Citation with persistent identifier:
Lienemann, Béatrice. “Aristotle’s Treatment of Force and Compulsion as Exculpatory Conditions for Moral Responsibility.” CHS Research Bulletin 2, no. 1 (2013). http://nrs.harvard.edu/urn-3:hlnc.essay:LienemannB.Aristotles_Treatment_of_Force_and_Compulsion.2013
§1 From history and the news, we are all familiar with difficult situations such as the following: the politician who has to decide whether or not to torture a terrorist to make him reveal where his group has planted the bomb which threatens to kill innocent civilians. This is Michael Walzer’s classical example of so-called “dirty hands” acts. Michael Stocker defines an act as a dirty hands act if “(1) it is right, even obligatory, but (2) is none the less somehow wrong, shameful, and the like”. When the politician decides to torture the terrorist, he chooses, on the one hand, the right action in saving the lives of innocent people. But on the other hand, he acts wrongly, since torturing is considered an inexcusable action. Such dirty hands acts cause difficult moral dilemmas for contemporary moral philosophy which also have been discussed in the history of moral philosophy. Aristotle was challenged by such intricate situations of decision-making, and some scholars even go as far as attributing a theory of dirty hands to Aristotle. I will come back to this in what follows.
§2 In this text, I focus on Aristotle’s treatment of force and compulsion as exculpatory conditions for holding an agent responsible for her action. The text is part of a larger project investigating Aristotle’s conception of moral responsibility; and the discussion of force and compulsion is a representative example for Aristotle’s development of this conception. The aim of my larger project is to show that Aristotle’s scattered remarks on the practice of responsibility in his ethical writings rely on a detailed, but hardly elaborated theory of responsibility. A theory of moral responsibility is a conception of the conditions under which an agent is morally responsible for her actions and their consequences (and maybe also for her character and emotions). The goals of the project are, firstly, to reconstruct systematically Aristotle’s theory of responsibility and, secondly, to analyze the reconstructed theory by applying it to problematic cases of actions dealt with in the text. In the first part, I reconstruct Aristotle’s treatment of the notions of hekousion (voluntary) and akousion (involuntary) in his ethical writings, especially in the Eudemian Ethics II 6-11 and in several books of the Nicomachean Ethics (especially books III and V). Here, Aristotle examines the conditions of what is hekousion and akousion, and he determines the involuntary negatively by two separate exculpatory conditions – first, external force (bia) or compulsion (anagkê) and, second, non-culpable ignorance (agnoia) of particulars – that let us refrain from praising or blaming an agent for her action. An action has to be considered as involuntary if at least one of the two possible exculpatory conditions is fulfilled, so that the agent is not responsible (aitios). In the second part, I examine the Aristotelian conception of responsibility by applying it to different specific types of actions that are discussed in the text such as akratic actions (i.e. actions undertaken by an agent in a state of akrasia (incontinence)), sudden and spontaneous actions (i.e. actions undertaken without preceding deliberation), actions in anger (thumos), or different types of “mixed actions”. A detailed study of such problematic cases of actions is instructive in order to sharpen our understanding of Aristotle’s views on responsibility and possibly to locate shortcomings and immanent tensions of his theory.
§3 By focusing on Aristotle’s treatment of force and compulsion in Eudemian Ethics II 8 and Nicomachean Ethics III 1, I touch on questions belonging to both parts of my project. The case is revealing, at the same time, for Aristotle’s specific views on the appropriate reactions to different kinds of actions done under conditions of force or compulsion, for his methodological approach to demanding questions of moral responsibility, and for uncertainties and challenges he leaves (intentionally or unintentionally) open to his audience and readers for further discussion. Eventually, it shows that Aristotle’s views are original and provide an illuminating perspective also for contemporary debates on responsibility in philosophy as well as for a broader discourse.
§4 Starting with Aristotle’s determination of clear cases of forced actions in the Ethics, I mainly investigate Aristotle’s analysis of difficult cases which he labels “mixed actions” due to the fact that they seem both voluntary and involuntary. Aristotle presents in both Ethics different types of such actions that respectively deserve different reactions. In discussing these various cases, I argue that he has reasonable answers in some cases, but brings forward more vague answers in other cases. This leads finally to the question whether Aristotle had certain reasons for offering various kinds of answers and leads to the insight that he consciously left specific reasoning and decisions to his addressees.
Clear cases of actions undertaken under force
§5 The context of Aristotle’s discussion of force and compulsion in both Ethics is his attempt to examine the voluntary and the involuntary. He analyses the voluntary and the involuntary in the broader context of his treatment of the character virtues that are essential to achieve eudaimonia (excellence). Noticing that persons deserve praise for their virtues (aretai) and blame for their vices (kakiai), Aristotle determines that praise and blame are accorded to the voluntary, whereas the involuntary receives sympathy (sungnômê) or even pity (eleos). Defining the voluntary and the involuntary serves therefore to become clearer about the conditions under which an agent acts virtuously or viciously. Aristotle names two separate conditions, force or compulsion on the one hand and non-culpable ignorance on the other, that let us refrain from praising or blaming an agent for her actions. In this paper, I will focus on the first of these two exculpatory conditions.
