Terminating in the Body: Concerning Some Errors of Action and Intention Christopher Tollefsen I. The New Natural Law theory offers a distinctive account of the nature of intention and human action, and, accordingly, of what aspects and consequences of a human agent’s performance should be considered outside the intention (praeter intentionem). The most thorough articulation and defense of the NNL theory of human action is to be found in John Finnis, Germain Grisez, and Joseph Boyle, “’Direct’ and ‘Indirect:’ A Reply to Critics of our Action Theory,” The Thomist 65 (2001): 1-44. In part, the distinctive features of the account follow from a methodological decision to consider human action from the perspective of the agent of that action, or, as I have called it, the first-person perspective. Christopher Tollefsen, “Is A Purely First Person Account of Human Action Defensible?” Ethical Theory and Moral Practice 9 (2006), pp. 441-460. I think it is more accurate to refer to it as the first-person agential perspective. That methodological decision preceded, in the theory’s genesis, but is fully consonant with, remarks made by Pope Saint John Paul the Great in his Encyclical Veritatis Splendor. John Paul II, Encyclical Letter Veritatis Splendor, hereafter VS. There the Pope made clear that the morality of certain kinds of acts, those picked out by so-called moral absolutes, depends “primarily and fundamentally on the "object" rationally chosen by the deliberate will.” But in order to “grasp the object of an act which specifies that act morally, it is therefore necessary to place oneself in the perspective of the acting person (VS, 78). This standpoint, it is perhaps not too obvious to note, is the standpoint of an agent oriented towards action, and thus towards what makes action possible: human goods, aspects of human well-being that promise some potential benefit. The NNL view, as is familiar, grounds all practical deliberation in an apprehension of these basic goods, an apprehension that it sees as foundational to practical reason, and not derived from any truths of theoretical reason. For general statements of the NNL view, see Germain Grisez, Joseph Boyle, and John Finnis, “Practical Principles, Moral Truth, and Ultimate Ends,” American Journal of Jurisprudence 32 (1987), 99-151. In this issues, E. Christian Brugger argues that the NNL refusal to derive normative from theoretical truths is both sound and true to the thought of St. Thomas: “St. Thomas’s Natural Law Theory,” National Catholic Bioethics Quarterly. All human action is thus carried out with a view to the attainment of human good: we do not engage in human action except insofar as we apprehend cognitively some possible benefit, and recognize as well some way of realizing that benefit through action. The recognized way might be quite complex, a sequence of nested means that serve as proximate ends ordered to a more distant end in which the benefit sought will be found; or it might be quite direct in the sense that an immediately carried out action itself realizes a basic good. The paradigm case of all fully human action involves free choice. In such a case, the benefits offered by two or more options (including the option of doing nothing) are incommensurable. Each offers some aspect of good not offered by the other; thus, a choice must be made. That choice is neither caused by factors in the natural world, nor is it caused by the goodness of the option chosen, as it would be were that option best in every possible respect. So it is free. See, for a helpful account, Joseph Boyle, “Free Choice, Incommensurable Goods and the Self-Refutation of Determinism,” American Journal of Jurisprudence, 50 (2005), 139-163. As we shall see later, this has consequences for what an agent may, and indeed must, know about her own intentions. The claims of the previous two paragraphs may be articulated in terms of the notion of a proposal. What is intended is what is proposed by the agent to him or herself and chosen. The proposal, so understood, encompasses both the end that the agent seeks, and all that the agent seeks as desired in order to bring about the sought after end. A fully formed proposal to travel somewhere, for example, includes within it not just the end point, but also an awareness of how the end point is to be attained: travelling to Dallas by driving from Columbia to Atlanta, then driving from Atlanta to Birmingham, and so on. As the proposal includes both the end and the nested sequence of means, so, on the NNL account, does the intention. Because the notion of a proposal specifies our understanding of an agent’s intention, it likewise specifies our understanding of the agent’s action. What is proposed and chosen, and thus intended, is what is done by the agent insofar as what is done is successful according to that proposal. The nature of the action is determined by the intention, and its reality is thus, like the reality of intention itself, radically first personal and agential. Neither what is intended nor what is done as a full human action is a function of what merely happens, though happenings consequential to, and even simultaneous with, what is intended and done, are inevitable. Put another way, while the proposal, and thus my intention, and thus my action properly speaking, includes everything that I take to be needful in order to achieve some benefit I set out to attain, the proposal, and hence the intention, and hence the action, does not include many states of affairs bound up with what I choose that are nevertheless not needful for what I am pursuing. Such states of affairs are accepted, if I go forward with implementing my proposal; but they are neither chosen nor intended. For example, and non-controversially: when my proposal is to alleviate my toothache by taking ibuprofen, the proposal does not include the inevitable brief experience of an upset stomach that inevitably accompanies my taking the ibuprofen. The upset stomach is not intended, but is rather accepted as a side effect of my action. The NNL theory rigorously maintains this first-personal agential account in discriminating what is intended from what is accepted as a side effect, even when what is accepted is a foreseen consequence of what is done, as is the upset stomach; even when it is an inevitable and foreseen consequence of what is done, as are certain other side effects to medical interventions; and even when it is an inevitable and foreseen accompaniment of what is done, simultaneous with the doing. Inevitable and foreseen stuttering, for example, accompanies the very thing the speaking agent does – it is not a downstream causal consequence. Yet it is not