ANNOUNCEMENT Brian J. Shanley, O.P., Editor of The Thomist since 2002, became President of Providence College on July 1, 2005. Because of these new responsibilities, Fr. Shanley has stepped down as Editor. We are pleased to announce that he will remain on the editorial board as an Associate Editor. In addition, Fr. Lawrence Donohoo, O.P., has been named an Associate Editor. The new Editor is Joseph Torchia, O.P., Associate Professor of Philosophy at Providence College. Father Torchia holds doctorates in Philosophy (Fordham University) and Early Christian Studies (The Catholic University of America), and an S.T.L. from the Dominican House of Studies in Washington. The Thomist 69 (2005) EDITOR'S NOTE JOSEPH TORCHIA, 0.P. I consider it a great honor to be named editor of a review which has defined itself as "Steward of the Thomistic tradition." As I begin my tenure in this capacity, I wish to comment briefly about my own sense of our mission, not in the language of some pronouncement, but merely as one man's opinion about the task implicit in our somewhat lofty motto. In my reckoning, this task encompasses two complementary roles. On the one hand, we seek serious scholarly investigations into the sources and development of Christian wisdom that focus primarily (but not exclusively) upon the thought of St. Thomas Aquinas. But good stewardship involves more than a custodial role. Reflection on the riches of the past must also inspire and sustain a response to contemporary problems. This publication has consistently proceeded from the premise that Thomism is a living tradition that can speak to those issues that run to the core of our being as rational creatures made in the image and likeness of God. Shortly after assuming the editorship, I examined The Thomist's 1939 "Announcement of Publication." I was immediately struck by the enthusiasm for a new venture "launched on the postulate that Dominicans have something very special to offer this twentieth century of ours. " 1 As I consider these words some sixty-six years later, I propose the following question: what can a distinctively Dominican speculative review of theology and philosophy offer at the beginning of this new millennium? Indeed, the vision that inspired and guided our founders must be ours as well-nothing less than a contribution "to originality of thought, to solutions rather than compromises with immediately pressing questions. " 2 These goals, in fact, are wholly consistent with the Dominican charism and its commitment to learned preaching 1 2 Excerpt from The Thomist's "Announcement of Publication," 1. Ibid., 15. EDITOR'S NOTE according to the ideal of contemplare et contemplata aliis tradere, that is, "to contemplate and bring what is contemplated to others." This ideal rests on the conviction that the life of the mind must translate into a passion to share its fruits in a concrete context. Our model in this endeavor is St. Thomas Aquinas himself. Aquinas fixed his contemplative gaze on truth. But ever the teacher, he sought to communicate truth in conversance with all the intellectual currents of his time. In the process, he forged a bold new synthesis of faith and reason versatile enough to speak to a wide range of outlooks and horizons of inquiry. This is why he could adapt and harmonize traditions that his contemporaries perceived as wholly incommensurable with such creativity and imagination. If Aquinas is the "perennial philosopher," it is because his thought affirms an openness to truth wherever it can be discerned. By the same token, he shows us that any search for truth must be guided by that Divine Wisdom whose ultimate source is in God, the ground of meaning and intelligibility. This is the Light that guided him throughout his inquiries, and it must be the Light that guides us in our own, as we confront the unique moral dilemmas that the present century opens to Catholic thinkers. This Light never undermines human reasoning; it only perfects it, and thereby raises the bar of our rational efforts to new heights. We live in an age that presents great challenges to those committed to the intellectual enterprise of the Catholic Church. And for those who endorse the Thomistic vision of reality, the need for "originality of thought" and "solutions rather than compromises" is more compelling than ever before in the history of this review. At a time when pluralism is widely valued above objectivity, it is all too easy to take refuge in one's own conceptual framework or ideological perspective, and thereby to exclude the possibility of genuine dialogue with one's opponents at the outset. But in the spirit of Aquinas, we do not perceive disagreement as an insurmountable obstacle. Rather, we welcome it as the occasion to advance the debate, and in the process to explore the possibility of consensus and a resolution of problems, even while EDITOR'S NOTE upholding those absolute principles that must be presupposed in any rational discussion. In his encyclical Fides et Ratio, John Paul II spoke of a hermeneutical crisis in which "many people wonder whether it still makes sense to ask about meaning. "3 By way of a response, he encouraged a recovery of philosophy's sapiential dimension in the search for the ultimate meaning of human existence. From this standpoint, the medieval understanding of theology as {ides quaerens intellectum provides a contemporary incentive to challenge the uncritical presuppositions of a postmodernist culture and its nihilist bent with the considerable dialectical resources at our disposal. In my estimation, The Thomist offers an excellent forum for engaging such issues and their wide-ranging implications with cogency, incisiveness, and depth. It is incumbent upon us to assume an active and even confrontational role in this endeavor, not necessarily in an adversarial manner, but in the best sense of Dominican disputatio. This year, the Eastern Dominican Province of St. Joseph celebrates the Bicentennial of its founding in the United States. The Thomist stands as one of the most salient legacies of its intellectual apostolate. As we move toward our seventieth year of publication, we seek to maintain this legacy, and in the words of our founders, to furnish material "fitting the dignity of human nature and . . . the joyous work of Wisdom. "4 But the ongoing excellence of this review depends, as always, on the input of our contributors, including that of my fellow Dominicans. As sons of St. Dominic, our common commitment lies in the pursuit of Veritas, with all its intriguing textures and nuances, wherever that pursuit may lead us and whatever the cost, "of one heart and mind in God." 3 Encyclical Letter Fides et Ratio of the Supreme Pontiff John Paul II to the Bishops of the Catholic Church on the Relationship Between Faith and Reason (September 15, 1998), para. 81. 