The great discussion on the previous posting got so much off
the topic that I thought I’d start another threat just on the interesting issue
we were discussing. This is Parfit’s account of state and object given reasons.
I’m sure we’ve touched in this in our virtual-reading group too but I’ve
forgotten what we concluded then.
Start from the desires for some things for their own sake.
That is, on occasion we desire something for a property which that thing has
and not because we have another, more fundamental desire for something else
that requires us getting the thing in question.
One view Parfit introduces would say that there are two
kinds of reasons for desires for things for their own sake. First would be the
object-given reasons. In this case, the object we desire has some property that
favours our desire for the object in question for the features it has. Second,
there are state-given reasons. So, for instance, when an evil dictator commands
me to want to be tortured for its own sake or she will actually torture me, you
might say that the state of affairs I am in gives me a reason to want to be
tortured for its own sake. In this case, the object, torture, of that desire to
be tortured for its being-torturedness does not give me a reason – the
favouring force for that state comes from elsewhere, the threat of the dictator.
I’m starting to warm up to this proposal but Parfit (and
Sven and Doug) argue against it. Parfit’s argument begins from the premise that
the reaction of rational agents to reasons is automatic and non-voluntary.
However, rational agents only react with the appropriate desires to
object-given reasons and not to the alleged state-given reasons. In a case
where a rational agent is given an option of eating a lifesaving pill, he will
automatically desire to take it in the same way as being aware of some decisive
evidence for a belief makes us directly have the belief in question. Yet,
rational agents do not react in the same way to state-given reasons. When a
rational agent is told about the dictator’s threat, she does not begin, by that
token, to desire to be tortured for its own sake. This is because torture
itself lacks the favouring features.
Parfit’s argument against the state-given reasons then is
that there is a ‘response requirement’ for reasons. In order for there to be a
reason, substantially rational agent must be able to react directly to it by
getting the appropriate desire on a basis of a belief about the reason. As we
saw, this does not happen in the dictator’s threat case, and therefore the
state I am does not give me a reason to want to be tortured for its own sake.
The same is supposed to go for all alleged state-given reasons, and therefore
there are no such things. Of course having the desire to be tortured in the
dictator case would be good for me. This feature of the desire gives me a
reason to want to have the desire to be tortured. This reason is something a
rational agent can respond to but it is not a state-given reason but rather an
object-given reason like all reasons.
My first problem with the argument is that it seems to beg
the question. If substantially rational agents are defined as those who desire
what they have reason to desire, then the claim that a substantially rational
agent does not react to the threat assumes that she has no reason to do so.
But, this was supposed to be the conclusion not the premise. If there is such a
state-given reason, then substantially rational agent reacts with the desire to
be tortured. Granted many of us would not probably be able to do, but this only
shows that we are not substantially rational.
I’ve also started to think that there is a bigger problem
looming which Parfit himself acknowledges elsewhere. It’s based on Korsgaard’s challenge
for Humeans in the “Normativity of Instrumental Reason”. The question is how
can I have reason to do something that is a means to an end that I have no
reason to have. Parfit makes the same challenge against the desire-based
accounts of reasons. According to those views we have reason to satisfy desires
even though fundamentally there are no reasons for the basic desires. Now, I
think Parfit’s object-given account of the dictator’s threat faces the same
problem. I’m told that I have no reason reason to desire being tortured at all.
I’m also told that I do still have a reason to desire to desire to be tortured.
Yet, in this situation then, I do have a reason for doing something (desiring a
desire) that is merely means for an end (desire to be tortured) I have no
reason to have. I’d rather say that we
have a reason to desire to be tortured for its own sake and having this reason
gives us a further reason to desire to have that desire. The latter reason just
is easier to respond to.
First, I’d like to apologize for re-posting what I just posted in the other thread here. Since Jussi just started this new topic while I was writing my last post for the other thread, and I am very interested in these questions about state-given reasons, my thought is that it would be alright and a better idea if I re-posted this material here. So, here goes.
Jussi,
While you seem to take it for granted that, if our having some attitude would be good, then we have reason to have this attitude, I myself feel no inclination to believe this. My sense, as I have said in the other thread, is that the pragmatic view which you embrace is based on a mistaken conflation of reasons for first-order attitudes and reasons to have certain second-order attitudes: attitudes about our own actual, or possible, present or future attitudes.
