Against Thought-Experiments
At my personal blog, I’ve got a post up putting forward seven propositions that pretty much define my present thinking about torture and the prosecution thereof. Some may find it strange that nowhere in that list does a professed (if not yet professional) philosopher have anything at all to say about “ticking time-bomb” scenarios and the other kinds of hypotheticals that Conor, for instance, ably takes up here. But this is actually because, perhaps contrary to what Conor says in his follow-up post, I think that such abstract considerations pretty much are little more than a distraction from the really crucial issues.
The reason for this is simple: the fact that a given sort of behavior can reasonably be counted as legal or moral in one circumstance says nothing at all about whether it can be so counted in another. There are countless behaviors that share this characteristic: sexual intercourse, for instance, is obviously one, as are cursing and slapping and swearing and humiliating and waking people up in the middle of the night or even trying to scare them with dogs. And we all can see, I think, when it is that such behaviors are appropriate, and when it is that they are not; being able to make such distinctions is just what it is to be appropriately responsive to moral demands.
Given this recognition, the important discussion about torture – which is to say: the discussion about torture that we need to be having, as opposed to the one we’d be in a position to have if not for all the things that happened on our government’s watch – is a discussion about whether, given the circumstances that actually obtained, the things that agents of our government did to prisoners and detainees were warranted. Whether there are some other possible circumstances in which some of those behaviors might have been warranted or even morally required is an entirely separate question, and while it’s of some philosophical interest it clearly ought to be far less important to us than the question of whether what we did constituted torture; and as such, it’s generally quite hard to see the insistence on posing wild counterfactuals rather than dealing with real-life cases as anything but a ruse.
Again, that’s not to say that counterfactuals are never relevant to moral deliberation. (For a helpful account of why they sometimes are, see Bryan Caplan’s post here – though for a note on why they sometimes aren’t, see his post here, too.) But the problem is that when you’re faced with an actual situation that call for a moral evaluation, thinking instead about some alternative possible world is at best a luxury you can’t afford; in practice, it’s often like discussing the trolley problem when faced with a charge of vehicular manslaughter. The key is to get the actual circumstances right, and understand what they demanded, not to dream up some imaginary alternative in which the behaviors in question would arguably have been okay.
One more point: even when we are entitled to go in for thought-experiments, there’s every reason not to make too much of their conclusions. In the first place, our intuitions about what to do in far-away circumstances are often pretty corrupted; moreover, there’s a tendency in designing thought-experiments to idealize the imagined circumstances to a far greater degree than worldly situations can ever achieve. (The standard “ticking time-bomb” scenarios are cases in point.) Perhaps most crucially, an excessive focus on what to do in idealized situations can easily corrupt our judgments about what to do in ordinary ones, and there’s every reason to have our laws written in a way that doesn’t permit even the handful of exceptions that abstract reflection might suggest are appropriate. The decision to harm, shock, or terrify a person, like that to go to war, can never be a “slam dunk”; in the real world, it always needs to be made with deep trepidation and concern for the seriousness of what’s involved. If such a mindset had been in place in the wake of 9/11, no doubt many of the subsequent evils would likely have been avoided.
John: “The key is to get the actual circumstances right, and understand what they demanded.”
That’s obviously not true. Given any “actual” circumstance, you have a lot of variability on how it might be digested by a human mind, none of which are preferred a priori.
Thinking any given circumstance “demands” something is just one of these ways.
— Sargent · May 5, 04:33 PM · #
Can’t these kinds of thought experiments be used to justify just about anything? Let’s take Linker’s thought experiment (super-powerful terrorists who can only be defeated through the use of torture) and take it a step further. What if the terrorists are simply too diabolical to be defeated at all? They’ll just murder and murder and murder until their demands are met. I guess, given a choice between the complete annihilation of the United States and knuckling under to terrorist demands (sharia law in America, etc.), we might have to choose surrender.
I don’t think that makes a lot of sense, and I certainly don’t think it has much to do with real-world counter-terrorism and foreign policy. Similarly, I don’t think concocting bizarre and implausible worst-case scenarios where torture is the only feasible response to terrorism is particularly helpful.
