Internal Reasons That Cannot Motivate

Bernard Williams begins an influential essay [1] by defining internalism about reasons as follows:

Internalism   There is a reason for person A to φ only if A has some motive which would be furthered by his or her φ-ing.

Plenty of philosophers have found something intuitive about this idea, but there has also been no shortage of disagreement over the exact sense in which A must “have some motive” which φ-ing must further. In the introduction to a recent anthology of literature on internal reasons, Kieran Setiya [2] seems to think that the most attractive versions of internalism are those which satisfy the explanatory constraint. Bernard Williams gives it best:

EX   If something can be a reason for action, then it can be someone’s reason for acting on a particular occasion, and then it would figure in an explanation of that action. (p. 106)

There are at least a few reasons for adhering to EX. You might think that there is a unified account of explanatory and normative reasons, and that EX is a link in that unification. You might think that what it is for A to have a motivation which would be furthered by A’s φ-ing is just for there to be some p such that A is disposed to make p A‘s motivating reason for φ-ing. If you’re inclined to believe either of these, you’ll probably think with Setiya that the broadest plausible version of internalism is IR:

IR   The fact that p is a reason for A to φ only if A is capable of being moved to φ by the belief that p. (p. 4)

However, I don’t see how EX could possibly be true, as I don’t think it can overcome the kinds of cases which motivate so-called “advice models” of reasons. I argue that an agent can have a reason for action which, qua reason, could not possibly motivate them. This undercuts the motivation for thinking that IR is the correct way of understanding internalism.

Understanding Ecumenical Expressivism

Call a metaethical view a hybrid if it allows that moral sentences sometimes express cognitive states of mind and sometimes non-cognitive states of mind. Then one kind of hybrid view is an ecumenical one, which requires that moral sentences always express both cognitive and non-cognitive states of mind. I’ve been interested in hybrid theories for a while now, so it’s been a pleausure to find Michael Ridge’s ecumenical expressivism (EE), which is  subtle and deftly defended. However, I’ve got some reservations about how it’s cashed out.

The basic idea of EE is that the non-cognitive attitudes expressed by a moral sentence have priority in that the cognitive beliefs which are also expressed do not determine the truth-conditions of the sentence. To color things in a bit, it’s much as if we thought that speakers had hazy necessary and sufficient conditions in mind for “good”, but we also stipulate that the meaning of “good” doesn’t commit a speaker to using that term whenever those conditions are satisfied (hence preserving the Open Question intuition).

Now here’s a first development of EE, which Ridge calls Plain Vanilla EE (concentrating just on the notion of a reason for action):

(PV)   “There is a moral reason to X” expresses (a) an attitude of approval of a certain kind toward actions insofar as they have a certain property and (b) a belief that X has that property. (Ridge, 2006)

Note that in PV Ridge intends there to be a unique property I approve of in all my assertions about reasons, though this property may be disjunctive. Ridge (2006) argues that this feature of PV excludes particularists and pluralists, so he suggests the following dispositional account instead:

(D)   “There is a moral reason to X” expresses (a) an attitude of approval of a certain kind towards actions insofar as they would garner approval from a certain sort of subject and (b) a belief that X would garner approval from that sort of subject (Ridge, 2006),

where it’s understood that the subject in mind is of the ideal observer variety.

But the curious thing about (D) is that Ridge (2006, n. 47) denies a very natural interpretation: that (D) is just (PV) coupled with a first-order normative view, i.e. that in talk of moral reasons I’m expressing a certain pro-attitude toward actions with the property that the ideal observer would approve of them (as well as a belief that this particular action has that property). One problem Ridge raises for this view is that it doesn’t really avoid the problem of monism above. But there are others as well. For instance, while dispositional theories might say something important about certain normative concepts, I doubt they can plausibly be folded into the contents of the associated states of mind. 

Consider the attitude of belief towards a proposition P. Suppose I believe P; if I were to learn that my epistemically-ideal self does not believe P, it seems I would ought to drop belief in P. But that normative relationship does not imply that my belief was all along about what my epistemically-ideal self believed. The content of my belief was just P. And if it wasn’t just P, then the content of my belief must be even less accessible to me than I thought, and presumably can’t play the kind of role in guiding conduct that many expressivists think states of mind do play.

Now suppose I have a pro-attitude toward all actions which are φ. Once again, were I to learn that my normatively-ideal self wouldn’t approve of all φ-ing, it seems I would ought to modify my attitude. But I don’t think this means that my pro-attitude was all along towards what my normatively-ideal self approved of. I just approved of actions that were φ – or so it seems. Once again, if I’m in the dark about my own states of mind, how can states of mind guide my actions?

