Reevaluating The Historical Evolution of Double Effect:
Anscombe, Aquinas, and The Principle of Side-Effects[1]

Eric Rovie (Georgia State University)

(Published February, 2006)

PDF Icon

Printer-friendly PDF version
(34 pages, 216Kb)

The doctrine of double effect[2] is a principle of reasoning common to moral philosophy and pervasive in contemporary applied ethics that argues that, under certain conditions, an agent can perform an act that aims towards good ends while also knowing that the action will bring about some bad results.[3] Some understand it as a release from an absolute prohibition on acting in a certain way: in other words, the doctrine allows an otherwise prohibited consequence that follows as a side effect of some other act. It is utilized frequently in discussions of the morality of killing (most notably for use in wartime) but is almost as often misunderstood and misused. I will examine the genesis of the doctrine in the works of St. Thomas Aquinas and offer a historical comparison to the evolution of the doctrine in the works of Elizabeth Anscombe, one of the doctrine's most prominent contemporary defenders. I will argue that the DDE, as presented by Aquinas, is most closely replicated by the later works of Anscombe, notably in the 1982 "Medalist's Address" as the "principle of side effects." Other interpretations of the doctrine, including the seminal interpretations presented by the Jesuits and Salmanticenses, as well as in earlier papers by Anscombe, are not as true to the spirit of double effect as expressed by Aquinas himself. I will leave it open for debate whether Aquinas' initial formulation is, in fact, the most useful interpretation: it is certainly the case that later interpretations offer a very clear depiction of the requirements surrounding the doctrine, and this is why the standard interpretation is so frequently invoked in the contemporary debates. The paper proceeds as follows: in part one, I examine the history of the doctrine, focusing on its development from Aquinas through the Salmanticenses and the Jesuits. In the second part, I analyze the early view of the doctrine in the work of Anscombe, from 1957 to 1961, connecting the views of her early work with the standard view of double effect. In the third part, I connect the doctrine of the textual Aquinas with the later view of Anscombe through an analysis of her 1982 "Medalist's Address." One thing that will be absent from this analysis is any substantive moral criticism of the doctrine itself, although it has been challenged by moral philosophers with some frequency for, I would think, at least plausible prima facie reasons[4].

§ 1: The History of Double Effect

1.1. The Structure of the Doctrine.In this preliminary section, I will first offer an outline of the structure of the DDE and an explanation of how it is supposed to work. For those who are familiar with the doctrine, this will be a recognizable summary of the historical evolution of DDE. Following that, I will examine its historical roots in Catholic moral theology, from Aquinas through the Salmanticenses and into the application of the theory to a whole field of moral theology by Jean Pierre Gury in the middle of the nineteenth century. The primary thrust of this historical analysis of the doctrine is to point out that the currently held version of the doctrine is, in fact, quite different from the doctrine as presented in Aquinas' own work.

The principle was developed and articulated by Catholic theologians, beginning with St. Thomas Aquinas in the Summa Theologiae, and was used primarily in Catholic moral theology until it was adapted for use in more secular discussions in moral philosophy and applied ethics.[5] As it is currently understood to be best expressed, the principle itself offers a four-pronged test[6] to evaluate whether or not the action in question would be considered morally permissible: first, the action itself must be either morally good or (at least) morally indifferent; second, that the bad consequence of the action not be directly intended but only merely permitted; third, that the good consequence not be directly caused by the bad one, or, put in terms of the so-called Pauline principle, "One should never do evil so that good may come"; and fourth, that the good result be sufficiently desirable and performed without unreasonable excessiveness to allow appropriate balance for the bad consequence. I will abbreviate these four principles as: the indifference requirement, the mere permission requirement, the simultaneousness requirement and the proportionality requirement.[7]

To set the debate up in the appropriate way, I offer one of the canonical examples of the DDE at work: the case of abortion to save the life of the mother, notably in the case of a cancerous uterus. These cases find a non-viable fetus present in a cancerous uterus with the imminent threat of death for the mother. It is in these cases that, even for those who believed (as Aquinas did) that the fetus was an ensouled human being[8], removal of the cancerous uterus (and, simultaneously, the abortion of the fetus) is justified. In these cases, it is argued that the bad effect (the deliberate ending of the life of the innocent fetus) is not directly intended.[9] The good result is not caused by the bad result (they occur simultaneously, rather than in a causal chain). The proportionality requirement is allegedly met (the good results seem to outweigh the bad results because, if the pregnancy continued, it would be likely that both mother and child would die) and the act itself (the removal of the uterus) is a morally neutral one.[10] This four-pronged approach is the current standard philosophical and theological understanding of double effect. I will call this the current or standard understanding of DDE, although there are deviations from this standard view.

