Aside from nihilism, there are the aforementioned multitude of first-order normative views. Certainly I cannot pick which one of those is right. But, I do have separation of cases available to me. And, if I can show that any of those views will give me my desired conclusion, then the first-order debate is irrelevant. To be sure, there is a near-infinitude of ethical theories floating around these days, and I cannot discuss them all in this paper. But these ethical theories tend to be variations on central themes, and I can certainly discuss archetypical examples of those themes and hope to accommodate other (close) versions. As mentioned above, ethical theories can typically be classified in three different categories: ethical monism, ethical pluralism, and ethical particularism. Ethical monism is the most dominant tradition in moral philosophy and includes its most-recognized theories: consequentialism, deontology, virtue ethics, and social contract theory. While there are certainly other versions, consequentialism achieves its most auspicious presentation through John Stuart Mill’s hedonic act-utilitarianism. Similarly, Immanuel Kant is usually flaunted as the flag bearer of deontology, though contemporary versions of deontology probably bear less similarity to Kant than contemporary versions of consequentialism do to Mill. Plato’s virtue ethics might have been the first, but Aristotle’s has been the most lasting, though it again has contemporary faces. Social contract theory similarly has a multitude of versions, and there are important differences among them, though we shall survey representative cases. Moral pluralism and moral particularism have been more recent alternatives to the aforementioned monisms; they have been espoused by, for example, W. D. Ross and Jonathan Dancy, respectively.
Returning to the notion of separation of cases, if one of these theories is right, and if all sanction the moral permissibility of torture (in some cases), then it follows that torture is morally permissible (in those cases). Unfortunately, things will not be quite as easy as this since one of our candidate moral theories will prove especially recalcitrant to endorsements of torture. As might be expected, this is going to be Kantian deontology. And this, of course, could throw a wrench in the whole project: if one of our candidate moral theories does not secure our conclusion, then separation of cases will fail and the conclusion will not be secured. Yet, I will not despair, and for two reasons. First, if only Kantians can object to torture, this really would be a substantive result. To wit, we will have now surveyed the entire landscape of moral philosophy and reduced the opposition to torture to a single (and probably lonely) outpost. But second, and more satisfyingly, I think that we can render the Kantian’s position implausible, thus removing it from the candidate moral theories. And, if we can eliminate its candidacy, then separation of cases is back in business and we can get to work securing our desired conclusion. But more on that later.
(Fritz Allhoff, "A Defense of Torture: Separation of Cases, Ticking Time-bombs, and Moral Justification," International Journal of Applied Philosophy 19 [2005]: 243-64, at 245 [italics in original; endnotes omitted])
Note from KBJ: I don't want to be mean to Professor Allhoff, who was kind enough to send me a PDF version of his essay several weeks ago, but I'm stumped by his argument. It's clearly deductive, in the sense that its conclusion is supposed to follow necessarily (as opposed to merely probably) from its premises. It goes like this:
1. At least one of the following normative ethical theories is correct: A, B, C, &c.
2. Each normative ethical theory on the list implies that torture is sometimes permissible.
Therefore,
3. Torture is sometimes permissible.
The argument has the form of a constructive dilemma, which is valid. But even Allhoff admits that the argument is unsound, for he admits that the second premise is false. Kantianism, which is one of the normative ethical theories on the list, does not imply that torture is sometimes permissible. Kantianism implies that torture is never permissible. So Allhoff has made an argument that, by his own admission, is unsound. With arguing skills like that, who needs critics?
Note 2 from KBJ: Professor Allhoff "solves" his problem by excluding Kantianism from the list specified in premise 1. This defeats the purpose of the argument, which was to show, without having to determine which normative ethical theory is correct, that torture is sometimes permissible. It turns out that he does have to determine which normative ethical theory is correct, or at least which is incorrect. I might add that it's strange to think of normative ethical theories as correct or incorrect. Normative ethical theories are sorting devices. One subscribes to a theory and then uses it for guidance. Professor Allhoff has done nothing to dissuade Kantians (of whom there are, for better or worse, many) from subscribing to their theory. All he has done is show them something they already knew, namely, that they are unconditionally opposed to torture. They are happy to accept that result. They are not the least bit embarrassed, surprised, pained, or troubled by it. I might add that what I just said about Kantianism is true of any type of absolutist deontology, including Catholic natural-law theory. (See Section 80 of this document; but see the final paragraph of this.) G. E. M. Anscombe, for example, would agree with Kant that torture is never permissible, though not for the reasons he gives.
Keith Burgess-Jackson, J.D., Ph.D. 