Some Of Our Books


« Program for the 2012 Arizona Workshop on Normative Ethics | Main | Two studentships in philosophy »

November 02, 2011


Feed You can follow this conversation by subscribing to the comment feed for this post.

My intuition is that the fact that performing x would have beneficial effect B1 is indeed a reason for you to perform x&y&z (it is outweighed).

Suppose one act, w, would have all three effects. Doesn't it seem like the fact that it has effect B1 is an outweighed reason to do it? If so, why should things be different just because your case involves a conjunctive act?

I do think this is slightly tricky. It might depend on more specific features of the case. E.g., if B1 and D1 and D2 are very closely related, my verdict could change.

Suppose one act, w, would have all three effects. Doesn't it seem like the fact that it has effect B1 is an outweighed reason to do it? If so, why should things be different just because your case involves a conjunctive act?

Good question. I don't have a good answer, but intuitively they seem different to me. Perhaps, my intuitions are idiosyncratic.

In any case, let me try a little more intuition pumping. I may have a reason to eat the leftover pizza in the fridge, but from this does it follow that I have a reason to perform any conjunctive act that both includes my eating the leftover pizza and doesn't negate the benefit that is to be had by my eating the leftover pizza? I'm inclined to say 'no'. I don't have, say, a reason to perform the conjunctive act of eating the pizza and then killing my entire family? Why would the fact that my eating the pizza would curb my hunger give me a reason to both eat the pizza and kill my family? It seems that gives me a reason to do the former but not a reason to do both.

So maybe the difference is that, in my case, B1 can be had without performing x&y&z but, in your case, (where w produces B1, D1, and D2) B1 can be had only by performing w. But, of course, this is more of just a statement of the difference between conjunctive and non-conjunctive act cases than explanation of why the difference matters.


What do you think of the following principle?

The fact that performing x&y would have effect E1 could constitute a reason to perform x&y (that is, a reason to perform both x and y) only if the same effect, E1, could not be just as well achieved by performing either x or y alone.

Hi Doug,

I guess my intuitions differ from yours here. I don't have anything more interesting to say; I'm just adding a data point, I guess. I like the "reason to perform the conjunctive act, but outweighed" view.

Thanks, Dale. It's not looking good for me.

Best, Doug
Sent from my iPhone

Hi Doug,
I'm inclined to cite features of the conjunctive act as the reasons to do it. So instead of saying "That B1 would bring you 19 utiles on Wed is a reason to do some (any?) conjunctive act that includes B1" I would say "That the conjunctive act would bring you 19 utiles on Wed is a reason to do it." Putting it the former way is risky because a conjunctive act can silence or reverse the reason you would have to do some singular act by itself, without the other acts in the conjunction. E.g., the fact that drinking red wine would taste yummy is a reason to drink it, but not a reason to do any conjunctive act involving it, like drinking the wine and eating artichokes (which completely alters the taste of the wine). Could this be the source of your intuition?

Once we cite features of the conjunctive act that are reasons, it's harder to see the motivation for your position. If your original is case 1, and case 2 is like your original save B1 brings 21 utiles on wednesday, why you have *no* reason to do the conjunctive act in case 1, where it would bring you 19 utiles on Wed, but some reason to do the conjunctive act in case 2, viz., that it would bring you 21 utiles on Wed? Better to posit some reason in each case and let the weights sort things out. I think.

Doug, I probably don't think of acts the same way you do. As I think of them, there isn't really any such thing as a conjunctive act. An act description can be conjunctive, but not an act. Acts can be more or less specific, and you can make an act more specific by adding a conjunct to the description of it. But you haven't made the act itself conjunctive.

I don't have a well-worked-out theory of any of this, so it's possible that the way I think about it has big problems.

To a philosophical amateur of highly questionable competence such as myself...

It would seem that the strict summation of benefits associated with each element of the conjunctive comprises the “reason to do some (any?) conjunctive act …” The accumulation of negative utiles from ‘component’ acts (D1 + D2) form the reasons not to take the conjunctive act.

Humanly speaking, it seems we run into two flavors of problem. First, even though the scenario was posed as “assume that these are the only effects of these actions and that the timing of the effects is irrelevant,” we still tend to cognitively weigh the earlier positive heavier than the later negative. This looks like a form of the availability heuristic that creeps up so often in studies of decision making and cognitive bias. The benefits of X sound so good, we really want them to triumph over the downsides of D1 and D2 that are also a part of the conjunctive act.