1.2 Aquinas. It is not entirely clear, however, that St. Thomas Aquinas either formulated or endorsed the doctrine as it is currently understood. It has been argued both by theologians and historians that his view was far less nuanced than the standard view is understood to be.[11] The standard understanding of the doctrine is broached specifically in Aquinas' discussion of injustice, specifically as the question "is it legitimate for one man to kill another in self defense?"[12] Aquinas' reply brings at least a germinal version of the doctrine to the fore:

A single act may have two effects, of which one alone is intended, whilst the other is incidental to that intention. But the way a moral act is to be classified depends on what is intended, not on what goes beyond such an intention, since this is merely incidental thereto, as we have seen already. In light of this distinction we can see that an act of self-defense may have two effects: the saving of one's own life and the killing of the attacker. Now such an act of self-defense is not illegitimate just because the agent intends to save his own life, because it is natural for anything to want to preserve itself in being as far as it can. An act that is properly motivated may, nevertheless, become vitiated if it is not proportionate to the end intended.[13]

This explanation of dual effects seems to explain the third and fourth prongs of the current understanding, noting that the violence must be proportional to the end of self-defense going "no further than due defense requires"[14] and also that the two descriptions of the act seem to be occurring simultaneously: the act of killing one's assailant does not precede the saving of one's own life or, alternately, saving one's life does not follow directly from the killing itself. The two effects seem to occur simultaneously in the causal chain, with one not being directly caused by the other, or at least that is the claim of the defenders of double effect. One must only intend her own defense and not intend the death of the unjust aggressor, although if the latter occurred under the former description it would be morally justified.[15] However, it seems quite feasible to argue that the preservation of one's own life seems to follow directly from the killing of the aggressor and that, short of incapacitation or forcing a retreat of one's aggressor, a truly unjust aggressor may only be adequately defeated by direct slaying and, in all likelihood, a directly intended slaying. So, while we clearly get the proportionality requirement from Aquinas, it is not clear we meet the simultaneousness requirement, at least in the examples given by Aquinas himself.

This reading of this opening statement by Aquinas on double effect seems to cover only the latter two (and possibly only the last) requirements for the current understanding of the DDE. What of the first two? He follows the initial statement of what might be construed as the doctrine with a similar statement:

It remains nevertheless that it is not legitimate for a man actually to intend to kill another in self-defense, since the taking of life is reserved to public authorities acting for the common, as we have seen. Killing in self-defense in this sort of way is usually restricted to somebody who has the public authority to do so; such a man may intend to kill a man in self-defense but he does so for the general good. This is exemplified by the soldier who fights against the enemy, and the official of the court who fights against robbers. And even such men sin if and to the extent that they are moved by some private passion.[16]

Here we see a more direct implementation of the second requirement, the principle of mere permission. The act that has two effects must see the effect with bad consequences as being merely permitted and not directly intended. One cannot "lawfully will the evil effect as a means to the good effect," Mangan argues. One is permitted to kill one's assailant only if the killing is not the intended goal of the defense, but merely a side effect of the necessary force required to repel the assault. This seems to be corollary to the prior claim of simultaneousness (the third requirement) and the two may not be easily separable. The fact that the act with multiple effects must see those effects occur simultaneously to be justified might be explained in the following way: the act of defending oneself from an assailant (call that act A) can lead to the consequence B (killing one's assailant) and/or consequence C (successfully defending oneself from the assailant). Obviously, if one can achieve only C as a result of act A, this would be the best scenario: the assailant flees or is incapacitated, or apologizes, or something of the sort. However, if one can ONLY achieve C because B and C come together as a package deal (the assailant will only stop if killed), C can still be permitted with B as an unfortunate side-effect. This, of course, will only apply if the goal of the fatal self-defense is to obtain C: if one's goal is B, this cannot be a lawful killing.

