytical Externalism are false. And why wouldn’t this be the right view? It’s plausible to think, after all, that some of the considerations that raise trouble for Analytical Internalism might also, mutatis mutandis, raise trouble for Analytical Externalism. In other words, just as the expression ‘normative reason’ does not (as Parfit puts it) mean ‘internal reason’, the expression ‘normative reason’ also does not mean ‘external reason’.”
I absolutely agree that Parfit’s definition of ‘externalism’ and the way he draws the distinction between internalists and externalists fails to adequately distinguish between (1) external reasons in the “standard normative sense” and (2) external reasons in the “standard normative sense” that also exist independently of whether there is a sound deliberative route from an agent’s subjective motivational set to the conclusion that the agent has said reasons.
I also agree that the considerations that(supposedly) raise trouble for Analytical Internalism might also, mutatis mutandis, raise trouble for Analytical Externalism and so, if Parfit’s arguments work, then we should reject Analytical Internalism AND Analytical Externalism. After all, wouldn’t Analytical Externalism be a concealed tautology on Parfit’s view?
Now I wonder how problematic this is for Parfit. Given Parfit’s arguments against analytic views, why not understand Parfit as arguing for Non-Analytical Externalism? Why would he need to provide an argument that “it’s (perhaps non-obviously) analytic that reasons in the standard normative sense are external reasons”?
If this is right, then couldn’t Parfit’s discussion of Williams’ (in Chapter 24 of volume II) rely on the Early Death case (in the same chapter) and Parfit’s arguments about Future Tuesday Indifference in order to establish that there are normative reasonEXT instead of just normative reasonINT or normative reasonNEU. Of course, whether Parfit’s arguments work is another question entirely. But *if* they work, then wouldn’t they vindicate Parfit’s appeal to normative reasonEXT instead of just normative reasonNEU?
Apologies if I am misremembering my Parfit or misinterpreting your argument.
————————————————-
I have another, more minor question. You suggest that the realist and antirealist can do an equally good job at capturing the intuitively plausible judgments in all real-life cases and only diverge with respect to esoteric cases that make use of ideally coherent eccentrics.
This is a very minor point, but aren’t there a number of real life cases of ideally coherent eccentrics. I am thinking psychopaths would present us with many such cases. Ignoring psychopaths and focusing on normal people, I have found that doing philosophy with my peers has made me suspect that there is widespread disagreement about lots of common moral judgments as well. In such cases your view and Parfit’s view would diverge when it comes to the reasons these agents have too, right?
Here is a different question about the same issue. In one sense it seems right to me these two views only diverge in a few cases. In another sense, it seems as if the disagreement between you and Parfit ranges over every single reason claim. Just to make sure that I understand what is going on, you and Parfit might agree about any claims that take the following form…
(a) That agent has a normative reason(neut) to Φ at time t.
but disagree about every claim that takes either of the following forms…
(b) That agent has a normative reason(ext) to Φ at time t.
(c) That agent has a normative reason(int) to Φ at time t.
Here’s an analogous case where there seems to be deep disagreement. Suppose some eccentrics accept a theory I just made up called Derek Command Theory (DCT), where all that it means for an agent to have a reason to Φ is that Derek commanded that we have a reason to Φ. Also, suppose that Derek makes the right kind of commands so that the first order normative judgments are extensively equivalent with a plausible constructivist view. In spite of the fact that DCT and this constructivist view can generate the same verdicts about agent’s reasons, constructivism strikes me as much more plausible than DCT. Moreover, constructivism seems to do a better job of capturing a huge swath of our intuitive judgments about everyday cases because my intuitive judgments about such cases extend to certain metanormative considerations not captured by first order normative judgments.
From what you wrote in section five, it seems like the only thing that matters in capturing our intuitive judgments about everyday cases is the first order verdicts. But I am not sure if that’s right given what I want to say about constructivism versus DCT.
Does this seem right to you? If so, what’s the difference between that case and the difference between your view and Parfit’s view? Is it simply because our “thin” concept of a normative reason rules out DCT, but not normative reason(ext) or normative reason(int)? Although maybe the “thin” concept of a normative reason has to be neutral about DCT too. I don’t know what to think about this exactly.
