Some hope to cash out claims about goodness in terms of claims about
reasons to respond. One such account is the Buck Passing Account of
Value (BPV), which says that to be valuable is to have other properties
that provide
reasons to take up an attitude in favor of it. I am thinking about the
merits and originality of an objection to BPV. Consider the following
claim BR:
(BR) One can have a reason to take up some attitude, a reason not to
take up that same attitude, and one of these reasons is better than the
other.
BR is overwhelming plausible. But BR uses evaluative terms: it says that some reason is better than another reason. So, BR uses evaluative language. Can BPV be reconciled with BR?
It can if we can express the thought BR expresses by substituting some
nonevaluative term for "better". We might be tempted to go with
"stronger". But we are not talking about explanatory reasons here; we
are talking about justifying reasons. So "stronger" is at best merely
metaphorical. It won't do.
We might be tempted to rely on reasons-talk again. To say that reason
A is better than reason B, it might be suggested, is to say that we
have reason to do what reason A recommends rather than what reason B
recommends. Let's call this reason about reasons a meta-reason.
Will this solve the problem? I don't think so. Meta-reasons
themselves can be better or worse than other meta-reasons. To escape
this regress, we would need a reason that itself is not better or worse
than other reasons. I don't relish the taste of that bullet.
Alternatively, we could embrace the infinite regress. Yuk.
Instead, it seems better to reject BPV.
I'll delete this post if it merely rehearses a well-known (but unknown to me) objection.
Eric,
I’m not sure that I see the problem. Buck-passers want to say something like the following:
For all states of affairs p and q, it is better that p is the case than that q is the case if and only if the set of all the right kind of reasons to prefer its being the case that p to its being the case that q is weightier than the set of all the right kind of reasons to prefer its being the case that q to its being the case that p.
And you want to retort: “But to say that one reason is weightier than another is just to say that the one reason is better than the other.”
If that’s right, then the buck-passer is analyzing “a better state of affairs” in terms of “a better reason.” But why does this entail any sort of regress?
Thanks, Doug!
So a buck passing account of “better” may dodge the problem allegedly had by the buck passing account of “good”. Good point…I mean…that’s a better point than most.
To answer your question, I don’t think BR points to any problem for those who want to analyze “a better state of affairs” in terms of “a better reason”. But then we are still stuck with an evaluative term: better. We haven’t yet cashed out evaluative terms in reasons terms. What’s the next move?
The next move would be to account for ‘good’ in terms of ‘better’, and account for ‘better reason’ in terms of the strength of a reason and the strength of a reason in terms of the justifying and requiring roles that reasons play, a la Josh Gert. That’s just one possibility. I’m sure that there are others.
I actually don’t see what the problem is supposed to be. I know there is supposed to be a regress, but I don’t see it.
So, some reasons are better than others; this must be (according to BP) because there is reason to (place appropriate attitude verb here) some reasons rather than the others. And something similar is true for higher order reasons, and again for even higher order reasons. But this isn’t a regress. It’s just a general feature of all the elements in this infinite collection of reasons.
I have no stake in buck-passing, myself, but if I had I don’t think I’d be worried about this regress. Yet.
I’m not sure I see the problem either. To say that one reason is better than another seems to be to say that one consideration counts more in favour of an action than some other consideration. Or, in another words, that consideration can require an action more from from you than some other consideration. None of these expressions use evaluative terms and they seem to tell us what is going on.
I guess I’m puzzled why the talk about strength of reasons would be metaphorical (and, btw, what’s the difference to explanatory reasons). If reasons are things that are in some kind of normative, beind a reason for relations to actions and attitudes, why could not these relations be stronger or weaker normatively? One way to think about this is with the analogy of forces such as gravity. Masses are in relations of gravitational attraction which can vary in strength depending on the size of the masses. In the same way, one could think that pains can normatively speaking call for helping and alleviating. This is a relation in which the pains are to helping actions, and how much calling for is done depends on the seriousness of the pain and so on. Anyway, the point is that if reasons are like normative forces some of them seem to be able to be stronger, more requiring than others. If this is a basic feature of reasonhood, then we don’t seem to need evaluative terms to make sense of this.
Final point, another deontic way to make sense of one reason being better than another is to think that better reasons are the ones that explain what one ought to do (this is the Broomean line). Here oughts are basic and the strength or goodness of reasons amounts to to what degree the consideration explains what one ought to do.