4 Excerpt from The Thomist's "Announcement of Publication," 15. The Thomist 69 (2005): 341-70 AQUINAS ON DEFENSIVE KILLING: A CASE OF DOUBLE EFFECT? GREGORY M. REICHBERG International Peace Research Institute, Oslo (PRIO) Oslo, Norway T HENOW-STANDARDREADINGofAquinas'saccountof private self-defense in STh 11-11, q. 64, a. 7 holds that it precludes any kind of deliberate killing. Defensive action that causes the death of an assailant will be morally permissible only when this lethal outcome has the character of a pure side effect. Hence, when Aquinas asserts that in self-defense "the killing of the attacker" is "beyond the intention" (praeter intentionem) of the defender, he is taken to be formally denying that a private defender can justifiably choose (however reluctantly) to kill his assailant in order to protect himself from grave harm. Article 7 is thus said to exemplify what has since come to be called 'the principle of double effect' (PDE). 'Double effect' is the heading under which the ethical quandaries surrounding side-effect harm have traditionally been discussed in philosophy. This term is shorthand for the two different kinds of effects that can emerge from our actions. On the one hand, there is the very state of affairs that our actions are meant to produce. This goal we will succeed at achieving more or less well, depending on our skill. On the other hand, there are the side effects that result from this deliberate intervention in the world. The idea that we are answerable for these side effects, yet in a manner different from the way we are accountable for our intentional projects, has been dubbed the 'principle of double effect' (PDE). This principle holds that while we can never be justified in deliberately willing a wrong, we may sometimes have 341 342 GREGORY M. REICHBERG moral warrant for allowing harmful side effects to eventuate from our otherwise good actions. 1 My argument here is that Aquinas did not in fact appeal to a version of PDE in formulating his theory of justifiable defensive killing in STh II-II, q. 64, a. 7. The broader context of his other statements on the morality of participation in violence, as well as the twelfth- and thirteenth-century canon-law teaching on this topic, indicate that he did not take self-defense to be an act from which harmful side effects (including the death of the assailant) might flow. He viewed self-defense instead as an aim that might justify the application of proportionate force against an assailant, even to the point of deliberately causing his death, if this were the only effective measure available at the time. Aquinas recognized that lethal force could also be applied with a very different purpose in mind, namely, to avenge a wrong. It was precisely by way of contrast with this latter aim that Aquinas described defensive killing as "praeter intentionem." I It is undeniable that the name PDE was originally taken from a sentence in the responsio to STh II-II, q. 64, a. 7 ("from the act of self-defense there follows a double effect"). Yet whether a doctrine of double effect may be found therein is a matter of some dispute. When Aquinas states that a private individual who engages in lethal self-defense must aim solely at preserving his own life, such that the death of his assailant would lie outside of his intention (praeter intentionem), does he mean to assert that an agent of this kind may never deliberately inflict death upon another as a means of saving himself? In other words, does this article restrict the scope of private self-defense to actions which, although they may foreseeably result in the attacker's death, can never be chosen precisely with that outcome in mind? 1 For further elucidations on PDE, see Gregory Reichberg and Henrik Syse, "The Idea of Double Effect-In War and Business," in L. Bomann Larsen and 0. Wiggen, eds., Responsibility in World Business:ManagingHarmful Side-effectsof CorporateActivity (Tokyo: United Nations University Press, 2004), 17-38. AQUINAS ON DEFENSIVEKILLING 343 Drawing on Cajetan's Summa commentary, Joseph Boyle has argued for an affirmative answer to this question. 2 His argument is built around an analysis of the term praeter intentionem (as it is used in various works of Aquinas), which, he maintains, must be taken to exclude any deliberate selection of an item, either as end or as means. This leads Boyle to conclude that lethal self-defense will be allowable only under condition that the assailant's death is not what ends the threat, but is rather a consequence of what stops the attack. In such a case one is not saved because the assailant is dead but the assailant dies because one has stopped the attack. 3 In a more recent treatment, 4 Boyle, with co-authors John Finnis and Germain Grisez, argues that this theory of lethal self-defense not only holds for private individuals, but, more broadly, should also apply to the killing of enemy combatants in war. 5 For even there military action must be directed toward stopping the enemy's unjust use of military force, not toward killing those who are bringing that force to bear. By requiring that the death of an enemy soldier be brought about only as a sideeffect of a military act having a different appropriate object, our moral theory would limit warfare as stringently as possible to the pursuit of the good purposes which can justify it. 6 Underlying this reasoning appears to be a twofold assumption: first, that punishment is no longer an acceptable war-aim under current international law; and second, that any deliberate killing in war will necessarily be punitive in character. The exclusion of the first (punitive war) thus entails the exclusion of the latter (the 2 Joseph Boyle, "Praeter intentionem in Aquinas," The Thomist 42 (1978): 649-65. 3 Ibid., 661. 4 John Finnis, Joseph M. Boyle, Jr., and Germain Grisez, Nuclear Deterrence, Morality and Realism (Oxford: Clarendon Press, 1987). 5 In this regard, the authors note (ibid., 315) that their "just-war theory is more restrictive than traditional versions, even that of Thomas Aquinas." In a more recent text (Aquinas: Moral, Political, and Legal Theory [Oxford: Oxford University Press, 1998]) Finnis clarifies that this application of PDE to the targeting of enemy combatants in war is within the logic, but not the letter, of Aquinas's teaching, since the latter judges "acceptable some intentional killing in capital punishment and war" (287). 