I am reviewing (our own) Mark Schroeder’s Slaves of the Passions for Ethics. As the advance praise from Michael Smith and Stephen Darwall indicates, anyone interested in reasons and rationality will profit from reading this book. In the book, Schroeder defends a novel Humean theory of reasons and maintains that his book is, “an existence proof of a viable reductive view of the normative.” (82) He also has interesting things to say about moral motivation, moral epistemology, and virtue. The book is full of thought-provoking arguments that break new theoretical ground.
I will be posting a few main worries I have about his view and would be very grateful if you would help me be charitable and assess the seriousness of my worries.
This post will focus on MS’s response to what he calls “the too many reasons problem.” Crudely put, the worry is that Humeans are committed to claiming that we have many more reasons than we actually do.
A stereotypical Humean holds a bold view about reasons: if an agent has
a desire and some action will promote the satisfaction of that desire,
then there is reason for the agent to perform the action. But, the
anti-Humean claims, there are actions that will promote the
satisfaction of our desires but which there is no reason for us to
perform. Schroeder mentions a case in which I have a desire to have
more iron in my diet – eating my car might promote the satisfaction of
this desire, but it sounds counterintuitive to say that there is reason
for me to eat my car.
To avoid this problem, many Humeans tweak the account of reasons;
they might start, for example, by claiming that only coherent desires
that would survive some reflection test generate reasons; they weaken
the bold account of reasons to avoid the counter-intuitive results.
But, with admirable resoluteness, Schroeder argues that it is (at the
least) unnecessary to retreat from bold Humeanism in any such manner.
Instead of weakening the account of reasons to avoid the purportedly
counter-intuitive consequences, Schroeder (92-97) casts doubt on the
intuitions behind the too-many-reasons worry; he argues that the
relevant “negative existential reasons intuitions” lose force under
reflective scrutiny. In the eating-the-car case, for example, he hopes
to convince us that there is some reason for me to eat my car; he hopes to do so by (i) weakening our confidence
in the intuition that I have no reason and (ii) getting us to conclude
on that basis that negative reasons intuitions are unreliable and
should not be taken as a guide in theory construction.
More specifically, Schroeder predicts that any relevant initial
intuition that A has no reason to P will weaken (twice) when you attend
to the following facts: (1) the reason for A to P and (2) the fact that
the reason for A to P is relatively weak. And, on the basis of this
general fact about how negative intuitions respond to the two-step
procedure, he maintains we should not rely on them when assessing
theories of reasons.
A sample application of the procedure: if you tell someone that
there is reason for her to eat her car, she will likely find this
reason claim very counter-intuitive. But if you then say what the
reason is – that the car has the daily recommended dose of iron in it –
the claim will not seem as counter-intuitive. And, Schroeder predicts,
if you then go on to point out that this reason is very weak and
massively outweighed by other reasons, the claim that there is reason
for her to eat her car will seem even less counter-intuitive; it might
not be destroyed, but it will be twice weakened (see 96-97).
I have two worries about this defense of bold Humeanism.
(#1) Even if we accept the claim that the two-step procedure will
regularly weaken the relevant initial intuitions (twice), I am not sure
we should accept Schroeder’s conclusion. He admits, for example,
that the intuition that there is no reason for you to eat the car might
persist in the face of the two-step procedure – he holds only that its strength will be reduced
(twice). But why conclude, from that fact, that such intuitions are
“misleading” and that we can’t rely on them “to decide between out
theories” (96)? On the contrary, an anti-Humean could argue that
insofar as the intuitions survive the double test Schroeder
introduces, we should be more confident that they are veridical and
rely on them when deciding between theories. Why, they might ask,
isn’t passing a tough test something that counts in favor of the
intuitions?
(#2) I am unsure that Schroeder’s two-step procedure will have the
predicted results in all relevant cases. He discusses the
eating-the-car case and a case in which his Aunt wants to build a
spaceship in her living room. I think his procedure does
reduce the strength of the intuitions in these cases; I would even be
sympathetic to his making a stronger claim, namely that the procedure
destroys the intuition in these cases. If that is right, it would
help avoid worry #1 above.
But I am worried about the general claim about negative-reasons-intuitions. Consider, for example, a case in which the agent has a
self-destructive desire: an evil hypnotist instills a desire in Billy
to eat a plate of broken glass that is spray-painted yellow. He finds
himself wondering where he could get a plate of glass…perhaps by
smashing the coffee pot in the break room? And he wonders where some
yellow spay paint might be…perhaps in the utility closet on the first
floor? Billy has a desire to eat a plate of yellow glass shards. And
his smashing the coffee pot in the break room, breaking into the
utility closet, and spray-painting the glass shards will help him
satisfy that desire. But it seems counter-intuitive to say there is
reason for him to do these things, e.g. to spray-paint the glass shards
yellow.
Start with the bare claim that there is reason for Billy to spray-paint yellow a plate of glass shards that lies on his desk. Seems a bit counter-intuitive. Now let’s apply Schroeder’s test. First, imagine that I tell you
what his reason is – that spray-painting the shards will allow him to
eat a plate of yellow glass shards. In the face of that information, are you less confident that there is no
reason for Billy to spray-paint the plate of glass shards? I would not be; in fact, I think it would sound more intuitive than
it did before I attended to the purported reason. How about if I
tell you that I think that that reason is not a very weighty one? Here
I am not as sure, but I am tempted to say that the intuition at least
retains its strength in the face of this step of his test.
My worry, then, is that in some cases his test fails to reduce, and
might even increase, the strength of the relevant
negative-reason-intuitions. This suggests that some of our intuitions
will remain useful in theory construction and pose a persistent threat
to bold Humeanism, even if he is right about the cases he explicitly
discusses.
Hi Brad,
Nice post! I’m embarrassed to say that I haven’t read Mark’s book yet. Does Mark think that there is a reason to eat the whole car, windshield-wiper blades and all?
This question might be silly? But, perhaps, the following one is better. Suppose that I desire to consume 500 mg of vitamin C. Inside my house is a pillbox with a 500 mg tablet of vitamin C. Eating my entire house would, then, fulfill my desire to consume 500 mg of vitamin C. Does the bold Humean view imply that I have a reason to eat my entire house?
If so, I’m not sure that his test does a good job of reducing the strength of my negative-reasons-intuition in this case either. But I like your case just as well.
I have two suggestions.
1. I suggest that you change the variable P to φ. There is a good reason for doing so, which I will not divulge.
2. A nice spray can of yellow paint can be purchased at any automotive parts store or hobby shop.
Two comments about being charitable. First, I think that mentioning how Mark’s view is a view of the pragmatics of conversations concerning what to do is important. Second, I think that it’s important to go through the Tom Grabit example.
Sorry, I should say that the relevant argument depends on part of a view of the pragmatics of conversations about what to do.
Thanks for the comments!
Doug,
Good question. I suspect that his precise formulation will avoid the worry you raise – but I will check the book when I am at the office tomorrow and let you know. In any case, I like the idea of having a reason to eat the house!
Jamie,
Ha!