Ignorance of the likely negative consequences of an act is not an acceptable excuse according to Aquinas. In his description of voluntary actions, he describes them as those that issue from the will both as action and as omission but which are also accompanied by an act of cognition. So "just as not willing or not acting are voluntary on occasion, so also is not considering."[17] But, as we saw in the discussion of simultaneousness, it is not clear, from the cases presented by Aquinas, that one can merely permit and not directly intend the killing of an unjust aggressor when it is not clear what amount of force would be needed to make the aggressor stop the attack.[18] It seems at least possible to argue that, in many, if not most, cases of unjust aggression (wartime scenarios, robberies, assault), it would be fair to assume that the best way to preserve one's own life is to "shoot to kill."

And what of the troublesome first requirement, that the act be morally indifferent? Here we seem to need some sort of reasonably construed and independent depiction of the moral quality of each particular type of act. Much of the Summa Theologiae is dedicated to providing guidance on just these sorts of questions: which action types are morally good (or at least neutral) and which ones are morally bad? Recall that the first prong of the test requires that the action type in question be either morally good or morally neutral. It seems, however, that we will need to rely on the "action under a description" model to make this work for the textual Aquinas.[19] On the one hand, we are explicitly told that it is "natural for anything to want to preserve itself in being as far as it can[20]." But, we are also directly informed at another point that intentionally killing another human being is expressly prohibited: "Any private individual may legitimately do anything for the common good provided it harms nobody."[21] We can do this even though cases of the former may involve the latter as well. So, we must to be able to distinguish between situations of self-defense and situations of killing, because, on the surface the most basic description of either action will be something like "person A caused the death of person B." Further muddying the waters, we also learn that an accidental killing is not a homicide if it does not occur as a result of nefarious activities or as a result of failing to take due care. So, an act that could be simply described as causing the death of another human being could also be described, according to Aquinas, as:

  1. Self-defense (without any intent to kill)
  2. Intentional killing
  3. Truly accidental killing (not in self-defense)
  4. Accidental killing caused by failure to take due care (not in self-defense)
  5. Accidental killing caused while committing nefarious acts (not in self-defense)

These descriptions may not exhaust all of the possible ways causing the death of a human being could be described. It is also going to be very difficult to determine which cases are which (telling the difference between an intentional killing and a self-defense act that simply goes too far might require a view into the deepest recesses of one's mind) but doing this is crucial for the doctrine to be invoked. While Aquinas might give us a clear list of which actions, commonly described, would be morally good, bad, or neutral, we will still be left with the difficult practical question of determining which actions were in self-defense and which were wrongly intended as killing. We are still left with the problem of intention or action, to use the terminology made so common by Anscombe and Davidson, respectively. We face challenging epistemological problems determining which category (or categories) each action might fit into, and it is truly difficult for us to know how to describe each action without a God's eye view of all the relevant circumstances. Even one's own description of one's action (with one's own internal assessment of the reasons involved) can be misguided and incorrect:[22] external views of another's action are even more likely to be wrong.[23]

Making this case even more difficult is the previously noted conundrum of the unjust attacker. The cases that Aquinas presents as examples of legitimate use of violent force are all cases in which the attackers are clearly unjust, and if they are construed as being sinners, they become legitimate targets. Since "sin corrodes the common good and so justifies the killing of the sinner,"[24] and the unjust attacker is considered a sinner, the indifference requirement may be a moot point for their actions. If a sinner's actions are corroded by her tendency to sin, the killing of a sinner might be morally required of man by God himself: "anyone who kills an innocent person in obedience to God's command commits no sin."[25]

So, what is the upshot of this analysis of Aquinas' version of double effect? While it is clear that he offers proportionality and indifference (although I would argue he offers the latter in the problematic formulation of the "unjust" attacker), it is not clear that he offers either mere permission or simultaneousness in the cases that he presents. But his is certainly not the only version of double effect, although his is the most frequently referenced version. The question at issue for this paper is whether the versions of the DDE that follow Aquinas' own presentation stick clearly to the spirit of his formulation: for that, we must look at those who commented on Aquinas' work in the centuries that followed.