intended. I may, of course, be held morally accountable for what I have accepted, and am sometimes morally to blame for side effects. But the distinction between what is intended and what is accepted is crucial, on the NNL account, as on the Pope’s account above, for the consideration of moral absolutes. For what those absolutes in most cases rule out is intentional damage to or destruction of, a basic good. It is thus always and everywhere wrong, for example, intentionally to kill an innocent human being. But the death of such an innocent is not automatically recognized as impermissible if it lies outside the agent’s intention and is merely accepted as a side effect. See, for discussion of such absolute negative precepts, John Finnis, Moral Absolutes: Tradition, Revision, and Truth (Washington, D.C.: The Catholic University of America Press, 1991). This theory of action and intention, here only briefly described, but discussed and defended at much greater length on many occasions by NNL theorists, has nevertheless been subject to considerable criticism. The view is held by many to be too first-personal, and to provide inadequate “constraints” on what an agent intends, when her performance will inevitably and foreseeably be accompanied or followed by states of affairs in which individuals are harmed. The most recent, and in many ways most competent such criticism has recently been that of Steven J. Jensen, “Causal Constraints on Intention: A Critique of Tollefsen on the Phoenix Case,” National Catholic Bioethics Quarterly (2014), 273-293. I hope to respond to Jensen’s helpful article in the future. To give perhaps the most disputed example: NNL theorists have held that it need be no part of a physician’s proposal, and hence intention, to kill or harm the child whose head is crushed to remove an obstetric obstruction threatening the mother’s life. The proposal, and hence the intention, is to change the dimensions of the child’s skull, in order to remove the obstruction, so as to facilitate removal of the child, so as to save the mother’s life. Critics, by contrast, are likely to hold that crushing the skull of an infant child “just is” killing, or harming, or mutilating that child. The death, or the harm, in this case is simply “too close” to what is done for it to be outside the intention. See, for example, Luke Gormally, “Intention and Side Effects: John Finnis and Elizabeth Anscombe,” in John Keown and Robert P. George, eds, Reason, Morality, and Law: The Philosophy of John Finnis (Oxford: Oxford University Press, 2013), 93-108. Similar misgivings, and others, animate the recent critiques of the NNL theory of action found in the special issue of the National Catholic Bioethics Quarterly dedicated to criticizing the NNL theory as a whole. This essay addresses three of the articles in that issue. Those essays, by Steven A. Long, Matthew O’Brien, and Fr. Kevin Flannery, SJ, as well as a more recent essay by Edward Furton (in Ethics and Medics), eventually converge on a variation of the “closeness” criticism that I will address in the second part of the paper. In the first part, I address three further issues, raised, respectively, by Flannery, Long, and O’Brien. II. Fr. Kevin Flannery’s essay begins with an account of how “object” is used by Aquinas, and the bearing of this usage on the topics of interior and exterior acts. Rev. Kevin Flannery, SJ, “Thomas Aquinas and the New Natural Law Theory on the Object of the Human Act,” National Catholic Bioethics Quarterly, 13 (2013), 79-104. It is perhaps worth differences in the approach taken to the concept of the “object” of the human act by the NNL thinkers. Grisez writes, as noted by Flannery, that he thinks Aquinas’s use of the term “neither clear nor coherent;” Boyle eschews the concept, and with it the interior-exterior act distinction, in favor of more straightforward language of means and ends. By contrast, Finnis argues that St. Thomas’s use is, despite appearances, coherent. See John M. Finnis, “Intention and Objects,” in Collected Papers vol. II, Intention and Identity (Oxford: Oxford University Press, 2011), 152-172. He concludes the discussion of his first three sections by showing that for Aquinas, identifying the object of the act does not tell us everything morally important about the act. This sets up the critical fourth section of his essay in which Aquinas’s view on this point is contrasted with Grisez’s: contrary to Grisez’s view, writes Flannery, there “is no suggestion in ST 2-2. 64.7 that an act of killing in self-defense is licit simply because the death of the assailant is not within the self-defender’s choice or intention” (Flannery, 2013, p. 90). This point is repeated several times: in his discussion of active per accidens scandal (in which the scandal is not intended): “Aquinas’s remark implies the contrary [of Grisez’s view]: there is a sin – indeed a serious sin – in which the agent’s choice, the interior object of the will, is aimed at something quite different from that which (in active per se scandal) gives the act its species” (Flannery, 2013, p. 92); in his discussion of sacrilege, where his response to Grisez is that “The moral character of the act itself is not determined simply by looking to that which is intended (or chosen)” (Flannery, 2013, p. 93); and in his discussion of craniotomy: “Since, according to Grisez’s analysis, the moral significance of the act depends wholly upon what, according to his theory, is chosen…” (Flannery, 2013, p. 94). Why does Flannery spend so much time on this point? Apparently in response to Grisez’s claim that “Moral responsibility is to be found first and foremost in one’s choosing.” Germain Grisez, The Way of the Lord, Vol. 1, Christian Moral Principles (Chicago: Franciscan Herald, 1983), p. 239. But (a) “first and foremost” simply does not mean “determined solely by;” And (b), in the very same section from which the offending claim of Grisez’s comes, Grisez writes that “We are responsible for more than just what we aim at and choose” (Grisez, 1983, p. 239). And indeed, in the very paragraph from which Flannery quotes, Grisez says “one bears responsibility for foreseen side-effects,” a passage that Flannery also quotes (Grisez, 1993, 239). Finally, (c) the claim that moral responsibility is to be found first and foremost in one’s choosing is, in itself, a staple claim of Catholic moral thought. See, for example, John Paul II, Encyclical Letter Veritatis Splendor, no. 79: “The primary and decisive element for moral judgment is the object of the human act, which establishes whether it is capable of being ordered to the good and to the ultimate end, which is God.” Flannery notes (p. 88) that