Errol,
(1) That is a good point about the pragmatics. I can’t think of how mentioning the way MS motivates the procedure (pragmatics) will lead to a way he could diffuse the worries I raise, but you are tight that it might provide a more charitable picture of his procedure. Please let me know if you think he could diffuse my worries by mentioning the pragmatics story that motivates the procedure.
(2) I do not see how the Tom Grabit example needs rehearsing. Unless you think the post here and review will not get readers to see the attraction of the procedure unless I mention it to explain the pragmatics stuff. Again, let me know if my worries could be assuaged by appeal to that case. I can’t see that yet.
I’m kind of with Errol here. Mark’s example is in service of a general point — that people’s negative existential intuitions about reason claims are often effected by the pragmatic features of making such claims in a context. Normally we expect people to make claims that are as relevant as possible for conversational purposes, and in most conversations about reasons claims about the most weighty reasons are more relevant than claims about reasons which are weak enough to be outweighed to the point where they are practically irrelevant. Talking about a less weighty reason without mentioning the more weighty and perhaps also indicating their relevant weights is not being sufficiently cooperative in such conversations. So we note that making such claims is inappropriate. And Mark thinks this is reason to think that the best explanation of our thinking the claim false is that it is conversationally inappropriate. Thus, the intuitions underlying the thought that we have no reason in such cases are unreliable, especially when the reasons are weak.
I think of the car example as more illustrative than as being used to offer support for the hypothesis all on its own. We have a relatively well worked out theory of conversational cooperation which nicely explains how implicature works and which itself predicts that we will be unreliable in cases where stating the literal truth will be conversationally unhelpful. The car example is supposed to show how that might work in a particular case. It is part of that theory that implications can be cancelled, whereas the semantic content of an utterance cannot. So, in the car case, we can cancel the implication that the reason to eat the car is among the better relevant reasons by stating explicitly that it is a bad reason. And this makes it less inappropriate.
That there is another situation in which saying the parallel thing does not cancel an implication would make me want to look at the assumed conversational context of that new case. Is there a difference between the spray painted glass case and the car case that would make us want more in order to cancel certain implications that are false, for example? I’m afraid I’m a little tired to work out the details of such an analysis here, but I agree with Errol that you need to focus on the details of how implicature cancellation would work to allow us to expect a mistake about inappropriateness in this case, and how the alleged failure to cancel that implication in the example shows this mechanism is not at work in the parallel cases of most concern.
I’m going to hit “Post” now with some worry that I’m to tired to have said what I should have meant. I’m hoping it will be helpful even if confused when I look at it in the morning.
I’m wondering how the view will fit holism about practical reasons. The idea there is that a given consideration can in some contexts be a reason for an act and in others a reason against that act. I’m worried that Schroeder’s view will entail that such a consideration is in all contexts both a reason for the act and a reason against the act. This is because the consideration obtaining will in part explain both why doing the action and not doing it will promote some of the agents desires. So, it seems like holism and Schroeder’s Humeanism are incompatible. Schroeder recognises this later and says that the holism intuitions can be explained by looking at how good reason the consideration is for the action and against it in different contexts as this can change in his view. But that still leaves me wondering.
Hi Brad –
Nothing much helpful here, but I’d like to second (or third) the point about pragmatics. But also I’d like to note that I saw his defense against the Too Many Reasons objection also bound up with his rejection of proportionalism. I suppose I’m inclined to believe that it’s not so implausible to think that you could have a reason to spray paint the glass, so long as we’re talking purely about pro tanto reasons, and specifically pro tanto reasons with very little weight. The problem, it seems to me, is how to account for reasons like this having very little weight in all cases, which the rejection of proportionalism seems to address.
I’d also like to echo your general praise of the book, and further note that no review of this book would be complete without substantial praise for a truly awesome cover.
Brad,
I think that the fact that Mark uses part of a (very plausible, I think) pragmatic theory to generate the tests you mention is a very important part of the argument. And I think that for purposes of making the argument on his behalf, mentioning his take on the Tom Grabit example is giving him all he’s got. For he shows very nicely that the standard lesson taken from the example is not the right lesson to take from the example. And this is all in the service of making us skeptical of a certain type of intuition.
Just to report my intuition about your case: insofar as I have Mark’s intuition about the car, I have it about your case. Think of it this way: It might be that the reason why we think that the bare reason claim is false is because it would be such an unhelpful response when deliberating about what Billy should do. But once it’s pointed out that it certainly isn’t what the speaker thinks Billy should do, it doesn’t sound so strange anymore. Presumably, the speaker would think that Billy should get some psychological help so that he doesn’t act on his reason to eat the glass.
Mark (vR) and Errol,
Thanks for pressing me on this. I now think there are two reasons of charity that speak in favor of mentioning this:
(i) It will better motivate the procedure for people who have not read the book.
(ii) It is worth mentioning because it will bring out the fact that even if I am right to doubt that his argument works as an adequate defense of bold Humeanism, it is – as he notes – an argument that should interest people in other areas of philosophy.
Maybe I will add something like this:
“Schroeder thinks we are tempted to assent to negative existential reasons because assertions that there is some reason to phi *imply* that we know that the relevant reason is and that the reason is not piddling. So if we are unaware of what the reason is or think it is merely piddling we are reluctant to accept that there is some reason; we are tempted to think there is no reason.
This is a good worry to keep in mind whenever we are tempted to think someone lacks a reason; we should check that we have good reason to think that and are not being lead astray by our resistance to accepting things that are implied *but not entailed* by the relevant negative reason assertions. And Schroeder helpfully introduces a two-step procedure to help as avoid making this mistake.”
Dale and Errol,
Thanks. I will posting about the rejection of proportionalism and his alternative account of weighing next. I have just been granting that for the sake of argument so far.
Hi Brad,
A couple of thoughts about your worries.
In respect of worry (#1), I would think that Schroeder’s position is not necessarily that we should reject these reasons.
I would argue (and did recently with one of my seminar group) that reason to do something has nothing to do with practical intuition.
I would imagine (and due to my sparse knowledge of the subject and book in question, am happy to be corrected) that Schroeder’s aim is to show that there is a value based heirarchy of reasons, and while we cannot reject a reason on this two fold test, we can show that the reason under consideration is less valuable (or rationally attractive) than other reasons which can be presented.
I would agree with this position myself, and would say that we cannot truly criticise the notion of having a reason to eat my car, my house or a plate of yellow glass, but we can show that any the reaons highlighted in those examples can be valued at the lower end of the scale.
Secondly, I dont see how in your glass example we could have more intuitive support for the glass eating on hearing the reason for the paint search. Can you explain?
Thirdly, perhaps the house example from Doug can be challenged on the notion that eating the house is an action far in excess of the achieveing of my vitamin C requirements, and as such is perhaps a badly strung example against Mark?
Lastly, I would claim good reason to just eat the whole world in totality just to be on the safe side for my anemia and scruvy needs.
Matt
Mark Vr and Errol,
On the substantial issue (sorry this is long!):
On reflection, I think that the pragmatics stuff might help him evade my worry #1. There are other cases in which one can try to cancel an implication by saying a bit more, but in which the implication fails to be completely canceled – at least in the sense that listeners will still attribute something like the implied claim to you. I am thinking, in particular, of cases in which someone uses a racist term but tries to cancel the implications. Maybe Mark could argue something like that is what explains the persistent negative reason intuition in the face of the two-step test. I think this might be a good line of response to the worry I raise on behalf of anti-Humeans.