1.3 Cajetan and the Salmanticenses. Little of any philosophical consequence was written on the DDE for several hundred years following the Summa Theologiae. At the beginning of the sixteenth century, Cardinal Cajetan provided the wording of the doctrine as we understand it today in his translation and commentary on Aquinas. Mangan points out that, even if Aquinas himself did not understand double effect as we do presently, Cajetan attributed this version of double effect to Aquinas and Cajetan's commentaries on Aquinas were influential enough to be considered authoritative in many circles. One of the crucial interpretations made by Cajetan was to attempt to clarify the meaning of intention under Aquinas' doctrine. Aquinas himself argued that intentional actions (or "will-motions") often come in two stages: "first, as the very last in which it comes to rest, and the finish of the whole movement. Second, as a midway position, which ends one phase and starts another."[26] Since the first usually subsumes the second (meaning that, when moving from A to C one must pass through point B), Cajetan claims that both the ultimate end and the means to that end come under the rubric of one's intention, and moreover that the death of the assailant is merely an effect that is part of the defense of oneself. It should not be thought of either as a means to the ultimate end (self-defense) or as the end in itself (intentional killing), but only as a necessary part of the defense of oneself against an unjust attacker. Cajetan offers us, in his analysis and commentary on the Summa Theologiae, a more direct expression of the requirements of mere permission and simultaneousness than is offered, textually, by Aquinas himself.[27] Cajetan's commentaries on the doctrine mark the beginnings of the more familiar, modern view of the doctrine, but those that followed Cajetan offered further interpretation and clarification.

Following Cajetan's interpretation of the doctrine, instances of practical application began to become more and more common by religious thinkers and practical theologians. This move, of applying DDE to practical life, was at least partially responsible for the continued interest in reevaluating and reformulating the doctrine in a more explicit fashion. As a practical concern it was commonly debated (and remains debatable) whether or not the DDE is needed for cases of killing in self-defense, given that many (diverging from standard interpretation of Aquinas) hold it is lawful to intend to kill an unjust aggressor to preserve one's life. Regardless of the adjudication of that particular debate, the DDE was applied to cases of wartime violence,[28] exposing oneself to mortal danger for good cause,[29] passive scandal,[30] and the arousal that might occur when a censor reads work of an illicit nature.

The topic of illicit sensual pleasure was dealt with specifically by Domingo de Sta Teresa in his treatise "De peccatis" in 1647, part of the larger collection of works by the Salmanticenses.[31] In the Cursus Theologicus of the Salmanticenses, it was argued explicitly that, for a sufficient reason, illicit sensual pleasures may be permitted without the stain of sin. The Salmanticenses offer a more direct description of the DDE, making explicit notice (present only implicitly in Aquinas himself) of the requirement of indifference. The Salmanticenses clearly state that actions which are evil in themselves cannot be allowed regardless of the good that may follow from them.[32] The only actions that may be considered in cases like these are actions that are either good or morally neutral. Aquinas never offers this first prong of the DDE explicitly in his assessment, but most commentators argue that it is implicit in his lengthy and detailed descriptions of the moral character of various action types. For instance, murder is never to be allowed under Aquinas' schema regardless of the situation, but killing in self-defense (understood to be different from murder) is not to be confused with murder, provided the appropriate intentions are present (or absent, as the case may be). For his categories to hold, Aquinas is bound to a very specific depiction of how actions and intentions are to be understood.