Consider these two possible chains of events:
(1) I come to have a telic desire to be tortured and am, therefore, not tortured by the evil despot
(2) the war ends, and the killing of innocent civilians ceases
I think that these two possible ways in which things could go are alike in how I have reason to want things to go in these ways. And, these are both good things. My having some desire is just like some war’s ending. It is a possible event which, because of its consequences, I have reason to want, and to try, to achieve. What may be confusing here is that, while wars cannot have reasons to end, we can have reasons to have desires. But, this should not lead us to conclude that, for this reason, we have reason to want to be tortured in this case. My having this desire is, in this case, more like the ending of some war. It is a possible event which, because it has desirable consequences, I have reason to want to achieve.
As these remarks suggest, it seems to me that the kind of pragmatism that you hold rests on an odd view about what it is to have reasons to have desires.
Here is another argument for the same conclusion. If we say that, in cases like the despot example, we have reason to want these good things, then we seem to be using ‘we have reason to x’ to mean the same thing as ‘it would be good if x’. But, these claims clearly seem to have different meanings. Therefore, in these examples, we don’t have reasons to have these desires which it would be good if we had.
Here is yet another argument. Our supposed state-given reasons to want to be tortured would be self-interested reasons. We have some self-interested reason to want something if this thing would (i) promote our well-being, or (ii) be one of the things our well-being consists in. Our being tortured would neither promote our well-being, nor be one of the things our well-being consists in. Therefore, we have no self-interested reason to want to be tortured. Therefore, since our supposed state-given reasons to want to be tortured in the despot example would have to be self-interested, we have no reason to want to be tortured. But, our having a telic desire to be tortured would, in this case, promote our well-being. Therefore, we have self-interested reason to want to have this telic desire to be tortured.
Lastly, and as before, thanks for responding to, and challenging, my posted arguments.
Jussi,
You write,
How is desiring to desire that P a means to acquiring a desire that P? Desiring to desire that P doesn’t have the effect of desiring that P. I wish it did! If it did, then my desiring to desire to exercise might get me to desire to exercise. But it doesn’t.
Now suppose that I have a pill that will get me to desire to be tortured. Taking the pill is a means to an end: that of preventing the evil dictator from torturing me and thus avoiding pain. This end is an end that I have reason to have. So I don’t see that you’ve shown that Parfit’s view implies that I do have a reason for intentionally and voluntarily doing something that is merely means for an end I have no reason to desire. Desiring to desire to be tortured is not a case of your intentionally and voluntarily doing something. Taking a pill so as to acquire a desire that will shield you from punishment from an evil dictator is a case of doing, but here it is a means to an end (avoiding pain) that you do have reason to desire.
Jussi,
I’m interested in just how much common ground there is between us. So please tell me which of any of the following claims that you reject and which of any of the following claims that you accept:
(1) The only way that I can respond to the fact that the evil demon has threatened to torture me if I don’t desire to be tortured for its own sake is (1) to desire to desire to be tortured for its own sake and (2) to intend to do whatever will make it more likely that I desire to be tortured for its own sake. Beyond (1) and (2), there is no other way for me to respond to this fact.
(2) Given the fact that the evil demon has threatened to torture me, I have both a reason to desire to desire to be tortured for its own sake and a reason to intend to do whatever will make it more likely that I desire to be tortured for its own sake.
(3) In order for something to be a (normative) reason to X, it must be possible for it to be my (explanatory) reason for X-ing provided that I’m perfectly rational (both substantively and procedurally).
(4) In order for some fact to count as my (explanatory) reason for X-ing, it must be that I can respond directly to that fact by X-ing.
WOW – Doug and Sven. There’s so much to respond here. I’m not sure I can do it all. But, I won’t give up. I think I’m going to start with Doug’s posts and the go on to Sven.
Doug,
About the first post. If the desire for a desire to be tortured is not a part of the means to come to have the desire to be tortured, I don’t see what reason we would have for that second-order desire. It might be our reaction in that case, but what would actually speak in favour of it, if it doesn’t get us any closer to the first order desire that would be beneficial for us?
About:
(1) What sense of ‘can’ do you mean here? It doesn’t seem logically impossible to respond to the threat with the right desire. Might be psychologically difficult. There is a difference to the belief case. Beliefs aim at truth and therefore they must be only reactive to evidence. Desires are not like this. Nothing in desires conceptually implies that they must be responsive to only object-given reasons. So, there seems to exist a third way of responding to the reason. I doubt the reasons for having the second-order desire response and the intention to use the means to get to first-order desire seems to require a reason for the end – the first-order desire, as Korsgaard argued.