Thanks for participating in this discussion on PEA Soup.
Reply to Clare LaFrance’s post of 8/19
Hi Clare — Thanks again for your question! Yes, that paper you mention (“Nothing ‘Really’ Matters, but That’s Not What Matters”) is a response to Parfit’s critique of my view in On What Matters. I’d just qualify that by saying that (1) in the paper I respond only to a few of the objections Parfit raises (since he raises many, I had to be selective); and that (2) I tried to write the paper in such a way as to be about what I think is a bigger underlying issue (with the material defending my evolutionary debunking argument being more of an afterthought or illustration of a larger point). The bigger underlying issue I have in mind is a problem that I think is actually somewhat widespread in current metaethical discussions, namely a tendency to think that if one has given a metaphysical and epistemological defense of talk of normative reasons, one has thereby given a metaphysical and epistemological defense of talk of robustly attitude-independent normative reasons.
Dear Sharon,
It’s great to see you up as featured philosopher here at Pea Soup – thanks for participating and for for arguing so forcefully for Humean constructivism in your writing! I’d like to ask you about the conception of our normative judgments ‘withstanding scrutiny’ that you develop, in particular in CAR. There you offer three examples:
(i) judging ourselves to have conclusive reason to pursue ends constitutively requires (not just in a rational but in an analytic, conceptual sense) that we take ourselves to have reason to pursue what we recognize to be the necessary means to them.
(ii) judging X to be a reason to Y constitutively requires our not judging X to not be a reason to Y.
(iii) judging that only facts of kind X are reasons to Y, and recognizing that Z is not a fact of kind X, consitutively requires not judging that Z is a reason to Y.
So here goes with the questions:
1) Suppose that Smith’s values are not ideal in one of the above respects. For example, suppose that Smith(prima facie) takes herself to have conclusive reason to pursue some end Y and recognizes that X is a necessary means to Y, but does not take yourself to have reason to X. Then the upshot of your constitutivist account seems to be that Smith doesn’t *really* take herself to have conclusive reason to Y at all. She must have been doing something else altogether – perhaps she is having some other attitude toward Y: schmaluing. So what reason is there – if any – for Smith to stop schmaluing, and to start taking herself to have conclusive reason to Y and (at the same time) taking herself to have reason to X? Does your account really leave room for the possibility of defective or non-ideal normative jugments? You go on to say that ‘to make a normative judgment is to “give laws to oneself.” As soon as one takes anything whatsoever to be a reason, one thereby “legislates” standards according to which, by one’s own lights as a valuing agent, one is making a mistake, whether one know it or not, if one endorses certain other normative judgments.” (CAR, 230). But the way you couch your constitutivist account as analytic raises the question of how such a mistake is even possible. Perhaps, as your discussion seems to suggest, you think that mistakes of this kind are possible just in cases of ignorance. But this answer seems to me to raise a dilemma for you: Either (i) the constitutive standards of normative judgment *have nothing to say* about cases where the agent is relevantly ignorant (e.g. cases where the agent is ignorant that X is the necessary means to her end Y), so that the agent violates no standards of normative judgment in these cases (and therefore, it would seem, has no obvious reason to revise her judgments e.g. to take herself to have reason to pursue the means), or (ii) the constitutive standards of normative judgment *do* apply even when the agent is relevantly ignorant, so that an agent who, e.g. fails to take herself to have reason to pursue the necessary means to her end just because of her ignorance of the means in fact violates a constitutive standard. But if you take this option you must rewrite the standard of taking oneself to have conclusive reason to pursue an end as follows: judging ourselves to have conclusive reason to pursue an end constitutively requires that we take ourselves to have reason to pursue what *is in fact* the necessary means. But now we will get the result that an ignorant agent Smith does not *really* take herself to have conclusive reason to pursue her end, so she doesn’t seem to be making a mistake at all: she is just schmaluing. Why should she care? Why should we care? So what if (as a matter of psychological fact) she *will* start taking herself to have decisive reason to pursue the end if and only if she comes to recognize the necessary means to it?