I am less clear about my second worry. I take it that the pragmatic point is (crudely) that claims that there is no reason usually imply that the speaker does not know what the reason is or thinks there is no non-piddling reason. The fact that the two-step procedure reduces the force of the intuition is supposed to confirm that these implications exist and might be tempting the speaker to deny the reasons claim. I agree that, insofar as this is true, the denial is not worth taking into account in theory construction. This is because the relevant extra claims are implied and not entailed by the bare no reasons claim.
If we bracket the fact some (e.g. Errol) will not share my response to the case I introduce, I hope we can agree that if my reactions were typical for some set of cases, then the pragmatics story would no longer throw into question the theoretical value of the negative intuitions about the cases in that set.
My basic thought is that Mark might be right about cases like the car case and the Quinn case (or his aunt case). Those are indeed cases in which we are reluctant to say there is reason because we can’t initially see what the reason could be (Quinn radio case and Aunt case) or think the reason is piddling (car case).
Notice that we are tempted to say that in these case the desires are for something that is no good but not bad (Aunt and Quinn cases?) or for something that is good (car case). These facts about the desires “track” the relevant implications that lead us to be tempted to deny there is any reason. But I introduced a case where the desire is for something that is all bad. (I know this depends on a contentious account of bad, but you see what I am after, I hope) I think that is connected to why the pragmatics story he gives will not help diffuse my worry. Maybe Mark (S) could add a third question to deal with this sort of case, but I doubt it.
Hi Matt,
On your second Q: I think you are actually agreeing with me here. I think that hearing the purported reason for spray-painting makes us more willing to agree that there is no reason to spray-paint. Similarly, I predict we will think that the glass-spraying has less support once we hear the purported reason.
I will be raising doubts about Mark’s weighing story next. I will mention, though, that I do not think is based on the sort of heirarchy of values towards which you gesture. I, on the other hand, might be sympathetic to some such story but would resist the claim that there is any good in spray-painting or eating the glass (unless you can show there is some good in having a general disposition to take means to satisfy desires).
Given your last point: is there anything you think you have no reason to do?
Jussi,
I am not sure I see the worry here. If mark says that something can be a substantive reason in one context but a piddling one in another won’t that allow him, on the basis of the pragmatics story, to explain (and discount) our temptation to say the thing can be a reason in one context but no reason in another?
I think I am missing you point. Maybe you can explain more clearly how holism entials the denial of some part of his view and then explain why holism is true? (In three sentences or less)
Errol,
You wrote: “Think of it this way: It might be that the reason why we think that the bare reason claim is false is because it would be such an unhelpful response when deliberating about what Billy should do….”
If the reason does not “count in favor” in the deliberative context, then I am left wondering in which type of context (of what we might call action assessment) it would “count in favor”. The claim about thinking the person needs to see a therapist seems to imply that it would not count in favor in an advice context. And I doubt it would count in favor in an approval/disapproval context.
Mark is right that talk of counting in favor is “slippery” but I think you need to answer my question to support the claim that the reason is normative rather than merely explanatory. Hope that makes sense.
Brad,
Well, personally, my concept of rationality and what we have reason to do is rather clouded by my bleak nihilism and existentialism.
I do think that reasons can be provided for anything, whether morally right or not, and those reasons can be raised in heirarchy of value, with those good reasons over bad reasons. I dont think that there is anything that there is no reason to do. All actions can be rationalised, but it seems that my view of this extended and open rationality is in contrast with others.
In respect the second Q, I think that hearing the reason for the spray-painting, we can say that it is a reason for him to continue.
We maybe cannot say that it is a good reason, but hearing the reason wouldn’t make me think that it was less intuitive for him to spray paint the glass. The reason is in and of itself not normatively powerful to that degree, there has to be something else in the mix as well.
I suppose, to cut any legion of my bumfuddled prose, that I dont hold that a reason need be teleological exclusively. You cannot hold the reason alone and determine if it makes the act more or less intuitive, we have to look at other areas, perhaps the end, the desire, the context (I’m not sure myself) and from that wider basis we can make the relevant conclusions.
Matt
Good post (and thanks Jamie, for making me laugh!). Brad, your move strikes me as exactly the right way to raise trouble for Schroeder here. My worry is that quite a few people might, like Errol, be tempted to say that their intuition that there’s a reason to spray-paint isn’t weakened, because there is a tiny reason, namely that it is a means to satisfy the desire.
Now I am embarrassed to admit that I haven’t yet read the book, but I guess that Schroeder argues that merely satisfying a desire constitutes a reason, irrespective of e.g. any pleasure gained or nagging urge avoided in virtue of satisfying it. If so, you could rule out the influence of such extraneous psychological factors. You could strengthen your example against the bullet biters by making the desire one the satisfaction of which the agent will be unaware, or even better, of which the agent will be unaware even whether he has taken the means or not, in case he might gain some pleasure from that.
Here’s a way of rigging that up, no doubt you can think of a better way to do it: your agent has a yellow spray can and a red spray can on the desk, and has his fingers on the buttons. He now desires that his future self eat yellow glass shards. But he also knows that the moment he presses the button on either spray can, the fumes will cause him to forget which color of glass shards he previously desired to eat. Now, the question is: is there a reason for him to move to spray the glass yellow in particular, and not red?
Finally, I think sometimes we can get confused about the kinds of reasons Schroeder is thinking of as being pragmatically suppressed by calling them “bad reasons” or “weak reasons” (since “weak” can be read as meaning the same thing as “bad”). Almost anyone will concede that there are *bad* reasons in these cases, if what is meant is that someone might conceivably though mistakenly act on those grounds. They are not, on Schroeder’s view, bad reasons, but rather are good reasons that are outweighed. So when we have the negative existential reasons intuitions, can’t we straightforwardly cancel the pragmatic implicature and test for the existence of reasons like those Schroeder wants by asking whether there are any *outweighed* reasons to perform the relevant action?
Brad,
You wrote,
If the reason does not “count in favor” in the deliberative context, then I am left wondering in which type of context (of what we might call action assessment) it would “count in favor”. The claim about thinking the person needs to see a therapist seems to imply that it would not count in favor in an advice context. And I doubt it would count in favor in an approval/disapproval context.
I took it that one of the upshots of Mark’s argument was to show that this is a bad way to think about it. Figuring out whether it would be a good idea to eat the glass, or whether one should advise Billy to eat the glass, or whether one should blame or praise Billy for eating the glass is a good way of figuring out what Billy has sufficient/most reason to do, but it’s not a reliable way of figuring out everything Billy has reason to do.
So, in a perfectly good sense of counting in favor, I think that there are things that count in favor of him breaking the pot and getting the spray paint. In other words, the reason does count in favor in the deliberative context, and in the advice context, and in the approval/disapproval context. It’s just that those reasons are massively outweighed. And Mark’s argument is designed to show that the reason why we hear the bare reasons claims as false is because it makes good sense to ignore such massively outweighed reasons when talking about what to do. But the fact that it makes good sense does not show that those considerations aren’t reasons.