§6 In the Eudemian Ethics (EE) Aristotle defines an action carried out under force (bia) by two criteria. An action of a person X is carried out under force if and only if it is against a natural impulse (hormê) of X and if its origin is external (exôthen) to X. The first criterion is based on Aristotle’s view that every material thing has a natural place and accordingly moves naturally towards it unless something prevents it from doing so. So, for instance, a stone naturally travels downwards and fire upwards, and it happens only through violence that a stone travels upwards and fire downwards. In contrast to inanimate things, human beings have, as Aristotle says next in the EE, two internal impulses, namely inclination (orexis) and reason (logos) which not always are in harmony. Concerning the second criterion, Aristotle clarifies the meaning of externality further in addressing a possible objection of the continent or the incontinent: the incontinent, for instance, could invoke being forced, because the origin of his action, inclination, is external to his other natural impulse, reason. But Aristotle refuses this plea in specifying that the external origin of an action needs to be outside the whole soul (holê psuchê) of the agent which is equivalent to its being outside the person.
§7 In the Nicomachean Ethics (NE), Aristotle gives the following definition of a forced action:
The involuntary, then, seems to be what comes about by force or because of ignorance, and what comes about by force seems to be that whose origin is external, i.e. such that the person acting or the person to whom something happens contributes nothing, as for example if a wind is to carry him somewhere or men who have him in their power. (NE III 1, 1110a1-4)
§8 The Nicomachean explanation partly accords with the definition in the Eudemian Ethics, namely in acknowledging externality as one criterion for an action undertaken under force. To this, Aristotle adds as a clarification that the person acting contributes nothing to her action. I take this as a clarification, because it serves to exclude the case where someone, for example, grabs and lifts another person in the air when this person had ordered it or willed it. In this case, the person to whom something happens does contribute something to her action when she orders it. It is evident that this definition of actions undertaken under force is very narrow. The agent carrying out an action under force rather seems to be a victim of something that happens to him. Nevertheless, Aristotle initially distinguishes this class of indubitable forced actions and gives two clear examples for them: when someone is carried off either by a natural physical force like the wind or by other agents as for example kidnappers.
“Mixed actions” and the question of dirty hands acts
§9 After presenting the clear cases of forced actions, Aristotle turns directly to the more contentious cases carried out under force or because of compulsion. Aristotle labels these more complicated cases of actions in the Nicomachean Ethics “mixed actions” in order to indicate the fact that these actions seem both voluntary and involuntary. ‘Mixed actions’ are actions that people take as painful and bad, that they could avoid, but that they nevertheless choose to do because, for instance, flogging, imprisonment or death await them if they do not do it.
§10 Aristotle’s discussion of ‘mixed actions’ exhibits the most important difference among his treatments of actions undertaken under force and compulsion in both Ethics regarding their content. In the Eudemian Ethics, Aristotle considers ‘mixed actions’ involuntary relying on the fact that the agent acts, “in a way, under force” (1225a12). She acts under force, because she does not choose the very thing she does (i.e. the bad and painful action); yet, she does not act under force “without qualification” (haplôs) (1225a12), since she does choose the thing for the sake of which she carries out the action (i.e. avoidance of abuse, of imprisonment or of death). To clarify this idea, it is useful to apply the explanation to one of the two instructive examples that Aristotle gives in the NE for ‘mixed actions’: the captain who throws his cargo overboard in a storm to save his crew, himself, and the ship (NE 11108-9). According to the EE, the captain acts under compulsion and consequently involuntary, since he does not choose to throw overboard the cargo for its own sake. However, he does not act under compulsion without qualification, because he chooses to save the crew and the ship for the sake of which he throws overboard the cargo.
§11 In contrast to the EE, Aristotle considers ‘mixed actions’ in the NE as voluntary:
As for things done because of fear of greater evils, or for the sake of something fine – if, for example, a tyrant gave one orders to do something shameful, when he had one’s parents and children in his power, and they would be kept alive if one did it, but put to death if not -, such cases give rise to dispute whether they are voluntary or involuntary. Something similar also arises in cases of jettisoning in a storm; no one jettisons voluntarily without qualification, whereas if it is a condition of saving oneself and the others, every sensible person does it. Such actions, then, are mixed, but they more closely resemble the voluntary actions; for they are desired at the time of acting, and the end for which actions are done varies with the occasion. So we must assign both “the voluntary” and “the involuntary”, too, with reference to the time when one acts. And one acts voluntarily [in the cases in question]. For in fact in actions of this sort the origin of one’s moving one’s instrumental parts is in oneself; and where the origin of something is in oneself, it also depends on oneself whether to act or not. Doing such things, then, is voluntary, but without qualifications they are presumably involuntary; for no one would choose anything of this sort for itself. (NE 1110b4-19)
§12 As in the EE, Aristotle begins referring to the conflicting common opinions on ‘mixed actions’ due to the fact that they simultaneously appear both voluntary and involuntary. But, then, he continues with his own analysis of such actions leading to the result that they are “rather voluntary”. Aristotle’s argument for regarding them as voluntary is that they are voluntary under the actual circumstances of the action. Furthermore, the goal of the action for the sake of which the action is chosen is relative to the occasion of the action. Aristotle’s reason for calling them ‘mixed acts’ is that they allow for different descriptions: in abstraction from the actual circumstances (haplôs, “without qualification”, in Aristotle’s words), the action seems involuntary, because no one would choose such an action for its own sake. All relevant circumstances considered, however, the action seems voluntary, because the agent chooses the action for the sake of a specific end. So, for instance, the captain chooses to throw the cargo overboard for the sake of saving human lives and the ship; but he does not choose to throw cargo overboard for its own sake. Consequently, Aristotle allows both descriptions of the action, but he makes clear that the question whether an action is carried out voluntarily or involuntarily needs to be decided with reference to an action token, viz. to the specific action under its relevant circumstances (and not with reference to an action type neglecting relevant circumstantial factors). If one considers an action as involuntary based on its description abstracting from the occasion, that judgment is only provisionally valid. The abstract description is potentially incomplete and the evaluation of the voluntariness of the action needs to be revised if some further circumstantial factor turns out to be relevant for the appraisal.