2) Even if the constitutive requirements above can and do act as normative standards for revising some of our normative judgments in the light of others, they seem to me to set the bar for ‘withstanding scrutiny’ very low. Is there a basis, on your account, for criticizing e.g. the Nazi morality that would take one to have reason to respect everyone but Jews and gypsies, and locate that reason in everyone’s non-Jewish, non-gypsy human human nature? Could Nazis bootstrap into existence a decisive reason for themselves to send Jews to concentration camps merely by judging that Jewishness is a decisive reason for internment, not judging that Jewishness is not a reason for internment, not judging that only reasons of criminality are reasons for internment, and so on?
Thanks again, I very much look foward to reading your answers!
Reply to Daniel Muñoz’s post of 8/20
Hi Daniel — Thanks for your question! You summarized your question as “Can you explain to a non-naturalist how your modestly attitude-independent normativity is still normativity?” One way of trying to explain it is this. I take it that you and I can agree that if there are reasons in the standard normative sense (employing Scanlon’s term from What We Owe to Each Other), then there is normativity. Normativity hasn’t “gone anywhere,” or been “lost,” so to speak, if it turns out that there are reasons in the standard normative sense. What else would normativity be? But on a constructivist view, there are reasons in the standard normative sense. There are not, the constructivist says, robustly attitude-independent reasons in the standard normative sense, but that doesn’t mean that there aren’t reasons in the standard normative sense. There are tons of them — and in fact most of the ones we always thought there were (the latter point being one I try to argue for in “In Defense of Future Tuesday Indifference”). So how is that not “still normativity”? One might object by saying “But if there are no robustly attitude-independent normative reasons, then there are no normative reasons at all.” But why think this? Why does it have to be the case that in order for something to be a normative reason for an agent to do something, it has to possess that status in a way that is robustly independent of that agent’s attitudes? It is not built into the concept of a normative reason that this is so, in the sense that one is not demonstrating incompetence with the concept of a normative reason if one thinks that there are normative reasons, but that nothing is a normative reason except ultimately in virtue of its relation to us and our attitudes. On a constructivist view, there are plenty of normative reasons. In particular, there is a normative reason whenever the conclusion that there is follows from the global evaluative perspective of the agent whose normative reasons are in question (in combination with the non-normative facts). So there is still normativity. (Notice that I am understanding constructivism itself as a substantive normative claim about what reasons there are.) I hope that helps, though I’m not sure it will — I’m not sure it gets to the root of what is bothering you. Thanks again for your question!
Sorry Sharon, of course I mean withstanding, not surviving scrutiny. I know you know that, but it’s been bothering me that I said surviving. I’m looking over what you say about it in CAR again, and unless there is somewhere else I should look for that notion’s explication, I’ll ask you about it based on what you say there.
Hi Sharon (if I may),
I was wondering what implications you take the Darwinian dilemma to have for normative moral theories rather than metaethics.
One popular idea seems to be that evolutionary considerations can help undermine (i) “Kantian” moral theories (this is the point Singer and Joshua Greene have made over the past couple of years) but (ii) do not affect an ethics of impartial benevolence in the same way (this seems to be the point Singer and de Lazari-Radek argued for in a recent Ethics article).
My question is whether you agree that this gets the implications of your dilemma right. In particular, as (I think) Guy Kahane has noted somewhere, if our belief in the objective value of benevolence is unlikely to be justified, then removing arbitrary distinctions between strangers and the people we are familiar with doesn’t make an attitude of benevolence any more justified than removing “arbitrary” distinctions between my and my neighbor’s backyard makes the intention to count blades of grass justified.
Moreover, suppose one agrees that constructivist approaches to normative ethics go beyond the judgment that it is ok to push the fat man off the bridge. Then it might turn out that that they are in some sense better justified than utilitarian approaches, because of the fact that “constitutive rules of agency”-focused moral theories are further removed from the gut evaluative intuitions which can plausibly be explained by an evolutionary debunking story.
Thanks!
Hi Professor Street,
Thank you for taking questions. Mine may amount to no more than a request for references.
In a footnote to “Coming to Terms with Contingency” (n. 6 in the version from your website) you mention that a full discussion of the attitude of valuing would have to engage with the literature on agency, especially the debate between Frankfurt and Watson. That debate interests me a great deal, as does your discussion of the attitude of valuing. I wonder if you have followed this comment of yours up anywhere in writing. If not, have you developed any further thoughts you would be willing to share here?