Hopefully that is helpful and not merely foot-stomping.
Brad,
I’m still thinking about what the worry is… Sorry. But here is a second pass.
Take Dancy’s example of the fact that my action is against the law. Intuitively, in most contexts, this is a reason against the action. But, it seems like we can think of contexts in which it is a fact that counts for doing the action (say, when one can do the action as a protest againt an unjust law). In these cases, we can find considerations that disable the fact’s (that my action is against the law) ability to count against
the action and enable it to count for doing the action.
This way of describing the case sounds natural. Schroeder, in contrast, has to find a more complicated explanation. He seems to forced to say that in all contexts that fact is invariably a reason both for and against the action. He has to explain our talk about the reason in one case counting against the action and in another for it as talk about saying that in one case the fact is a strong against-reason (and a weak for-reason) whereas in other cases it is a strong for-reason (and a weak for-reason). On this picture, reasons never really change valency but rather their invariable strength just changes from context to context.
Now, this may not be objectionable as such (even though I do find holism highly attractive – more so than Humeanism about reasons) but I do worry at least about this entailing weak reasons to do very evil things. I guess one of my basic beliefs is that there are some things that are so evil that there is no reason to do them at all even if some of our desires might be satisfied as a result.
Simon,
I like the idea of tweaking my example in some such way. I was initially thinking of stipulating that there was no pleasure and even suggesting that the agent is also alienated for the desire. But your example has me thinking of a new direction; thanks.
I never thought of that point about ‘weak’. You might be right, so maybe I should consistently use ‘piddling reason’ since it seem to lack the possible confusion with ‘bad reason’.
I’m not sure he should be ok with canceling (as the second step in the procedure) by asking about any outweighed reasons. There might be cases in which there is a piddling reason to phi and no strong reason against doing so, but about which people will be reluctant to assert ‘there is reason to phi’ because that assertion would, in the given context, imply that there is some *non-piddling* reason to phi. It is relevant to note that (on page 95) Mark predicts that positive reasons assertions will, “reinforce the standing presumption that the speaker has a relatively weighty reason in mind.” So the context for the case I mention would need to be one in which there is such a standing presumption. Hope that makes sense.
Simon,
Fill in a detail in your glass guy example.
Does he think that if he sprays the glass yellow and then forgets which color glass he wants to eat, he will be more likely to eat some yellow ground glass than if he sprays the glass red and then forgets which color glass he wants to eat?
If not, then he has (so far as we’ve been told) no reason to spray the glass yellow instead of red.
Suppose that you now prefer that tonight you eat some zucchini. (That is, courgettes.) You know that when meal time comes, you will be utterly indifferent between eating zucchini and eating yellow squash. Do you now have any reason to pass over the yellow squash at the grocer in favor of the zucchini?
I say yes. It’s a bit difficult to imagine being in this situation — the problem is that when we try we are apt to imagine ourselves in a slightly different, more ordinary situation, in which we prefer the zucchini dinner to the yellow squash dinner only conditionally on continuing to prefer it at dinner time. But I think I understand the situation as I described it, and I think in that situation there is a (presumably weakish) reason to pass over the yellow squash at the grocer in favor of the zucchini.
Errol,
Fair enough — I mistakenly thought you were granting the no reason claim in the deliberative context.
I fear we just have different intuitions.
Errol,
You wrote: “And Mark’s argument is designed to show that the reason why we hear the bare reasons claims as false is because it makes good sense to ignore such massively outweighed reasons when talking about what to do. But the fact that it makes good sense does not show that those considerations aren’t reasons.”
A *first* go at a response:
Just think of a case in which no good will come from the relevant action, serious bad will come from the action, and the actor does not think any good will come from it. He has a desire that would be served by it though.
You explanation of *my* reluctance to say the person has reason to preform the action is that the reason is outweighed. But I maintain that my resistance is best explained by my beliefs about the relevant values and the thought that those value claims entail that there is no reason in favor AND lots of reasons against his acting – a crude slogan: no reason if no (value or perceived value).
Since your explanation is revisionary (relative to my self-understanding) I would think the burden is on you to show you are right.
Jamie:
I intended that he will retain the desire to eat shards of glass after spraying, he’ll just forget which color he previously desired to eat. So spraying the glass yellow is an effective means to satisfying his present desire. It’s just that the agent never learns whether he has taken the means to that desire or not.
I share your intuition about your zucchini example, I think, but there is of course an important difference between that example and Brad’s – namely, that you did not introduce the relevant desire as an essentially ungrounded and irrational one. The natural way to fill out your example might be: I know that zucchini is more nutritious than yellow squash, so I desire now that I eat it tonight. I also know that the smell of the damn stuff cooking will cause me to lose my appetite for it later, and in consequence I know I’ll be equally desirous *then* to have the yellow squash. It’s easy to conclude from this that my later self will be a bit irrational or imprudent and that there’s a weakish reason for me to now choose the zucchini. I know you didn’t say any of this, but this kind of implicit filling-in *may* still be the source of the intuition.
There’s a second disanalogy between your example and my modification of Brad’s example as well: in the supermarket case, the agent gets to know that he is taking the means to satisfy his present desire if he chooses the zucchini. It’s not implausible to think that this might cause him some weak pleasure or give him some kind of relief, and that this is what gives him a weak reason to make that choice.
I mostly want to keep my nose out, for now, but I thought I should respond to Jussi’s question about holism.
According to Hypotheticalism, R is a reason for X to do A just in case R is part of the explanation of why X’s doing A promotes P, where P is the object of one of X’s desires. So if the same consideration, R, can in some circumstances explain why X’s doing A promotes P, where X desires that P, and in other circumstances can explain why X’s doing ~A promotes Q, where X desires that Q, then the very same consideration can be a reason to do A in some circumstances and a reason to not do A in other circumstances, as holism requires.
In fact it is easy to see how this could be the case. Suppose that X’s doing A promotes P. Then plausibly, X’s doing ~A promotes ~P. So if X desires that P, then R is a reason for her to do A, whereas if X desires that ~P, then R is a reason for her to not do A. Difference circumstances, different ‘valence’. It is also easy to see how differences in background conditions other than the agent’s desires can also engender holistic behavior like this.
What Hypotheticalism can’t get you, however, is the absence of any robust generalizations about the conditions under which R is a reason for X to do A. On the contrary, Hypotheticalism is such a generalization. As the discussions of Jackson, Pettit, and Smith and of McKeever and Ridge have illustrated, Dancy seems to want something stronger like this, rather than the bare claim of holism as you articulated it.
Mark,
thanks. I did get that. I guess I was more worried that the agent will have in all circumstances *some* desire such that R will be part of the explanation why X doing the given action will promote an object of that desire. And, likewise for not doing the action, there will be *some* desire such that the same R will be part of the explanation why X not doing the action will promote an object of that desire. This seems to follow unless we in some way limit the desires that are relevant for some consideration counting as a reason for the agent. But, it doesn’t seem like the Humean will do that. So, all in all, I can see how your view allows holism of sorts in principle but not really actually given how many desires we seem to be allowed to have.