§13 From a systematic point of view, Aristotle’s Nicomachean argumentation for considering ‘mixed actions’ as voluntary is convincing. The assessment of an act with regard to its voluntariness should be based on the most complete description in terms of encompassing all relevant circumstances of the action. Furthermore, in comparing the Eudemian or the Nicomachean judgment about the voluntariness of ‘mixed actions’, the two texts also contain some evidence for which appraisal is probably to be preferred. The text in the NE continues as follows:
And for actions of this sort, people are sometimes even praised, i.e. whenever they endure something shameful or painful in exchange for something great and fine; while if it is the reverse, they are blamed; for enduring the most shameful things as a condition of getting something not at all fine or something only moderately so, is the mark of a bad person. (NE III 1, 1110a19-23)
§14 Here, Aristotle distinguishes two types of ‘mixed actions’: first, mixed actions for which the agent is praised, and, second, mixed actions for which an agent is blamed. The EE contain a similar distinction between different types of mixed actions deserving different reactions. A mixed action is praiseworthy if the agent does a shameful or painful act in order to achieve something that is good and fine (respectively, better and less shameful than not doing it) or to prevent something that is a greater evil. On the other side, a mixed action is blameworthy, if the agent does something shameful in order to achieve something that is less good and fine or to prevent something that is a lesser evil and less painful. These passages show that in both Ethics Aristotle assumes that some types of ‘mixed actions’ deserve praise, whereas others deserve blame. Though, only in the NE the distinction seems to be justified. Attributing praise and blame to an agent because of his action presupposes, as Aristotle repeatedly makes clear, that the agent acted voluntarily. Thus, since the EE considers all actions undertaken under force or compulsion as involuntary, it lacks the necessary justification for founding the distinction between praiseworthy and blameworthy forced actions.
§15 Before discussing two other types of ‘mixed actions’ that Aristotle presents, I will now come back to my introductory remark on so-called “dirty hands acts”. Praiseworthy mixed actions seem to comply both conditions for dirty hands acts: they are right, even obligatory, because an agent does it in order to achieve something good and fine; but none the less they are somehow wrong and shameful, since the agent chooses to do something shameful and bad. Does this mean that it is legitimate to attribute a theory of dirty hands to Aristotle as, for instance, Stocker has argued? I think it does not because a defender of an Aristotelian dirty hands theory neglects Aristotle’s distinction between the two different action descriptions and the preeminence of the description encompassing all relevant circumstances with regard to the question of the voluntariness of the action argued for in the Nicomachean analysis. Although Aristotle holds the view that an agent performing a praiseworthy mixed action carries out an action that is bad in abstracto (haplôs), he does not consider this action as bad under the specific circumstances of the action. In a situation where the normally bad action is the necessary mean in order to achieve a better or less painful end, it is indeed the right action to do. For example, jettisoning is a bad and blameworthy action under normal circumstances. But under the specific circumstances of a storm, it is not a bad, but actually the right and praiseworthy action to do due to the fact that it is the action that has to be chosen in order to achieve the better end, viz. saving the crew and the ship. Thus, Aristotle is able to reject the attribution of a dirty hands theory and to challenge the objection that he allows for bad actions of the virtuous which would be a fatal concession for his ethical conception and his view of character virtues.
Actions carried out due to an overwhelming inclination
§16 In contrast, two serious interpretative problems are involved in my eyes in the descriptions that Aristotle subsequently gives in the NE of two other types of ‘mixed actions’. I begin with his characterization of the first that also finds its equivalent in the EE:
In some cases it is not praise we accord someone, but excuse, when one does some things one should not do, because of what is such as to overstrain human nature and which no one would endure. (NE 1110a23-26)
§17 With this sentence, Aristotle begins to introduce two other types of ‘mixed actions’ which vary from the first two with regard to the kind of coercion that forces an agent to do something. Here, the coercion draws its exculpatory force not from the unavoidable but nevertheless controllable choice between two bad alternatives, but from an overwhelming inclination that overpowers human nature. In the EE, Aristotle gives several examples for such actions. At first, he points to actions carried out under the influence of love (erôs), in certain cases of anger (thumos) or due to the other natural things. Furthermore, Aristotle cites agents possessed by divine inspiration (enthousiôntas) and prophecy (prolegontas). In these cases, the pressure is so strong that no one would endure it, or, as Aristotle adds in the EE, that no one is able to withstand. This suggests the natural meaning that the agent does not have it in his power to act otherwise. Moreover, actions undertaken under overwhelming coercion differ essentially from the first two types of ‘mixed actions’ in the appropriate reaction they receive. Whereas agents deserve either praise or blame for their actions belonging to the first two types, excuse (sungnômê) is accorded to ‘mixed actions’ of the third type when the coercion overpowers human nature.
