1 Of course, if one were prepared to argue (as perhaps H. L. A. Hart was in his article, “The Ascription of Responsibility and Rights”) that behaviour should count as a human action only where it is attributable to an agent in such a way as to make him fully responsible for its principal features and consequences, then doing something under duress would be at most a qualified case of action.

2 This, and related remarks, at least hold good of the normal or standard case of acting under duress. There may be cases of actions done under duress where the just-mentioned intentionality and knowledge requirements are not met. For example, such a case might occur where an agent believes with good reason, but does not actually know, that it is in order to get him to do an action of such and such a description, A (for example, delivering a package of heroin), that he is being subjected to threats; we may suppose that the action expected of him has never actually been described to him in precisely this way. However, in these sorts of cases there often is a resultant uncertainty either about the admissibility of a plea of duress or about the description of the action which was done under duress, e.g., would it be fair to regard him as having Ad under duress?

3 Aristotle, Nicomachean Ethics, Bk. III, Chap. 1. Cases of duress seem to be distinguishable from the larger class of Aristotelian mixed actions by the fact that in the former case there is a deliberate attempt by one individual to manipulate, by the use of threats, the actions of another individual. And since actions done under duress are made “choiceworthy at the time they are done” as a consequence of the use of threats, it is the case that actions done under duress are typically chosen as being the lesser of two evils.

For a discussion which puts less restrictions on the notion of duress than I have argued for above, see Raab, F. “Free Will and the Ambiguity of ‘Could’.” Phil. Rev., LXIV, No. 1955.Google Scholar

4 One obvious implication is that the idea of acting under duress presupposes an understanding of a good deal of the “machinery of acting,” including its mental components. For instance, one needs to know what it is to have a reason for acting, to act for reason, to govern one's action by a specific intention, to act after deliberation, etc.

5 Complexities of the sort raised here are not confined to the case of duress, but apply to Aristotle's mixed actions generally, and also beyond the area of excuses. See, for instance, a similar complexity mentioned by H. L. A. Hart in his discussion of the idea of being obliged to do something, in The Concept of Law, Vol. 2, pp. 80–81. Hart concludes that the idea of being obliged to obey someone is “in the main a psychological one referring to the beliefs and motives with which an action is done.” This conclusion may underrate the importance of a condition of type (5) in the analysis of being obliged. However, Hart's conclusion is understandable in view of the fact that his principal concern at this juncture was to emphasize certain broad differences between the ideas of being under an obligation to, and being obliged to, do something.

6 J, L. Austin, “A Plea for Excuses,” reprinted in Philosophical Papers, p. 124. A minimal characterization of what is ordinarily understood by an excuse is the following: excuses are (1) those considerations relating to an agent or his conduct that could be proferred with the intention of establishing, and that typically would be successful in establishing, that the agent's (otherwise blameworthy) conduct was not blameworthy but merely unfortunate, or at least less blameworthy and more unfortunate than it would have been had these considerations not obtained, and (2) those considerations having as their characteristic or standard function, the sort of role described in (1).

7 The idea of a person's moral account is developed by J. Feinberg in his article, “Action and Responsibility,” in Philosophy in America, ed. M. Black (1965), and reprinted in Feinberg's Doing and Deserving (1970).

8 H. L. A. Hart has undertaken such a classification of the principal responsibility locutions in “Postscript: Responsibility and Retribution,” in Punishment and Responsibility: Essays in the Philosophy of Law (1968). I think that Hart's classification is in large measure, although not in its entirety, correct; but limitations of space do not make it possible to detail here my disagreements with his account.