Before I get to the results, let me say what issue I’m interested in and what I think about the cases in the survey. The question that interests me is:
(Q) Does the fact that some future agent, F, will actually perform a certain future action bear on what some present agent, P, is obligated at present to do?
On the one hand, actualists (who hold that S is obligated at t to do x at tꞌ [t being earlier than tꞌ] if and only if S can at t do x at tꞌ and what would happen if S did x at tꞌ is better than what would happen if S did not do x at tꞌ) answer “yes” to (Q) regardless of whether or not F’s future action is under P’s present control—that is, regardless of whether F’s doing x at tꞌ depends on P’s present actions, intentions, and other attitudes. Thus, the actualist holds that, in Case 5 (Phil and the Pill), the fact that Phil is not going to administer Drug C tomorrow is relevant in determining whether Phil is obligated at present to administer Drug C later today. Indeed, it is in virtue of this fact that actualists claim that Phil is not obligated as of the present to administer Drug C later today. And they claim this even though whether or not Phil is going to administer Drug C tomorrow depends entirely on whether he now intends to do so. So, on actualism, Phil avoids an obligation at present to administer Drug C today in virtue of not intending at present to follow up by administering Drug C tomorrow. And this seems implausible.
On the other hand, possibilists (who hold that S is obligated at t to do x at tꞌ if and only if S can at t do x at tꞌ and what could happen if S did x at tꞌ is better than what could happen if S did not do x at tꞌ) answer “no” to (Q) regardless of whether or not F’s future action is under P’s present control—that is, regardless of whether F’s doing x at tꞌ depends on P’s present actions, intentions, and other attitudes. Thus, the possibilist holds that, in Case 3 (Phil and Dr. Evil), the fact that Phil is not going to administer Drug C tomorrow is not relevant in determining whether Phil is obligated at present to administer Drug C later today. What’s relevant, according to the possibilist, is that Phil could, even after be injected with Dr. Evil’s serum tomorrow, still decide not go to the bar and to continue on to the pharmacy and then give Pat the second dose of Drug C. The fact that Phil is no more presently able to prevent his future self from going to the bar and drinking too much in this case than he was, in Ashton and Dr. Evil, presently able to prevent the future Ashton from going to the bar and drinking too much is neither here nor there. And this too seems implausible.
I think in contrast to both the actualist and the possibilist that whether the fact that some future agent, F, will actually perform a certain future action bears on what some present agent, P, is obligated at present to do depends on whether F’s performance of that future action is under P’s present control—that is, whether F’s performance of that future action depends on P’s present actions, intentions, and other attitudes. So I think that in those cases in which what the person in charge of Pat’s care tomorrow will do tomorrow is under Phil present control (cases 2, 4, and 5), Phil is obligated at present to administer Drug C later today. And I think that in those cases in which what the person in charge of Pat’s care tomorrow will do tomorrow is not under Phil present control (cases 1 and 3), it is not the case that Phil is obligated at present to administer Drug C later today. And it appears from the survey that most people agree.
As of 7:40 AM on March 14, 2011, there were 61 people who had completed the survey. Like me, most of them thought that, in cases 2, 4, and 5, Phil is obligated at present to administer Drug C later today. More specifically, here is the percentage of respondents who responded that Phil is obligated at present to administer Drug C today:
- Case 2: 82.0%
- Case 4: 82.0%
- Case 5: 62.3%
And, like me, most respondents thought that, in cases 1 and 3, it is not the case that Phil is obligated at present to administer Drug C later today. More specifically, here is the percentage of respondents who responded that it is not the case that Phil is obligated at present to administer Drug C today:
- Case 1: 78.7%
- Case 3: 75.5%
Hi Doug,
I didn’t see this in time to fill out the survey, but I want to run a question by you. Do you think it matters that in cases 2, 4, and 5 we are not told whether Phil actually takes the pill (or administers it to Ashton) presently? If we are told that he won’t take/administer the pill, though he could, maybe respondents will treat this as they treat stipulated future facts, viz., as exogenous to the question whether Phil ought to administer C/P. (Don’t know how well this comment applies to 5, as there you say he intends to refrain from taking the pill, so presumably he doesn’t.)
Also, if respondents think it still open whether Phil takes/administers the pill presently, maybe the initial thought is that Phil ought presently to take/administer the pill and give a dose of C. Respondents might then be inclined to say that Phil ought to give dose C (even if, on reflection, they’d deny that ‘O(p and q), therefore Oq’ is a valid rule of inference). Just a thought.
Hi Matt,
Good question. I personally don’t think that it matters whether in fact Phil now intends to (and therefore will) administer the pill to himself/Ashton later today (clearly, he ought to, and that’s what seems to matter to me). But it might have mattered in what the results of survey would have been. The fact that people responded as they did in 5 suggests to me that the majority would have still gone the same way but that that the majority probably would not have been as large as it was and would have been around 62% rather than 82%.
I’m sure that the survey was poorly designed in a number of ways. Given the emails that I received, it’s clear that there were a number of confusions as result of my poor design. But what matters to me is only that what I say above sounds plausible and that these examples have some hope of pumping people’s intuitions about this, as I think that the survey shows that it does. It’s not that I think that people’s responses to the survey support my contention that the answer to (Q) should be that it depends on whether P has present control over whether F will do. It’s just that I want to see if these sorts of cases are capable of pumping in others the sorts of intuitions that I have about them.
I answered the survey as I think an actualist would, since that was my understanding of what “objective obligation” is supposed to mean given both the definition of it prefacing the scenario, its application of the wire-cutting scenario, and other discussions I’ve heard of it. Hence I answered that Phil should give C in those cases where, in fact, doing otherwise will result in Pat’s death, and “can’t say” in all the other cases where whether Pat will get C or P or nothing depend on facts which have are yet established by the fact of whether Phil gives C or P or nothing. If I’m not mistaken, all the questions fell into one of those two categories.
However I don’t think this says anything interesting about my “intuitions” since I have no intuitions regarding “objective obligation.” I generally see it used, and used in the definition & example, to mean “X is objectively obliged to A just if X’s Aing will actually lead to the best results,” so once I’m told what leads to the best results, or am not given enough information to answer that question, I answer accordingly. If anyone thinks it means something other than that, then frankly I’m tempted to conclude that the concept is more confusing and potentially misleading than I already thought it was. This is part of why I think the more useful and morally significant concept is subjective obligation: not only is it based on facts accessible to agents, but it does not duplicate the already perfectly clear phrase “actually produces the best results.” Just my 2 cents.