Wednesday, September 28, 2011

Why I was wrong about Triple Effect

Frances Kamm has proposed something called 'Triple Effect', in which an agent can bring about a desired effect, which is essential for his overall plan, without intending it. She says this shows we can use something as a means which we do not intend. I agreed with her analysis, though I said it should not be called a means of the agent.

Kamm's examples. The party: I will only throw a party if I don't expect there to be a lot of clearing up. I think my guests will do the clearing up for me, so I throw the party on that basis. But I don't intend that they clear up.

Looping Trolley: I can direct a trolley away from one set of people tied to the track only the basis that lives will be saved overall. In fact there are even more people tied to the alternative track. But luckily (?) there is a very fat man tied to the track as well, who will stop the trolley before it gets to these latter people. I can direct the trolley in that direction without intending the squashing of the fat man.

These have some plausibility, particularly the party example (or so it seemed to me), but I have realised that this approach is subject to powerful counterexamples, like this one.

The wicked uncle. My uncle is very rich and very wicked; a whole community is suffering under his exploitative sway. I fancy scratching my finger on the trigger of a loaded gun I am pointing at him, which will obviously go off and kill him. Normally such finger-scratching would be wrong, because the unintended effect of killing an innocent (non-aggressor) would outweigh the good of relieving my itch in this way. But in this case the good consequences of his death far outweigh the badness of the death in itself. So the balance of non-intended consequences is actually positive.

Now this seems absurd, and if we allow this then any action with overall good consequences which violates a deontic constraint (a common-sense moral prohibition like 'don't kill the innocent) could be done with a little morally irrelevant posturing. So the moral structure of deontic constraints would effectively collapse into Consequentialism.

The case is indeed absurd because when I say that the good consequences of the action make it morally possible to do, I am taking cognizance of them in a way which implies that I intend them. One intends things which motivate one to act as one does. The good consequences of the death are motivating me, in part, and I would nit act without them; thus I must be intending the death of the uncle as well, as a means to my intended end.

What I failed to see was that on Kamm's examples the agent must be intending the good foreseen results because they are essential to his plan. If he did not expect them then he'd have to call it off, and he'd better make sure they happen, by adapting the plan if necessary. They are indeed his means, but by the same token they are intended.

This admission also effect another example I came up with: the railway enthusiast. He is so keen on railways that he wants to build one really as an end in itself. Someone points out that railways are dangerous things and people are bound to be killed in accidents in the years after it is built. This seems to rule out the project. Then someone else points out that railways are safer than roads and by shifting traffic away from roads it will have an overall positive effect on the number of accidental deaths. I DID say that the enthusiast can proceed with the building with no intentions about accidental deaths, happy in the knowledge that the balance of unintended consequences is positive. I NOW say that since the improved overall safety is essential to the moral viability of the project he must intend it, if only as a means to the end of building a railway.

Changing my position in this way brings underlines the principle found in many discussions of the Principle of Double Effect, that the good consequences of an action must not flow causally from the bad foreseen side effects. At least, if the bad side effects are such that it would be wrong to intend them, one cannot justify the action on the basis of further, good, effects which flow from them, for to do this is to bring them, and the bad cause of them into one's intentions.

- Posted using BlogPress from my iPhone


Daniel Hill said...
This comment has been removed by the author.
Daniel Hill said...

Joseph, I don't see why you need to insist that the agent does intend the causal means in Kamm's examples. Why isn't it enough to assert, as in your final paragraph, that `the good consequences of an action must not flow causally from the bad foreseen side effects', or, `[a]t least, if the bad side effects are such that it would be wrong to intend them, one cannot justify the action on the basis of further, good, effects which flow from them'? Or why not adopt Kamm's own suggestion for PDE, viz. that the action must be undertaken for the sake of the outweighing good? Or, more weakly, as Alexander Pruss has it, undertaken for the sake of a non-trivial good?

Unknown said...

this is the best blog and you are doing very well job. This is one of the most informative information I've read. It really helps a lot. Thanks for sharing this and teaching some of your Idea's. also check communicable diseases.