for Grisez, “‘proposal adopted for choice’ has more or less the same meaning as Aquinas’s expression ‘the object of an action.’” So it seems Grisez and Pope John Paul II are making exactly the same point here. For it is our choices that in the first instance shape us as the people we are, people whose lives are lived in accordance with God’s plan, or not. So Flannery’s argument seems directed against a claim not made by Grisez, and indeed denied by him; while the claim that Grisez does make appears to be on solid ground. Steven A. Long asserts of the NNL theory that it is “no improvement whatsoever” over proportionalism and consequentialism save in the fact that it “piously obscures the enormity of the derogation of the moral law that it represents.” It is further “deceptive”, and its errors “seismic in their gravity.” Steven A. Long, “Fundamental Errors of the New Natural Law Theory,” National Catholic Bioethics Quarterly, 13 (2013), 105-131, p. 130. Underlying these claims is Long’s belief that the NNL countenances, as an account of an agent’s choice, whatever he might say about it, and that it “becomes enough that one seeks a good end” (Long, 2013, 124). Accordingly, Long holds, the NNLT fails to recognize that an action that, in his words, “terminates in the body” of an innocent human person “in such a way as to harm or destroy” that person, ought not to be considered “indirect” (Long, 2013, 126). Or, as Furton puts it, “The central moral question, as I understand it, is whether it is possible to directly strike the body of an innocent person, and so cause that person injury or death, without intending to cause the injury or death.” Edward Furton, “Tollefsen on the Phoenix Case,” Ethics and Medics 39 (2014), 3-4, p. 3. “Terminating in,” and “directly striking” are descriptions here, in the order of nature, of natural causality. It is the fact that my axe, or sword, or scalpel, or, no doubt, bullet, comes directly and lethally into contact with your body that is thought to be definitive by Long, Furton, and many others. For in such cases, the lethality is part of the “integral nature and per se effects of the action chosen” (Long, 2013, 124). They thus constitute direct killing. As will be clear from my summary on the NNL view in Part I, the NNL theorists deny that the “integral nature and per se effects” of an action considered in the order of natural causality, are determinative of the nature of the action considered in the order of the will, that is, in the order which reason introduces into the acts of the will. See Brugger, “St. Thomas’s Natural Law Theory,” for a discussion of Aquinas’s “four orders” And why should it be? In the paradigm case of intentional action, will, a rational appetite, is responsive to the judgments of reason, a spiritual power, and a particular course of action is chosen under the description that made it attractive to reason. Of course, and again as noted above, insofar as what is chosen is realized in action there will be many other descriptions true of what happens, some of which will be foreseen by the agent, some not. The reality of what happens insofar as we act is far from under our full control. But that is not to say that the reality of what we judge and choose is not under our control, much less that that reality could be determined by the shape of what happens in the world. That fails to do justice to the nature of reason and will and spiritual powers. Now none of this implies that in morally assessing the permissibility of action it “becomes enough that one seeks a good end.” What one chooses as a means and what one wills as an end are willed by the agent together, and the willing of each, end and means, is available for moral assessment. One may not will death as an end, but the NNLT are well known for their view that one may not will death as a means either, and indeed, their view on this is quite a bit less permissive than that of most Catholics who believe that death my be willed as a means, by those with lawful authority, for the bringing about of justice. Nor is any of this to hold, as Long alleges, that an agent may say whatever he likes about his choices, nor may he describe them to himself as he wishes. More accurately: while agents may say what they like, the saying makes nothing so. Proposing, willing, and choosing are realities, albeit spiritual realities. They do not get their shape from our wishes or self-rationalizations, nor from what we might say about them. They cannot be gerrymandered in the way that critics suggest. This is made clear by the NNL theorists: What counts for moral analysis is not what may or may not be included in various descriptions that might be given by observers, or even by acting persons reflecting on what they have done, but what is or is not included within a proposal developed in deliberation for possible adoption by choice. Only the truthful articulation of that proposal can be a description that specifies an act for the purposes of moral analysis. John Finnis, Germain Grisez, and Joseph Boyle, “’Direct’ and ‘Indirect:’ A Reply to Critics of our Action Theory,” The Thomist 65 (2001): 1-44, p. 29. So the NNL make no claims to the rightness of what is said. But they do make a claim about an agent’s awareness of what she intends in choosing some course of action over others: they claim that what is so chosen is necessarily knowingly chosen. Critics of the NNLT seem inclined to deny this, as in the following passage from Michael Pakaluk: What someone intends is not subjective but rather objective.  Hence, he can be wrong about what he intends.  Hence, a person’s self-report about his own intention may be mistaken, if he lacks self-knowledge.  Indeed, the road to hell is paved with subjectively good intentions.  The doctor who says that he intended in administering the lethal dose only to release the soul of the patient from suffering also intended, actually, the direct killing of an innocent human being.  