6 Finnis, Boyle, and Grisez, Nuclear Deterrence, 315 344 GREGORYM. &EICHBERG deliberate killing of active enemy combatants on the battlefield). Recognizing the continued moral viability of defensive killing in war, the authors are compelled to justify it by an appeal to PDE. The scope of this article does not allow for more detailed discussion of this issue in military ethics. For our present purpose, it should be noted nonetheless that it is the second of the two assumptions which stands out as most problematic. Must one assume that deliberate killing (whether public or private) must always have a punitive purpose? Or, on the contrary, is it possible for killing to be deliberate yet purely defensive? Taking up this last point, Gareth Matthews 7 argues that neither the textual evidence nor sound reasoning warrants a reading of STh 11-11, q. 64, a. 7 in which deliberate killing is wholly excluded as a means of private self-defense. It is with respect to the second aspect-the philosophical plausibility of applying double-effect reasoning to licit self-defense-that Matthews's analysis is most trenchant. Unlike the standard examples of double effect (as, for instance, in medical ethics, when it is argued that causing the death of a fetus can be allowed as a side effect which results from the removal of a pregnant woman's cancerous uterus), where harm is certainly foreseen but is in no way intended, self-defense, by its nature, involves the deliberate infliction of some harm on the opponent (striking him with one's hands or a weapon). Thus, in his rebuttal of Boyle's thesis, Mathews notes that even in a case in which the assailant was hit and knocked unconscious, but not killed, there would again be a bad effect-hitting him and knocking him unconscious-as well as the good effect of defending my life against an attack. And in this case, too, the bad effect would be the means by which the good effect was achieved 8• Indeed, the very notion of side-effect harm presupposes the existence of some prior act which is not itself a side effect (otherwise an infinite regress would result) but which is itself something deliberately posited in the world. In the case of lethal self-defense 7 Gareth Matthews, "Saint Thomas and the Principle of Double Effect," in Scott MacDonald and Eleonore Snunp, Aquinas's Moral Theory (Ithaca and London: Cornell University Press, 1995}, 63-78. 8 Ibid., 69. AQUINAS ON DEFENSIVE KILLING 345 this prior act will be nothing other than the application of force against an assailant, an application which is chosen precisely as a means of "stopping the attack." PDE teaches that agents will be absolved from the liability which ordinarily attaches to the production of harmful side effects only if (at a minimum) the deliberate activity which gave rise to the said side effects was itself not blameworthy. By contrast, the negative side effects that follow from the commission of a crime (or, more generally, from any morally wrongful deed) are ascribable to the agent as an aggravating condition for which he will be held accountable, regardless of the fact that these were wholly undesired by the agent and in no way contributed to the commission of his crime or the enjoyment of its illicit gains. 9 Thus, in line with Matthews, when we apply this reasoning to Boyle's analysis of lethal self-defense in STh II-II, q. 64, a. 7, a manifest problem emerges. As an instantiation of PDE, defensive killing will be allowable only when it has the character of a pure side effect which flows from an ostensively good act. Therein lies the rub. The deliberate application of force by persons in a position of authority (private individuals) is, on Aquinas's account, prima facie wrong: "It is not permitted for a man to strike another," he writes in the subsequent question (STh II-II, q. 65, a. 2), "unless he have some authority over the one whom he strikes. " 10 On such an act (private violence) he confers the name rixa (strife) which, like the kindred acts helium (war) and seditio (sedition) denote kinds of sin. 11 My point is not that, for Aquinas, the application of force by private individuals can never be justified. On the contrary, in STh II-II, q. 41, a. 1 ("De rixa"), he distinguishes sinful striking from the repelling of injury by force that is proper to the individual who engages in sinless self-defense. It is essential to note, however, that defensive striking requires justification no less than See STh I-II, q. 73, a. 8. aliquem non licet nisi habenti potestatem aliquam super ilium qui verberatur." 11 For an account of Aquinas's conception of "sins against peace," see Gregory M. Reichberg, "Is There a 'Presumption Against War' in Aquinas's Ethics?" The Thomist 66 (July 2002): 337-67. 9 10 "Et ideo verberare 346 GREGORY M. REICHBERG the parallel case of defensive killing. Indeed, Aquinas employs nearly the same argumentation in his treatment of these two types of defensive reaction (see section III, below). Hence, if his aim in STh II-II, q. 64, a. 7 was to show that defensive killing will only be justified when it issues as the pure side effect of an otherwise good act, he would need to mount much the same argument in favor of defensive striking, thereby denying that it too could ever be the legitimate target of deliberate action by private individuals. But if defensive striking will thus be justified only when it has the character of a pure side effect, we cannot by the same token appeal to it in our justification of the defensive killing without giving rise to an infinite regress and hence a vicious circularity in our reasonmg. One could of course attempt to get around this objection by limiting private self-defense to some sort of mere blocking motion, yet it would be hard to find support for this minimalist construal of self-defense in Aquinas, or in the canon-law writings of his day, which provided the intellectual background for his comments on this topic (see section II, below). It also cuts against widespread contemporary moral and legal intuitions about justifiable self-defense, which generally allow for using hands or weapons to strike at an aggressor, as long as the response is proportionate to the scale of the attack. Another strategy would consist in arguing that lethal selfdefense is allowable only in cases where the repelling motion can reasonably be thought to succeed without killing the attacker, yet where the defender acts with some margin of incertitude with respect to the lethality of his response. Not having a great deal of practice at stopping attackers by hitting them with rocks, I might not have been able to judge accurately how much force would be needed to stun but not kill my assailant. 