— Charlie · May 5, 04:52 PM · #
Sargent: I don’t understand your remark at all. Are you saying there’s no such thing as an “actual circumstance”? That’s not obviously true; if it’s obviously anything it’s obviously false. Or are you saying that interpretation is always required to characterize or “digest” it? Well, yeah, but why does that tell against John’s point at all? If the point is that it won’t always be obvious or non-contentious how to characterize what actually happened, that’s fine, but again, that doesn’t tell in favor of talking about hypothetical situations rather than actual ones. Some work has to be done to pin down what the actual situation was and how to talk about it – yes. Therefore we have to talk about non-actual situations instead – no. And your final remark is still more baffling: given a situation involving genocide, for example, why on earth would we not think a certain response and judgment is demanded? Why is the default attitude one of withholding moral judgment, as you seem to suggest? That is obviously not how we typically operate, and it’s far from clear that it’s how we should operate.
— Joe · May 5, 04:58 PM · #
In one respect it seems appropriate to bring up hypotheticals like this: when we are trying to get clear what people are arguing. If, like Andrew Sullivan, for example, one holds that torture is always wrong and questions of utility are irrelevant to that, then someone who thinks we need to talk about utility can properly have recourse to counterfactual situations to try to reopen the question of utility. There has been a lot of argument over efficacy and whether we should even talk about that, and in that specific context it’s perfectly appropriate to address counterfactual cases that draw out the contrast between, say, consequentialist and deontological positions on the matter. I agree, however, that often this is just avoidance of the nasty reality and a distraction from the pressing matter at hand.
— Joe · May 5, 05:04 PM · #
This is an excellent point and a useful principal.
— cw · May 5, 05:18 PM · #
Joe, have you ever tried to depose witnesses?
The problem with John’s post is twofold: one, he uses the definite article (the key when it’s really his key), and two, the picture is worth a thousand words problem.
Reaching consensus on an actual circumstance is functionally no different than a thought-experiment: you must decide what information to keep, what information to emphasize in what order, etc. And that’s after your mind has filtered out some 95% of the data, after your subterranean processes have keyed on qualia and emotionally accented their memories.
Add to this each person’s neocortical image of “Me and the Universe” — a paradigm which may or may not be informed by reality-corresponding knowledge and extensive self-reflection.
The upshot: there’s no real difference between reality’s moral, and a story’s moral. It’s all narrative, all accented information constructed according to folk physics and folk morality.
— Sargent · May 5, 05:25 PM · #
FYI, I’ve updated the most recent post on my blog in response to this post.
http://blogs.tnr.com/tnr/blogs/linker/default.aspx
— Damon Linker · May 5, 05:56 PM · #
John:
Great post. One quick point is that I think there is a limited role for thought experiments in this discussion, because some participants argue that torture is never the right choice to make. It seems to me that testing such an assertion with thought experiments is valid.
— Jim Manzi · May 5, 05:57 PM · #
Sargent – ok, I think I have a better grasp on your point. But John’s point was fairly specific: we shouldn’t act as if the important scenario to talk about is a story that everyone admits (or should admit) bears little relationship to what happened in this particular case, rather than a story that makes a serious attempt to get at what happened in this particular set of cases. By contrast, your complain is little more than general skepticism about human objectivity. If someone says, “It’s better to believe in the generally accepted results of the scientific community than to listen to witch doctors,” the appropriate response is not “Well, it’s all just narrative anyway, so it all depends on which narrative you happen to start with or choose or whatever.” Similarly, if an historian asks whether a certain very recent administration did certain things to terrorist detainees, the appropriate response is not “well, who can really know anyway, since we’re hopelessly doomed by our severe cognitive biases and limitations.” A specific comparative point about two different ways of proceeding towards a conclusion is typically not appropriately answered by blanket skepticism. That’s like taking a Sorites problem as justification for ignoring the difference between one’s neighbor and a block of concrete.
— Joe · May 5, 06:22 PM · #
Jim:
I agree to an extent, but then the further point is just that the universal prohibition is itself a red herring in this context, and distracts attention from the issue of whether torturing in this circumstance was or wasn’t justified.
And Damon: Thanks. I’ve got a response to your response over at my own blog: http://johnschwenkler.wordpress.com/2009/05/05/thought-experiments-ctd/.
— John Schwenkler · May 5, 08:20 PM · #
“some participants argue that torture is never the right choice to make. It seems to me that testing such an assertion with thought experiments is valid.”
I think most people who oppose torture absolutely would be willing to amend their position from, “never the right choice to make,” to “never the right choice to make in any scenario that has, or ever could, actually happen.” So there may not be all that much value in imagining highly improbable scenarios.
If torture sometimes has benefits so great that its use is justified, then we should be able to find historical examples backing up that proposition. The resort to thought experiments seems like an admission that there’s no real-world justification for torture. I’m sure there are abstract philosophical implications to the proposition that torture could theoretically be justified in an alternate universe, but I think focusing specifically on our own reality is a more productive way to proceed.
