Here is an inconsistent triad of propositions:

  1. There are certain pure “ticking bomb” cases (e.g. of the sort that I considered in my last post on this topic), in which torture (even on the part of a state official) is all-things-considered morally permissible.
  2. It is right for the law to impose an absolute prohibition
    on torture (at least as practised by state officials) and to make all cases of torture liable to criminal punishment.
  3. It is not right for the law to make an act
    liable to criminal punishment if that act is all-things-considered morally permissible.

If (like me) you find each of these three propositions prima facie plausible, how should you resolve this inconsistency?

As I said, each of these three propositions seems to me prima facie plausible. Still, I think that the one that we should amend is the third. (For a powerful defence of this third proposition, see Douglas Husak’s
recent article “The Cost to Criminal Theory of Supposing that Intentions are
Irrelevant to Permissibility”.)

So I suggest that we should replace the third proposition with
someone like the following:

3’.   It is not right for the law to make an act
liable to criminal punishment unless the relevant agent had a moral obligation not to perform that act.

As I understand the notion of an “obligation”, moral obligations
can conflict. (E.g. suppose that you
have made two important promises, and it becomes impossible for you to keep
both. Then I would say that you have two conflicting obligations.) Very
roughly, a moral obligation is a strong
moral reason
, of the sort that at the very least makes it appropriate for other
people to demand an explanation of an
agent who fails to comply with this reason.

Now, in these cases where obligations conflict, one
obligation may be overridden by
another. In such a case, violating the overridden obligation will be
all-things-considered morally permissible.

Even if there are cases in which torture is all-things-considered
permissible, I would still say that you always have an obligation not to torture anyone. If I’m right, then (1), (2), and (3’) can all be true.

Why should it be that (3) is false but (3’) is true? Let me
leave that question for another occasion…

13 Replies to “My Problem with Torture: The Solution

  1. What is gained by employing a terminological practice where ‘there is always an obligation not to torture’ is compatible with ‘it is morally permissible to torture given pure ticking bomb conditions’?
    Isn’t the question most people have to start with, ‘is torture always morally impermissible’? If the answer is no the next question is under what conditions is it permissible and why? Your use of the term obligation amounts to saying ‘that it is only OK to torture if one has very very good reasons XYZ, and normally it is not OK to torture’. In other words, the ‘always’ in your use of ‘always an obligation’ does not seem to speak to what is centrally at issue in the torture debate.

  2. I agree that (3) is the one to drop, but I’m not keen on (3′) either. Why think that law is even that closely connected to morality? Law is a social institution, and as such there are a range of considerations about what laws should be made, especially to do with what can in practice be done with the techniques of law, with collective action problems, and with the unintended effects of laws. So, for instance, a law mandating driving on the left side of the road (in the UK) is entirely fine, even though there’s no moral obligation not to drive on the right. In the present case, there are good reasons to make torture illegal under all circumstances, regardless of whether there are imaginable cases where a particular act of torture is the morally right thing to do.

  3. Interesting proposal!
    A worry that perhaps you can assuage…
    I worry your solution will generate problems in other domains – it might threaten to license state infringements on morally permissible exercises of personal autonomy.
    Consider, for example, euthanasia. Given your conception of obligation, we have an obligation to refrain from helping others kill themselves. But, defenders of euthanasia argue, helping another to kill themselves is all-thing-considered morally permissible in some cases (in which you would have the people say that the obligation is overridden).
    Now defenders of euthanasia might also argue that there should be no law against euthanasia because (i) there are euthanasia cases in which it is permissible to help someone kill themselves, and (ii) your (3).
    If we accept (3′) as a replacement for (3) this argument will collapse.
    The defender euthanasia might reject your proposal by make general complaint I mentioned above: your proposal licenses state infringements on morally permissible exercises of personal autonomy (in cases in which this is the wrong thing for the state to do).
    Is this off the mark? Does it worry you?

  4. An amendment/correction:
    On second thought, the general “infringement on personal autonomy” charge could be better illustrated. Perhaps I should have considered the obligation to not kill yourself, which is, defenders of euthanasia maintain, overridden in relevant cases.
    But I also now worry that the case I describe provokes a second objection to your view – the general objection in the case I described above (the person *helping* someone kill themselves) would be better put this way: your proposal licenses state infringements on morally permissible benevolent acts (in cases in which this is the wrong thing for the state to do).