Hi, Jussi.
It’s quite likely to follow from Hypotheticalism that for each agent X and each action A, there is at least some reason for X to do A. I hope it doesn’t follow from Hypotheticalism that for each agent X, each action A, and each consideration, R, R is a reason for X to do A. If it did, the problems would be a lot deeper than simply the issue about holism that you mention.
Simon,
So, is your point about the glass example that the guy doesn’t even have a weak reason to spray the glass shards yellow? I must say I don’t find this any more compelling than any of the other examples.
About the vegetables: sorry, I chose zucchini and yellow squash because they are so similar. I didn’t say explicitly that you don’t believe there is any nutrition difference; I just assumed that you would assume it. So the case is supposed to be one in which you just brutely prefer to be eating some zucchini, rather than some yellow squash, tonight, just as your glass eating guy prefers to be eating yellow and not red glass shards.
Jamie: well I am surprised. I find your intuitions strange, I must admit – they strike me as resulting from a kind of fetishism of desire satisfaction.
Let me take your zucchini case and modify it a little. Suppose that our subject knows that every single week while he shops in the afternoon, he strongly desires for no reason at all to eat zucchini later on, but that by the time dinner comes around, he is always utterly indifferent between zucchini and yellow squash. Suppose also that he knows that he gets no pleasure from buying zucchini, and would suffer no pain, no nagging desire, and so on, if he bought yellow squash instead. Finally, suppose that zucchni is just a little bit more expensive than yellow squash. Do you want to say that it is most rational for him to keep spending extra money on zucchini every week, rather than economizing on the yellow squash?
Simon,
Part of the problem appears to be that you’re insisting on changing the subject. Jamie said that ‘there is a (presumably weakish) reason to pass over the yellow squash at the grocer in favor of the zucchini.’ You respond, ‘Do you want to say that it is most rational for him to keep spending extra money on zucchini every week, rather than economizing on the yellow squash?’
Never mind that the detail that the zucchini is more expensive just got added at this point; Jamie’s claim wasn’t that it was most rational to buy the zucchini at all, but just that there is a ‘weakish’ reason to do so. Presumably the more weakish the reason, the easier it would be for it to be outweighed by even weak reasons to buy the yellow squash instead – and you’ve added a number of strong reasons to the case, including the fact that by the time dinnertime comes around, he will enjoy the yellow squash more, and the fact that the yellow squash is cheaper. Obviously those reasons will outweigh a ‘weakish’ reason to buy the zucchini, making it most rational to buy the yellow squash, even on Jamie’s own terms. If the reason to buy the zucchini is ‘weakish’ enough, presumably it could even be outweighed by the fact that you have to reach farther up the greens rack in order to grab it.
I think that in responding to Jamie in this way, you’re therefore proving the point of my response to the Too Many Reasons problem in chapter 5 of Slaves. It’s clear that what is animating you is judgments about what it is rational to do – but bare existential claims about reasons are much weaker than claims about what it is rational to do – unless we pragmatically interpret them as carrying more import.
I should also note, since you appear to be of the school of thought that enjoyment or pleasure can ground reasons but desires cannot, that except in chapters 8 and 9 of the book, I use ‘desire’ in a stipulative sense for that state of mind, whatever it is, which plays the role of explaining reasons in paradigmatic cases like the normal zucchini case, in which it is intuitive that the agent has some reason to buy what he prefers (you’ll say that it is a fact about what he would enjoy, or something like that).
So since you can think of the Too Many Reasons problem as arising no matter what you think this state of mind is – whether it is something about enjoyment, for example, or desire in a more conventional sense – the question of what sort of mental state the best theory will take desire in the stipulative sense to be simply cross-cuts the Too Many Reasons problem that Brad introduced in this thread. If you want to see why I prefer the view that desire is to be understood in something much more like the conventional understanding of desire, I recommend that you look at chapter 8 of the book.
Mark: thanks for the clarifications.
I think you’re a bit unfair in saying that I’m adding “a number of strong reasons” to Jamie’s case – I didn’t say that the subject would prefer the yellow squash by dinnertime, but rather accepted Jamie’s description that he would be indifferent between the two by dinnertime. I stipulated only that the zucchini was “just a little bit” more expensive, and I meant this price difference to be small enough to make the reason it provides a very small reason, weaker than the supposed reason generated by the (strong) bare desire at shopping time. It can be as little as 1 (Zimbabwean) cent, if you like, so long as you think the amount provides *some* reason.
I don’t think of myself as changing the subject by talking about what’s most rational, since I think it’s most rational to do what you have the most obvious overall reasons to do, on balance. But if you want to dispute this, or dispute its relevance, I can anyway keep my question in terms of reasons: do you think he has most reason to keep spending extra money on zucchini every week, rather than economizing? (My intuitions stay the same: I don’t think he does.)
Finally, I excluded enjoyment because I assumed (without evidence) that you wanted to ground reasons in desire defined in a conventional sense. Since you in fact do, it still seems reasonable to exclude it as I did. I don’t myself want to commit to the view that enjoyment *always* grounds reasons, and the Too Many Reasons problem won’t arise unless one says something of that sort. In this zucchini case I think enjoyment could provide a reason, but in many other cases I would be inclined to resist the claim that it does (e.g. if the source of enjoyment were deeply immoral). I’m open to persuasion on this, but that’s another argument. I look forward to reading your book!
Simon,
Jamie: well I am surprised. I find your intuitions strange, I must admit – they strike me as resulting from a kind of fetishism of desire satisfaction.
Hm, if you mean by ‘fetishism’ what I mean, then you have misunderstood. Can I recommend “Backgrounding Desire”?
I think what’s rational has to do with what is coherent, not with what is economical. You think it’s clear, from the details already provided, that the squash-buyer has most reason to save the money, is that right? Because I think that is obviously wrong; it’s obvious that we cannot tell from the details provided what he has most reason to do. Some people don’t care about money. Do your intuitions tell you that those people are irrational?
Simon,
Jamie: well I am surprised. I find your intuitions strange, I must admit – they strike me as resulting from a kind of fetishism of desire satisfaction.
Hm, if you mean by ‘fetishism’ what I mean, then you have misunderstood. Can I recommend “Backgrounding Desire”?
I think what’s rational has to do with what is coherent, not with what is economical. You think it’s clear, from the details already provided, that the squash-buyer has most reason to save the money, is that right? Because I think that is obviously wrong; it’s obvious that we cannot tell from the details provided what he has most reason to do. Some people don’t care about money. Do your intuitions tell you that those people are irrational?
Hm. I’m sure I only typed in the verification letters once. How did the comment appear twice? And with the same time stamp on both?
Simon, two quick things –
First: if you wait for Brad’s next post, you will see that I do not believe that having a stronger desire for zucchini over yellow squash needs to give you a stronger reason to buy zucchini. (Under the further conditions that you mention, I in fact think that the strength of the desire is probably irrelevant.) So I am not committed to any claims about what your agent has most reason to do. In fact, the basic structure of my answer to the Too Many Reasons Problem (just to reiterate) is precisely that existential claims about reasons are simply being confused (for predictable reasons) with false claims about the balance of reasons or at least about the relative weightiness of reasons. The fact that you are continuing, despite this, to make claims about ‘most reason’ simply proves my point.