§18 Here lies, in my opinion, the first interpretative problem of the description that Aristotle offers in the NE for ‘mixed actions’ undertaken under overwhelming pressure. I have argued that it is reasonable that Aristotle considers ‘mixed actions’ of the first two types as voluntary supposing that the agent deserves praise or blame for them. But, then, assuming that excuse is to be accorded to ‘mixed actions’ of the third type causes a problem for the consistency of Aristotle’s analysis in the NE. Either he has to consider these actions as involuntary (as he does in the EE), but this conflicts, firstly, with the overall reasoning in the passage, because aside from the third type Aristotle apparently regards ‘mixed actions’ in this context as voluntary; and, secondly, this is opposed to subsequent statements when Aristotle explicitly denies that actions carried out under the influence of certain emotions such as desire or anger are to be considered as involuntary. Or he holds on the assumption that all types of ‘mixed actions’ are voluntary, but this view conflicts with other statements in the NE that make clear that excuse is appropriately directed to involuntary actions. In light of this dilemma, I think it is more reasonable to prefer the first alternative, because it has the advantage of converging with the overall treatment in the EE that seems, this time, to offer the more suitable (and systematically more convincing) analysis of this kind of actions. Assuming that an agent acts under a compulsion that overpowers human nature intuitively suggests that the agent has no choice to act otherwise, and this means that he acts involuntarily.
§19 The second interpretative problem arises from the way in which Aristotle continues his presentation of ‘mixed actions’ forced by overwhelming inclination in the NE:
But perhaps in some cases there is no such thing as being compelled, but one should rather accept death having suffered the most terrible things; and indeed the things that compelled Euripides’ Alkmaion to commit matricide appear ludicrous. It is sometimes difficult to determine what sort of things one should choose in exchange for what and what one should endure in exchange for what – and more difficult still to abide by what one has determined, since mostly what one is expecting in such cases is something painful, and what one is necessitated to do is shameful, which is why there is praise and blame in the case of those who are or are not compelled.
§20 The passage reads as the direct sequel to the description of the third type of ‘mixed actions’, but the newly introduced fourth type – which is not mentioned in the EE – seems to be inconceivable given the literal meaning of the preceding type that I have proposed. For, when it is not in the power of the agent to act otherwise due to an overwhelming coercion, this reasonably forbids assuming that in some cases one should not be necessitated by the overpowering compulsion, but rather accept enduring the most terrible things. When the compulsion is a physical necessity, then it is not possible to accept and endure it; and in this case, it is not legitimate to request that sometimes one should not be necessitated, but rather suffer the most terrible things (since ‘ought’ presupposes ‘can’).
§21 Several commentators have noticed this inconsistency and argued that Aristotle mixes different criteria in presenting the third and fourth type of ‘mixed actions’. While he, at first, assesses the psychological capacities of the human nature to resist an overwhelming pressure, Aristotle then refers to the moral gravity of certain types of actions and their consequences, e.g. matricide, that seem to rely on fundamental moral principles. In view of this problem, I do not believe that Aristotle meant his remarks in this place to be conflated into a unified consistent conception. Rather, I understand his discussion as tentative and conceived to expose in a reasonable way different legitimate criteria that can prove essential for the assessment of the voluntariness of an action and that have to be considered and weighed against each other. He refuses to give clear-cut answers, since he had argued before that every action needs to be evaluated with regard to its specific circumstances and it is not possible to offer definite and universally valid answers with regard to the voluntariness of certain action types. However, in presenting the third and fourth type of ‘mixed actions’, Aristotle exposes two important and apparently inconsistent criteria for the consideration of the voluntariness and justifiability of actions. This may be explained by the fact that he is thereby addressing, at the same time, different groups of addressees for which different evaluative standards (such as legal standards on the one hand and moral standards on the other) are pertinent and appropriate. But in order to explore this idea in more detail, one has to take other parts of the Ethics into consideration, but this is beyond the focus of this text.
§22 To conclude, I summarize the main insights that one gets from an analysis of Aristotle’s treatment of force and compulsion in terms of the aim to reconstruct the Aristotelian conception of moral responsibility.
§23 Firstly, the treatment is a representative example for appreciating Aristotle’s method in developing his view on moral responsibility: initially, he relies on common opinions and the actual practice of holding someone responsible; then, he defines clear-cut examples, proceeds in exhibiting a typology of various cases of more challenging examples and finally reaches limiting cases. Although he presents a distinct principle for evaluating the voluntariness of an action, viz. that it needs to be evaluated under the specific circumstances under which the action is undertaken, Aristotle avoids giving precise advice for how different relevant criteria have to be weighed against each other in case they may conflict. While Aristotle disclaims that it is possible to decide these cases in general, the positive answer he gives is that one has to examine every single case on its own and to deliberate which criteria are relevant and how they have to be assessed. In discussing the limiting cases, Aristotle alludes to different criteria that may be essential for the evaluation of an action without admittedly taking a clear stand to their mutual relation.
§24 This shows, secondly, why it is so important to apply Aristotle’s definition of the voluntary and the involuntary subsequently to specific cases dealt with in the text. The case studies help to sharpen our understanding of the different Aristotelian criteria and to become clearer about the individual challenges of their application.
§25 Thirdly, the treatment of force and compulsion contains evidence for some far-reaching insights that one, however, cannot sufficiently appreciate before investigating other passages of the Ethics in detail. One is the conception of so-called “stereotypically bad actions”, i.e. actions that are bad under normal circumstances that no one would choose for their own sake, but that count only provisionally as bad due to the fact that they may only be described incompletely due to neglect of relevant circumstantial factors. The second insight rests on Aristotle’s remarks on the fourth type of ‘mixed actions’ that seem to allow for actions which the agent carries out involuntarily, but for which he nevertheless is accountable. The third and last observation concerns Aristotle’s view on the influence of various emotions on human actions. The typology of ‘mixed actions’ has touched on the question whether Aristotle considers some natural influences as so strong that they overpower human nature to such an extent that no one can withstand them and no one has it therefore in his power to act otherwise than he is compelled to do. It is worth comparing this view on certain emotions with Aristotle’s discussion of the individual emotions in the Rhetoric, on the one hand, and with Aristotle’s treatment in NE VII 5 of the so-called “brutish dispositions” (hexeis thêriôdeis) of animals which may occur naturally, but which lie outside the limits of badness of character, on the other hand.