In Miss Anscombe’s example, Mr. Truman said that he intended only to end the war and save lives but, whatever he said, because of what he commanded to be done, he thereby intended the killing of women and children. Michael Pakaluk, “Some Simple Mistakes about Formal Cooperation,” Thomstica.net October 1, 2012, emphasis added. Available at: http://thomistica.net/commentary/2012/10/1/some-simple-mistakes-about-formal-cooperation.html. There is some ambiguity here, for Pakaluk writes about what the agent in question might say, and as noted, people can say anything they like. The important question is whether one must know what one is choosing, and intending, as one is choosing and intending, and the answer must be affirmative if we are indeed capable of making free choices. For choices are freely made only if the agent has before her two possible proposals for action – to do this for the sake of that, or the other, for the sake of something else. Whatever is opaque to the agent at the time of choosing simply is not available for choosing by the agent; and correlatively whatever is chosen must be chosen knowingly or it is not, after all, chosen. So agents do know what they are choosing (and intending) as they choose (and intend), even though they may, for a variety of easily identifiable reasons, quickly forget, confuse, or distort in their own recollections. And of course, they may simply lie. The form of access to the content of one’s choices envisaged by the NNL theory is not a truth serum. How should we think of that form of access, though? Should it be conceived on the model of a looking at, as if the act of will, the proposal, the intention – any aspect of the agent’s choosing and acting – were interior acts to be known through introspection? This is a question raised by O’Brien in his interesting essay; he holds that such a view would be mistaken, and holds too, I think, that the NNL embrace of the first person perspective commits it to some version of this error Matthew B. O’Brien, “Elizabeth Anscombe and the New Natural Lawyers on Intentional Action,” National Catholic Bioethics Quarterly, 13 (2013), 47-56.. Against this, he puts forth important claims by Anscombe concerning the identity of what is intended, chosen, and done, with what is observable to a third party. But it seems to me that O’Brien overstates the distance between Anscombe and the NNL theorists on this point. I address the issue, however, not insofar as it is a question of how close to or far from Anscombe the NNL theorists are; they do not claim her as an authority, but rather a source of ideas. Anscombe asks early in Intention “how do we tell someone’s intentions?” G.E.M. Anscombe, Intention (Cambridge, MA: Harvard University Press, 2000), p. 7. O’Brien notes that for Anscombe “the greater number of the things you would say straight off a man did or was doing [were you, say, observing him] will be things he intends” (Anscombe, 2000, 8). Thus, to use her example, if she is sitting in her chair writing, that will typically be what anyone who is asked says that she is doing, and in making that attribution, they are typically also identifying her intention. Why this convergence of intention, which earlier described as a first personal reality, with what happens, what can be seen? Surely it is because some part of what is intended by an agent is typically some part of what is visibly done by that agent. What the agent knows practically, and without observation, is thus the very same thing that can be observed and recognized by a third person: that she is sitting and writing. To this extent, it would be problematic to hold that all intention was unobservable, private, interior, an invisible “trying” to be inferred from what happened publicly. Such a picture would be dualistic, a charge which O’Brien has made elsewhere against the NNLT. Matthew B. O’Brien and Robert C. Koons, “Objects of Intention: A Hylomorphic Critique of the New Natural Law Theory,” American Catholic Philosophical Quarterly 86 (2012): 655-703. I respond to this in Christopher Tollefsen, “Response to Koons and O’Brien’s ‘Objects of Intention’”, American Catholic Philosophical Quarterly, 87 (2013), 751-778. For extended, and helpful, criticism of the dualistic picture, see David Braine, The Human Person: Animal and Sprit (South Bend, IN: University of Notre Dame Press, 1993). I do not think that the NNL theorists need deny this convergence of first and third person for two reasons. First, the convergence is in part a consequence of the identity of the person and his or her organic bodily existence. Actions are not the movements of bodies caused by previous interior and logically independent intendings; intention informs human action, which is the action of a bodily human person. And second, human beings are social, and their sociality has epistemic consequences: as mutual participants in shared social forms, with a shared set of concepts, human beings by and large understand what other human beings are doing across a wide range of activities just by looking. But both the ontological and the epistemological features that underlie the convergence are limited, and thus the convergence is limited too. For, first, the social conditions that underlie our epistemic access to what is being done by way of shared social forms are far from universal. It is natural for us to observe people intending to score a basket, to plead with a judge, or to perform heart surgery. But members of societies without these practices will be unable to “see” these intentions in action at all. And second, the identity of human persons with their bodily existence is likewise limited because of the transcendence of the person to her organic bodily existence: the person is both the same as and not reducible to the organism that sits and writes. And as transcendent, there is more to the person than what was made fleshly in the acts that realized what that person intended; indeed, there is more to what was intended than what was made flesh in action. Anscombe recognizes this last point in a way that makes clear some of these limits of the convergence of first and third personal: Now it can easily seem that in general the question what a man’s intentions are is only authoritatively settled by him. One reason for this is that in general we are interested, not just in a man’s intention of doing what he does, but in his intention in doing it, and this can very often not be seen from seeing what he does. Another is that in general the question whether he intends to do what he does just does not arise (because the answer is obvious); while if it does arise, it is rather often settled by asking him. And, finally, a man can form an intention which he then does nothing to carry out, either because he is prevented or because he changes his mind…(Anscombe, 2000, 9). And Anscombe goes on to note that these phenomena can prompt us to think of intention as a “purely interior thing,” to be discovered by probing “the contents” of our minds. This would be an error. But I do not think that Finnis was making this dualistic error in the passage cited by O’Brien to demonstrate his divergence from Anscombe: “But the actions, and the reasoning towards the choice(s) whose carrying out is what human action centrally is, are complex and often subtle. Transparent to a really clear-headed chooser, they are for observers always more or less a matter