I’m grateful to Zofia for the excellent summary and questions. Let me say something about her question in order:
On her (1): Zofia asks whether there is really much point to hopeless ideal theory if we really know it won’t be met. I wrote that one of the benefits of such theory is that we can surprise ourselves, but suppose we really do know that certain standards will never be met. She’s write that, since we’re limiting our concern here to standards that are not beyond people’s abilities, I would insist that it is usually pretty hard to really know that they won’t ever be met. But I hasten to emphasize that my defense of hopeless aspirational theory does not rest on this conjecture. The fact, if it is one, that a certain kind of normative political theory will never be met, is not a defect in the theory. More concessive theory is valuable and important too, of course.
Zofia’s second point seems to me to teeter between two interpretations. One point might be that even impossible standards might still be real standards, so why rule them out? A second point she seems also to have in mind is that what is impossible now might later become possible, so we shouldn’t denigrate theories that raise standards that are impossible now unless we also think they will never become possible. First, I only grant that we should reject theories that demand the impossible for the sake of argument. But I’m not sure we should accept them either. Zofia notably speaks of standards that are impossible “at present, and in the foreseeable future.” What is the point of this except to leave open a possibility in an unforeseen future? In that case, she’s not really giving us an impossible demand. What would she say about a truly impossible demand, such as “justice requires that people stop sleeping, in order to devote more time to productive labor.” It might be that, in a sense, that would be better. But I deny that it is a matter of any moral significance.
Jonathan points out that something might be impossible given current conditions at time 1, but not impossible if conditions change at time 2. He concentrates on the case where we might be able to change the conditions. In a case like that I think we must say that what’s being required at time 1 is not impossible at time 1. So there no inability to block the “ought.”
There are some interesting other cases in this vicinity. Suppose that our doing x (at time 3) is impossible given current conditions at time 1, but not impossible if conditions change, but that we cannot change the conditions. They might change at time 2, but we can’t change them. I think we should say that it is not impossible that we meet the requirement. But that’s not the relevant question if we’re trying to respect the principle that ought implies can. Just because something is possible at time 1 (which I think it is in this case) doesn’t mean we have the ability at time 1. That’s what oughts (on that principle) require. So, if we can’t change conditions, but we can do x if they change, can we do x? I don’t know the answer to this, but I think it’s a very interesting question. You might say, well, since we know it might turn out that we do x, that will prove that, all along, we could do x. I doubt that this would be a good argument. All actuality proves is possibility, not ability. We could probably say that at time 2, when conditions have changed, then we have the ability to do x. I don’t see any clear way to settle whether or not we have the ability at time 1. If we don’t, “ought implies can” would say we can’t, at time 1, be required to do x.
Here’s another interesting case: Suppose that we are at time (ii), and that now and forever we cannot do x at time (iii). But at time (i) it was the case that we could do x at time (iii). But we blew it, and the time has passed. We, at time (ii), might not be required to do x, since we can’t. But, should that determine that justice does not require x? I doubt it. I think we need to allow the possibility that we missed our opportunity for justice. So justice requires x, of us, though not of us at this time. Still, if we at time (ii) are responsible for our actions at time (i), then we have violated a requirement. The point is that we may not dumb-down justice just because it is not, from here, any longer possible. This seems to me an important limit on the uses of the “ought implies can” principle.
On Zofia’s (3): Recommending that people avoid a certain shortcut because it’s dangerous is certainly not very unrealistic. But I think any way of measuring realism in order to ground the claim that this is realistic will also be forced to say that some recommendations are more realistic yet. I take it that this proposal is quite realistic partly because there’s a high probability of compliance. But that directly implies that, other things equal, a higher probability of compliance would render a recommendation even more realistic. So recommending that people continue unsafely to use that shortcut is more realistic. As Zofia says, not much seems to hang on this.
I haven’t looked back at Parfit’s distinction between two kinds of impossibility, but a few remarks on Jonathan’s brief presentation of it: Something might be within the laws of nature and human nature but be beyond our abilities. It is not somehow outside of those laws for me to do competent heart surgery this afternoon. But I can’t do it, and so I’m not required to. Nothing can be within my ability if it is precluded by laws of nature, so “deep impossibility” entails no ability. What about a requirement that I, at some time of my own choosing, do heart surgery. I can’t do it now, but it may be that if I devoted all my effort and resources to it I could learn (though this is far from certain). We don’t need a special notion of impossibility here. My doing heart surgery this afternoon is not impossible, but it is outside my abilities. Doing it at some future time might not be outside my abilities. Incidentally, the concept of “impossible given current conditions” isn’t entirely clear. Does every set of conditions at a moment make everything “technically impossible” except what actually happens? Given those conditions nothing else could have happened. As I say, I’m not consulting Parfit’s text in the interest of time, and so I make no claim to be addressing what he actually says.
A Concluding Note
I had been scheduled to give a concluding round of comments next week. But none of us anticipated that I would insert myself into the discussion so loquaciously along the way, and so enough is enough. I won’t try to sum up any philosophical lessons here, but I will say that I have found these exchanges to be very illuminating. The quality of the summaries, criticisms, and discussion has been very high, and so it was impossible for me to resist the temptation to take full advantage by engaging in depth and in as much detail as I could on a weekly schedule. Unfortunately, this may have made the reading group difficult to keep up with for many, just out of the sheer volume of reading. Still, many of you stuck with it, and I’m grateful. This is a wholly new kind of scholarly exchange, I think. It doesn’t resemble anything I’ve ever done or seen before, and the differences are mainly advantages. I’m sure I’ll continue to draw on these exchanges for some time as I continue to think about the weak links and possible extensions of the theory in my book. Thank you all.