Jamie, that helps. If I’ve read you right, then you do see that it is a infinite regress, but you don’t see why the regress is worrisome. I don’t think it’s worrisome for those who merely maintain the BPV is true. But it still seems to pose some sort of problem for those who want to use BPV to reduce evaluative discourse to reasons discourse.
Jussi, I’m not sure I follow your point. Is your proposal to understand claims like BR as
1) The first reason more requires some response than the second reason.
or
2) The first reason requires some more response than the second reason?
The second interpretation seems most plausible when the competing reasons concern commensurable properties (e.g. two pains).
The first interpretation leaves it unclear what degrees of requirement are.
I had the first reading in mind. If I have one painkiller and A is in a lot pain and and B in mild pain, A’s being in pain requires from me that I give her the pain killer more than B’s pain does. Her pain likewise counts more in favour of giving her the pill than B’s pain does. I’m not sure what unclear there is about such degrees of requiring forces. For instance, that there are such degrees is exemplified by our practices to criticised those who are not pulled by the stronger requirements.
Not only does “stronger reason” not seem to me to be metaphorical here, but talking about having a “better reason” to do A than B doesn’t seem like ordinary reasons discourse. I think that I would normally tend to say that I have “stronger” or perhaps “weightier” reasons to do one thing than another. The only time that I can hear myself talking about “better” reasons is when I am comparing different reasons that all ostensibly favor the same course of action. If a colleague makes an argument for why our department should follow some course of action, for instance, I might well agree with her about what we should do but think that there are better reasons for this decision than the one she offered.
Eric,
Okay, so what is the problem it poses for those who want to use BPV to reduce evaluative discourse to reasons discourse? That’s what I do not see.
To me it seems analogous to the following. Suppose somebody thinks he has a reduction of humor discourse to something else (anything, you name it). Now you pick a term in his reduction, a referring expression: t. And you point out that t could be funny, too.
As far as I can see, this would not in itself be any problem at all for the humor reduction. And if it isn’t, then why is the fact that reasons themselves can be better or worse a problem for the reduction of evaluative predicates to reason predicates?
Dale, I think you are right about the comparative prevalence of “better reason” talk. People do often say that one reason to V is better than another reason to V. I’ll take that as another datum supporting my argument.
Jamie, I think that’s an interesting analogy. Since I’m also skeptical about reducing what’s funny to our responses, my worry is not immediately extinguished.(Which is not to say that what’s funny can’t partially depend upon our responses.) But I grant that the analogous version of my objection doesn’t show what (if anything) is indeed wrong with the humor reduction.
One possibly relevant difference is that in your example, it just so happens that the reduction basis of what’s funny is itself funny. The reduction basis isn’t itself characterized in terms of funniness. But reasons, I think, are different. I don’t think one can really understand what a reason is without grasping the difference between a good reason and a bad reason, or a better reason and a worse reason. This is why I think that biconditionals like BPV, even if true, don’t reduce value to reasons. But it’s obvious to me that I need to spend more time thinking about what reduction is.
Jussi, I do feel like I’m in the dark about what “degrees of requiring force” means. We might think that the law requires me to X more than it requires me to Y, because the penalty for not-Xing is worse than the penalty for not-Ying. But this can’t be a model for us, because 1) there’s nothing analogous to a penalty in the reasons case, and 2) it relies on the very idea of one penalty being WORSE than another. Alternatively, we might think that the Federal government can require more from me than the local government can, and that this is the correct model to understand how reasons can require responses of different forces. But I don’t see how that would go.
Eric,
If you think that the requiring force of reasons is somehow mysterious, then you won’t understand 90% of reasons-literature where this is taken for granted a primitive. Of course, it might be that everyone else is talking about something which really doesn’t make sense but I would want to suggest that we should be charitable. Anyway, it does seem like we have come to an end of communication here if you think that that notion does not make sense. What worries me now more is that your argument does not pose a problem for the buck-passer if she is allowed her assumption of reasons having requiring strengths. So, it seems like your more objecting to that notion, which would be much more interesting objection to much more general views about reasons.
But, let me ask this? If you have a normative reason to do some action such as that someone is in pain, what is that you think the reason does in relation to your action? Does is support the action, justify the action, count in favour of the action, make the action what you ought to do, or what? If you specify some relation between the reason and your action (surely there must be some such relation), then ask yourself can that relation come in different strengths? I find it hard to image what the relation could be such that it couldn’t.