  5. Suppose it is a very important norm in our society, right at the heart of our moral conception of ourselves, that nobody must use the word “bottom” in public. So we pass a law that says if you use the word “bottom” in public you must pay a £100 fine. Being clever philosophers however we can think of cases where the gains and losses at stake in whether you say “bottom” in public are so high that we worry maybe you could in such circumstances permissibly do so. These cases while conceivable are pretty fanciful and have never in fact occurred. Whoever in such cases the stakes are so high that the risk of a £100 pound fine is not, relative to whatever else is at stake, such a big deal. Given that and given how very seriously we take this norm we reasonably decide not to muddy the waters of the message our law sends by hedging it in with exceptions.
    After all any moral duty to obey the law is surely only prima facie (or should one say pro tanto these days? I can never remember.) And surely inevitably so. For, while I think some moral philosophers have rather overcooked claims about uncodifiability, particularism, the contours of the morally permissible and impermissible are surely too complex for it to be a realistic goal for the law to capture them precisely and trying to do so would be too costly with regard to keeping our laws clear and comprehensible.
    Compare rape. It’s not hard for moral philosophers to cook up cases where it could credibly be urged that rape is morally permissible. Dr Evil will destroy the word unless you rape Esmerelda. Or whatever. But if the stakes are that high and Dr Evil really will blow up the world so that concern for Esmerelda’s moral rights is legitimately overridden, we long passed the point where your concern to avoid a prison sentence is not also properly overridden. Given that and given how serious a crime rape is and given the importance that the law should affirm this with some clarity, I think the best thing for the law to say about rape is not, “Don’t, unless…” but simply, “Don’t.” In the case of torture similar considerations surely rather plausibly apply.

  6. I’m not sure if I agree with either 3 or 3′ (I’m inclined to deny 1 instead) but maybe there is a middle ground between 2 and 3 (or 3′) that would allow us to have a LEGAL punishment while still allowing the concept of moral permissibility for the action? In his book Terrorism and Counter-Terrorism, Seumas Miller argues that those who torture (morally rightfully) should be tried and convicted, but should receive 1-day symbolic prison sentences. This is based on Miller’s acceptance of 1 and 2, but it seems to alter the ‘punishment’ of 3 to something more symbolic.
    I don’t buy Miller’s argument for 1, but he does seem to be taking on the apparently inconsistent triad.

  7. I also agree that the three claims seem initially plausible, yet they cannot all be true, and that the third seems like the one that should go given that one has to go. How about something like this? It is not right for the law to make an act liable to criminal punishment if that act is all-things-considered morally permissible unless all other ways for the law to be would be morally worse.

  8. Most legal systems have affirmative defenses with “necessity” or “choice of evils” being one of those. The idea as I recall it, is that when it is proven beyond a reasonable doubt that you have done the crime as defined by the elements of the defense, you can still be found not guilty if you can show that what you did is excused by one of the affirmative defenses.
    As Jimmy Lenman’s comment points out, the issue seems really quite general. To what extent do can you give exception-less sufficient conditions for doing something wrong in terms that could be adjudicated by applying the sort of criteria that make up the elements of criminal offenses given the possibility of overriding considerations? Apparently some of the people who write model legal codes are pessimistic that this can be done, so they build in such overriding defenses. And they think it is better to do it in this general way rather than build “unless” clauses into the elements of the offense.
    Ralph, I take it that you want your principle to be read in such a way that it allows making something illegal which a person only has an obligation not to do in virtue of its being illegal, and not independently. Otherwise I think you won’t be able to prescribe cheating on one’s taxes. But that may raise a further problem; those who have the view that one always has an obligation not to break the law will think that the clause is always trivially satisfied.