of inference or belief in testimony…” John Finnis, “Introduction,” in Collected Essays, vol. 2, Intention and Identity (New York: Oxford University Press, 2011), 13. “Always more or less” seems entirely in keeping with Anscombe’s point: we do, given shared social forms, generally know what someone is doing by observation but we almost never know the whole of what they are doing simply by observation. And of course, we can on occasion be quite wrong. Consider as an example of a shared social form that makes available generally non-inferential knowledge the possession of common language. Often, perhaps typically, what an agent intends to communicate can be known straight off without inference by others who share that language. But not always: irony and metaphor, for example, can stymie such understanding, and the eventual recognition of what is really being communicated can be precisely a matter of inference or testimony. Two final points should be made here. The first returns to the issue of self-understanding and transparency. The NNLT holds that, as Finnis says, intention is transparent to a clear-headed chooser. As noted earlier, this strikes many as objectionable, perhaps because “transparency” is taken to conjure a privileged “looking at,” a scrutinizing of the invisible contents of one’s mind. But this, I think, does not do justice to the nature of first-personal awareness. Even when not spiritual, but sensory, such awareness is not observational – a point familiar to readers of Anscombe from her discussion of, e.g., our knowledge of the position of our limbs. But in some further cases, such first person awareness is not only non-observational, but non-sensory and reflexive: we know that we know without that act of knowing being a further independent and interior act behind the initial knowing. Such is also the case in the paradigm instance of intention, what is freely chosen, for here we will that we will and know that we will in a reflexive and non-observational way. Our transparency to ourselves in such acts is not at all like observational privileged access. The second point concerns what may or may not be inferred from the unity of first and third persons, to the limited but real extent that that unity obtains. Why is there this unity? Because in acting the agent realizes in what she does her proposal for action. But for precisely this reason it would be a mistake to conclude that the unity implies any inference in the other direction from the “matter” or “nature” of what was done to the content of the proposal or choice. For the proposal propositionally identifies those aspects of a possible performance that serve our end(s), and it is as that (or those) propositionally identified aspect(s) – under that (or those) description(s) – that our intention is realized in an event that admits of many more descriptions. That intentions are realized in actions which are public events in the world thus does not imply that intentions are shaped, nolens volens, by public events in the world. Yet this is where O’Brien’s essay ends: “human agents cannot act without choosing a concrete, particular kind of means that is within their power to perform” (O’Brien, 2013, 56). And so the proposal to reduce a child’s skull is a proposal which, given the state of the world and medical science, must include the lethality of the reducing. It is an observable third personal fact of the matter, against which the agent cannot protest. Here, an entirely legitimate point about the convergence of the first and the third personal standpoints starts to move in the wrong direction, with the third-personal taking the reins. But this, as I argued above, is to fail to reckon with both the spiritual nature of intention, and with the way in which intention shapes action. While intention is not divorced from the third personal, it is transcendent to it, and not at its mercy; and what happens is human action precisely insofar as it is shaped by an agent’s proposal. II. Here we come, though, to the central worry of the various critics, a worry the focal point of which is the concern that the “intentionalist”, “logicalist” approach of the NNLT justifies intentional killing of the innocent. Indeed, the accusation that the NNL theorists are justifiers for baby-killing lurks barely beneath the surface, as can be seen from the following, by Long, the special issue’s editor: That New Natural Law Theory is now harmful to the common good both of the Church and of civil society insofar as its proponents oblige themselves to justify, under the aegis of a false account of intention, acts that are in fact directly destructive to innocent life. This is an account that is not only false, but false in such a way as invites censure by the Church. One wonders how many therapeutic abortions, how many wrongful homicides of innocent children, must occur under the erroneous account they propound — an account in which whether an action directly harms is merely a function of intention — before that account will be finally corrected. Steven A. Long, “’Goods’ Without Normative Order to the Good Life, Happiness, or God: The New Natural Law Theory and the Nostrum of Incommensurability,” Thomistica.net, September 18, 2011. Available on-line: http://thomistica.net/news/2011/9/18/goods-without-normative-order-to-the-good-life-happiness-or.html. Long’s view here might profitably be compared with Richard Doerflinger’s judgment on the role played by Germain Grisez in the pro-life movement. Speaking of Grisez’s book Abortion: The Myths, the Realities, and the Arguments (New York: Corpus Books, 1970), Doerflinger writes “If any book in the pro-life movement deserves to be called prophetic, it must be this one… It is not much of an exaggeration to say that Germain Grisez's book put the intellectually respectable case for the pro-life position on the map.” Richard Doerflinger, “The Rational Case for Life,” available on-line: http://www.nrlc.org/archive/news/1998/NRL7.98/book.html. Long’s essay in the special issue has the virtue of making clear what the sticking point is: where an action “terminates directly in an innocent so as to harm that person,” then the agent must be understood to have chosen that harm, regardless of her end, or her proposal for action (Long, 2013, 128). As noted above, Furton puts the point in a similar way: “The central moral question, as I understand it, is whether it is possible to directly strike the body of an innocent person, and so cause that person injury or death, without intending to cause the injury or death” (Furton, 2014, 3). I will in this essay consider no specific cases. The critics have in mind the NNLT’s views on craniotomy, and