12 Here killing would have the character of an accidental outcome. It would be akin to the risky adventure of a mountaineer, who, despite all precautions to the contrary, is nevertheless killed by an avalanche: he thereby succumbs to an eventuality that he knew to 12 Matthews, "Saint Thomas and the Principle of Double Effect," 68. AQUINAS ON DEFENSIVEKILLING 347 be a distinct possibility from the outset, but which he had hoped to avoid. It is questionable, however, whether Aquinas ever intended to limit defensive killing praeter intentionem to the sort of case just envisioned. Indeed, the fact that Aquinas made accidental killing the express topic of the next article (STh II-II, q. 64, a. 8) 13 suggests that this kind of occurrence was not his particular concern in article 7. Another twist on this argument would hold that defensive killing by private individuals is allowable only when it occurs under conditions of strict urgency, such as when a defender has little or no time to consider whether his actions will result in the death of the attacker. In this instance, the press of danger, hence the necessity to take immediate action in the interests of selfpreservation, will render void any meaningful attempt to foresee the lethality of one's defensive response. The death of the attacker could thus be described as beyond the defender's intention, for the good reason that the latter never had sufficient time to deliberate about the likelihood of a lethal outcome. While not a wholly implausible interpretation of Aquinas's use of praeter intentionem in STh II-II, q. 64, a. 7, it remains that this (much like the argument based on uncertainty of outcome) would be a lessthan-suitable illustration of the principle of double effect, since this principle is meant to explain, most paradigmatically, situations where an agent foresees, with clarity, that a deadly effect will flow from his otherwise good action. Moreover, those like Boyle who read STh II-II, q. 64, a. 7 as an illustration of PDE do in fact construe this article as teaching that "the death of the attacker is praeter intentionem even though it can be foreseen with certainty to follow. " 14 Thus, when Boyle asserts that "the death of the attacker is not a means in those cases where Aquinas regards it as praeter intentionem, " 15 this proposition is clearly meant to cover a situation in which the defender is fully aware that the lethal effect will inevitably flow from his action, as for example, when a man spears an assailant's heart in self-defense 13 "Whether one is guilty of homicide through killing someone by chance [causaliter]." in Aquinas," 660. 14 Boyle, "Praeter intentionem 15 Ibid., 661. 348 GREGORY M. REICHBERG without intending to kill him (to borrow Finnis's provocative example). 16 And it is precisely the conjunction of these two propositions ("death of the attacker is ... foreseen with certainty to follow" and ". . . is not a means") that Matthews and others like him 17 consider highly dubious when applied to the special case of self-defense. Is the Boyle/Finnis reading of Aquinas accurate? Can one deliberately spear the heart of an assailant without intending to kill him? The difficulty in question does not arise because PDE is thought to be incompatible with a condition of definite foresight regarding the negative side effect, or because causing another's death with such foresight would, given its special gravity, not be covered by the principle. In military ethics, for instance, it is generally recognized that some missions will be justified, on grounds of PDE, in spite of a recognition that civilian casualties will ineluctably follow. Thus, during the Second World War, Norwegian resistance fighters sabotaged a ferry that was carrying their countrymen across a lake. Their goal was to sink the ship, in order to prevent the occupying German force from transporting a cargo of heavy water from the Vemork plant to the railway lines 16 "If, as Aquinas seems to assert and never denies, one can spear an assailant's heart in selfdefense without intending to kill, it is possible to wage war, too-lethally and often successfully-without that intent" (Finnis, Aquinas, 287). 17 For a similar position, see Steven A. Long, "A Brief Disquisition regarding the Nature of the Object of the Moral Act according to St. Thomas Aquinas," The Thomist 67 (2003): 45-71, see especially 51-53, 56-68. A number of Aquinas's commentators (e.g., Vitoria, Vasquez, De Soto, Lessius) have likewise interpreted Sfh 11-11, q. 64, a. 7 as teaching that a defender can lawfully will the death of his attacker as a means of self-preservation. Hence they have (implicitly) rejected the double-effect reading of this text. For an historical survey in support of this position, see Vincentius M. Alonso, S.J., El principio del doble efecto en los comentados de Santo Tomas de Aquino (diss., Rome, 1937). For the opposing position, see Joseph T. Mangan, S.J., "An Historical Analysis of the Principle of Double Effect," Theological Studies 10 (1949): 41-61. This author erroneously attributes to Vitoria the idea that Sfh 11-11, q. 64, a. 7 represents a teaching on double effect. In the text alleged (Commentarios a la Secunda Secundae de Santo Tomas, q. 64, a. 7 [Latin text and English translation in J. P. Doyle, ed., Francisco de Vitoria, O.P., On Homicide (Milwaukee: Marquette University Press, 1997)]) , after distinguishing between intention (intentio) and choice (e/ectio), Vitoria states that while it is not licit (for a private individual) to intend as an end the death of the attacker ("non licet propter se intendere mortem alterius"), this can nevertheless be willed (i.e., be chosen as a means) when it is necessary for self-defense (§2, Sed quia [Doyle, ed., 194-95]). AQUINAS ON DEFENSIVEKILLING 349 on the other side of the lake. Heavy water was an ingredient then vital to the construction of an atomic bomb. The resistance fighters knew in advance that many of the civilians on board would die, yet there was no way to warn them of the danger, as the mission would thereby have been endangered. As illustrated by this example, it