Aspasius, Aspasii in Ethica Nicomachea Commentaria, ed. G. Heylbut. In: Commentaria in Aristotelem Graeca. Vol. 19. Berlin).
EE (Ethica Eudemia). Aristotelis Ethica Eudemia. Recensuerunt brevique adnotatione critica instruxerunt R.R. Walzer [et] J.M. Mingay, praefatione auxit J.M. Mingay. Oxford et al.: 1991. (OCT)
Susemihl, Franz: Aristotelis Ethica Eudemia. Eudemi Rhodii Ethica. Adiecto de Virtutibus et Vitiis libello. Recognovit Franciscus Susemihl. Leipzig: 1884.
NE (Ethica Nicomachea). Aristotelis Ethica Nicomachea. Recognovit brevique adnotatione critica instruxit I. Bywater. Oxford: 1894. (OCT)
Translation and Commentaries
Bonitz, Hermann. 1844. Observationes Criticae in Aristotelis quae feruntur Magna Moralia et Ethica Eudemia. Berlin: Bethge.
Broadie, S., and Ch. Rowe. 2002. Aristotle: Nicomachean Ethics. Oxford.
Dirlmeier, F. 1956. Aristoteles: Nikomachische Ethik. Berlin.
———. 1962. Aristoteles: Eudemenische Ethik Berlin.
Gauthier, R.A., and J. Y. Jolif. 1970. L’Ethique à Nicomaque. Introduction, Traduction et Commentaire. 2 Vol. Louvain/Paris.
Konstan, D. 2006. Aspasius: On Aristotle’s ›Nicomachean Ethics 1-4, 7-8‹. Ithaca, New York.
Taylor, C.C.W. 2006. Aristotle: Nicomachean Ethics. Books II-IV. Oxford.
Woods, M. 2005. Aristotle: Eudemian Ethics. Books I, II and VIII. Oxford.
Echeñique, J. 2012. Aristotle’s Ethics and Moral Responsibility. Cambridge.
Heinaman, R. 1988. Compulsion and Voluntary Action in the Eudemian Ethics. Nous 22/2:253-281.
Kenny, A. 1979. Aristotle’s Theory of Will. London.
Mele, A. 1992. Springs of Action: Understanding Intentional Behaviour. New York/Oxford.
Meyer, S. Sauvé. 2011. Aristotle on Moral Responsibility: Character and Cause. Oxford/New York [first published in 1993].
Nielsen, K. M. 2007. “Dirtying Aristotle’s Hands? Aristotle’s Analysis of ›Mixed Acts‹ in the Nicomachean Ethics III, 1.” Phronesis 52:270-300.
Nussbaum, M. 1986: The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy. Cambridge.
Rickert, G. 1989. ΕΚΩΝ and ΑΚΩΝ in Early Greek Thought. Atlanta, Georgia.
Stocker, M. 1990. “Dirty Hands and Conflicts of Values and of Desires in Aristotle’s Ethics.” In Plural and Conflicting Values, ed. Stocker, 51-84. Oxford. Orig. pub. 1986: Pacific Philosophical Quarterly 67:36-61.
Walzer, M. 1973. “Political Action: The Problem of Dirty Hands.” Philosophy and Public Affairs 2:160-180.
 EE 1224a20-23: “Similarly with animate things, i.e. animals, we see them doing and undergoing many things under force, when something external moves them against their internal impulse.” ὁμοίως δὲ καὶ ἐπὶ ἐμψύχων καὶ ἐπὶ τῶν ζῴων ὁρῶμεν βίᾳ πολλὰ καὶ πάσχοντα καὶ ποιοῦντα, ὅταν παρὰ τὴν ἐν αὐτῷ ὁρμὴν ἔξωθέν τι κινῇ. [The translation is mine, partly following Woods’.]
 NE 1109b35-1110a4: δοκεῖ δὴ ἀκούσια εἶναι τὰ βίᾳ ἢ δι’ ἄγνοιαν γινόμενα× βίαιον δὲ οὗ ἡ ἀρχὴ ἔξωθεν, τοιαύτη οὖσα ἐν ᾗ μηδὲν συμβάλλεται ὁ πράττειν ἢ ὁ πάσχων, οἷον εἰ πνεῦμα κομίσαι ποι ἢ ἄνθρωποι κύριοι ὄντες. Translations from the NE are mine, but closely based on the translations of Rowe and Taylor.
 Cf. Aspasius 59, 16-17: “For in fact if someone should grab me and lift me in the air when I had ordered it and willed it, the origin would be in me.” καὶ γὰρ εἰ λαβών μέ τις μετέωρον φέροι ἐμοῦ προστάξαντος καὶ βουληθέντος, ἐν ἐμοὶ ἡ ἀρχή. [I quote Konstan’s translation]
 Aristotle too seems to admit the limitedness of his definition when he adds ὁ πάσχων (the person to whom something happens) in 1110a2-3 as an alternative to “the person acting”.