Thanks in advance!
Reply to Sven Nyholm’s post of 8/20
Hi Sven — Thanks for writing; that’s a great question. You are absolutely right to point out that (expanding on your examples just slightly) the experience of the fact that the restaurant serves chocolate mud cake (yum!) as “counting in favor of going there tonight” versus the experience of the fact that I promised him I would do so (aargh!) as “counting in favor of helping him move today even though it’s 120 degrees and I’m swamped with work” have an extremely different character to them. So you are absolutely right to ask: In virtue of what do such experiences even deserve to be lumped together? Is there supposed to be a distinctive element common to both of them, and indeed common to all experiences of various features of the world as “counting in favor of” this or that response on our part?
I agree entirely with what you seem to be suggesting, which is that it is very implausible to think that there is a common element present in each of these experiences in the way that there is (say) a common element present in a hundred photographs of the Washington Monument — in other words, such that one can look from one photograph (or experience) to the next and point to the same element showing up over and over again in every photo (or experience). Instead, it seems to me that the color analogy is again helpful here, with the idea being that it’s not the case that one can look across “this experience of yellow” and “that experience of blue” and “this experience of vermilion” and “this other experience of vermilion” and point to a common element that is present as part of the intrinsic character of each of these experiences, such that each counts as an experience of color (or even a second experience of vermilion) in virtue of having that element embedded in it (again, the way the Washington Monument might show up again and again in thousands of photos).
Instead, if we’re asked to say what it is in virtue of which they all count as experiences of color — and if all we’re allowed to talk about is the intrinsic character of those experiences — then it seems we have no choice but to invoke the same concept in a way that is not informative — saying, well, they’re all experiences of color. Similarly, it seems to me that we are left with equally little informative to say if we are asked to say what it is that the experience in the restaurant case and the experience in the promise case have in common such that both count as experiences of something as “called for” or “to be done” or of one thing’s “counting in favor of” another thing. They are utterly different experiences in the same way the experience of red and the experience of blue are utterly different, we might say; the only thing the experience of red and the experience of blue have in common (again, if we’re restricted to talking about their intrinsic character) is that they are both experiences of color. Similarly, the only thing our “normative experience” in the restaurant case and our “normative experience” in the promise case have in common is that they are both experiences of one thing as “counting in favor of another”. It’s the good old irreducibility of normative concepts confronting us again. Anyway, those are just some preliminary thoughts — not at all meant to be definitive. Thank you so much again for your question!
Many thanks for the reply, Sharon! That’s a very interesting extension of the color-analogy. Again, thanks!
Reply to Eric Campbell’s post of 8/21
Hi Eric — Thank you so much for your questions! There are many interesting issues here. Let me respond with three points.
First, you say at one point that I “seem to conceive of [the question whether a given normative judgment withstands scrutiny] as being answerable by the person’s own assessment.” You then write, “But it seems to me that people can easily be wrong about which of their values are ‘deeper’ than others, and even wrong about whether they value something at all.” I totally agree with you that people can be, and often are, wrong about whether (or how much) they actually value something. People can be very deceived or ignorant of their own deepest concerns and commitments. So I agree with you that it would not make sense to take the view that agent A’s normative reasons are a function (in thus and such a way) of what she thinks are her most strongly held values; the constructivist position is rather that agent A’s normative reasons are a function (in thus and such a way) of what, as a matter of fact, are her most strongly held values. That is one of the points of the passage where I say “when we ask whether the judgment that X is a reason to Y (for A) withstands scrutiny from the standpoint of A’s other normative judgments, we are not asking what A or anyone else thinks withstands scrutiny from that standpoint” (p. 231). We are asking what does withstand scrutiny from the standpoint of A’s values, and the agent herself is not the final arbiter of this; it depends on what her values are, not what she takes them to be. In other words, to locate the “standards” that determine what A’s normative reasons are, we look not to what A might think is her evaluative standpoint on the world (as you point out, she could be wrong or self-deceived about what her deepest values consist in); rather, we look to A’s actual evaluative standpoint.