The Salmanticenses go further in their evaluation of double effect, pushing it considerably closer to the form in which it is commonly understood today:

It is lawful to actuate the cause of that evil effect, if a proportionately serious good effect follows equally immediately from the cause or proceeds the evil effect. And although the evil effect may be foreseen, it may not be intended…[33]

In this description, we get three of the four requirements from the standard understanding (missing only the indifference requirement, noted earlier). We then get a more nihilistic statement of the possibility of providing explicit rule guidance:

Again, the necessity must be proportionately greater when the evil permitted is greater in itself, when the cause is closer in nature to the evil, when it is more certain that the evil will result, and when there are fewer ways of preventing the evil after the cause is once actuated. Because of all these differences it is impossible to lay down more particular general rules.[34]

We are not offered an absolute prohibition on the permitting or causing of evil, but are told to measure the evil that will follow against the necessity of the good results that will also be generated. We get something akin to a consequentialist model here (or possibly a casuist model) in which particular rules are difficult to present because of the infinite complexities of each possible situation.[35] While we are still faced with some absolute prohibitions (we can't perform acts that are evil in themselves and we can't intend the evils that follow but merely permit them to occur), there is a relaxing of the rigid moral standards of Christian ethics at work with the doctrine. The Salmanticenses offered a far more specific explication of the DDE and how it was meant to be understood, but even more was done in terms of clarification by those that followed in the wake of the "Salmanticenses Morales", first published in 1665. In both the explicit discussions of the Salmanticenses and in later works that do not mention them, more explication and further clarification of the DDE were offered.

1.4 Jean Pierre Gury. Nothing seems to be closer to the standard modern understanding of the doctrine, however, than the presentation of DDE in Jean Pierre Gury's Compendium Theologiae Moralis.[36] At no point is this more clear than in Gury's treatise on human action where he states:

In the same act, there may be a direct will in itself, indirect in its cause and involuntary.[37]

This is followed by an analysis of the acts of indirect will:

That will, as it has been said, has not a direct aim in itself but results from something else directly thought of, as the effect from the cause. Who wants the cause, wants the effect which follows it, if that effect has been foreseen.[38]

Gury then moves to analyzing the effects of the indirect will:

A bad effect proceeding from indirect will, that is to say, from a cause indirectly willed, must not always be imputed as a fault from its author. In order that there should be fault, these conditions are necessary: that the author has foreseen the effect, at least confusedly; that he many not have been able to produce the cause; that he has been detained from producing the cause, or to have suppressed it, if it already existed…[39]

Finally, we get the direct utilization of the familiar terminology:

It is permitted to produce a cause good or indifferent, from which follows immediately a double effect, a good one and a bad one, if the cause is serious and the intention honest, not directed toward the bad effect. Because it this was not allowed, the author would sin, either by the intention of a bad effect, by the production of the cause, or by the prevision of the bad effect. Nothing of this can be sustained.[40]

In this abbreviated translation of the Compendium we get a rough and ready depiction of the DDE. Mangan provides a more explicit version of the doctrine, translated from the full German edition. Here we see the four part analysis of the DDE offered in its most explicit form, and in the form in which it is still most often referred to today:

It is lawful to actuate a morally good or indifferent cause from which will follow two effects, one good and the other evil, if there is a proportionately serious reason, and the ultimate end of the agent is good, and the evil effect is not the means to the good effect.[41]

The action is not unlawful because

First of all, it is not unlawful on account of the end intended, because the end is good. Secondly, it is not unlawful from the very actuating of the cause itself, because the cause in the supposition is either good or at least indifferent. Thirdly, not on account of the foreseeing of the evil effect, because in the hypothesis the evil effect is not intended but merely permitted; and fourthly, there is a proportionately serious reason for permitting the evil effect.[42]

Gury then finally enumerates the restrictions on the principle in the standard way:

1. The ultimate end of the author must be good, that it, the author may not intend the evil effect, because otherwise he would intend something evil and consequently sin. Hence, too, he may not consent to the evil effect in any way.

2. The cause itself of the effects must be good or at least indifferent, that is, as an act the cause must not be opposed to any law[43]. The reason is evident. For, if the cause is evil in itself, of itself it makes the action imputable as a fault.

3. The evil effect must not be the means to the good effect. The reason is that, if the cause directly produces the evil effect and produces the good effect only by means of the evil effect, then the good is sought by willing the evil. And it is never lawful to do evil, no matter how slight, in order that good may come of it. Therefore, one may never tell a lie even to save a man's life.