 Although Aristotle uses two different terms, βία and ἀνάγκη, he does not observe this terminological difference in order to distinguish systematically between vis maior and vis compulsiva, i.e. a physical force that forces the agent to do something and excludes that the agent could act otherwise, whereas the second indicates that the agent is compelled by the structure of the choices, but has it in his power to do or not to do it (cf. Meyer 2011: 93). Only in the MM, the terminological difference seems to correspond consistently to a systematic distinction, but both the EE and the NE exhibit distinct counterexamples (cf. EE 1224a13-18, NE 1110b9-11) and show at most arguable hints for such a distinction (cf. NE 1110a32-b5, the lines EE 1225a17-19 are controversial due to a disputed negation in line 1225a17 contained in all MSS, but bracketed by Fritzsche, following Bonitz, and others). For a comparison between Aristotle’s use of βία and ἀνάγκη and the use in a larger corpus of ancient Greek texts from Homer to Thucydides, see the study of Rickert (Rickert 1989: 6-34).
 Although the label “mixed action” does not occur in the EE, I regard it as unproblematic to use it equally for the cases mentioned in the EE, because they largely correspond to the cases in the NE and it is convenient to have a name referring to all discussed cases in general.
 EE 1225a2-8: “In another way, men are said to act under force and to be compelled to act although reason and inclination are not in disharmony, and when they do what they take to be both painful and bad, yet, if they not do it, abuse or imprisonment or death await them. They certainly say they are compelled to do these things. Or is it not so, but do they all do the thing itself voluntarily? For it possible for them not to do it, but to endure the pain.” λέγονται δὲ κατ’ ἄλλον τρόπον βίᾳ καὶ ἀναγνασθέντες πρᾶξαι, οὐ διαφωνοῦντος τοῦ λόγου καὶ τῆς ὀρέξεως, ὅταν πράττωσιν ὃ καὶ λυπηρὸν καὶ φαῦλον ὑπολαμβάνουσιν, ἀλλ’ ἂν μὴ τοῦτο πράττωσι, πληγαὶ ἢ δεσμοὶ ἢ θάνατοι ὦσιν. ταῦτα γάρ φασιν ἀναγκασθέντες πρᾶξαι. ἢ οὔ, ἀλλὰ πάντες ἑκόντες ποιοῦσιν αὐτὸ τοῦτο; ἔξεστι γὰρ μὴ ποιεῖν, ἀλλ’ ἐκεῖνο ὑπομεῖναι τὸ πάθος. [The translation is mine, closely following Woods’.]
 EE 1225a8-19: “Alternatively, someone might say that some of these things are [voluntary], but others not. For all things of that kind that are such that it is within someone’s power whether they come not about or come about, one always – without wishing to do them – does these things voluntarily and not by force; but things of that sort which are not within one’s power one does, in a way, through force, but not without qualification because one does not choose the very thing one does but the thing for the sake of which one does it, albeit in these, also, there is some difference.” ἔτι ἴσως τούτων τὰ μὲν φαίη τις ἂν τὰ δ’ οὔ. ὅσα μὲν γὰρ ἐφ’ αὑτῷ τῶν τοιούτων μὴ ὑπάρξαι ἢ ὑπάρξαι, ἀεὶ ὅσα πράττει ἃ μὴ βούλεται, ἑκὼν πράττει, καὶ οὐ βίᾳ· ὅσα δὲ μὴ ἐφ’ αὑτῷ τῶν τοιούτων, βίᾳ πώς, οὐ μέντοι γ’ ἁπλῶς, ὅτι οὐκ αὐτὸ τοῦτο προαιρεῖται ὃ πράττει, ἀλλ’ οὗ ἕνεκα, ἐπεὶ καὶ ἐν τούτοις ἐστί τις διαφορά. εἰ γὰρ ἵνα μὴ λάβῃ ψηλαφῶν ἀποκτείνῃ, γελοῖος ἂν εἴη, εἰ λέγοι ὅτι βίᾳ καὶ ἀναγκαζόμενος, ἀλλὰ δεῖ μεῖζον κακὸν καὶ λυπηρότερον εἶναι, ὃ πείσεται μὴ ποιήσας. οὕτω γὰρ ἀναγκαζόμενος, καὶ [μὴ] βίᾳ, πράξει, ἢ οὐ φύσει áγε,ñ ὅταν κακὸν ἀγαθοῦ ἕνεκα ἢ μείζονος κακοῦ ἀπολύσεως πράττῃ, καὶ ἄκων γε· οὐ γὰρ ἐφ’ αὑτῷ ταῦτα.
 For a critique of this argumentation see: Heinaman 1988. Echeñique presents an objection of Heinaman’s argumentation relying on the conception of hypothetical necessity in Phys. II 9 (Echeñique 2012: 119-123).