BTW, your example doesn’t requite taking being worse as more basic. Even there one might think that capital punishment is a stronger reason for me to avoid certain actions than a small fine is. In what sense stronger? Well, it counts more against certain actions than fines do. If this is mysterious, then it is.
I wonder if some of the confusion between Eric and Jussi is the result of an ambiguity. “Requiring force of reasons” can mean one of two things. If we say of two options A and B that the reasons in favor of A have greater requiring force than the reasons in favor of B, we might mean:
(1) that the reasons in favor of A require A to degree-X; the reasons in favor of B require B to degree-Y; and X is greater than Y. So, “requiring force” means force-which-yields-requirement, where greater force yields a “bigger” or “stronger” requirement. [Jussi’s comment at 9:32 suggests this reading, and Eric’s at 9:36 attributes it.]
Or we might mean: (2) that the reasons in favor of A count in favor of requiring A more than the reasons in favor of B count in favor of requiring B, and that therefore A is in fact required. On this view, “requiring force” is force-which-counts-in-favor-of-a-requirement, or (very roughly) force-which-would-require-in-the-absence-of-opposing-forces.
It seems to me that (1) is indeed incoherent, or at least mistaken. Being required is like being pregnant, in that it’s a binary property, so talking about degrees of requirement doesn’t make sense. But I think that most people writing in the reasons literature mean something like (2), and (2) doesn’t invoke the idea that some actions are more required than others. Prima facie requirements or pro tanto requirements aren’t actual requirements. The only thing actually required is the option supported by the greatest balance of pro tanto requirements. (I think that everything Jussi said is at least compatible with the (2) reading, right?)
Andrew,
I think that’s right, but I worry that the way your phrase (2) is slightly too complicted. I think you can make the same point by distinguishing between (doing) requiring on one hand and something being required or a requirement on the other. The first, requiring, is a notion on the contributory, pro tanto, level. And, it seems that a consideration can do that more or less. Of course, once you get all the requiring done, the balance of requiring creates a requirement or makes it the case that the action is required.
I’m still not seeing what is the problem with requiring things more or less strongly or seriously.
Another way to make (2) more simple would be just to talk about counting in favour of. On this view, a reason is better than some other reason if it counts in favour of A more than the other reason does for B. I think this is the best formulation and it is not evaluative. But, Eric doesn’t seem to be happy with that either.
I understand the thought that a reason to V counts in favor of V-ing, but to me “counts in favor of” seems, to borrow a term from Andrew, binary. I also understand the thought that one reason to V is better than some other reason to V or some third reason not to V. But I don’t fully understand the thought that a reason to V requires V-ing. Thomas Pink’s work on reasons and obligation has shaped my thought here. Today’s a nice day, and that is a reason for me to go for a walk outside. But the fact that it’s a nice day requires nothing of me, and this is not because I have better things to do.
And even if this all were clearer to me, I still need to think about and understand better what degrees of requirement means. Andrew’s comments here I find helpful.
Thanks for being patient with me. Time to go for a walk!
Good luck for the walk! One thing that now puzzles me is what does it mean for a reason to V be better than a reason not to V if the first reason counts just as much in favour of Ving as the second reason counts in favour of not Ving? I’m not sure I can understand betterness here if these reasons count just as much for and against for Ving and not V-ing.
And, it’s true that not all reasons are taken to be requiring. Some people think that the weaker reasons (it’s a nice day outside) can do is to entice to do actions. But, once we get strong enough reasons the normative force of reasons changes from enticing to requiring.
Jussi,
I’m not sure how significant this is — it may just be terminological — but I’d resist saying that ultimately defeated reasons “do requiring”. I get that so long as we distinguish that from being required, we don’t obviously commit ourselves to saying that two actions are required, but that one is “more required” than the other. But this kind of language isn’t what we use in other contexts. In a basketball game, points “contribute to winning” in something like the way pro tanto requirements contribute to requirements. But we wouldn’t say that if I score a basket I’ve “done some winning”. Rather, I’ve done something that could end up contributing to a victory. If my team finishes with less points, though, then I didn’t win, nor did I “do any winning”.
The reason I don’t like sticking with the simpler “counting in favor” locution is, as you note later, many philosophers (myself included) think that there are ways reasons can “count in favor of” that don’t involve requirements. I’d also resist the assumption that it’s the strength of reasons that makes the difference between requiring and non-requiring ones. (On some views, for example, it may be that reasons of certain types, or concerning certain domains are intrinsically incapable of yielding requirements.)