is neither the ex ante certitude of a negative result, nor the special gravity of human death, that raises a doubt about whether defensive killing belongs within the category of side-effect harm. Rather, the problem arises because in this particular case the person against whom the defender deliberately directs his blows and the one who suffers death (purportedly as a side-effect harm) are the self-same subject. Moreover and most importantly, the defender clearly stands to benefit from the harm that he has done to the assailant, since this is precisely what stops the latter's wrongful attack. Thus it is disingenuous to speak of this harm, and the resulting death, as though they were pure externalities that were of no benefit whatsoever to the defender. The blow that stops the attack has the character of a means which, at least when self-defense is virtuously carried out, will be reluctantly chosen (the defender takes no pleasure in the harm done), but is a means nonetheless. By contrast, when PDE is used to justify collateral damage in wartime, as in the Vemork example given above, there is a clear differentiation between the target of the intentional action (the German military personnel and their cargo of heavy water) and the innocent passengers on board who died as a consequence of the attack. No harm whatsoever was intended against the latter. Their death was in no way beneficial to the military operation; the Norwegian saboteurs would have much preferred to adopt an alternative course of action had one been available. One could still argue that defensive killing will be legitimate only under condition that, counterfactually and on the level of intention, the agent would have chosen a nonlethal means of response, had one been available at the precise moment when the aggression took place. Under such a supposition, it would be correct to say that the defender did not actually set out to kill the attacker. He did not choose this particular means precisely 350 GREGORY M. REICHBERG because of its lethal qualities, but simply because at the time it was "the only available effective defensive measure." 18 Imagine a man being pursued up a steep mountain path by an armed assailant. The former spots a loose boulder, and, seeing that this represents his only chance of survival, nudges it downwards in the direction of his pursuer, who will undoubtedly be killed by the impact. Lethality was not the motivating factor in this defender's choice of means; he acted despite this factor, rather than because of it. On this basis,19 Boyle concludes that the death of the assailant, although a direct consequence of the defender's forcible response, should nevertheless be described as praeter intentionem. Boyle recognizes, however, that on at least one occasion the term praeter intentionem is used by Aquinas to signify an agent's choice of an unwanted means. This occurs in the Summa contra Gentiles, book III, chapter 6, where, reflecting on Aristotle's example of the sailors who cast their cargo overboard so as to avoid capsizing (Nie, Ethics 1110a8-10), Aquinas comments that although this act is not willed or intended per se, it is nevertheless willed for the sake of something else ("vult propter aliud"), namely, protection from drowning, and in this sense it must be described as a regrettable expedient or means. 20 This sense of praeter intentionem he most often discusses under the heading of the "mixed voluntary"; according to this concept, an act that one ordinarily finds repulsive, hence 'nonvoluntary' in the sense of being antithetical to the will, can nevertheless be rationally desired (chosen) under circumstances of immanent danger. The similarity with the case of lethal self-defense is indeed striking, 21 18 Finnis, Boyle, and Grisez, Nuclear Deterrence, 312. 19 At this juncture, Boyle quotes Cajetan, who (in his commentary on SI'h II-II, q. 64, a. 7), distinguished between killing an assailant as a means ("ut medium ad finem") of selfprotection and killing him as a consequence which follows from the necessity of the end ("ut consequens ex necessitate finis"). The second would be praeter intentionem, and the first not. For an analysis (with a translation of the key passage) of this text, see Long,, "Object of the Moral Act," 53-56. 20 Summa contra Gentiles ill, c. 6 (Marietti edition, S 1907). 21 Interestingly, in one of his treatments of the mixed voluntary, Quodl. 5, q. 5, a. 3 ("Utrum ea quae per timorem fiunt, sint voluntaria"), Aquinas mentions killing as an example of an act that can become desirable in particular circumstances, even though in universali consideratum it is an unqualified evil. The special case of self-defense is not mentioned. AQUINAS ON DEFENSIVE KILLING 351 yet Boyle resists the comparison on grounds that in later texts Aquinas utilizes praeter intentionem solely to designate the will's relationship to side-effect harm. Thus, according to Boyle, in his more mature writings Aquinas appears to have avoided any identification of this term with the related category of the mixed voluntary. This analysis is plausible, yet it is not quite compelling, since it is always possible that Aquinas, for reasons left unstated, simply reverted in STh 11-11, q. 64, a. 7 to the usage he had earlier established in ScG III, c. 6. In any event, whether described as a side effect or as a repugnant means, the mode of defensive killing just described (in which a nonlethal method of resistance would have been chosen had it only been available at the time) must be distinguished from a defensive situation in which someone puts his assailant to death precisely to ensure that he will never be able to attack again. Under this scenario, less than lethal means of defense, although available, would not be chosen, since these would provide only ineffective protection against attacks that the aggressor would likely renew in the future. Imagine a woman trapped on a boat with a serial killer. Escape being impossible (the boat is on the high seas, and its radio connection with the outside world has been severed), she decides that her safety can be assured only by killing the assailant. No other means will protect her during the time (conceivably quite long) that they will live together on the boat, since, even if she manages to incapacitate him temporarily, he will likely recover to threaten her anew. She finds a gun, and shoots him dead. In this case, the defender directed her choice precisely at the death of her assailant; this tactic she consciously adopts in order to secure herself from the threat of ongoing and future harm. If this were the sort of defensive killing that Aquinas sought to exclude in STh 11-11, q. 64, a. 7, it would indeed make good sense for him to describe, by way of contrast, the other sort of defensive killing as praeter intentionem, even if, at the limit, he did not consider it a pure and simple side effect, but rather a regrettable means that one might adopt in the heat of the moment. 