 NE 1110a4-19: ὅσα δὲ διὰ φόβον μειζόνων κακῶν πράττεται ἢ διὰ καλόν τι, οἷον εἰ τύραννος προστάττοι αἰσχρόν τι πρᾶξαι κύριος ὢν γονέων καὶ τέκνων, καὶ πράξαντος μὲν σῴζοιντο μὴ πράξαντος δ’ἀποθνήσκοιεν, ἀμφισβήτησιν ἔχει πότερον ἀκούσιά ἐστιν ἢ ἑκούσια. τοιοῦτον δέ τι συμβαίνει καὶ περὶ τὰς ἐν τοῖς χειμῶσιν ἐκβολάς× ἁπλῶς μὲν γὰρ οὐδεὶς ἀποβάλλεται ἑκών, ἐπὶ σωτηρίᾳ δ’ αὑτοῦ καὶ τῶν λοιπῶν ἅπαντες οἱ νοῦν ἔχοντες. μικταὶ μὲν οὖν εἰσιν αἱ τοιαῦται πράξεις, ἐοίκασι δὲ μᾶλλον ἑκουσίοις× αἱρεταὶ γάρ εἰσι τότε ὅτε πράττονται, τὸ δὲ τέλος τῆς πράξεως κατὰ τὸν καιρόν ἐστιν. καὶ τὸ ἑκούσιον δὴ καὶ τὸ ἀκούσιον, ὅτε πράττει, λεκτέον. πράττει δὲ ἑκών× καὶ γὰρ ἡ ἀρχὴ τοῦ κινεῖν τὰ ὀργανικὰ μέρη ἐν ταῖς τοιαύταις πράξεσιν ἐν αὐτῷ ἐστίν× ὧν δ’ ἐν αὐτῷ ἡ ἀρχή, ἐπ’ αὐτῷ καὶ τὸ πράττειν καὶ μή. ἑκούσια δὴ τὰ τοιαῦτα, ἁπλῶς δ’ ἴσως ἀκούσια× οὐδεὶς γὰρ ἂν ἕλοιτο καθ’ αὑτὸ τῶν τοιούτων οὐδέν.
 NE 1110a19-23: ἐπὶ ταῖς πράξεσι δὲ ταῖς τοιαύταις ἐνίοτε καὶ ἐπαινοῦνται, ὅταν αἰσχρόν τι ἢ λυπηρὸν ὑπομένωσιν ἀντὶ μεγάλων καὶ καλῶν· ἂν δ’ ἀνάπαλιν, ψέγονται· τὰ γὰρ αἴσχισθ’ ὑπομεῖναι ἐπὶ μηδενὶ καλῷ ἢ μετρίῳ φαύλου.
 Cf. EE 1225a14-19: “Thus if someone were to kill in order to prevent someone from catching hold of him, it would be absurd if he said that he did so under force, and because he was compelled to do it; the evil which he is going to suffer if he does not do the thing has to be greater and more unpleasant. For a man will, in this way, be acting because he is compelled and [not] under force, or not naturally at any rate, whenever he does evil for the sake of a good, or the removal of a greater evil, and he will be acting involuntarily, as those things are not within his control.” εἰ γὰρ ἵνα μὴ λάβῃ ψηλαφῶν ἀποκτείνῃ, γελοῖος ἂν εἴη, εἰ λέγοι ὅτι βίᾳ καὶ ἀναγκαζόμενος, ἀλλὰ δεῖ μεῖζον κακὸν καὶ λυπηρότερον εἶναι, ὃ πείσεται μὴ ποιήσας. οὕτω γὰρ ἀναγκαζόμενος, καὶ [μὴ] βίᾳ, πράξει, ἢ οὐ φύσει áγε,ñ ὅταν κακὸν ἀγαθοῦ ἕνεκα ἢ μείζονος κακοῦ ἀπολύσεως πράττῃ, καὶ ἄκων γε· οὐ γὰρ ἐφ’ αὑτῷ ταῦτα. I quote Woods’ translation with minor changes (Woods 2005). Two comments on the passage: First, Fritzsche and others bracketed the negation “μή” in 1225a17, although all MSS contain it. Second, commentators (e.g. Rackham, Dirlmeier, Woods) suggest that “ψηλαφῶν” refers to an innocuous game like the blind-man’s bluff.
 Cf. e.g. NE 1109b30-35: “Since virtue is concerned with emotions and actions, and since praise and blame are accorded to those that are voluntary and excuse, and sometimes pity as well, to those that are involuntary, it is presumably necessary for those who are investigating the topic of virtue to define the voluntary and the involuntary, and it is also useful for legislators with a view to honors and punishments.” (Taylor’s translation with minor changes.) τῆς ἀρετῆς δὴ περὶ πάθη τε καὶ πράξεις οὔσης, καὶ ἐπὶ μὲν τοῖς ἑκουσίοις ἐπαίνων καὶ ψόγων γινομένων, ἐπὶ δὲ τοῖς ἀκουσίοις συγγνώμης, ἐνίοτε δὲ καὶ ἐλέου, τὸ ἑκούσιον καὶ τὸ ἀκούσιον ἀναγκαῖον ἴσως διορίσαι τοῖς περὶ ἀρετῆς ἐπισκοποῦσι, χρήσιμον δὲ καὶ τοῖς νομοθετοῦσι πρός τε τὰς τιμὰς καὶ τὰς κολάσεις. See also EE 1223a9-20.
 Some commentators tend to prevent the conjecture that Aristotle allows for dirty hands acts by suggesting only slightly and not morally bad actions as the actions that an agent chooses to do under force in order to achieve a better end (for discussion, see Gauthier/Jolif 1970: 175). Aspasius, for instance, gives the example of a man who obeys a tyrant’s order to parade in public in women’s clothing, in order to save his city from destruction and his family from death (Aspasius 61, 26-28). Choosing this example allows arguing that the mixed action is not a morally bad action and therefore does not comply the first condition of dirty hands acts. But this maneuver is unavailing in my opinion, since it is indeed a conceivable case that the tyrant orders a virtuous person to carry out a morally bad action and Aristotle, consequently, cannot avoid assuming situations that require a choice between two morally bad actions.
 Nielsen uses – following a suggestion of Jennifer Whiting – the telling phrase “stereotypically” shameful or bad acts (Nielsen 2007: 282) depicting cases of actions that are shameful and bad under normal circumstances (kath’ hauto) and that no one would choose for their own sake, whereas their bad character may be cancelled out by specific circumstances of the action which turn it into the right action on this occasion.