Second: if you refer back to Errol’s points, above, you will note that the whole point of my discussion is to point out the way in which contrary reasons cloud the issue. If introducing the contrary reason created by the greater cost of the zucchini is not clouding the issue, then why did you need to introduce it? I have a simple explanation: you needed to introduce it, because it was precisely what it took to cloud the issue. Moreover, I have a detailed explanation of exactly why that is exactly the kind of further feature of the case that you would need to introduce in order to cloud the issue. My explanation appeals to pragmatics, as several others have noted above, and it makes specific predictions, predictions which Brad began by acknowledging seem to be confirmed at least in the cases that I discuss. And as Errol notes, my Tom Grabit argument seems to show that a pragmatic explanation of our negative existential intuitions about reasons must be right – even in some cases in which the reasons are reasonably good ones. If my explanation works even in cases in which there are reasons that are reasonably good ones but are simply obviously outweighed, it is certainly on much better grounds in cases like your zucchini case, in which the reason is only ‘weakish’ and is obviously (with due respect to Jamie, giving up money has opportunity costs even if you don’t care about it, and I’m happy to assume for the purposes of the example that the agent does care about it) outweighed.
Mark,
thanks again. This:
“It’s quite likely to follow from Hypotheticalism that for each agent X and each action A, there is at least some reason for X to do A. I hope it doesn’t follow from Hypotheticalism that for each agent X, each action A, and each consideration, R, R is a reason for X to do A.”
clarifies a lot. My worry is in the middle of these options. I wonder whether for each consideration R such that it is sometimes a reason for action A for some agent X, R will always be a reason for X to do A. So, the question is will, for instance, the consideration that it contains the daily allowance of iron always give a reason for me to eat the car. And, given that in 6.2 you say that these reason-relations are overdetermined and the promotion relation can be very weak, I suspect that it will be always a reason to eat the car that it contains the daily allowance of iron. As you note in the footnote, also other desires than the desire to get the daily allowance of iron could explain why this consideration is a reason.
Anyone who cares,
This is getting away from the comments so far, but I just noticed something. I suspect my case would be better put if I removed the *evil* hypnotist bit. The claim about him being evil might well trigger undue attention on the all things considered issue (esp given Mark’s view of weighing which I will post about on Monday). I now think I should say that an alien, who does not know that eating glass would harm Billy, is the one to puts the desire in him. I can’t see that that change would weaken my case.
Jamie: I have enjoyed your elegant paper “The Negative Reason Existential Fallacy” on this though.
I was describing you, not our hypothetical subject, as fetishizing desire-satisfaction by treating it as reason giving. So unless you had something in mind I’m not seeing, I don’t think (the excellent) “Backgrounding Desire” is relevant. However, I now regret that remark as mere name-calling.
For the record, I didn’t think it obvious that the squash buyer has most reason to save the money, I just thought it obvious that he *doesn’t* have most reason to take the means to satisfy his fleeting zucchini desire instead. If the desire by itself provides any reason at all, that point shouldn’t be obvious.
Mark:
Thanks again, I appreciate your engagement here. I am somewhat hampered by not having your book to hand and having access to only limited pages of it on Google Books – ironically, I ordered it online a week ago but it’s not here yet.
I didn’t introduce the cost differential in the zucchini example because *I* thought it necessary to provide a countervailing reason, I introduced it because Jamie said *his* intuitions still hadn’t changed even in the red/yellow spray cans example, where mine clearly differ. I introduced the price differential a bit carelessly, but I intended for the reader to understand it as an amount providing an exceedingly tiny reason, tiny enough that it would be (or you could make it, by adjusting the price further) smaller than any supposed reason generated in the other direction by the presence of the fleeting desire. For this reason I find it puzzling that you – even after my earlier clarification – say the reason generated by the desire is “obviously outweighed” by the difference in cost. Can you explain why this is so?
Your Tom Grabit argument, if it succeeded, would show that it would be in principle wrong to introduce any countervailing reasons of great weight, if I were committed to doing that to prove my case. But I have reservations about that argument. You say that when you introduce the third sibling one has even less reason to believe that Tom stole a book, so the reason couldn’t have gone away entirely when the second sibling was introduced. I’m not convinced: it seems to me plausible to say that in the two sibling case, you have reason to believe that (Tom stole a book v Tim stole a book), but no reason to believe simply that (Tom stole a book). And in the three sibling case, you have reason to believe that (Tom stole a book v Tim stole a book v Tam stole a book), but now no reason to believe that (Tom stole a book v Tim stole a book) or that (Tom stole a book). You might also have reason for some beliefs about probabilities. Now it is true that you know that, in the two sibling case as opposed to the three sibling case, you are objectively more likely to have a true belief and less likely to have a false one if you believe simply that (Tom stole a book). But it’s not clear to me that this entails your having a weightier reason to believe it, or any reason to believe it at all. Maybe you just have reason to believe the relevant disjunction in either case?
Simon,
Hm. I see the difference, but now I don’t understand the remark. (I thought fetishizing something involved making it an object of one’s desire.)
The only alternatives are: save the money; buy the zucchini. Unless the reasons for doing these alternatives are exactly tied, then he has most reason to do one of them, or most reason to do the other. So if we can ignore (what I think is) a very remote possibility, that the two options are exactly tied, then if it’s obvious he doesn’t have most reason to buy the zucchini, how can it not be obvious that he has most reason to save the money?
Maybe you think it is likely that the reasons really are exactly tied. Or, come to think of it, there’s another possibility – do you think the reasons are incommensurable?
Simon –
Let me repeat myself one more time. When I say that there is a reason for the guy to buy the zucchini, I do not mean to be committing myself to any degree to which that reason is weighty. Hence there is no amount of money, no matter how small, which the zucchini could cost in excess of the yellow squash, which is guaranteed to generate a reason of lower weight than the reason I say there is for him to buy the zucchini. It doesn’t matter what tiny decimal value you pick for what trivial unit you pick in what hyperinflated currency you pick. For any small value you pick, there is still an infinite range of smaller values, and for all I’ve said, the weight of the reason to buy the zucchini falls in that range.
I do agree that it is possible to get off of the boat in the Tom Grabit argument. You seem to want to do it by accepting some principle according to which if there is not conclusive reason to believe P, then there is no reason to believe P. Now that’s an interesting principle, but I believe that I’ve offered a more attractive alternative, which doesn’t involve postulating principles like yours which apply to reasons for belief but not to other kinds of reasons, and can explain everything that your principle can explain. Like most arguments, the Tom Grabit argument doesn’t force you to accept my conclusion; it merely invites you to see the attractiveness of my conclusion, and to see your resistance to it in the context of some of its costs.
Jamie:
One can fetishize something by being preoccupied with it, or by giving it reverence that’s not due. On the zucchini case, I thought it possible that the reasons were tied, and I didn’t see any need to commit to them not being.
Mark:
First, I fully understand that you were not committing yourself to the existence of a reason of any particular strength. Neither was I; that’s why I invited you to set your own price. It seems to me that unless your claim is that the reasons provided by desires are in fact infinitely small, you should be able to set the price (in a hypothetical hyperinflated currency, if you like) such that the reasons seem at least roughly to intuitively balance out.