(2) As I say above, the threat itself does not seem to be a good reason for the desire to desire to be tortured. The second-order desire will not get you out of being tortured and you say that it is not a means to the first-order desire that would. Why would we then have the second-order desire?
(3)Yes, but if there is such a reason, then substantially rational agent reacts to it with the appropriate first-order desire. Now, *you and I* may not be able to do this, but this shows that we are not substantively rational in that case. I’m not sure why this is a problem. There are other gambling cases where almost all actual persons are substantively irrational.
(4) Even more – in order for some reason to count as my explanatory reason, I must already have responded to it. I don’t see the point though.
Sven,
first, I don’t see why we should say that the desire to be tortured is a possible event I should desire to want to happen. Rather, it being tortured is what I should desire in the case. Desires are not events in any case but mental states. And, they don’t usually happen – rather you actively desire things. Some things there are more reasons to desire than others. In this case there are perfectly good reasons to desire being tortured. I don’t see the problem. And, by saying that we have a reason to desire being tortured, I don’t definitively mean that ‘it would be good’. Rather it being good for me, gives me a reason to desire being tortured.
About this:
“Our supposed state-given reasons to want to be tortured would be self-interested reasons. We have some self-interested reason to want something if this thing would (i) promote our well-being, or (ii) be one of the things our well-being consists in.”
I don’t see the reason to put *this thing* to the second sentence. That just assumes that there are only object given reasons. But, you cannot assume that in your argument against object-given reasons. Rather, a neutral way of putting the sentence would be:
‘We have some self-interested reason to want something, if either wanting that thing or the satisfaction of that desire was good for us.’
And, then the argument does not follow.
Jussi,
You write: “If the desire for a desire to be tortured is not a part of the means to come to have the desire to be tortured, I don’t see what reason we would have for that second-order desire. It might be our reaction in that case, but what would actually speak in favour of it, if it doesn’t get us any closer to the first order desire that would be beneficial for us?”
Is it your view, then, that there is a reason to desire that P only if desiring that P is a means to some good end (such as self-benefit)? That seems to be what you are suggesting. But isn’t what’s good just what we have reason to desire? So to say that self-benefit is good for its own sake is to say that it is something we have reason to desire independent of whether or not desiring it has any good effects. In which case, don’t you have to accept that there could be a reason to desire something without it being a means to some good end?
Regarding (1), it seems impossible (I’m not sure in what sense) for me to respond to the evil demon’s threat by desiring to be tortured for its own sake. How could one respond to such a threat by desiring to be tortured for its own sake and thus apart from its being a means of avoiding what is being threatened? In what sense, would I be responding to the threat by desiring to be tortured independent of whether I’m being threatened? It just seems pychologically impossible for me to desire to be tortured for its own sake as result of some threat if I don’t. I can’t see how such a threat would move me to desire to be tortured for is own sake. I can only see being moved to desire to have this first-order desire and to do what I can to acquire this first-order desire.
Regarding (2), I have a reason to desire to desire to be tortured, because the object of this second-order desire is desire-worthy, something that is good as a means to avoiding the evil demon’s threat.
The point is that if you can’t respond to the threat by desiring to be tortured for its own sake, then, given (4), it can’t ever figure as the explanation for why you desire to be tortured for its own sake if you do, and so, given (3), it can’t be a normative reason for you to desire to be tortured for its own sake.
Doug,
regarding to the first point I would not accept the ‘only if’. But, it seems like *a* good reason to desire something that that desires is a means to some good end. What I wanted to know was what you thought was the good reason to have the desire to desire to be tortured. Only way I could think of to rationalise that desire was the instrumental story. If you reject that one, then I’d like to hear an alternative. And, in many cases I accept that there are such alternative stories – I just couldn’t imagine one here for the second-order desire. But, you too seem to accept the instrumental story so more of it below.
I think in (1) you seem to confuse the content of the desire, the reasons to adopt the desire with that content, and the causal story of how having that attitude comes about. Of course, without the threat you would not have a reason to adopt the desire [be tortured because of what torture is like]. But, in this case you do. Adopting that attitude for the threat-provided reasons would be responding to the threat. But, doing so for that reason doesn’t change the content of the desire though if you succeed in getting yourself to that state.
I still can’t make sense of the direct response requirement. I just can’t accept that there are no reasons for actions for completion of which you need to do something else first. That’s just plain unintuitive. But, there we are going in circles. This however is very revealing. You write that:
“Regarding (2), I have a reason to desire to desire to be tortured, because the object of this second-order desire is desire-worthy, something that is good as a means to avoiding the evil demon’s threat.