Second, as the illustrious Doug Portmore seemed to imply, the difference between the conjunctive act and the non-conjunctive act (and treating the net utiles of each act separately) is a tricky one for our minds. We want to be able to separate the positive yielding component act (X giving us B1) from its negative siblings. Even though the definition of the conjunctive says they all come together or not at all, I believe we mentally start wishing/seeking for a disjunction that would allow us to separate their consequences (reminiscent of the fable that concludes ‘and maybe the horse will learn to sing’).
We are time-oriented sequential and experiential for the most part, and it is tough to move away from that.

This points back to the challenge of the utilitarian premise. If our value structure has truly been normalized (so that 1 positive utile is exactly = to 1 negative utile), the net value leaves us stuck with forsaking the highly desirable X because of its inescapable Y and Z siblings-in-conjunction. But suppose the values were reversed (leading to a net gain), then if we have a value structure that forbade us to ever dip below an even level. We would have to forgo the good net (of +1 in this case) for fear that the unknowably sequenced X+Y+Z might occur in a way that caused us to have a temporarily negative balance.

Thanks for the thought provoking post. Please excuse my comments if they are in any way unwanted.

Doug - "Why would the fact that my eating the pizza would curb my hunger give me a reason to both eat the pizza and kill my family? It seems that gives me a reason to do the former but not a reason to do both."

I wonder if Gricean implicature is influencing your intuition here? Suppose I explicitly cancel the implicature as follows: "You have a reason to [eat the pizza and kill your family], namely that by doing so you would get a tasty meal -- but of course you have no reason to kill your family. The aforementioned reason is just as well served by [eating the pizza and leaving your family alive], which is a more advisable course of action in every respect."

Does it still sound to you like I said anything false in that spiel?

P.S. Like Jamie I'm not sure that there are conjunctive acts, in which case perhaps this is merely a terminological question of whether "reason to do [A and B]" should be understood to mean "reason to do some act in the sequence [A, B]" or "reason to do every act in the sequence [A, B]". Could you say more about what turns on this?


I share your intuitions in some cases, but I think, like Jamie, I'm having a bit of trouble with conjunctive acts. Also like Jamie, I do not have a fully worked-out theory about this, so what follows is going to be pretty rough and exploratory, but perhaps it will be useful.

Presumably we want to individuate acts by intentions. But I'm not sure it makes sense to talk about merely conjunctive intentions. In most cases if I say of you that you have an intention [to eat pizza and kill your family], I don't think I've said anything beyond that you have an intention to eat pizza and that you have an intention to kill your family. So the fact that it would be tasty is a reason to eat pizza, but it's not a reason to eat pizza and kill your family, because in order to be so it would have to be that the pizza's tastiness was a reason to eat the pizza and a reason to kill your family, and thus a reason to form both intentions. But obviously it's not.

Now, I said that I don't think there are merely conjunctive intentions. But I do think there might be conjunctive intentions of a derivative sort. For instance, suppose you have a conditional intention to kill your family if you eat pizza. In that case, it seems to me that were you to form an intention to eat pizza, you would thereby form an intention to kill your family. In that case, your having an intention to [eat pizza and kill your family] is something over and above the two intentions, but that's because there's an antecedent link between them. And in that case, I think I would say that the pizza's tastiness is a (heavily outweighed) reason to eat the pizza and kill your family.

Hi Everyone,

Thanks for talking me out my strange intuition. It has been successfully exorcized. I think that Richard is probably right about the source of the intuition. And I accept Jamie and David's point about conjunctive acts.

Hi Doug,

sorry for being late on this. I guess I am a holist about these things. Given all the enabling and disabling conditions that can be present, I'm fairly certain that we can create cases in which the intuitions are the same as your initial ones and others where they go the other way (if we try to ignore the worries about conjunctive acts).