Second, I hardly need such an improbable principle as the one you attribute me – “if there is not conclusive reason to believe P, then there is no reason to believe P” – to dispute your Tom Grabit argument.
Suppose I tell you that Jamie is in Boston now, and Jussi tells you at the same time that Jamie is in England. Then I accept that in the absence of other information, you’ll have some reason to believe that Jamie is in Boston, and some reason to believe that Jamie is in England, but no conclusive reason to believe either proposition.
Yet this case seems to me very different from the Tom Grabit case, which is more like the following one. Suppose I say to you: “I just tossed an ordinary coin.” I think I’ve given you reason to believe that Simon just tossed a coin. You also will have (independent) reason to believe that an ordinary coin nearly always lands either heads or tails, so I will have given you reason to believe, by inference, that Simon just tossed a coin and it landed either heads or tails. But I don’t accept that I have given you any reason, by inference or otherwise, to believe that Simon just tossed a coin and it landed heads. You might take a reason for one belief (a disjunctive belief) to be a reason for a quite different belief. But I think this would be a mistake.
I have a draft paper on the reasons for action version of the principle I’m using here and I’ll be pleased to send it on request.
I shall try not to say any more on this thread, but I’ll check back later to read any responses.
Simon,
Here is a reason you have in the Tom Grabit case: It seems as if Tom stole the book (note: you don’t have to think the seeming is the reason). The fact that it seems as if Tom stole the book is a reason to believe Tom stole the book. This reason is weakened considerably by the existence of Tim. But it’s weakened even more by the existence of Tam, and so on as you add more siblings. If these plausible claims are true, then there is a reason to believe Tom stole the book–even in the case when he has an identical twin. But this is not the common lesson taken from the example, as you know.
Hi Simon,
Yes, it is possible (barely, I would have thought) that the one Zimabwean cent counted exactly as much as the shopper’s preference for zucchini. You are in a rather peculiar situation with respect to what you find obvious. It’s like finding it obvious that John Doris is either taller than or exactly the same height as Michael Bratman, but not obvious that John is taller.
Let me steer back to Brad’s original point about my response to the Too Many Reasons problem, since it’s only fair that he get a response.
Brad had two worries: (#1) The first was that even if negative existential intuitions about reasons weaken in the face of my “two-step procedure”, he wants to know why we should conclude from this that negative existential intuitions about reasons are misleading?
The answer, I think – which I think Brad may have already accepted above – is that my “two-step procedure” wasn’t really designed to “weaken intuitions”. It was designed as a test of the predictions of an independent pragmatic explanation that I gave of why we would have negative existential intuitions about reasons even when they aren’t true. There were two parts to my pragmatic reasoning, each of which led to its own prediction. So I tested both parts in the ‘reason-to-eat-your-car’ case, and believe the predictions were confirmed. Since what was confirmed was an independent pragmatic explanation of why under certain conditions we would have negative existential intuitions about reasons even if they were not true, it confirms the thesis that negative existential intuitions cannot be relied on, under those conditions.
So which conditions are these, under which negative existential intuitions about reasons are not reliable? There are two parts: first, when the only reasons to do something are very weak. And second, when the only reasons to do something are obviously and clearly outweighed. The weaker the reasons to do some thing, and the more compelling its competitors, the easier it will be to elicit the intuition that there is no reason to do something, even if there is.
This brings us to Brad’s second worry: (#2) he thinks that the case he discusses is unlike the cases that I discuss, in that the predictions of the pragmatic explanation seem to work in the cases that I discuss, but not in the case that he discusses.
Now, I’m open to the possibility that this is so, but I’m not yet persuaded. The main problem, is that I don’t think I understand what is really supposed to be different about Brad’s cases than the cases that I do discuss, except that they are set up to involve fewer and weaker other reasons to do the thing that it is controversial whether there is any reason to do it, and better reasons not to do it.
So whereas I discuss Aunt Margaret, who deeply desires to reconstruct the scene depicted on page 78 of the November 2001 issue of Martha Stewart Living – on Mars – and hence has a reason to build a Mars-bound spacecraft, Brad discusses the guy who deeply desires to eat a plate of broken glass that is spray-painted yellow – and who has that desire only because he was hypnotized. So what I want to know is: what are the differences between these two cases? I might simply be being dense, but it really looks to me like the main differences come from the fact that eating a plate of glass is more harmful than building a Mars-bound spacecraft, and that the desire was instilled by a hypnotist. (The “painted yellow” thing doesn’t seem to do any work – if anything, it’s less arbitrary than Aunt Margaret’s desire.)
Now, the addition of the added harmfulness to the case is precisely one of the things that I had already predicted could cloud things up. I don’t think Brad thinks that is what is going on; he thinks not just that it is very clear that he still has the negative existential intuition in his case, but also that it is just as strong after I try to cancel the implicature as it was before. So let’s try that: I say, ‘the fact that there is yellow spray-paint in the closet is a reason for him to open the closet door’. Sounds weird, doesn’t it? Now I say, ‘I don’t mean that he has very good reason to open the closet door – in fact, I think he has as close to no reason at all as it is possible to get while still having some reason to do something’. I do still think this doesn’t sound as bad. Maybe this is just a clash of intuitions, but I was really doing my best to take the intuitions out of it – if the strength of negative existential intuitions about reasons can be affected by all of the things which affect those intuitions in the first place, then by concocting cases in which the reasons are even weaker and even more outweighed, that is the sort of thing to make it harder to pull apart the bare reasons existence claim from all of the other false things.
Hi Mark,
On the first worry: you are right; I find that response compelling.
On my second worry:
(1) I agree that the yellow paint part of the example is inessential and I will leave that out.
(2) You ask me to spell out the difference between the Aunt Margaret case and my case. I agree this is key.
It might help to distinguish some different cases:
(Pure Good) A has a desire and acting to help satisfy it will bring benefit to someone and no harm to anyone.
(Swamped small good) A has a desire and acting to help satisfy it will bring small benefit to someone and large harm to someone.
(Pure Bad) A has a desire and acting to help satisfy it will bring no benefit to anyone but some harm to someone.
As you describe it, I assumed that the Aunt Margaret case is a swamped small good case – the benefit of her acting on the desire might be that she will not be frustrated or something like that. My case, on the other hand, is a pure bad case.
(3) Now for the dialectic: We have negative intuitions about each case.
You claim that negative intuitions are misleading in both sorts of cases because large harms make us confuse weak,swamped reasons with the absence of reasons. I agree this is true in swamped good cases, but deny it it true in pure bad cases.
Your support for the general claim about all negative intuitions is that the two-step procedure will have the predicted results in all these cases. I agree it has that result in swamped good cases, but deny it does in pure bad cases.
Here is how I should put my case, I think:
A has a desire to eat a plate of glass. You say she has reason to smash the coffee pot, pour the glass shards onto a plate, and walk it back to her office. I give you the incredulous stare. You now tell me that it is only a very tiny reason, which is massively outweighed. I am now am a bit less incredulous and curious (as you predict!). But now you tell me what that reason is: it will allow her to eat the plate of glass shards! At this point I want to simply deny that this is a reason and swing back to thinking there is no reason for her to smash the pot, etc.