Two points. One, the object of the second-order desire is the first-order desire to be tortured. Your claim then implies that the first-order desire is ‘desire-worthy’. I agree. But, how can something be worth of desiring unless you have a reason to desire that thing? What explains the worth of that desire unless it is being favoured by some consideration, i.e., a reason? Parfit cannot say that it is favoured by anything. But, then the desireworthiness is a mystery too.
And, you say yourself what that reason is – it is a means to a good end to avoid the threat. That final reason of avoiding the threat gives you a derivative reason to desire being tortured for its own sake.
The (4) is still confusing. If you *do* desire to be tortured in the case like you say, then surely you *have responded* to the reason you have for wanting to be tortured. And, we would have the explanation and the normative reason. Actuality would guarantee possibility.
Jussi,
I’m afraid that I don’t have time to respond to all of what you wrote. But here’s something quick. You write, “If you *do* desire to be tortured in the case like you say, then surely you *have responded* to the reason you have for wanting to be tortured.”
Now you also think that my three-year-old daughter (only 36 inches) has a reason to be taller, right? Suppose she does grow taller. Does it follow that she has responded to the reason she has to be taller? No, right? But then that quote of yours above is false.
Anyway, thanks for interesting discussion. After reading your response, I’ll have to move on to other things.
Doug,
that’s good. I think it depends on the mechanism that response to reasons. In the case of your daughter, it seem like she would grow whether or not she had a reason to grow. In that case, she doesn’t have a reason-responsive mechanism that response to height reasons.
The other case seems different. In that case if you *do* have the right desire as a result of the threat and you would not have had that desire without the threat, then your having the desire seems to show that you do have some mechanism that responds to evil-dictator threat given reasons. And that seems to show that it is possible for you to respond to the threat in the right way. And, that the explanatory reason is the right one that can also be the normative reason.
I accept that if you do have a desire to be tortured for its own sake as a result of your deliberations concerning the evil demon’s threat, then that shows that you have responded to the evil demon’s threat by coming to possess the desire to be tortured for its own sake. But even supposing that you do have a desire to be tortured for its own sake (and thus you desire to be tortured independently of whether or not having such a desire is a means to avoiding the demon’s threat), I just don’t see how you could have come to possess this desire as a result of your deliberations concerning the demon’s threat. That just seems (psychologically) impossible, at least to me.
Here’s another way of making the same point. Let “X” be “to be tortured,” and suppose that my wanting X for its own sake would have certain good effects. How could contemplating the good effects that my wanting X for its own sake would have lead me to want X for its own sake? After all, how could my contemplating the good effects of anything lead me to want something for its own sake.
To be clear, I can imagine that things might be such that the physical state that underlies one mental state causes the physical state that realizes some other mental state. Maybe the physical phenomena on which the one mental state supervenes on cause the physical phenomena that realize the other mental state. But I take it that responding to some fact that provides me with reason to X by X-ing involves something more robust than just underlying physical states causing each other. After all, we could imagine that God sets things up such that some short person contemplating the benefits that would come with her being taller causes her to grow through some purely physical and completely unconscious process. But I wouldn’t think that would be a case of her responding to reasons.
I don’t see the impossibility of beginning to want something for its own sake from thinking about the good consequences. Friendship is a good example. I can think of the good consequences for me from being a friend to others and by doing so form a desire to be friends with them. It does not follow that when I have formed that desire to be friends it is a desire to be friends merely for the sake of the consequences. I do want to be friend for the reasons by friends and the relationship to them provide even though this desire is got through thinking about the consequences.
Much of such reason-responsive mechanisms do not work through explicit reasoning. Arpaly is good on this. How can the alcoholic become a converted Christian who believes in God? Certainly not through weighting evidence and the object given reasons in deliberation. Rather, she reacts to the state-given reasons for the belief that is beneficial for her by forming the right belief in some other way.
I take it that the disagreement comes down to the direct response requirement. Here’s how I can form the required desire to be tortured for its own sake. I recognise that I have such a reason for the desire. Unfortunately, this does not get me to the required state by itself. So, I try to brain-wash myself into forming that desire. Even though self-delusion is a difficult art, this seems very possible to me. Then I have the required desire to be tortured for its own sake. Have I responded to the reason I have by acquiring the right desire? Yes. The desire was even sensitive to my judgments about reasons albeit in an indirect way. Have I *directly* responded to the reason by adopting the desire? Maybe not. You say that this implies that I didn’t have the reason in the first place. I don’t see much that could be said for that move. It just proves way too much – that there are no reasons for most actions at all.