352 GREGORY M. REICHBERG Thus far, we have considered four different ways in which someone, acting in self-defense, might praeter intentionem cause the death of his assailant: (1) mere blocking motions where no harm is directly intended, (2) actions where some harm is intended (e.g., striking an assailant with one's fists) but where a lethal outcome is unforeseen or uncertain, (3) actions chosen under conditions of urgency with little or no time for reflection on the lethal consequences, and (4) actions in which the assailant's death was foreseen with certainty, yet not chosen with that precise end in view, as the defender would have adopted other, nonlethal means, had they only been available at the time. In this last case, the defender may thus seek to disable but not kill his assailant; should he intentionally move beyond this limit, his defense will be illicit. 22 In a brief discussion of STh II-II, q. 64, a. 7, Hugo Grotius, the Dutch theorist of just war, proposed an additional way in which defensive killing might be described as praeter intentionem. It has been well said by Thomas [Aquinas]-if he is rightly understood-that if a man in true self-defense kills his assailant the slaying is not intentional [ex intentione]; not that, if reason supplies no other means of saving oneself, it is not sometimes permissible to do with set purpose [destinato] that which will result in the death of the assailant; rather it is that in such a case his death is not chosen as something primarily intended [quiddam primario intentum], as in a judicial punishment, but is the only resource available at the time. Even under such circumstances the person who is attacked ought to prefer to do anything possible to frighten away or weaken the assailant, rather than cause his death. 23 On this interpretation, when Aquinas distinguished two sorts of killing, intentional and nonintentional, 24 his aim was not exactly 22 Boyle makes it amply clear that, on his reading, Aquinas's theory of private self-defense does not allow for this sort of intentional killing: "So Aquinas admits that the killing of another may be within the intention (presumably as a means) but contrasts it with the kind of deadly self-defense which can be morally justified" ("Praeter intentionem in Aquinas," 661). 23 De iure belli ac pacis, II, I, IV. English translation in G. M. Reichberg, H. Syse, and E. Begby, The Ethics of War: Classic and Contemporary Readings (Oxford: Blackwell, forthcoming 2006). I am grateful to Blackwell Publishing for allowing me to quote from this volume in advance of its publication. 24 In STh II-II, q. 64, a. 7, speaking of private citizens, Aquinas initially refers only to killing praeter intentionem. Later, in the next to last sentence of the responsio, he mentions the case of soldiers and "ministers of a judge" (policemen), who have moral license to kill AQUINAS ON DEFENSIVE KILLING 353 to set up a contrast between harm deliberately inflicted, on the one hand, and side-effect harm, on the other. The contrast was meant instead to differentiate between inflicting harm instrumentally, as a means of self-preservation, versus causing harm as the very goal of one's action. In the latter instance, the agent's primary intention is to visit harm on another, either to avenge a past wrong (i.e., retributive punishment, as in Grotius's example), as an expression of hatred, or because of the pleasure it procures (cruelty). Grotius's primario intentum would then correspond to Aquinas's intentio, an act of the will that is directed to some end. 25 Thus, within the context of STh 11-11, q. 64, a. 7, the killing that s designated as praeter intentionem would formally exclude all noninstrumental rationales for terminating another's life, retribution in particular, which, by the logic of this article, will legitimately be exercised only by public officials acting on behalf of the common good. Killing by private individuals would be restricted strictly to defensive actions, in the sense that all forms of revenge killing would be emphatically ruled out. Yet, so construed, the scope of legitimate defense would be quite broad, since, at the extreme, it would allow for even a deliberate act of killing, if, under the circumstances, this is reasonably viewed as the only effective (and proportionate) response to an attack that is underway or imminent. The limit set by (4) above would be breached, since here (on Grotius's reading), the lethal outcome could itself be chosen, if the attack could not otherwise be repelled. Does Grotius offer a plausible reading of STh II-II, q. 64, a. 7? His proposal is worth considering, as it sums up a view which (as we have seen) 26 a number of Aquinas's Scholastic commentators had already entertained. My assessment of this reading will be two-pronged. First, I will review several key texts from medieval canon law, as these texts would likely have served to frame Aquinas's own treatment of legitimate defense. Grotius himself was well versed in this literature, and it is possible that his intentionally (intendens hominem occidere). 25 See SI'h 1-11, q.12, a. 2. 26 See note 17 above. 354 GREGORY M. REICHBERG interpretation of Aquinas was informed by a reading of the standard canon-law texts on self-defense. Then, against this backdrop, I will scrutinize the logic of STh 11-11, q. 64, a. 7, as well as some related texts (in particular STh 11-11, q. 41, on strife ["De rixa "]) to see whether any support can be found in them for the view of self-defense that Grotius attributes to Aquinas. II If the writings of St. Augustine were Aquinas's sole point of reference in elaborating his teaching in the Summa Theologiae on defensive killing, it could easily be assumed that this teaching was intended by him to be as restrictive as possible, in line with the very strong reserves formulated by his eminent predecessor on the permissibility of killing in self-defense. Augustine, it will be recalled, argued that while it could be virtuous to use lethal force as a means of protecting others from assault, it would be best to abstain from such a measure-and thereby suffer death if necessary-when faced with an attack upon one's own self.27 However, beginning with the Decretist commentaries that appeared in the second half of the twelfth century (the most significant of which was the gloss Qui repellere possunt) the moral standing of forcible self-defense underwent a significant reappraisal in the law schools of the Latin West. The high-water mark in this development may be found in the commentary (ca. 1241) of the Dominican William of Rennes on Raymond of Pefiafort's Summa de casibus poenitentiae (ca. 1235). Given the importance of these texts, and their wide distribution, it is unlikely that Aquinas's treatment of moral questions in the Secunda Secundae (written ca. 1270) would not have been informed by the new and very robust conception of self-defense that had emerged in the canon-law teaching of the preceding decades. 