I’m skipping a bunch of steps here, but I want to suggest another reason for agreeing with Ridge on how not to interpret (D): that it might make the notion of a moral state of mind so anti-transparent that it (a) strains credulity and (b) may not do a lot of the simple work in explaining behavior expressivists want it to. But, that done, I’m not sure how to interpret it in a way that avoids these problems. It can’t be that moral sentences express the ideal advisor’s approving state of mind. I am not my ideal advisor, and I cannot express another’s state of mind unless I am acting, or perhaps quoting that individual. But in those contexts I generally don’t mean to endorse those attitudes. Everyone finds Hitler reprehensible, but no one finds actors who play Hitler reprehensible. (At least, not for that reason.)

At this point I wonder whether the solution is to drop the uniqueness requirement in (D). Why not think that some moral sentences express approval of property φ and others of property ψ, for ψ ≠ φ?

Bibliography

Ridge, M. (2006). “Ecumenical Expressivism: Finessing Frege”. Ethics 116, pp. 302-336.

Ridge, M. (2007). “Ecumenical Expressivism: The Best of Both Worlds?”. Oxford Studies in Metaethics, Vol. 2, pp. 51-76.

Can an OUGHT follow from no ISs?

Suppose murder just is wrongful killing. Then it seems that Sally ought not murder Bob follows from no premises, the empty set of premises. Trivially, the empty set of premises is a set containing only descriptive premises, in Hume’s sense. But then, Sally ought not murder Bob, a substantive normative claim, follows from a set of purely descriptive sentences. So, you can derive an ought from iss. Take that, is-ought gap.

I have some half-baked potential responses in mind, but let’s see what you think.

What does consistency have to do with reasons?

I asked Allan Gibbard this post’s titular question earlier this year, and he patiently noted that many people value consistency because if one has inconsistent beliefs, one or more of one’s beliefs must be false. I responded that since he and I believe that all normative beliefs are, in some sense, false, this sort of consideration wouldn’t give us — or other anti-realists about value — any reason to want our “ethical beliefs,” whatever form they might take, to be consistent. I can’t remember where the conversation went from there, but I don’t think that he was particularly moved by my response.

But I was moved by it, and I still am. Sure, if I believe P and not-P, I must have a false belief. But if there aren’t any normative truths, then if I believe P and not-P with respect to some normative matter, I should probably just stop having beliefs of that sort, but if I’m unable to stop (as some anti-realists seem to hold), what harm could follow if I have inconsistent beliefs about such matters?

I recognize that only certain varieties of “anti-realism” about ethics hold that there are no normative truths. Perhaps Allan’s “quasi-realism” isn’t one of them — I can never keep track of what he does or doesn’t think can be true — which might explain his not being impressed with this line of thought. But if one doesn’t think that there are any truths of the form “X is good,” “X is right,” “X is reasonable,” etc., then it seems one doesn’t have any truth-related reasons to value consistency among one’s normative commitments, beliefs, statements, etc.

Leaving that aside, here’s another little pseudo-problem: Suppose, like Sharon Street, you hold 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, … if one endorses certain other normative judgments.” (This is from Street’s “Constructivism About Reasons,” pp. 229-30.)

I take Street’s account of reasons and normativity to depend largely on an appeal to internal consistency of some sort. According to Street, one’s judgments about reasons can be judged as “correct” or “incorrect” from the standpoint of all one’s other judgments about reasons. Of course, since Street also holds that there are “ultimately” no normative truths or truths about reasons, I’m inclined to ask why we should continue to make “judgments about reasons” or take ourselves to have reasons if we agree with Street that there are ultimately no truths about these matters. She has an answer to this, of course: We can’t help but occupy the “practical standpoint,” and occupying this standpoint necessarily involves making normative judgments and taking oneself to have reasons. I’m not convinced that she’s right about that, but I’ll grant it for the sake of argument.

Prescriptive Metaethics

By and large, metaethicists have focused on descriptive questions about the nature of our moral discourse. For instance, is it in the business of stating facts, or of expressing affective states? If the former, are there such facts? If the latter, how is this reconciled with the role that moral language plays in reasoning?…

There is one clear exception. Some who are interested in error theory have shifted their attention from the descriptive question to a prescriptive/ practical one–namely, “Are we to retain moral language? And if so, how are we to treat it?” (I give the practical variant since it might be thought problematic for an error theorist to ask a question framed in terms of “should.”) The reason for this shift is that the prescriptive/practical question seems quite pressing in the case of error theory. Error theory seems to force the question. The two most prominent answers discussed by error theorists, to my (limited) knowledge, are eliminativism and moral fictionalism (of the prescriptive variety).

A Troublesome Quartet

Based on the following line of thought, it seems to me that if one is a buck-passing existence-internalist who accepts a certain thesis about motivational diversity, then she or he should give up on one of those commitments or give up thinking that anything has intrinsic value. (There are other ways to frame the matter, of course.) But I’m not yet sure that the following line of thought is unproblematic. Feedback is welcome!

P1: (Intrinsic Value) Something has intrinsic value. [In other words: Something is worthy of being valued (by anyone) for its own sake.]

P2: (Fitting Attitudes Reduction) Something’s being worthy of being valued (by anyone) for its own sake just is for it to be such that there is reason for anyone to value it for its own sake.