Second, you say that “in CAR you seem to equate valuing with conscious valuing.” With all the emphasis in that paper on “judgments,” I can see why it might come across that way. But I did have unconscious judgments in mind as included among the states of mind that “set the standards” for what counts as a reason for a given agent. There’s a little hint of that on p. 233, when I talk about normative judgments that we implicitly endorse, and also on p. 245, when I mention unreflective versions of normative judgments. In my earlier paper “A Darwinian Dilemma…” I’m more explicit about the fact that I have in mind consciously or unconsciously held normative judgments; see p. 110 of that paper. Still, you are right that the point receives virtually no emphasis in the “Constructivism about Reasons” paper.
Third, you raise the interesting question whether “discovering what our values are should be an important, even central part of normative inquiry.” This issue deserves a much larger discussion than is possible here, but one way of summing up what I think about this is that if the inquiry is into “what our values are” (with the idea being that we are trying to “discover” what our values are by doing an empirical or descriptive psychological inquiry) then that inquiry is not “normative inquiry” at all, at least not at that moment — pursued in that way, under that description. Of course — and this is why things get complicated — according to metaethical constructivism, these facts about what we do value are highly relevant to what is normatively the case, in the sense that they are part of what determines, in the end, what our normative reasons are. There is much more to say about this, but one of the most useful discussions I know of in this context appears in Scanlon’s piece “Rawls on Justification” in The Cambridge Companion to Rawls — see especially Scanlon’s suggestion on p. 147 that what he calls the “deliberative” version of the method of reflective equilibrium is primary. (Rawls’s “Outline of a Decision Procedure for Ethics” is also very relevant here.) One way of putting it is that there are two highly distinctive ways of, as you put it, “discovering what our values are.” One is to observe ourselves from a descriptive/”sideways-on”/empirical psychological standpoint; the other is to engage directly in normative inquiry — asking not what do I value, but what is valuable. The constructivist suggestion is that the latter is ultimately a function of the former, but that doesn’t mean that there isn’t still a sense in which the latter question is primary. (With this last comment I mean to be gesturing at the idea that constructivism itself may be viewed as a normative discovery.) Many thanks again for your questions!
Reply to Travis Timmerman’s post of 8/21
Thank you so much for your questions about my draft on Parfit! Regarding your first question, you write “why not understand Parfit as arguing for Non-Analytical Externalism?” I think we may be talking past each other here due to some terminological issues: I agree with you that we can understand Parfit as arguing for Non-Analytical Externalism. That’s what I mean to be saying when I say on that same page (p. 7) that “Make no mistake: It’s easy to see how one might argue that there are external reasons; on one reading, this is largely what Parfit is concerned with in Part I of OWM.” In other words, I agree with you that the Early Death case, etc., may be regarded as arguments for Non-Analytical Externalism — which, in my terminology, is the position that there are reasons in the standard normative sense (what I also call reasonsNEUT) such that agents have those reasons even though the conclusion that they do doesn’t follow from their “S”. Turning to the terminological issue: I notice that earlier in your question you summarize my view as saying that Parfit fails adequately to “distinguish between (1) external reasons in the ‘standard normative sense’ and (2) external reasons in the ‘standard normative sense’ that also exist independently of whether there is a sound deliberative route…,” etc. But that is not how I would put my view. The way I’m using the terminology of “external reasons” — which is not the way Parfit is using it, since I find his usage highly problematic — your expression (2) contains a redundant phrase. In my terminology, we don’t need to say “external reasons in the standard normative sense” and then add “that also exist independently…,” etc. External reasons just are reasons in the standard normative sense that exist independently of…etc.
Regarding your second question, it’s hard to know what to say about “psychopaths” without hearing more about how you are envisioning them. I doubt that real-life psychopaths are “ideally coherent” any more than real-life anorexics are ideally coherent or the real-life Caligula was ideally coherent. (I say more about these kinds of issues in “In Defense of Future Tuesday Indifference.”) That is, presumably real-life psychopaths are mistaken about many non-normative facts, and, more importantly, it’s not at all clear (to me, anyway) that their evaluative attitudes are internally consistent with one another. (Oftentimes they avow moral norms at one level, while showing no motivation to comply with them at another, as I understand it. This raises questions about how we should even understand what their evaluative attitudes are, but it doesn’t on its face look like ideal coherence.)