 EE 1225a21-23+26-30: “That is why many classify even passionate love as involuntary, and certain cases of anger and the natural things, because they are too strong and overwhelm the (human) nature; and we regard them as excusable, since they naturally overpower the (human) nature. […] What someone’s nature is not able to withstand, and what is not naturally in the scope of one’s natural inclination or reasoning, is not in one’s power. That is why even with those who are possessed by divine inspiration or utter prophecies, though they produce a work of thought, we say that it was not under their control to say what they said or to do what they did. “ διὸ καὶ τὸν ἔρωτα πολλοὶ ἀκούσιον τιθέασιν, καὶ θυμοὺς ἐνίους καὶ τὰ φυσικά, ὅτι ἰσχυρὰ καὶ ὑπὲρ τὴν φύσιν· καὶ συγγνώμην ἔχομεν ὡς πεφυκότα βιάζεσθαι τὴν φύσιν. […] ὃ δὲ μὴ οἵα τε, μήδ’ ἐστὶ τῆς ἐκείνου φύσει ὀρέξεως ἢ λογισμοῦ, οὐκ ἐφ’ αὑτῷ. διὸ καὶ τοὺς ἐνθουσιῶντας καὶ προλέγοντας, καίπερ διανοίας ἔργον ποιοῦντας, ὅμως οὔ φαμεν ἐφ’ αὑτοῖς εἶναι οὔτ’ εἰπεῖν ἃ εἶπον, οὔτε πρᾶξαι ἃ ἔπραξαν. ἀλλὰ μὴν οὐδὲ δι’ ἐπιθυμίαν· [Translation is mine closely based on Woods’.]
 NE 1110a23-26: ἐπ’ ἐνίοις δ’ ἔπαινος μὲν οὐ γίνεται, συγγνώμη δ’, ὅταν διὰ τοιαῦτα πράξῃ τις ἃ μὴ δεῖ, ἃ τὴν ἀνθρωπίνην φύσιν ὑπερτείνει καὶ μηδεὶς ἂν ὑπομείναι.
 When Aristotle cites these examples, he refers to the common opinion, perhaps in order to dissociate himself in a way from this view. Besides, Taylor points to the apparently paradoxical fashion of the idea that someone’s nature is overpowered by his own natural states (Taylor 2005: 135). But this may be adjusted by saying that the human nature can produce states that evolve in a way that they become such that they are no more in the control of the agent.
 Cf. NE 1111a30-33: “Further, how do cases where we go wrong in rational calculation differ in respect of being involuntary from those where we go wrong through anger? Both are to be avoided, and the non-rational feelings appear no less proper to humans, so that actions from anger and desire are no less human actions. So it is absurd to class them as involuntary.” [ἔτι δὲ τί διαφέρει τῷ ἀκούσια εἶναι τὰ κατὰ λογισμὸν ἢ θυμὸν ἁμαρτηθέντα; φευκτὰ μὲν γὰρ ἄμφω, δοκεῖ δὲ οὐχ ἧττον ἀνθρωπικὰ εἶναι τὰ ἄλογα πάθη, ὥστε καὶ αἱ πράξεις τοῦ ἀνθρώπου áαἱñ ἀπὸ θυμοῦ καὶ ἐπιθυμίας. ἄτοπον δὴ τὸ τιθέναι ἀκούσια ταῦτα.] For the voluntariness of emotions: see also NE 1109b-30-32, quoted in note 18.
 See e.g. NE 1109b-30-32, quoted in note 18. The meaning and the translations of συγγώμη ranges from “excuse” and “pardon” to “mitigation” and “sympathy”. Authors, e.g. Kenny and Taylor (Kenny 1979: 28; Taylor 2006: 135), have brought out the ambiguity of the Greek term which is relevant for the interpretation, because some meanings (e.g. excuse) establish that the agent acted involuntarily and therefore did no wrong, whereas other meanings (e.g. sympathy) presuppose that the agent acted voluntarily and did wrong, but the punishment should be withheld due to mitigating factors.
 NE 1110a26-b1: ἔνια δ’ ἴσως οὐκ ἔστιν ἀναγκασθῆναι, ἀλλὰ μᾶλλον ἀποθανετέον παθόντι τὰ δεινότατα· καὶ γὰρ τὸν Εὐριπίδου Ἀλκμαίωνα γελοῖα φαίνεται τὰ ἀναγκάσαντα μητροκτονῆσαι. ἔστι δὲ χαλεπὸν ἐνίοτε διακρῖναι ποῖον ἀντὶ ποίου αἱρετέον καὶ τί ἀντὶ τίνος ὑπομενετέον, ἔτι δὲ χαλεπώτερον ἐμμεῖναι τοῖς γνωσθεῖσιν· ὡς γὰρ ἐπὶ τὸ πολύ ἐστι τὰ μὲν προσδοκώμενα λυπηρά, ἂ δ’ ἀναγκάζονται αἰσχρά, ὅθεν ἐπαινοι καὶ ψόγοι γίνονται περὶ τοὺς ἀναγκασθέντας ἢ μή.
 Meyer, for example, draws a distinction between “psychological compulsion” and “rational compulsion” (Meyer 2011: 98-100) and Taylor distinguishes between a “psychological thesis” and a “moral thesis” meaning that the psychological view applies to the third type, while the rational/moral approach is related to the fourth type.