the recent controversy over the Phoenix abortion case, but they need not concern us here. I address the Phoenix case at greater length in “Double Effect and Two Hard Cases in Medical Ethics,” American Catholic Philosophical Quarterly, forthcoming 2015. It will suffice to address the critics and their views philosophically. I intend to show that their emphasis on “directly terminating in the body,” or “directly striking” is entirely misplaced and makes no good argument against the NNLT account of human action and intention. It is clear enough what these critics hope to accomplish with this emphasis on “directly terminating in” or “directly striking.” The dialectic here returns to the problem of “closeness.” How ought we to identify just what kinds of “closeness” are such as to render some aspect of an agent’s behavior, or some consequence of what an agent does, something that is intended by that agent? Long and Furton are here attempting to identify at least some of what is thus sufficiently “close” by means of the notion of “directness”, “directly terminating,” or “directly striking.” There are other attempts to demonstrate the meaning of “sufficiently close”; in this issue, Sherif Girgis criticizes Matthew O’Brien and Robert Koons’s claims that the nature of social practices can help to fix an agent’s intention in “Intent to Kill: Always Wrong?”, National Catholic Bioethics Quarterly. In what follows, I argue that this strategy must fail: “directly striking” and “directly terminating in” on their own do not accurately identify the dividing line between what is intended and what is accepted as a side effect in the standard case of self-defense. So this core notion of Long and Furton must be supplemented, and supplement they do: it is directly striking, or terminating in the body of, the innocent that they identify as the, or perhaps only a, form of closeness that marks some feature of an agent’s performance as intended. But, as I shall show, incorporation of the notion of “innocence” into act analysis is both un-Thomistic, and is philosophically deficient: in the end, it renders their analysis viciously circular. Let me begin the argument, then, with the following well-known point. Aquinas believed that no private citizen may ever permissibly intend the death of another person. I will return below to the further Thomistic claim, that public authorities, or their deputies, may intend the use of lethal force on miscreants. For now, the point about private citizens is the important one. Must private citizens therefore refrain from defending themselves if they are attacked? No: the critical passage is ST 2-2, 64.7. There Aquinas notes that “nothing prevents there being two effects of an action, only one of which is intended…So, the act of self-defense can have two effects: one the preserving of one’s own life; the other, the killing of the aggressor.” This passage marks the beginning of Catholic moral thought about the so-called “principle of double effect.” And we should note that Aquinas’s concern is to argue that the attacker may be prevented with a lethal response only if the lethality is outside the defender’s intention. What sort of defense might we envisage? Perhaps the defender has a long piece of piping with which he can, by striking forcefully, repel the attack. Perhaps he has a powerful waterhose, which by aiming at the attacker will repel him. Perhaps he has a gun; shooting the attacker will repel him. Let us suppose that in each case, the attacker will be lethally harmed. In each case, the defender’s act will certainly “terminate directly” in, and “directly strike” the body of the attacker. Yet the tradition equally certainly defends the permissibility of such defense, even when the attacker’s death can be foreseen. So a lethal act terminating directly in the body of another, or an act of directly and lethally striking another is not just as such an instance of intentional killing. How can Long, or Furton, overcome this difficulty? Their strategy, it seems, is to supplement their core concepts with the concept of “innocence”: repeatedly, they focus on actions directly terminating in, or directly striking the bodies of, innocent persons. But this supplement is inadequate to solve the problem, for some actions that appear to terminate in the body of, or involving direct strikes upon, innocents also need not involve an intention to kill those innocents. “Innocent” in the natural law tradition has two meanings, which might easily be confused. One meaning is that the agent in question poses no threat (Long calls this “performatively innocent”). The second meaning is that the agent in question is morally innocent. Is it the case that actions lethally terminating in the body of an innocent in one or the other or even both senses must always involve intentional killing? The answer is no. Consider first the case of agents who pose a threat but are morally innocent. The threat might come from a child, strapped with explosives, or from a madman, unknowing of what he does, or from a person whose bodily momentum, as a result of an act over which he had no control, currently propels him towards a possibly fatal collision with several infants. This last agent is a threat because of his body and its momentum. And so that threat may be responded to by physical action intended to repel, thwart, or prevent, the eventual and fatal collision his body’s momentum threatens. He may be shoved, if he is close to hand and I am strong enough; but equally he may also be struck with a long pipe, shot with a gun, or blasted with a lethally strong blast of hot water. The intention here is to save the children by repelling his body, by means judged sufficient to repel, even if they are also recognized to be lethal. His moral innocence is irrelevant, and has no bearing on what I intend. Nor is it the case that actions terminating in the bodies of those who are “performatively” innocent because they pose no threat must involve intentional killing on grounds that they are morally wicked. For the collateral damage of a legitimate bombing campaign against enemy military targets is damage done, albeit as a side effect, on human beings entirely innocent in Long’s “performative” sense – even if some or all are not morally innocent or indeed are quite wicked. But supposing, more realistically, that at least the children are morally innocent, we here have actions terminating lethally in the bodies of those who are innocent in both senses that involve no intentional killing. We are thus in a bind: actions directly striking, or terminating directly in the body of an innocent do not necessarily involve an intention to harm those