Regarding the last part of your post, I think the terminological issues I mentioned above may be implicated in those claims (a), (b), and (c) that you list. Perhaps one thing to keep in mind is that I think claims like what you call “Derek Command Theory” and constructivism itself may be understood as substantive normative claims expressed in terms of the concept normative reasonNEUT. I’m sorry if that’s not very helpful; there are lots of things going on here and this is something it would probably be easier to sort out in conversation. I hope there is a chance to do that sometime; in the meantime, thanks again for your questions and your interest in that paper! Your questions will be very helpful to me when I go to edit it some more.
Reply to Ben Mitchell-Yellin’s post of 8/22
Hi Ben — Thanks for your question. No, I’m afraid I haven’t followed up on that comment anywhere in writing, though I think it’s a very important issue. But one extremely helpful essay that is addressed directly to such questions is Michael Bratman’s “Constructivism, Agency, and the Problem of Alignment,” which appears in the collection Constructivism in Practical Philosophy, edited by James Lenman and Yonatan Shemmer (Oxford: Oxford University Press, 2012). I highly recommend this article if you’re interested in these issues as they pertain to constructivism. Many thanks for your question!
Hi Professor Street,
Thank you very much for the reference!
Thanks a lot for your responses Sharon! I had understood that the constructivist position holds that a person’s reasons are a function of what in fact withstands scrutiny and not what the person thinks does, but I had not gotten the impression that there was an appreciation that a person could be very mistaken about what their values are, as opposed to what does or doesn’t withstand scrutiny from the perspective of (some subset of) those values. Thanks a lot for clearing that up for me, and for the references too.
Dear Professor Street,
Thank you for the thoughtful reply! From what you’ve written here and in your papers, it seems that constructivism does much less violence to commonsense morality than I’d initially thought.
Reply to Hanno Sauer’s post of 8/22
Hi Hanno — Thank you so much for this question, which is a really important one. My view is that Darwinian considerations do not have implications for debates in normative ethics. For example, I strongly disagree with the idea that evolutionary considerations could somehow tell in favor of consequentialist theories while telling against Kantian views (or vice versa).
In my view, the genealogy of a normative judgment is, in and of itself, completely irrelevant to the question whether that normative judgment is true. What evolutionary considerations undermine is not this or that particular normative judgment — e.g., the judgment that we have normative reason to care for our offspring — but rather realism about normativity. The realist and antirealist can agree, for example, that we have reason to care for our offspring; the question is whether that is true, ultimately speaking, in virtue of robustly mind-independent normative truths or in virtue of what follows from the standpoint of our own evaluative attitudes. Evolutionary considerations can help to shed light on this latter debate, in my view, but they shed no light on whether the first-order judgment is true. What decides whether a normative judgment is true or not, on my view, is how that judgment coheres with the totality of our evaluative point of view on the world, in combination with the non-normative facts; this could be so, or not, regardless of how one originally came to have that judgment.
Since I think that the causal genealogy of a normative view is, in and of itself, totally irrelevant to its truth, I think, a fortiori, that it is a mistake to think that the fact that a given normative judgment (or better, “basic evaluative tendency”) does have an evolutionary explanation provides any reason whatsoever to doubt its truth. I also think it is a mistake to think that the fact that a given normative judgment does not have an evolutionary explanation provides any reason whatsoever to think it more likely to be true. So I disagree with the criteria set forth on p. 23 of Lazari-Radek and Singer’s recent article, concerning “three elements in the process of establishing the highest possible degree of reliability.” I strongly disagree with the suggestion that “if an intuition that met the first two requirements but not the third were to clash with an intuition that met all three, we would have a ground for preferring the intuition for which there was no evolutionary explanation” (p. 24). I therefore also disagree that with the help of these criteria, we can resolve the standoff between rational egoism versus the principle of universal benevolence. To resolve the first-order questions, we’re stuck doing the same thing we’ve always done, namely assessing the competing theories on their merits. Of course, in so doing, we inevitably have to rely on our larger, background, evaluative standpoint on the world, but the mere fact that a given evaluative tendency does or doesn’t have a certain causal explanation is neither here nor there in and of itself.