Here are two cases:

a) beer drinking
Imagine that if I drink one beer, I get nicely happy and not experience a hangover tomorrow (this is act a, and +19 utils). Imagine that, if I drink a second and a third beer, I don't get any happier but rather get a hangover (two beers - a short one, three beers - a longer one (ie., -10+(-10) utils). Do I have a reason to do the act of drinking three beers (actually this is somewhat plausible case of a conjunctive act)? It's not going to make me any happier or have any other positive results. Yet, my intuition in this case is that I still do have some reason to do that act of drinking three beers. I might have less reason to do so than just to drink one beer. So, in this concrete case, my intuition seems to be different even if the case has the right structure.

b)Pain killers
I have a headache that is going to last for 10 hours if I do nothing. Taking one pain killer will make this go away (+19 u). Taking pain killers 2 and 3 will also help the headache to go away for the same 10 hours but they will then make my head hurt again just as much for the next 11 hours after that 10 hours (-20 u together).

Do I have any reason to take three pain killers in this case (again not that implausible case of a conjunctive act)? My intuition here is no - the same as your original intuition.

So, it seems like there are different enablers and disablers present in these cases. It seems that in some cases the reason remains but not so in others. I acknowledge that these intuitions too are a bit shaky and people might have different ones.

Hi Jussi,

My previous intuition concerned cases in which the effects of x, y, and z are all independent of each other. I guess this just shows how crazy my previous intuition was. Your cases are cases in which the effects that x has depends on whether y and z are also performed. So I think that your cases are plausible, but not what I had in mind. I think that Jamie was eluding to your type of case when he wrote: "I do think this is slightly tricky. It might depend on more specific features of the case. E.g., if B1 and D1 and D2 are very closely related, my verdict could change."


well, I tried to formulate the cases so that the consequences of x does not depend on whether y and z are also performed but maybe I failed.

At least consequences of x are independent of y and z. In the first case, the consequence of me drinking one beer is same whether or not I drink the second and third beers - I'll just be happy for the evening. Likewise, in the second case, the consequences of taking the first pill are the same - next ten hours pain-free whether or not I take the second and the third pill.

We could also make the consequences of y and z independent of x in these cases. Maybe y and z are special beers that would not make me happy by themselves if I hadn't drank x, but just give me a hangover. Maybe pills y and z would be lousy pain killers that, by themselves, only extended my headache from hours 10 to 21. I think my intuitions remain the same.

Hi Jussi,

If y and z are special beers that would not make me happy by themselves if I hadn't drank x, then doesn't the effects of y and z depend on x (and its effect of enabling y and z to make me happy)? Of course, I wasn't sufficiently clear, but I was thinking of cases like eating pizza Monday and killing my family later that week on Friday. Eating pizza on Monday doesn't alter the effects of killing my family on Friday. (Or so I'll assume. Assume, then, that the pizza is completely out of my system by Friday so that when I throw up after killing my family, what I throw up is not affected by what I ate on Monday.) Also killing my family on Friday doesn't alter the effects that my eating pizza on Monday had.


I share your difference in intuitions at first pass, but only because (I think) it's natural for me to imagine cases in which "drinking three beers" is a derivatively conjunctive intention (a la my comment above). For instance, I might know that I am disposed such that when I have one drink I tend to have three, and so the intention to drink one beer is relevantly linked with the intentions to drink the others such that it makes sense to talk about my having a single intention to drink three beers.

The same is not true in the pill case because I am not familiar with the disposition to take three pills if I take one (i.e., it's not natural to imagine someone's having this disposition), and thus I tend to hear "intention to take three pills" as merely conjuctive—as no more than a description of three separate intentions to take a single pill.

This seems to me to explain the difference in intuitions. And, indeed, if I force myself to think of the beer case as like the pill case—if I imagine that it is as open to me to drink just the one beer as it is to drink the three because I have no relevant linking disposition—then the intuitions line up and it then seems to me that the happiness produced by the first beer is no reason at all to drink three (again, I propose, because it will not be a reason to take each act of drinking a beer).

At first, my intuition was different from Doug's. But after thinking about it for a bit, I share Doug's intuition. (Since my intuitions tend to be unusual, Doug probably won't feel much comfort.)

But if there is some action description that aggregates the specific acts x,y, and z, such that in normal circumstances to A just is to x,y, and z, then I think you do have some reason to A. E.g, if A-ing is having the prix fixe menu, wherein the appetizer is fantastic but the last two courses are awful. That the appetizer is fantastic is indeed some reason to have the tasting menu.

But if the three acts are completely independent of another, e.g. x=scratch an itch, y=watch American Idol, and z=talk to my neighbor...then I have no reason to do x&y& fact, I'm not even sure I understand what such a reason could be.