28 27 See Augustine, De libero arbitrio 28 Very little has been written on Aquinas's relationship to medieval canon law, despite the fact that he makes ample reference to Gratian's Decretum, the Decretalsof Gregory IX, etc., in his various writings (see, for example, the long list of his references to canon-law sources in vol. 16 of the Leonine edition of the Summa Theologiae:Auctorum et operum index primus, AQUINAS ON DEFENSIVE KILLING 355 In the Decretum, part 2, causa 23 (which is devoted to the theme of war and violence), 29 Gratian opens the first quaestio with an objection that he seeks to refute in what follows: It would seem that it is contrary to the teaching of the Gospel to serve as a soldier, since the point of all soldiering is either to resist injury or to carry out vengeance; but injury is either warded off from one's own person or from one's associates, both of which are prohibited by the law of the Gospel. 30 The contrast here enunciated between the two chief aims for which Christians might wage war, defense and punishment, became a staple in subsequent canon-law discussions of this topic. The first of these aims (resistance to injury) was taken as the express topic of the gloss Qui repel/ere possunt, 31 which articulates one of the first explicit theories of legitimate defense. The gloss follows ancient Roman law in emphasizing the limits that apply to self-defense: it is legitimate only when exercised in the heat of the moment (incontinenti) against an attack that is imminent or in progress. Moreover, the defender must show due moderation or proportionality: he is entitled to use only so much force as is "ius canonicum," pp. 220-22). He would, in particular, have had an especially close knowledge of Raymond of Pefiafort's Summa de casibus (from which he quoted extensively in book 4 of the Sentences commentary; for documentation, see Jean-Marie Aubert, Le droit romain dans !'oeuvre de saint Thomas [Paris: Vrin, 1955], 32 n. 6), which, as noted by Leonard Boyle ("The Setting of the Summa Theologiae of St. Thomas-Revisited," in S. Pope, ed., The Ethics of Thomas Aquinas [Washington: Georgetown University Press, 2002], 3), "Thomas would have used ... as his springboard for occasional casus and for the regular conventual 'Collationes de moralibus" [at the Dominican studium in Orvieto, during his stay there in 1261-1265]. "By 1261," Boyle adds, "and indeed long before that date, the Summa of the former General of the Dominicans (1238-40) had become an integral part of Dominican training. William ofRennes, Lector of the house of Orleans, had written a valuable Apparatus to it about 1241. ... " See the remainder of the discussion, where, after speculating that "Thomas had a healthy respect for Raymund as both a fine legist and an able moralist," Boyle indicates at least one question in the Summa Theologiae that contains "large and unsuspected borrowings from the Summa de casibus" (ibid.). 29 The work appeared ca. 1140. 30 Decretum Magistri Gratiani, in E. Friedberg, ed, Corpus Juris Canonici, pars prior (Leipzig: Tauchnitz, 1879); translation by Robert Andrews and Peter Haggenmacher in Reichberg, et al., eds., The Ethics of War. 31 The text may be found in Decretum Divi Gratiani una cum glossis & thematibus prudentum, & doctorum suffragio combrobatis (Lyon, 1554), p. 840a-b; this gloss has been attributed to Johannes Teutonicus. 356 GREGORY M. REICHBERG necessary to ward off the attack. The gloss further distinguishes between defense of persons and defense of property. The former allows for some forward-looking (preemptive) action, while the latter generally does not. Moreover, the notion of self-defense also serves to characterize a special kind of war, one that is undertaken precisely as a response to actual or imminent attack. The gloss acknowledges that this sort of war could be undertaken not only by private individuals (for whom no appeal to a higher authority is in principle required) but by polities well. It thereby follows Gratian in positing self-defense as one of several possible just causes of war (alongside punishment and recuperation of stolen property). In this respect, Qui repellere possunt moved well beyond the conception of legitimate defense that had been articulated in ancient Rome, where the relevant laws applied solely to the actions of private individuals, and not to the public domain of war. 32 Very significantly, for our purposes, the gloss raises the question whether it is "permissible for anyone to resist violence by hitting back [repercutiendo]" or "only by obstructing [impediendo] the attack?" In response, the author makes clear that Christians, both clerical and lay, who use force to defend themselves are entitled to engage in more than simple blocking motions. They are also permitted to back, even to the point of killing an assailant, either preemptively (e.g., to ward off an ambush) or, after the attack has already been initiated, to prevent its renewal. This active resistance to injury the author sharply distinguishes from revenge. Defense and revenge are thus construed as two contrasting reasons for the sake of which someone might choose to return violence for violence. But certain people have contended that no one ought to resist force before it strikes; yet it is permitted to kill [licitum est occidere] an ambusher and anyone who tries to kill you .... If, however, someone returns violence [vim factam repel/at], this should be done with the assumption that it is for defense [defendend1], rather than for revenge [ulciscend1] • • • and only if the first attacker intends to strike once more [volebat percutere denuo]; otherwise, if the attacker does not intend to strike once more and the other person still returns Haggenmacher, Grotius et la doctrine de la guerre juste (Paris: Presses Universitaires de France, 1983), 94-95. 