Stephen Kershnar is Wrong About Consequentialism
A (short) response to (part of) Morality Collapses
I’m no consequentialist. But I am a philosopher, which means I can’t stand bad arguments against consequentialism. So, here we are.
In his recent book, Morality Collapses — and in his even more recent conversation with
—, philosopher offers the following argument against consequentialism:If consequentialism is true, the right maximizes the good.
If the right maximizes the good, then for any right action, there’s a fact of the matter about how much good that action would bring about.
If there’s a fact of the matter about how much good an action would bring about, then there’s a fact of the matter about what would happen were that action not performed.
It’s not the case that, for any right action, there’s a fact of the matter about what would happen were that action not performed.
Therefore, consequentialism is false.
This is my formulation of the argument, not Kershnar’s, but it’s a faithful formulation.
One of these premises is false. But before saying which, I’ll briefly explain why Kershnar thinks each premise is true.
Premise (1) is a standard, reasonably innocuous way to express (maximizing) consequentialism.
Premise (2) is true, according to Kershnar, because consequentialism requires ranking actions according to the good they would bring about. An action that maximizes the good is an action that sits atop such a ranking. But in order for an action to sit atop such a ranking, it must be true of that action that it would bring about at least as much good as any alternative. And for this to be true, there must be a fact of the matter about how much good it would bring about.
Premise (3) is true, by Kershnar’s lights, because the amount of good an action would bring about is just the difference between how good the world would be if the action were performed and how good the world would be if the action weren’t performed. And in order for there to be such a difference, there must be a fact of the matter about what would happen if the action weren’t performed.
Premise (4) is true, in Kershnar’s mind, because there’s no principled way to determine which of the many alternatives to an action would occur if that action weren’t performed. Suppose, for example, that I’m deliberating about whether to slash my friend
’s tires. In order to determine how much good this would bring about, according to Kershnar, I must ask myself what I would do if I didn’t slash Amos’s tires, check how good the world would be then, and compare that to how good the world would be if I did slash Amos’s tires.But what would I do, if I didn’t slash Amos’s tires? Would I kick him instead? Would I cut his brake lines, rather than slashing his tires? Or would I steel myself against temptation and do nothing? (Fat chance.)
There’s no principled way to say, in Kershnar’s view. There’s no correct baseline world to compare to the world in which I slash Amos’s tires. Hence, there’s no fact of the matter about what I would do if I didn’t slash Amos’s tires, and so no fact of the matter about how much good slashing his tires would bring about.
This claim of Kershnar’s — that there’s no fact of the matter about what I would do if I didn’t slash Amos’s tires — is very controversial. But I won’t dispute it here. I don’t need to, because premise (2) is false.
To rank actions according to their consequences, a consequentialist needn’t say, of any action in the ranking, how much good that action would bring about full stop. They need only make a series of pair-wise comparisons between actions in the ranking: e.g., action A is better than action B, which is better than action C, which is equally as good as action D, which is better than action E, and so forth. An action A maximizes the good (sits atop the ranking) just in case no action in the ranking is better than A. In order for A to meet this condition, there needn’t be a fact of the matter about how much good A would bring about full stop. There need only be a fact of the matter, for any alternative action B in the ranking, about whether B is better than A.
“Hold on,” you might protest, “how is this supposed to help the consequentialist? Why is it any easier to say whether B is better than A than to say how much good A would bring about? Mustn’t there be a correct baseline world in order for there to be a fact of the matter about whether B is better than A?” Well, yes and no. Comparing A to B requires comparing the world in which A is performed to a baseline world in which A isn’t performed. But in this case, it’s obvious which world is the correct baseline world: it’s the world in which B is performed rather than A. And even if there’s no fact of the matter about what would happen if A weren’t performed, there is a fact of the matter about what would happen if B were performed rather than A. By singling out an alternative to A (in this case, B), one thereby singles out a correct baseline world.
This applies nicely to the tire-slashing case. Although there may be no fact of the matter about what I would do if I didn’t slash Amos’s tires, it’s still perfectly sensible to say that it would be better to bake him a cake than to slash his tires, since the world would be better if I baked him a cake than if I slashed his tires. If the actions being compared are tire-slashing and cake-baking, there’s no mystery about which worlds should be compared. Specifying an alternative action amounts to specifying an alternative world.
Bentham’s Bulldog gestured toward this point in his conversation with Kershnar. Kershnar replied that the relevant pair-wise comparisons must be made on the basis of cardinal values. And the only cardinal values that can serve this purpose are the amounts of good that the relevant actions would bring about. Hence, the consequentialist can rank actions in the way described above only by “smuggling in” facts about how much good those actions would bring about full stop.
This is mistaken: the amount of good an action would bring about isn’t the only cardinal value on the basis of which it can be compared with other actions. Another is the amount of good the world would contain if that action were performed.
These quantities are distinct. Recall that the amount of good an action A would bring about is the difference between the amount of good the world would contain if A were performed and the amount of good the world would contain if A weren’t performed:
Good A would bring about = (Good the world would contain if A were performed) - (Good the world would contain if A weren’t performed)Kershnar’s argument for premise (2) is that there’s no fact of the matter about how much good the world would contain if A weren’t performed, since there’s no fact of the matter about what would happen if A weren’t performed. In other words, one of the terms on the right-hand side of this equation is ill-defined.
But this argument does not show that there’s no fact of the matter about how much good the world would contain if A were performed, nor does it show, for any alternative action B, that there’s no fact of the matter about how much good the world would contain if B were performed rather than A. And these quantities — the amount of good the world would contain if A were performed and the amount of good the world would contain if B were performed — are cardinal values that can form the basis of a pair-wise comparison of A with B: A is better than B just in case the world would be better if A were performed than if B were performed. And, as we’ve already seen, such pair-wise comparisons are enough to determine which actions maximize the good.
Nice article!