C1: So, there is something such that there is reason for anyone to value it for its own sake. (from P1 and P2)

P3: (Existence-Internalist Thesis) For any thing X and any agent A, there is reason for A to value X for its own sake only if a (specified) X-related motivational fact about A obtains.

P4: (Motivational Diversity) For any thing X, there is some possible agent A such that the (specified) X-related motivational fact about A does not obtain.

C2: So, for any thing X, there is some possible agent A for whom there is not reason for A to value X for its own sake. (from P3 and P4)

C3: So, it is not the case that there is something such that there is reason for anyone to value it for its own sake. (from C2)

C4: So, C1 & C3 (Contradiction)

Color-judgment expressivism?

Hey guys,  

I don’t know too much about the philosophy of color, except for what you learn from examples people use in discussions about other topics. I’ve noticed that some people, to take one example, think that we should be error theorists about color: we falsely believe that surfaces have these mind-independent color features that they don’t really have, and therefore our color-judgments attributing these properties to surfaces are all systematically erroneous.  This, if I am not mistaken, is Velleman’s view. He brings this up in a discussion about whether people’s belief in free will is some kind of systematic illusion of a similar sort.  Other people, like Gibbard, think that an error-theory about color-judgments is too extreme; we should not attribute systematic errors to the folk unless no other, friendlier theory is available.  One type of theory of a friendlier sort takes color-judgments to be judgments about dispositional properties: we judge things to be such that they tend to look a certain way to people.   But, what are some other possible views?

Here’s what I am specifically interested in knowing: do you guys know if any philosophers have given expressivist views about color judgments?  The idea would be something like this.  In saying, for example, “the sky is blue” what people are doing is expressing their mental state of seeing the sky as blue.  Now, what is their seeing it as blue supposed to mean here?  Well, perhaps I would have done better to say something like “they express their mental state of its looking a certain particular way to them”.  So, the state of judging something to have a certain color is closely tied, on this view, to the state of something looking/appearing a certain way to you.  And, if you say that the thing has the color in question, then you are expressing this state of mind, rather than reporting it.  (This, by the way, seems right: in calling something blue it seems better to say that I am expressing my state of mind of seeing it as blue than to say that I report its looking to blue to me.)

Some experimental philosophy on Happiness

Hey you all,

here’s a link to a blog post of mine on the Experimental Philosophy blog that describes experiments that I think suggest that the folk concept of happiness is a normative one:

Moral Judgments and Happiness

According to the hypothesis I am testing in my experiments, if the folk thinks that somebody is living a bad life—perhaps by being a morally bad person—then they are unlikely to, or will not, attribute happiness to this person even if they believe that she is in the kinds of mental states which we usually associate with happiness.

Most psychological research on happiness uses definitions of happiness that are wholly non-evaluative. This means that, when some psychologist judges that somebody is happy, the folk might not. That, I think, is an interesting result. Why? Because it means, I think, that when we give a philosophical account of happiness, then this will be a normative project at least in the following respect: we will have to give reasons for favoring either the normative concept of happiness used by the folk or the non-normative one used by psychologists and many philosophers. (That is, if, as I think, the kinds of experiments that I’ve been running show that the folk are using a partly normative concept of happiness. Again, see the link for descriptions of these experiments.)

sentimentalism and fit

Sentimentalism is an account about the meaning of some evaluative terms in natural language. In general, the sentimentalist about some evaluative term E proposes to analyze E in terms of the notion of fittingness of some emotion. So, for example, ‘x is shameful’ gets analyzed as ‘it is fitting to feel shame at x‘. Fittingness in turn gets analyzed in terms of the truth of the representational content of the relevant emotion. That is to say: shame at x is fitting just in case x-directed shame represents x truthfully. (And, in general, an emotion with representational content p is fitting just in case p.)

The sentimentalist analysis of ‘shameful’ avoids making a mistake that other accounts of the meaning of ‘shameful’ have made. Say that we propose to analyze ‘x is shameful’ as ‘x‘s bearer ought to / should be ashamed of x‘. This is clearly wrong. It is possible that John should be ashamed of smoking — say Osama will blow up India unless John’s ashamed of smoking — even though his smoking is isn’t itself shameful.

This is all seems pretty much correct. Still, the following sentences sound pretty odd to me.

Identity and agency

1. Sam, is there any way to get a log-in thingy on this page rather than having to go to the wordpress home page?

2. Scarily, I woke up this morning and the first thing I thought of was something that might help me solve a problem in my dissertation. I’m wondering whether this idea can fly.

How far can I go with a distinction between “identity” and “agency”? I mean ‘identity’ in the sense of who you live your life as (not yet sure if this is the best formulation for my purposes)—intended to be something related to but not completely conceptually identified with the usual metaphysical sense of personal identity. And I mean ‘agency’ as the thing the action theorists are after. These two things have in common the possibility of describing “where the agent really is”, but I think this phrase is ambiguous and I’m exploring whether it can be disambiguated.