You are right that the presence of the harm is what is driving the reaction here, but I think that is exactly the right reaction to have. And I find myself thinking that your two-step procedure re-enforces that conviction. It does so because it ensures that I know what I am saying: I am disagreeing with you about whether something is a reason for you to act, and not simply being lead astray by *other, additional* reasons that are in play.
I do fear, there is a parting of the intuitions here. But I hope this brings out how my case differs from the Aunt Margaret case.
Hmmm. I didn’t mean to be building in that my Aunt Margaret would be frustrated by not achieving her goal of reconstructing the scene depicted on page 78 of the November 2001 Martha Stewart Living catalogue on Mars. If she is, we can build that into your case, too. And if that’s the only difference between the cases, then it seems more perspicuous just to consider two versions of the Aunt Margaret case, in one of which she would be frustrated and the other not – or two versions of your glass-eating case.
Similarly, suppose that I tell you that Aunt Margaret has a reason to build a spacecraft in her backyard. You seem incredulous, so I note that it is not a very good reason. And then I go on to tell you what it is: it will allow her to reconstruct the scene depicted on page 79 of the November 2001 Martha Stewart Living on Mars. This helps you out, but the glass answer doesn’t? I feel like there must be some further difference that you’re assuming between the cases.
Let me try and summarize where I think we agree and disagree: my argument isn’t supposed to show that negative existential intuitions are necessarily wrong – just that in cases in which reasons are weak and competing reasons are obvious and strong, they are at risk to be wrong. I’m understanding you as taking this point and suggesting that my pragmatic predictions can give us an alternative intuitive test that we can apply instead. My position, in contrast, is that we should rely less on intuitions, and look for other factors to decide the issue, instead. Once we see how easily intuitions can be led astray in this domain, I think it’s better to stay away from trying to pump them about cases at all, and look for other considerations which can decide for us how we should expect things to fall out.
Mark,
Thanks for the response(s).
Sounds like we are indeed pretty much in agreement – about how we disagree.
Good to know that you intended the Aunt case to be like Quinn’s radio-turning-on case – one in which there is no benefit to be gained such as evading frustration of pleasure. I do not think that the car case or Aunt case comes off that way – because it is easy to assume that getting the iron is good for the agent and that your aunt will get something out of building the Mars-lander – some pleasure or intellectual engagement, for example.
I am glad to know that you intend for there to me no good in the offing and will emphasis that the Aunt case is supposed to be like the Quinn case, where there is no temptation to assume benefits will flow from acting on the relevant desire.
On the substantive issue, I think we agree about where things stand. I think your argument shows that we have reason to not rely on some inutions, but rely on others, and that your test helps us separate the wheat form the chaff; you think the lesson is that we should not rely on intuitions at all. I should mention I also think we need to go on to consider more that intuitions, of course – hence my gestures towards an account of value.
I think each view is reasonable here, and should mention that, consequently, I do not think my argument provides a strong challenge to the reasonableness of your reductionist aspirations – although I will raise independent worries about that when I turn to the weighing (and correctness).
Brad,
I’ve come to this a bit late and I must confess I have not yet read Mark’s book, but there’s something I noticed from the beginning that seems to be coming out in the final comments between you and Mark and so I hope what I say here will be relevant.
You write: “A stereotypical Humean holds a bold view about reasons: if an agent has a desire and some action will promote the satisfaction of that desire, then there is reason for the agent to perform the action. But, the anti-Humean claims, there are actions that will promote the satisfaction of our desires but which there is no reason for us to perform.”
It seems like there are two ways of running this anti-Humean objection which track the differences between the car-eating and glass-eating cases: The anti-Humean might claim that not all desires are reason-providing. Alternatively, he might grant that all desires are reason-providing, but deny that they provide reasons for all actions sufficient to satisfy the desire in question.
This latter seems like what is going on the in the car-eating example. You can easily grant that my desire to get more iron in my diet provides me with some reason to act but deny that it provides me with a reason to eat my car. And this seems like precisely what Mark is talking about: it is a question of too many reasons being provided by a desire, not too many desires providing reasons. The glass-eating case, though, seems like it rests on the other objection, namely that the person’s alien-instilled desire to eat yellow shards of glass doesn’t provide him any reasons for action whatsoever.
Anyway, it seemed to me that this explains the relevant differences between the cases. Since I haven’t read the book, I don’t know what Mark says in defense of the idea that desires provide reasons in general, but it looks as though whatever he says to that end is what should be used to deal with your case in #2, because once we grant that the person’s desire to eat yellow shards of glass provides him some reason for action, then it looks as though there really is no difference between affirming or denying that he has reasons to eat his car in one case or open the closet door to get to the spray paint in the other.
Hi David,
Thanks for the comment. I think that might be a good way to lead into describing our disagreement. My main worry is that (I now think) the cases I tried to pick out are best identified by the harm or benefit that follows from the relevant action.
There *are* cases, e.g., in which the desire to eat a plate of glass would constitute a reason to eat it – imagine that someone is going to blow up the world unless you eat the plate of glass and your desire to eat is explained by awareness of that fact. This makes me reluctant to put the dispute in terms of general claims about desires instead of actions. But there are complications with the action way of putting it too.
I’ll have to think about how to best do this without needlessly complicating the review. Thanks for getting me to think about it!
Brad,
I’ll say a little bit more that I hope will be of use:
Certainly, cases like that (exploding world) make it such that we can’t simply stipulate which desires do and do not provide reasons for action without knowing more. But acknowledging this is a broader denial of Humeanism that must be separated from denying that Mark is right that once we acknowledge that a desire provides us with a reason, it might provide us with all sorts of strange reasons (like to eat a car).
You write: “But I maintain that my resistance is best explained by my beliefs about the relevant values and the thought that those value claims entail that there is no reason in favor AND lots of reasons against his acting – a crude slogan: no reason if no (value or perceived value).”
This seems right to me: Desires can provide us reasons to act only if they are the sorts of things that can count in favor, namely that they either are themselves or further things of value. But we don’t think that (non-outweighed) harms are of value, and so since some desires are for such harms, they are not themselves of value and do not further anything of value. Thus, we ought to deny the broad Humean thesis that desires are always (or even usually) reason-providing. But this is compatible with a part of Mark’s view, namely that once we acknowledge that a desire is reason-providing, we are committed to its providing reasons for all sorts of strange actions. Eating a car is probably never going to be a good idea, but since it furthers my desire to get more iron in my diet, which stipulatively we have agreed is of (either inherent or instrumental) value, we have some reason to do so.
What your exploding-world case does, I think, is to point out that many desires, if valuable at all, are going to be valuable only instrumentally, because we think we can only determine what reasons we have once we know the consequences of acting on the desire in question. Normally, the desire to eat glass is a desire for a pure harm, but not when the world is at stake.
Hi David,
Well put; sounds like we are in agreement. Thanks for suggesting a nice way to express what I should accept in Mark’s picture, along with his point about intuitions sometimes being misleading.
Brad