— Charlie · May 5, 09:14 PM · #
Question from the peanut gallery: Does justification make a decision moral? That is to say, suppose we can justify the use of torture (a la “ticking time bombs of improbability”). Is it possible to have a justified decision that is still immoral?
— James F. Elliott · May 5, 10:45 PM · #
I’m not quite sure what you’re asking, or how the terms are meant. By “justification” do you simply mean apparent justification, i.e. something that the person making the decision takes, perhaps not unreasonably, to be a justification? If so, then I’d say no: you can think that you’re justified but not actually be so, and your decision won’t be moral unless your justification is real and not merely apparent. If, on the other hand you mean real justification, then on a natural reading to have such justification just is for one’s decision to be moral, and so in that case the answer would be yes.
Or maybe that was entirely tangential to the what you were really after?
— John Schwenkler · May 6, 01:27 AM · #
Charlie,
“If torture sometimes has benefits so great that its use is justified, then we should be able to find historical examples backing up that proposition.”
Since torture is illegal and is likely to be carried out in secret, I don’t know why you think that. Even justified cases of torture are not exactly the kind of thing governments are likely to want to publicize.
The benefits of torture, like the benefits of policies like the death penalty, lifelong imprisonment, strategic bombing and economic sanctions, are highly uncertain and speculative. It would be very difficult to make a clear and compelling case that the benefits of these policies outweigh their costs (Does executing murderers really make us safer? Do the military benefits of strategic bombing outweigh the loss of life? Do economic sanctions do more good than harm?). But few people seem to categorically oppose all of them. So I don’t think your challenge to prove that torture is ever justified is terribly relevant.
— Anderson · May 6, 01:36 AM · #
I think James’ point is that a utilitarian justification is not necessarily a moral justification.
— Adam S · May 6, 02:17 AM · #
I guess what I’m after, which I failed to articulate, is what we mean by “moral.” If we can find justification within our moral intuition — a sliding scale of wrongness, such as Jim Manzi was exploring with the inadequacy of utilitarianism in arriving at a universal prohibition of torture — then does that make torture a “moral” response in some situations? Is our justification — our intent — sufficient to make something moral?
On the flip side, is it possible that a decision is justified but still immoral? That is to say, torture is still “immoral” — it is reprehensible, a violation of our intuition — but justified by the circumstance? Does one ever take a back seat to the other, morality trumping justification and vice versa, or does the justification necessarily imply morality?
— James F. Elliott · May 6, 02:22 AM · #
Since torture is illegal and is likely to be carried out in secret, I don’t know why you think that. Even justified cases of torture are not exactly the kind of thing governments are likely to want to publicize.
The benefits of torture, like the benefits of policies like the death penalty, lifelong imprisonment, strategic bombing and economic sanctions, are highly uncertain and speculative. It would be very difficult to make a clear and compelling case that the benefits of these policies outweigh their costs….
I agree completely with your second paragraph. That’s why I think it makes more sense to evaluate the inherent morality of an action (torture, massacres of civilians, the execution of criminals, etc.) rather than speculate about the possible benefits of doing things that are immoral and illegal (like torturing people).
As to the first paragraph, I don’t think government secrecy is so air-tight that covert torture could be kept secret over the long-term. It didn’t take long for the current torture program to come out. And there are examples from eras and from wars (e.g., the US in the Philippines, the French in Algeria, etc.) where more brutal tactics were less offensive to public opinion and therefore not kept all that secret.
I’m open to the possibility that there are good examples of torture “working”, depending on how “working” is defined. But if:
a) It’s difficult/impossible to assess whether torture works or not in real-world scenarios
b) There are very few (or no) examples of torture “working” (however that’s defined)
c) Thought-experiments are resorted to because they provide a theoretical justification for immoral behavior that cannot be justified by anything that’s ever happened in the real world
then I don’t think people who oppose torture categorically and on principle are in danger of losing the argument.
— Charlie · May 6, 02:56 AM · #
Well, I don’t think so. I’d say that intent can affect the goodness or badness of an action (/the gravity or lack thereof of a sin) in certain ways, but even the best intentions are never sufficient on their own to make an action moral unless it’s already got something to be said for it.
I suppose that this is pretty much what I was denying in the second half of my earlier comment, when I said that for an action to be justified just is for it to be moral. That’s not, however, to say that there can never be cases in which a genuine evil can be the best of the choices available, but in such a case I’d say that doing that evil thing is both justified and, therefore, the moral course of action.