Eric, when you say "if there is some action description that aggregates...", it sounds like you're saying that whether there is a reason for the conjunction depends on something linguistic -- as if it might turn out that in English there is a reason for the conjunction, but not in Japanese. This seems doubtful! And probably not what you meant, since the contrast is "if the three acts are completely independent of [one] another," which suggests that it's not linguistic. I have my doubts that "independent" can be cashed out in a plausible way, though.

Oh, and I'm sure you do understand what a reason to scratch an itch and watch American Idol and talk to your neighbor would be! You just don't think you have any such reason.

I share Jamie's worry about thinking any combo of acts can be amalgamated into one conjunctive act. But still, although it is questionable whether there is a single act of "scratch an itch and watch American Idol and talk to your neighbor" under ordinary circumstances, there could be a reason for one if there were such a conjunctive act: you get an award exactly if you meet the odd combo specified. And you might do all three at once: talk to your dance partner while scratching his hard-to-reach itch while you both watch the show together. (You want him to concentrate on the dance moves to learn them for what would be a rewarding upcoming joint performance, but you know he will be too distracted by the itch he can't reach just as the move begins unless you scratch it while you direct him to watch closely and remind him why.) And we do in teaching children or dance classes, teach one element and then tell people to do both at once.
Positing conjunctive acts that are simultaneous might be deemed more plausible than temporally scattered acts treated as amalgamated into one act merely by fiat, and you can get many of the same utility effects at a single moment by having switches controlled by different hands, feet, fingers, toes, etc. I'll switch to this framework in the next post. Ultimately, I don't think the underlying issue hinges on conjunctive acts per se.

Some support for Doug's leanings and perhaps a generalization of the issue:
I will ignore sequencing and talk of sets of switches flippable at a given moment. Is the fact that flipping switches x&y&z at once will have the same impacts specified, and the impacts are independent as specified mean that the favorable impact flipping x has provides me with a reason for flipping x&y&z? I think it is not obvious that the only plausible route is to speak of transfer of reasons to conjunctions and cover the rest by defeat, and there is something that pulls in the other direction. It seems strong to say that the fact that flipping x will have a good outcome gives me a reason for doing *each* thing available to me that includes x? Suppose I have no clue what impact, if any, y and z will have (in isolation or conjunction), it is not obvious that the fact that x will have a good impact will provide me with a reason to *do* x&y&z. Don't I need more reason for doing not only x but x,y and z than I now have for doing x if all I have is the positive impact of x? Isn't the broader underlying question at the level of reasons and guidance: if I have a reason for doing x, then do I thereby have a reason for doing each thing that includes doing x? Why not say if I have a reason for conducting myself in some way that includes my doing x, but not thereby a reason for each such x-inclusive way of conducting myself? Seen yet another way, why not say I have a reason for doing x, but not a reason for doing x&y else I would have a reason for doing x&y rather than not doing x&y, which latter I can fail to do by simply doing x and not doing y, which is as good an option as any for all I seem to have reason to do? There seems to be a difference at least at this comparative level that needs marking as we move to the various conjunctions, whether it is Doug's way of marking it or not. The favorable impact of x directs us to do something that includes x, but does it direct us to do each of the things that includes x even if it demarcates something favorable about each? Should it so direct us if it is a reason to do the conjunction?
The answer might be that in order to do what you have reason to do, namely flip x, you need to either flip x and not y, or flip x and y (assume only two switches), and so that fact--to avoid paralysis, gives you a reason to do x and y (and a reason to do x and not y), but perhaps not a reason to flip x and y rather than not flip both x and y (not flip x or not flip y or both); whereas the reason given to do x is is also a reason to do x rather than not do x, and in that respect bears a stronger relationship to doing x.

Hi Paul,

Thanks for these interesting comments. I, of course, share the intuition that "It seems strong [wrong or too strong] to say that the fact that flipping x will have a good outcome gives me a reason for doing *each* thing available to me that includes x?" But I worry, thanks to Richard, that the intuition is due to Gricean implicature and not to the fact that that there is no reason for my doing *each* thing available to me that includes x.

Of course, there's a lot in your two comments to think about. This is just one initial reaction.

The comments to this entry are closed.

Ethics at PEA Soup

PPE at PEA Soup

Like PEA Soup

Search PEA Soup


  • Unless otherwise indicated, the views expressed in any given post reflect the opinion of only that individual who posted the particular entry or comment.