innocents. What to do? One possibility would be to return to the notions of “directly striking” and “terminating directly” to see if further refinement might provide more adequate analyses. So let us consider some possible ways of restricting the meaning of these phrases. Consider what it means to “directly strike” something. Perhaps my counterexample of the strategic bomber misses the correct sense of this phrase – after all, the strategic bomber is aiming at the military installation, and not the bodies of the citizens. Let us therefore adapt an example from Christian Brugger. A hostile terrorist might use an entirely innocent – in both the moral and the performative senses -- human being, a child perhaps, as a human shield, while threatening, say, to shoot down an airliner. Shooting through the innocent to get to the terrorist has, as its side effect, the death of that innocent. Yet it seems an example of directly striking and even aiming. Brugger’s example comes from Alien vs. Predator (not a court case). In this movie, an alien is incubating in the chest of a human who urges his comrade to “Kill it before it reaches the surface!” She takes aim at his chest and shoots, killing both her friend and the alien. See “Action, Intention and Self-Determination,” Vera Lex: Journal of the International Law Society, vol. 6, nos. 1-2 (Winter 2005), 79-106. Perhaps this example, however, fails to capture the sense of “termination” in Long’s oft repeated phrase “terminating in the body”, for my aim is that the bullet go through the child to hit the terrorist, and not that it terminate in the child’s body. So let us consider the cases of those soldiers, of whom there have been an extraordinarily and even awesomely high number, who have fallen on a grenade in order to shield their comrades from its blast with their own bodies, cognizant, of course, that they were headed for virtually certain death. We may even look at a case in which the grenade in question was not thrown by an enemy. Here is the case, of William “Billy” McFadzean, as told by the Dictionary of Ulster Biography: As day was breaking on 1 July, he was preparing his ammunition, which came packed in boxes. As he was cutting the binding rope on one box, two grenades fell out, and lost their pins – this meant that they would explode within seconds, and cause along with the other grenades an enormous explosion which would have been devastating in the combined space of a trench. A split-second decision – McFadzean threw himself on top of the box, thus absorbing almost all the blast, so much so that only one other soldier was injured. In the words of his VC citation in the London Gazette two months later:  “The bombs exploded blowing him to pieces, but only one other man was injured. He well knew his danger, being himself a bomber, but without a moment's hesitation he gave his life for his comrades.” “William McFadzean, vc (1895-1916) soldier”, Dictionary of Ulster Biography. Available on-line: http://www.newulsterbiography.co.uk/index.php/home/viewPerson/1949  Did McFadzean’s action “terminate in the body of an innocent”? Surely the answer is yes, in any relevant sense of the word, nor was there even an unjust aggressor in the vicinity. Yet McFadzean is not to be condemned as a suicide: his death, accepted with astonishing bravery, was outside his intention, an intention which was to save his friends by protecting them from the shrapnel by covering the grenade with his own body to as to absorb into himself that shrapnel. The case reveals clearly that the focus on “directly terminating in the body of an innocent”, just like “directly striking the body of an innocent”, is utterly misplaced and provides no adequate way to think about intention. Now if we go back to the original case of self-defense against an unjust attacker as put forth by Aquinas, Long has an alternative story about why a lethal intention is not present in just self-defense. The story is illustrative of some general differences between Long, Furton, and various other critics, and the defenders of the NNLT. Long writes, It is true that, as Aquinas teaches, where the object is per se ordained to the end, the species derived from the end is most formal and containing. Thus, in just defensive action one may harm an unjust assailant under the ratio of warding off unjust assault, insofar as such harm is the only proportionate means of defense. In this passage, Long seems to reduce intention to a willing of ends only, and to justify chosen harm as outside the intention so long as the person harmed is not himself innocent, in at least one, and possibly both senses of the term. Why? Because when one is engaging in just defensive action against an unjust assailant, “the object is per se ordained to the end, [and] the species [is] derived from the end [and] is most formal and containing.” The good end of saving oneself from unjust attack thus swallows up the description of the means. Thus, Long justifies cases in which “a deliberately and per se lethal means is chosen because it alone is proportionate to the end of moderate defense.” Steven A Long, The Teleological grammar of the Moral Act (Naples, FL: Sapientia Press of Ave Maria University, 2007), 50. Put another way, where one’s killing is per se ordered to just defense, one can, according to Long (and on Long’s interpretation of St. Thomas) choose to kill. Long’s analysis here mirrors his analysis of cases in which his judgment is that immoral means have been adopted. In those cases, what Long calls the “integral nature and per se effects” of an action are not included in the intention as “most formal and containing”, because their physical character is not teleologically ordered towards their end – towards what made it desirable to the agent to do the action in question. We should note that for Long, this “integral nature” or “the act itself” is something that stands apart from the agent’s choice as available to be chosen, but not constituted by that choice. The integral nature of the act itself thus constitutes an external and objective constraint on what an agent intends. Thus, if it is of the integral and essential nature of an act that it be an instance of poisoning, scraping, spearing, or crushing an innocent human being, then, because such acts are per se ordered towards death, insofar as they terminate directly in the body of the innocent person, they may never be licitly done. Long’s interpretation of Aquinas on self-defense is crucial to his entire account, for self-defense is clearly action that terminates directly in the body of another. If that by itself suffices to render it intentional killing, then Long must either abandon his emphasis