32 See Peter AQUINAS ON DEFENSIVE KILLING 357 force, this should be seen as revenge rather than resistance to violence. And this is what I understand when it is said that force may be resisted "on the spot" [incontinenti]. It is therefore required that a return blow be in defense, not in revenge ... and self-defense must be exercised in moderation. 33 In the Summa de casibus poenitentiae, 34 Raymond of Penafort proposes an expanded version of the main principles outlined in Qui repellere possunt. He begins his discussion with the remark that "anyone" may "repel force with force," and thus wage a defensive war "without the special authority of a prince or the Church," under condition that this be done (1) "immediately" (in continenti) and (2) with a "moderation of blameless defense" (moderamine inculpatae tutelae). 35 The latter expression was at the time the standard legal formula for designating the requirement that the exercise of defensive force should not exceed what is strictly necessary to resist an attack. 36 Such "defense or protection is blameless, that is without fault" comments William of Rennes (who expressly mentions the deliberateness of this forcible response), "because [it] does not harm the adversary more (especially [if it is] on purpose) than is required to repel his violence. "37 Some authors, Raymond notes, restrict this moderation to actions that consist merely in "impeding" (impediendo) an o •• 33 Decretum Divi Gratiani una cum glossis, p. 840 (Andrews and Haggenmacher, trans., in Reichberg et al., eds., The Ethics of War). 34 Raymundus de Pennafort, Summa de poenitentia, et matrimonio, cum glossis Ioannis de Friburgo [in fact William of Rennes] (Rome, 1603). Below I refer to this work under the abbreviated tide Summa de casibus.Page numbers are given in parentheses. 35 Ibid., II, § 18 (185). 36 The formula moderatio (or moderamen) inculpatae tutelae (which seems to have first appeared in the constitution Recte possidenti, promulgated around the time of the Emperor Diocletian) originally designated the limits that should be observed by someone who used force in just defense of his property. The medieval jurists subsequently applied this formula to the defense of persons (see Haggenmacher, Grotius, 278). 37 Guilielmus Redonensis, Apparatus, ad Summa de casibus, II, § 18 (moderamine inculpatae tutelae) (185): "[N]on excedit modum in defendendo, plus laedendo aduersarium, ex proposito maxime, quarn exigat violentia illata. »William's gloss on the Summa de casibus has sometimes been erroneously attributed to John of Freiburg, as for example in the 1603 edition referenced above. This error is explained by the fact that both the Summa de casibus and its gloss had long since fallen into disuse when they were first edited in print on the occasion of Raymond's canonization in 1601. This confusion was cleared up by 1715 (on this, see Johann F. von Schulte, Die Geschichte der Quellen und Literatur des Canonischen Rechts von Gratian bis auf die Gegenwart, 3 vols. [Stuttgart: Enke, 1875-80]). 358 GREGORY M. REICHBERG attack, 38 "for example," adds William, "[by] holding up an arm or a stick lest one receive an injury to the head or body. "39 "Others," Raymond continues, "say that it is permissible for the laity to strike back [repercutere],but not for clerics," while "a third group say that it is permitted not only to the laity but also to clerics to strike back, but draw a distinction between force directed against persons or against property. "40 The expansive view articulated by this third group of authors was, in William's firm opinion, "the better one." 41 Unlike the first view, which restricts legitimate 38 Summa de casibus, II, § 18 (185). This (as noted in William's Apparatus [§18, Dicunt quidam]) is undoubtedly a reference to Huguccio, whose very restrictive conception of selfdefense was "a survival of Augustine's view that denied to the private individual the right to private defense on the grounds that it entailed a loss of love" (Frederick H. Russell, The Just War in the Middle Ages [Cambridge: Cambridge University Press, 1975], 96 n. 23). Russell mistakenly attributes this position to Raymond of Pefiafort ("[l]ike Huguccio before him, Raymond harbored moral suspicions about attempts to justify the right of a private person to resort to violence on his own behalf, however licit it was in Roman and canon law"), on grounds that for Raymond "private self-defense justified neither wounding nor killing an enemy." The tight resttictions on self-defense that Russell refers to in Raymond's Summa, the violation of which merit a sentence of excommunication, bear only on the special case of persons who exceed the measure in defending themselves against clerics: ("I agree with the opinion which holds with respect to the sentence of excommunication that if someone defending himself against a cleric goes too far, if this is intentional, he is subject to excommunication" [quantum ad sententiam excommunicationis concordo cum eis, ut, si aliquis defendendo se contra clericum excessit modum, incidat in excommunicationem]") (§ 18 [186b]). It is true, however, that Raymond did give a nod in the direction of Augustine when he affirmed that the permissibility of violent self-defense "applies [only] to the imperfect, the perfect fpeefect1] are not allowed to strike back, which the Lord shows in Matthew 5 [:39], 'If anyone strikes you on the right cheek, turn to him the other also' ... "(§ 18 [187]). In this context the perfecti signified those individuals who, by their professed state of life, were committed to living out the gospel counsels of perfection. William, characteristically, defined the category quite narrowly: "Who are these [perfect]? Does this include regular clergy who have renounced possessions and even their own will? Response: as perfection has many shapes, such as perfection in respect of sufficiency, order, empire, religion, and some other matters distinguished by the theologians, I believe that here only those are meant who are obliged by vow, precept or law not to retaliate; among these were the apostles and disciples in the primitive Church who were held to non-retaliation by the teachings of the Lord, since the Church was to be promulgated and enhanced by blood and martyrdom"