— John Schwenkler · May 6, 03:07 AM · #
Wow, reading this is quite encouraging to me, you’ve expressed something I find myself thinking a lot in this current situation.
Thank you, it’s good to see I’m not the only one wondering why we care so much about some situation that’s purely hypothetical.
— Thetortuoustruth · May 6, 04:32 AM · #
charlie,
If your “argument” is that, as a matter of principle, torture is always and everywhere wrong, regardless of consequences, then I think you’ve already lost the argument and will always lose it. I think most people are utilitarians, or at least consequentialists, not deontologists. They just don’t accept that the morality of torture is independent of its consequences. Moral absolutes are the domain of religious fundamentalists.
The point about your demand for examples of torture “working” is that it would be hard to produce such examples for various other policies too (the death penalty, strategic bombing, economic sanctions.) The absence of such examples has not led people to categorically reject those policies. I don’t think it’s likely to cause them to categorically reject torture, either.
As for secrecy, I think your idea that most acts of torture conducted by the government have been made public is just naive. It wouldn’t surprise if the CIA has tortured hundreds of people during its 60-year history.
— Anderson · May 6, 05:15 PM · #
John: for an action to be justified just is for it to be moral.
I have to say, I’m beginning to get frustrated. This is provably, demonstrably wrong; the worst kind of anachronism — the kind that has been definitely refuted by logic, moral philosophy, and published and peer-tested experiment.
I really can’t get over how incurious and unselfcritical you are on this subject, especially given the quality of your work on other things.
— Sargent · May 6, 05:53 PM · #
definitively, natch.
p.s. I don’t mean to be rude. But really, a little self-skepticism goes a long way.
— Sargent · May 6, 05:59 PM · #
Sorry Sargent, but I just don’t follow. I take myself to be making a specific sort of a priori claim about how I use, and how we in general ought to use, a certain set of terms. I’m willing to make the philosophical arguments in favor of these principles of usage. But how such a view could be “refuted by logic”, let alone “peer-tested experiment“, is simply, definitively beyond me.
— John Schwenkler · May 6, 06:17 PM · #
As for secrecy, I think your idea that most acts of torture conducted by the government have been made public is just naive. It wouldn’t surprise if the CIA has tortured hundreds of people during its 60-year history.
Maybe I am being naive. But I don’t see how your argument is advanced by the assertion that there might be secret examples of secret torture having secret positive effects at some point in time somewhere in the universe.
This is the same problem inherent in the thought-experiment approach: in the absence of any evidence to support the proposition that torture is sometimes justified on utilitarian grounds, people just make up scenarios that “prove” what cannot be proven with logic and real-world evidence.
Utilitarianism asks us to take into account real-world consequences affecting real people. It does not ask us to make up circumstances that may or may not have ever existed, and then make real-world decisions based on those fantasies. I’m entirely open to the possibility that torture is justified in some cases on utilitarian grounds. I’m just asking for a single example that isn’t made up.
But since you acknowledge in your last post that it’s difficult/impossible to assess the utility of torture, and since your best “examples” or torture “working” are things you imagine the CIA might have done, but which no one knows about, I think I’m going to bow out of this discussion.
— Charlie · May 6, 06:57 PM · #
Charlie,
Can you prove, using real-world examples, that the death penalty is justified on utilitarian grounds? Life imprisonment without parole? Strategic bombing? Economic sanctions? Laws against selling drugs to children? A top income tax rate above 30%?
The fact is, we have all sorts of laws and policies that impose demonstrable harms on people for social benefits that are uncertain, and difficult or impossible to measure. There is nothing unique about torture in this respect.
— Anderson · May 6, 11:05 PM · #
James Elliot,
In this context, I take “justified” to mean “morally justified.” “Justified but immoral” therefore seems to me a contradiction in terms. What do you mean by “justified” if not “morally justified?”
— Anderson · May 6, 11:09 PM · #
Anderson,
To just dip my toe in to very deep waters, a few … well, thought experiments, in circumstances where they ARE justified – will make it very clear that pure consequentualism leads very quickly to some very horrible conclusions. Yes, most people are (in part) consequentialists, but very few people are ENTIRELY consequentialists. And the few people who are, are moral monsters.
— LarryM · May 7, 03:45 PM · #
And I did not realize when I posted the above the extent to which the first Caplan link supported my point. :)
— LarryM · May 7, 03:56 PM · #