There is obviously a lot more that could be said about this, but I’ll leave it there for now. I say more about some of these issues in “Does Anything Really Matter or Did We Just Evolve to Think So?”. There I try to explain what it takes for an explanation to be undermining of a given belief that P, and suggest that evolutionary considerations, while debunking realism about normativity, do not debunk our first-order normative views. Thank you again for your question!
Hello Professor Street,
Thank you very much for your helpful response. I apologize for mixing up the terminology a bit. I knew that you accepted that Parfit’s Early Death case could be considered an argument for Non-Analytical Externalism. I also saw Parfit as arguing exclusively for Non-Analytical Externalism and, as such, was not sure why Parfit would need to provide an argument that “it’s (perhaps non-obviously) analytic that reasons in the standard normative sense are external reasons.” I suspect that I am just missing something again and will give your paper another close read.
I don’t know much about psychopaths, so the following assumption could very well be false. Although no actual psychopaths are likely ideally coherent, I was assuming that if they were to subject their beliefs to scrutiny and weed out any contradictions, some would discover that they had a radically different set of reasons than the average person.
I hope we get a chance to talk in person sometime too. I agree that we could probably sort these issues out in conversation much more quickly. Thanks again for your response and for all of your great work.
Hi Sharon, I think you may have missed my questions (posted August 21, 1:43pm).
thanks, Simon.
Reply to Simon Rippon’s post of 8/21
Hi Simon — Thank you very much for your questions! Regarding the first, you describe a case in which “Smith (prima facie) takes herself to have conclusive reason to pursue some end Y and recognizes that X is a necessary means to Y, but does not take [herself] to have reason to X.” The case is actually underdescribed, for my purposes, because on my view, it matters whether Smith is (allegedly) making all three judgments simultaneously, in full conscious awareness of all three at once. If that is how we are thinking about the case, then this is just like the Rome example I give on p. 227, and so right, my view is that this isn’t recognizable as a case of “valuing” or “normative judgment” at all. So assuming that’s how we’re thinking of the case, yes, we could call this thing she is doing — allegedly holding all three judgments at once, in full conscious awareness — “schmaluing.” It’s a little hard to get a grip on what “schmaluing” is — if my account is right, then the language of “taking herself to have conclusive reason” seems like it can’t even count as a proper description of what she is doing — but in any case, you are right that whatever this state of mind is, on my view she has no reason to stop “schmaluing” (assuming all her attitudes are like this, anyway). But this is a consequence I embrace. In my view, talk of normative reasons doesn’t even get purchase until the being in question has a certain kind of attitude up and running — and in particular, the attitude of valuing (though it need not be fully reflective). Schmaluing doesn’t cut it. That doesn’t mean Smith is making a mistake in schmaluing, any more than a rock is making a mistake in not valuing. You ask “Does your account really leave room for the possibility of defective or non-ideal normative judgments?” But there seems to be an ambiguity here. What do we mean by “defective”? My account leaves room for false normative judgments. But when confronted with a case like Smith’s, the account isn’t going to say that this is a false normative judgment, or for that matter a “defective” normative judgment; it’s going to say it’s not a case of normative judgment. And to me, this seems an intuitive result. (The analogy is that a parent who doesn’t have children isn’t a defective parent; rather, she isn’t a parent.)
Regarding the possible dilemma you describe, the constitutive standards of normative judgment describe what standards there are from the point of view of the creature who is already engaged in normative judgment. They describe what it takes to count as being in the state of mind of normative judgment (not a normative claim, but a conceptual one). In so doing, though, they articulate what the agent herself takes to be the case regarding the possibility of error. So, regarding the second horn of the alleged dilemma: You say that “if you take this option you must rewrite the standard of taking oneself to have conclusive reason to pursue an end as follows: judging ourselves to have conclusive reason to pursue an end constitutively requires that we take ourselves to have reason to pursue what is in fact the necessary means.” But this isn’t right. (Or rather, there is a sense in which it is, and a sense in which it isn’t, and the objection trades on the sense in which it isn’t.) Suppose I take myself to have reason to get to Rome immediately, but I am ignorant of the fact that to do so it is necessary to get on a plane. Counterfactually, though, let us suppose, it is true of me that were I to become aware of this fact, I would immediately take myself to have reason to get on a plane, on the grounds that it is a necessary means to my end (or else give up the end, if I think getting on a plane is too high a price to pay — obviously there are complications here). Okay, good, so I count as genuinely making the normative judgment about my reason to get to Rome. So we don’t get the result you talk about — i.e., that “the ignorant agent…doesn’t really take herself to have conclusive reason to pursue her end.” (Your imagined agent “Smith” is not the relevant case here. In her case, as you have described it, it is not counterfactually the case that she would, upon learning that getting on a plane is the necessary means, immediately judge herself to have reason to get on a plane — or else abandon the end, etc.)