Eric,
“Counts in favor of” is binary in the sense that something either counts in favor of X or it doesn’t. But (at least to me) it doesn’t seem to be binary in the way pregnancy or requirement-status is. The latter two are purely on-off: once one is pregnant or once an action is required, that’s all there is to say about its pregnancy-status or requirement-status. But “counts in favor of” invites a scalar component. So, “point-yielding” is binary in that every basketball action either yields points or it doesn’t, but it’s nevertheless true that a 3-point basket yields more points (or is more point-yielding) than a foul shot. In the reason case, we could understand “counts more in favor of” in terms of the (hypothetical) reasons the reason in question could outweigh. So, to say that reason Q counts in favor of A more than reason R counts in favor of A is to say that there are more hypothetical situations (or, more hypothetical sets of reasons for not-A) in which A would be required when supported (only) by Q than in which it would be required when supported (only) by R. (There are of course lots of complications here, and I myself don’t like this type of analysis, but I think it’s roughly what a lot of people have in mind.)
Apologies for arriving late to this party, but…
Chapter 7 of Slaves of the Passions defends a ‘buck-passing’ account of what it is for one reason to be weightier than another. This account does have the feature that the weight of reasons depends on the weight of reasons to place weight on them, but there is no regress; it’s simply a recursive account. The details are complicated, but they’re there.
I wouldn’t identify ‘is a better reason than’ with ‘is a weightier reason than’, however; I think that’s like identifying ‘is a better can opener’ with ‘opens cans more efficiently’. I do think that one reason is better than another just in case it is weightier, but I don’t think it’s being a better reason is just it’s being a weightier reason, any more than being a better can opener is just opening cans more efficiently. In both cases, ‘better’ doesn’t just have a special meaning restricted to reasons or restricted to can-openers; it’s a compositional meaning generated by the meaning of ‘better K’ together with ‘reason’ or ‘can-opener’ stuck in for ‘K’.
I’ve also endorsed a buck-passing account of ‘better K’ as well, and I also believe that my account of ‘better K’ and my account of ‘weightier reason’ together predict that one reason is better than another just in case it is weightier. This, I believe, is the right result. (My account of ‘better K’ is in ‘Value and the Right Kind of Reason’, not in Slaves.)
Thanks for your help so far.
Maybe it would be good to figure out some way to test whether the goodness of reasons really is equivalent to the strength of reasons. Here’s a hunch. Suppose agent A has reason R1 to V, and also has a reason R2 to V, R2 being completely independent of R1. Let’s say that R1 is just as good a reason as R2. Let’s also say that R1 is just as strong a reason as R2. If R1 and R2 are of equal strength, and they are independent of each other, then it is plausible that the set of the two reasons is twice as strong as either reason individually. Strength seems to be additive in this way. But it isn’t _as_ plausible that the set of reasons R1 and R2 is twice as good as either reason individually. This may be most obvious in cases where either reason alone is good enough to be decisive. I’m tempted to flip through my Dancy books to find a good example to illustrate this point.
What would be the consequence of this? It doesn’t yet show that the goodness of reasons is basic. That goodness of reasons is related to decisiveness intuitively in this context would seem to suggest that we should understand the talk about good reasons in these locutions as referring to the sufficiency of the reasons. That two reasons as just as good seems to tell us that either one of them would be sufficient. This makes me wonder if we say that something is a good reason to x unless it is a sufficient reason to x. In any case, the buck-passer can make sense of sufficiency of reasons in non-evaluative terms by saying that sufficient reasons make it the case that one ought to do something.
And, Dancy’s book will also give plenty of cases where two just as strong reasons do not make twice as strong reason together.
Hi Eric,
I have not thoroughly read through the thread, but you seem to be arguing that a reason’s relative comparative quality can diverge from its relative strength.
A friendly suggestion popped into my head – maybe “wrong kind of reasons” cases will help you:
(1) A threatens B by saying he will kill B’s son unless B testifies that 2+2=4 when asked by C.
(2) B knows that 2+2=4
(3) A’s threat is a stronger reason for B to testify that 2+2=4, than his knowing that 2+2=4. But his knowledge is a better reason to testify, than the threat is.
I am not confident we should accept this, but thought you might be interested.
What do you think?