  9. Thanks for all your comments! Here, in what I hope will be the most helpful order, are my responses to these comments:
    Jimmy —
    You’ve anticipated most of the points that I was going to make about why (3′) is true (but (3) is isn’t). I.e. it probably shouldn’t be the business of the law to adjudicate the ultimate balance of reasons in all these cases. Moreover, it also seems in practically all the cases where it is permissible to commit torture (or rape or whatever), there are such extraordinarily massive countervailing reasons that they also not only outweigh the reason against disobeying the law, but also the self-interested reason against suffering the criminal punishment. So criminalizing acts of this kind doesn’t seem wrong, even though the acts in question are in fact permissible.
    So, I gratefully acknowledge that Jimmy’s comment has made — clearly and elegantly — all the points that I was going to make myself about the advantages of (3′) over (3). But still, something about Jimmy’s tone makes his comment sound as if it’s somehow intended as an objection. (Perhaps he thinks that these points are too obvious to merit being discussed on PEA Soup? Or perhaps Jimmy and I just can’t believe that yet again, we are agreeing about something?!)
    Mark —
    First, I should clarify: as I was understanding proposition (2), I was assuming that it entailed that the law should not allow any “necessity defence” for torture (at least as practised by state officials). I guess it’s a good legal question whether, in current international law, there could ever be a successful “necessity defence” for torture, but I wasn’t assuming any answer to this question.
    Secondly, on your final point — Yes, of course I did mean want my “principle to be read in such a way that it allows making something illegal which a person only has an obligation not to do in virtue of its being illegal, and not independently.”
    You are quite right that someone who believed that we always have an obligation to obey the law will think that my principle (3′) is trivially satisfied. But not many people accept such a sweeping version of “political obligation”: it seems to me that there are certain laws (anti-sodomy laws, for example) that do not generate any obligation to comply. So for most people (even if not for adherents of this extreme version of the doctrine of political obligation) my clause (3′) is not trivial.
    David —
    I suppose that I am a bit worried that your proposed alternative to (3′) will be trivial. Your proposed alternative seems to come perilously to saying no more than that the law should criminalize a certain type of conduct — unless it shouldn’t! So I think that my formulation is more illuminating and explanatory (though also more liable to counterexamples of course….)
    Sam —
    You ask, Why should we think that the law is connected to morality at all? Well, we should all agree that the law is driven by all sorts of policy considerations, by the evidentiary concerns of courts and prosecutors, and so on. But let me repeat what I said to Mark: I wasn’t claiming that one must have an antecedent obligation (quite independently of the law) not to do any act that counts as a criminal offence. I was claiming that it is a fatal objection to a criminal law if it makes a certain act liable to criminal punishment, but (even after the law has come into existence) the agent has absolutely no obligation not to perform that act.
    Part of the idea behind this principle is that it is a terrifying coercive thing to be made liable to criminal punishment. Surely it would be outrageously oppressve if the law made you liable to criminal punishment for doing an act that was (i) in itself quite harmless, and (ii) something that you had utterly compelling reasons to do. Surely that is quite a plausible idea? If so, isn’t it reasonably plausible that the underlying rationale is somehting like my principle (3′)?

  10. I offered the principle: It is not right for the law to make an act liable to criminal punishment if that act is all-things-considered morally permissible unless all other ways for the law to be would be morally worse.
    I agree that this version of the principle seems trivial. Yet its difference from your initial principle highlights, I think, where the exit from the inconsistent trio might be and where the false air of paradox in the initial trio might come from. The moral escape clause reminds us that, at least on some views of morality, it does not settle the moral status of a general law that that law has morally unappealing features in some cases.

  11. The problem is perhaps not so much with (3) as it is with (2): you’re equivocating the word “right” in (2) and (3), with the reading in (3) as the more natural one.
    (2) could be true, that “it is right” for a law to do X, provided you mean that the action is permitted under the state’s constitution. But if you mean that the action is morally permitted even if (1) is true, then you’re speaking from the assumption that natural law is correct, that laws necessarily are about morality, i.e. moral permissibility.

  12. I hope you don’t mind if I recommend a short bibliography on “torture” to those who may be interested, as many of the questions explored here have been treated by others in a manner that, at their best, artfully link the moral, legal and political dimensions. In particular, or for example, there are compelling critiques of the “ticking bomb” scenario that crop up again and again and yet hardly anyone seems aware of them (indeed, a couple of books devoted to the topic). The Israeli Supreme Court case cited in the list has an interesting take on the “necessity defense” in conjunction with recognition of the absolute prohibition on torture (I don’t agree with it, but I think it’s worth reading). Please see:

Comments are closed.