Regarding your second question, I don’t think it’s correct that the constructivist proposal “[sets] the bar for ‘withstanding scrutiny’ very low.” I don’t know whether you’ll find it persuasive, but I discuss such matters in “In Defense of Future Tuesday Indifference.” As I argue there, it takes a lot to be ideally coherent, and it is absolutely crucial that we not run together cases of actual human beings with the hypothetical characters imagined by metaethicists. Thanks so much again for your thoughtful questions!
Hi Sharon, thank you for your kind reply.
Your response to my first question is particularly helpful, as it corrects a misreading I fell into of what you meant in saying that your constitutivist account was “analytic or conceptual” (I now think your view is further from Korsgaard’s and closer to my own than I had previously thought). It seems to me that you give your view most clearly when you present it roughly this way: *What it is* to judge that one has conclusive reason to pursue an end is, in part, to judge that one has reason to pursue the necessary means to it(on the de dicto reading, rather than the de re reading, because one might of course be ignorant about the necessary means). This seems to me preferable to trying to explain your view by talking about what one *would* judge in a counterfactual state of “full awareness”. For example, in your reply here, as in CAR, you attempt to explain the difference between taking oneself to have conclusive reason to pursue an end and what we can call merely “schmaluing” in terms of its entailing what looks like a purely psychological *counterfactual* truth – e.g. the imagined fact that *if* I were fully aware that getting on a plane were necessary for getting to Rome, *then* I’d make the normative judgment that I have reason to get on a plane. But this way of explaining what the normative judgment amounts to seems to invite a “so what?” reaction. It may be conceptually true that I only hold the normative judgment ‘I have conclusive reason to go to Rome’ if I would, in a state of full awareness, judge ‘I have reason to get on a plane’. But there is nothing here to suggest that my making that normative judgment gives me any reason at all to try to get into the state of full awareness, e.g. by finding out what the means to getting to Rome actually is. Is there a reason for preferring this full awareness type of explanation that I’m missing?
On the second question, I had read your defense of future Tuesday indifference and found it by and large highly persuasive, but it didn’t entirely assuage my worries about actual or potential moral degenerates (some of which seem like more than mere figments of the fevered metaethicist’s imagination!). I’m not really worried about a sort of alien Ideally Coherent Nazi who just has an all things considered reason to persecute Jews, but rather about a common Nazi who – with ruthless consistency – just doesn’t recognize a set of moral reasons pertaining to Jews that the rest of us tend to accept. What reason do I have (perhaps relevantly, in the less than ideally coherent position I find myself in) to think that this Nazi is indeed subject to the moral reasons he denies – especially given that the only examples of “incoherence” you give (at least in CAR) are all apparent examples of either means-end or logical inconsistency? You seem to suggest in the Tuesday paper that a Caligula could be subject to moral reasons that are non-normative for him, but this looks like an inverted commas sense of “moral” and this is surely not what I want to think the Nazi is subject to. Could you say a bit more about your view on what makes a reason moral in character, and on how individuals become subject to moral reasons?
My spouse and I absolutely love your blog and find most of your post’s to be what precisely I’m looking for. can you offer guest writers to write content to suit your needs? I wouldn’t mind publishing a post or elaborating on a lot of the subjects you write with regards to here. Again, awesome site!
Yes! Maybe you could write about whether, regardless of what one prefers, it is irrational to starve oneself to death for the sake of a trim figure.
Ha!