Thursday, August 22, 2013

Clockmakers and Thought Experiments

Last time I wrote about a category of thought experiments in ethics where you are presented with a choice between performing some kind of ethical transgression or not, with the consequences of not performing the transgression being incredibly dire. There are a few things someone could be trying to argue for with this sort of setup--utilitarianism, intuitionism, or merely an alternate deontological ethics that doesn't forbid the act in question.

The thing I always want to know when someone brings up one of these situations is this: where's Snidley Whiplash?

Not to be confused with Dick Dastardly.

A bunch of people got tied to the traintracks, and the only way for me to save them is to kill an innocent bystander. Suppose also that no one else is in a position to save the people. Well, how the hell did they get there? Who's the clockmaker, so to speak? Because if Snidely Whiplash tied them to the tracks, then it seems like he's the one at fault if they die, and my refusal to murder an innocent doesn't make me the murderer of the tied-up people, even though I could have prevented their deaths by not refusing.

This is even more clear if we imagine that instead of setting up a situation that somehow mechanically requires me to commit murder, Snidely Whiplash retains control of the fate of the tied-up group. "Strangle this toddler, philosopher, or I'll kill everyone here!"

Look, Snidely: I'll do me, you do you. I'm not responsible for Snidely Whiplash's choices. If he want's to kill a boatload of people, that's on him. What I have under my control is whether or not I strangle the toddler, so I don't.

Just beneath the surface here is the fact that once Snidely has set up the trolley problem, we've entered upon a second-order problem of what to do about puzzles that only arise once the rules have already been broken (Set aside the possibility of some kind of trolley-problem natural disaster). So how can we reconcile our first order duty to "keep our hands clean" so to speak with this second-order problem?

I don't think we can, really, at least with respect to our behavior to bystanders. Someone might argue that because my set of permissible acts with respect to Snidely Whiplash changes once he's broken the rules, it is plausible that my set of permissible acts might change with respect to other people. I don't buy it. We might say that it would be noble for a bystander to "take one for the team," so to speak, but to require this of them because someone else has broken the rules would be to inflict an undeserved punishment on them.

Friday, April 26, 2013

Thoughts on Thought Experiments in Ethics: A Reply to Matt Zwolinski and George H. Smith

Therefore no one will be declared righteous in God’s sight by the works of the law; rather, through the law we become conscious of our sin.
-- Romans 3:20, New International Version
In a series of blog posts starting with this piece, Matt Zwolinski and several other discussants have been weighing the value of Rothbard's Non-Aggression Principle ("NAP") as a foundation for libertarian ethics. That led up to this piece, and the comments section thereof, where Zwolinski engages George H. Smith in an extended discussion of how the NAP holds up in hard cases where the choice a person faces is to violate the NAP or accept dire consequences, for example a large number of people being raped, or the person's own death, or an innocent third party dying, etc. Situations like air pollution where small NAP violations seem to result in large social benefits are also discussed, but I want to set that set of cases for purposes of my own discussion here.

Zwolinski seems to think that these sin-or-die cases he has constructed where it seems like a reasonable person would violate the NAP show the NAPs weakness as a moral principle. He describes himself as a moral pluralist, which means he thinks acting ethically involves weighing several prima facie plausible moral arguments and deciding on the right course of action using our (properly honed) moral intuition.

I wasn't completely clear on Smith's final position, but I will try to summarize it as best I can. Smith thinks that usually the NAP is both a rule about justice (meaning roughly what rights people have) and a rule about moral permissibility (meaning roughly what actions it is or is not wrong or right to take), but that the two concepts are distinct and in some cases they diverge. Smith holds that, for example, in the case where someone must break into a cabin they don't own or face death from exposure, it is morally permissible for her to break into the cabin and furthermore that this morally permissible action is a violation of the cabin owner's rights. The cabin owner would be entitled to expel the person by force should she so choose and would be entitled to compensation for the trespass.

Smith describes himself as an ethical egoist, meaning he thinks doing the right thing means acting in one's own interest, properly construed. This will be familiar to libertarians as Rand's position, though I am not sure whether Smith considers himself a Randian.

My own position on the NAP controversey is that while Zwolinski is right to point out that most of the heavy lifting in Rothbard's ethics is done by his theory of property, not the NAP, I think his objections to Rothbard's ethics in general are not very convincing. I think he does offer a fairly convincing case that its better to treat the NAP as a heuristic than a foundational rule, though he would be disappointed by this conclusion as his aim is a "paradigm shift" in how libertarians think about ethics, which I think would be a huge mistake.

That said, I think Smith and Zwolinski are both making a mistake in how they are thinking about the relationship between thought experiments about extreme cases and ethical systems. I don't think Zwolinski's contention that we might expect a person to act contrary to the NAP in certain circumstances, and might even be sympathetic to them acting that way, tells us much about the validity of the NAP as a moral system. I think Smith's view that both the cabin owner and the cabin invader can be in the right is plausible, but I don't think it's necessary for him to go through any contortions just to demonstrate that how we would expect a person to normally act in a situation--breaking into the cabin--is an ethically permissible act.

This discussion points at a category of mistakes in ethical theory I want to now discuss at some length.

Consider this proposition, P.

P: In some set of circumstances C, I believe I would break rule R of some given moral system S.

My thesis is that P is a proposition about psychology and has no bearing whatsoever on the correctness of S.

The implied argument here seems to be as follows.
  1. I am a good person.
  2. Good people never do bad things under any circumstances.
  3. I would break rule R in circumstances C.
  4. It is not wrong to break rule R.
  5. System S is incorrect.
Christianity, as per the quote I opened with, asks us to reject (1). I want us to reject (2). Sometimes people value things, like their own life or the life of a loved one, more than acting morally. This shouldn't strike us as odd, and the fact that we can imagine people acting  on those values isn't a problem for a proposed moral system.

Now, that isn't to say that we don't feel sympathy for the actions of the trespasser. I just don't think, like Zwolinski seems to, that this sympathy means the cabin owner was not wronged and that justice demands the trespasser either be punished or make restitution, etc., depending on what our thoughts on crime and punishment are.

One plausible view we might have here is Randy Barnett's position that when we're looking at the treatment of criminals, we ought to give a lot of weight to what their behavior signals about their likely future behavior. In the cabin-in-a-blizzard example, unless our trespasser makes a habit of getting lost in blizzards and breaking and entering, it seems like she's unlikely to be habitual rights violator--at least not moreso than the average person--and we ought to be fairly lenient. If, on the other hand, it becomes clear through a chain of incidents that this person isn't taking adequate preparations against the weather because she knows she can just break into a house if things go bad, then it makes sense to throw the book at her, life-or-death situation be damned.

Another way to look at the situation is to ask what we would think of a person who froze to death rather than break into the cabin, or to take a more extreme example from Zwolinski's thought experiements, lets the world be destroyed rather than scratches someone's finger against their will. We might very well have a hard time relating to a person who acted that way--but wouldn't we also see their actions as somehow admirable?

People have brought up the Joker in The Dark Knight in this context--"Tonight you're gonna break your one rule"--but this impulse goes much deeper in human culture. I am reminded of a vignette from Arthurian legend where a knight--I think Galahad--comes upon a tower of damsels, who tell him the only way for their lives to be saved is for him to commit a mortal sin. The knight refuses, and the damsels are revealed to be demons, the tower an illusion. I couldn't find this story specifically (I read it in The Once and Future King), but here is a similar story.

So, fiat justitia ruat caelum, I say. Just don't judge those cabin trespassers too harshly--they're only human.