Jussi,
You write,
Note that I never suggested that it was impossible to do certain things that make it more likely that you’ll have this desire. What I claimed was that it’s impossible to come to have the desire in question simply by contemplating the fact that the demon will torture you if you don’t {without doing anything else like trying to brain-wash yourself).
So I would agree with everything you said in the above quote if only you replaced the second and third sentence with: ‘I recognize that I have a reason to do what will make it more likely that I have this desire’.
So there really isn’t that much disagreement between us, I think. You agree that recognizing the fact that desiring to be tortured for its own sake “does not get me to the required state by itself.” So the only way to respond to this fact (and this is to respond indirectly) is for me to try to brain-wash myself or to take a desiring-inducing pill or to do whatever else might make it more likely that I have the desire. Now, surely, you also agree that the fact that brain-washing myself would have good consequences is a reason to so act. So what’s the point of saying that in addition to the (object-given) reason that I have to brain-wash myself, I also have a (state-given) reason to desire to be tortured for its own sake, which I can respond to indirectly by brain-washing myself? Why postulate this extra reason? Do I have more reason to brain-wash myself in this case than in the case where the evil demon threatened to torture me if I didn’t try to brain-wash myself into desiring to be tortured for its own sake? It seems to me that you’re committed to saying, “Yes.” For you think in addition to the object-given reasons, there is this state-given reason. And this seems implausible to me. You have no more reason to do what will make it more likely that you desire X in the case where that demon threatens to torture you if you don’t desire X than you do in the case where the demon threatens to torture you if you don’t do what will make it more likely that you desire X.
I would say that the object-given reason to brain-wash myself is the very same reason as the state-given reason that I have for the desiring to be tortured for its own sake. Avoiding torture favours both the desire itself and any means I have for getting myself to have that desire. No extra reasons added. I avoid awkward consequence that I have reasons to get myself to a state I have no reason to be in. And, the consequence of not being able to say that I have reasons to act. That too seems like a heavy prize to pay. You can’t say in the case above that I have a reason to brain-wash myself but only that I have a reason to want to brain-wash myself.
Okay, I think that we’ve made a lot of progress, but now the dispute between you and me (and Parfit as well) just seems to be a verbal one. We agree on what reasons there are and how we are to respond to them. We just disagree on what to call them. What you call a state-given reason to desire to be tortured for its own sake, we call an object-given reason to intend to do what makes it more likely that I have that desire. And, regardless of what we call this reason, we agree that the way to respond to this reason is to intend to do what makes it more likely that we will have this desire. Now the way to resolve a verbal dispute is to appeal to common usage, and thus you suggest that, on my/Parfit’s view, we can’t say certain things such as I have a reason to be healthier and that this is what explains why I have a reason to do what will result in my being in a healthier state.
Is this a fair characterization of where things stand?
Just to get things clear in my own mind, I wanted to articulate more formally what my argument is.
(I) I cannot directly respond to the fact that the evil demon has threatened to torture me if I don’t desire to be tortured for its own sake by desiring to be tortured for its own sake.
(II) In order for some fact to count as my (explanatory) reason for X-ing, it must be that I have responded directly to that fact by X-ing.
(III) So, the fact that an evil demon has threatened to torture me if I don’t desire to be tortured for its own sake could never count as my (explanatory) reason for desiring to be tortured for its own sake.
(IV) If some fact could never count as my (explanatory) reason for X-ing (not even when I’m perfectly rational — both substantively and procedurally), then it is not a normative reason for me to X.
(V) Therefore, the fact that an evil demon has threatened to torture me if I don’t desire to be tortured for its own sake is not a normative reason for me to desire to be tortured for its own sake
You’ve said previously that you accept (II) and (IV), which were previously numbered (4) and (3), respectively. You seem to accept (I), as you’ve said: “I recognise that I have such a reason for the desire. Unfortunately, this does not get me to the required state by itself.” (III) follows from (I) and (II). So you should accept (V). But I’m guessing that you reject (II), right? But what would you accept instead? Perhaps this: “In order for some fact to count as my (explanatory) reason for X-ing, it must be that I have responded directly or indirectly to that fact either by X-ing or by Y-ing, where Y-ing leads to my X-ing.”