on “terminating directly,” or allow intentional killing in defense. Long’s solution to the dilemma is to hold that for St. Thomas, chosen killing of an unjust aggressor is not intended: intention is only of the end when the aggression is unjust, and the end alone thus gives the moral species to the act of defense. This analysis does not accurately construe St. Thomas’s position. Christian Brugger, in an essay to which Long has not, to my knowledge, responded, has shown quite thoroughly that it is an error to think that Aquinas, in his treatment of permissible self-defense, and homicide more generally, uses “intention” only to refer to what is pursued as an end. For example: Aquinas defends the intentional killing of criminals and exterior enemies by those with public authority, for the killing is ordered to the public good. As Brugger then argues, …it is clear that this type of intentional killing (which is the normative type for justifiable intentional homicide throughout question 64) does not conform to Long’s strict “per se instance of human action.” If it did, the killing could not be said to be intended, since it is willed as a means to the end of preserving the common good. Yet Aquinas clearly calls the killing intended: “It is not lawful for a man to intend killing a man in self-defense except such as have public authority, who while intending to kill a man in self-defense, refer this to the public good (Aquinas, ST, II-II, Q. 64, a 7). It must therefore be the case that Aquinas means to include under his use of intention in q. 64, a. 7 the willing of both ends and means. E. Christian Brugger, “Praeter Intentionem in Aquinas and Issues in Bioethics,” in Bioethics with Liberty and Justice: Themes in the Work of Joseph M. Boyle (Dordrecht, the Netherlands: Springer Academic Press, 2011), 102-103. And one can see the same point at work elsewhere in the treatment of self-defense. In the first two objections, Aquinas adverts to St. Augustine’s view that killing by a private person in self-defense is impermissible. Aquinas’s solution is that The words quoted from Augustine refer to the case where one man intends to kill another to save himself from death. The passage quoted in the Second Objection is to be understood in the same sense. Hence he says pointedly, for the sake of these things, whereby he indicates the intention. Thus the authority of Augustine against the permissibility of killing in self-defense is limited by holding that what Augustine was ruling out was the defender’s intending to kill in order to save his life – i.e., to kill as a means. But this is precisely what Long believes St. Thomas allows. His reliance on the authority of St. Thomas at this point is thus an error. Let us leave aside the question of the Thomistic bona fides of Long’s view. A further problem is that the analysis is circular. What we want to know is when an effect – the death of a person – that would be immoral if intended, is not intended, for if not intended, its immorality is not guaranteed (though it might, as we saw in the discussion of Flannery, still be morally impermissible). However, we cannot give as an answer to that query: the death is not intended when the lethal effect is part of a just defensive action. There is the circularity, for we seek to know what is intended as a preliminary to determining the justice of the response. Nor can the circle be eliminated by enlarging it, with the claim that when a response is just, it is “per se ordained to the end,” and the lethal effect thus praeter intentionem in virtue of the end’s being “most formal and containing.” The circle is still there, for it is the justice of the response that is again determining what is and is not intended. But perhaps the explanation goes in the other direction: perhaps, when the means are per se ordained to the end, then the means are praeter intentionem, and, consequently, appropriate and just. This in fact appears to be Long’s position. But then how do we determine which means belong to this category and which do not? There are two points to make about this issue. First, what is meant by an action’s being “per se ordained to [its] end”? Here we must turn to Long’s frequently made assertions about the “act’s integral nature.” Long holds that a human action has its integral nature in some measure independently of the agent’s proposal, and rather in virtue of its “physical reality.” So an agent’s choice of an action is in some respects like a choice of, say, an apple, when offered more than one: it is a picking of something that has a prior existence, with certain objective features that make their mark upon the intention of the agent willy-nilly. But what exactly is envisaged here? Actions do not have such a prior reality; they are brought into being through an agent’s intention and choice, as those are realized in a performance undertaken by the agent. But secondly, I think, it is in fact the concept of “innocence” that does most of the heavy lifting when Long is pushed to give an account of an act’s integral nature. When the person in whom one’s lethal action terminates is innocent, then the terminating action is not per se ordained to its end, whatever that end might be, and the act’s integral nature is to be an act of killing. But when the person in whom the lethal action terminates is not innocent, then the act undertaken becomes in itself an act per se ordained to its end. Which is to say that when the lethal act is morally justified, then it is not intended; and we are once again back to the circle. Long’s view here seems to be an instance of the approach to double effect strenuously criticized by Alison McIntyre in her well-known essay Doing Away with Double Effect, in which double effect reasoning is used after the fact to justify moral judgments made on other grounds. Alison McIntyre, “Doing Away with Double Effect,” Ethics 111 (2001), 219-255. This is not the approach of the NNLT. Its defenders believe that intention is one thing, moral assessment another. Thinking clearly about intention should not in fact be guided by prior moral judgments, and McIntyre’s judgment seems correct: if what is really doing the work is not a judgment about intention but something else, then intention, and with it double effect, should be jettisoned. It is certainly questionable whether such an approach is consistent with the Catholic tradition of moral thought, of course. I conclude, then, that the approach advocated by defenders of the NNLT has not been impugned in the least by the essays of its critics as assembled in the NCBQ and Ethics and Medics. My thanks to Joseph Boyle and Ryan Anderson for helpful comments on earlier drafts. 28 27