4. There must be a proportionately serious reason for actuating the cause, so that the author of the action would not be obliged by any virtue to omit the action. For natural equity obliges us to avoid evil and prevent harm from coming to our neighbor when we can do so without proportionately serious loss to ourselves.

Furthermore, in these cases a more serious reason for acting is necessary, the closer the cause is to the evil effect, the more probable it is that the evil effect will follow from the cause, and the less right the authors has to perform the action looked at in itself.[44]

It is at this point (the earliest editions of Gury's work appeared in 1850) that we have a firm starting point for the current understanding of the DDE, and most of the authors that followed Gury (both theologians and secular moral philosophers) in his explicit formulation of the requirements of the doctrine. We can see a direct line linking the Summa through Cajetan's commentaries on Aquinas, to the Salmanticenses and finally to the explicit formulation of the modern version of the principle in Gury. While it remains an open question as to whether Aquinas meant for the DDE to apply to as many cases, or in the specific ways now commonly understood, the doctrine is, at the very least, strongly inspired by the commentary on homicide at Article seven of question 64 of the Summa. I argue that Aquinas' description of double effect clearly contains the requirements of proportionality and indifference, but that the requirements of mere permission and simultaneousness are only implied and that they were further developed (and become a part of the traditional view) through the historical analysis and exposition on the doctrine that followed Aquinas.

§2: Anscombe's Early Thoughts about Double Effect[45]

2.1 Anscombe's Impact on the Debate. At some point between the late 1950's and the early 1970's, the DDE began to grip the imagination of moral philosophers. Up to this point, the doctrine had primarily been utilized by theologians and religious moralists, but Elizabeth Anscombe is at least partially responsible for placing it square into the forefront of the broader philosophical debate. In this section, I will take a closer look at the evolution of Anscombe's approach to DDE. Because of concerns about space, I will not evaluate several other seminal pieces on the doctrine from this time period (roughly 1960 to 1980), including those by Bennett, Foot, Boyle and Thomson, nor will I examine Anscombe's responses to them, although each deserves careful exploration in the context of the historical evolution of the doctrine.[46]

Anscombe is famous (or infamous, in some circles) for her curious mix of analytic philosophy and Catholic dogma. She offers, for example, a famous philosophical defense of the phenomenon of transubstantiation.[47] But she seems conflicted about the status of double effect, offering both support and criticism of it at various points. Her view on the DDE changes somewhat drastically from the time of "Modern Moral Philosophy" (1958) to her "Medalist's Address: Action, Intention and Double Effect" (1982).[48] She also has crucial things to say about the doctrine in "Mr. Truman's Degree" (privately published in 1957) and "War and Murder" (1961) that play into the evolution of her view.[49] I will argue that the position Anscombe holds on DDE in the three earlier papers is consistent with the standard view of double effect, but that her reanalysis of the doctrine in the "Medalist's Address" seems to be closer to the textual Aquinas in spirit.

2.2 Modern Moral Philosophy. "Modern Moral Philosophy" is a rich and varied piece, offering both an analysis of double effect and also strong criticisms of the lack of an adequate philosophy of psychology as well as an attack on the dominant moral philosophy of the time: consequentialism. In this latter critique, Anscombe analyzes Henry Sidgwick's construal of intention and the consequentialist way he chooses to analyze the moral status of actions. She begins by challenging his definition of intention, that "one must be said to intend any foreseen consequences of one's voluntary action,"[50] stating directly that it is obviously incorrect. She replaces Sidgwick's "faulty" conception of intention with a more appropriate version that argues "it does not make any difference to a man's responsibility for an effect of his action which he can foresee, that he does not intend it."[51] This leads, says Anscombe, "to its being quite impossible to estimate the badness of an action except in light of expected consequences" and allows that "you can exculpate yourself from the actual consequences of most disgraceful actions, so long as you can make a case for not having foreseen them."[52] This is clearly not acceptable to Anscombe, who says "a man is responsible for the bad consequences of his bad actions, but gets not credit for the good ones; and contrariwise is not responsible for the bad consequences of good actions."[53]

Anscombe leaves the distinction between actual and expected consequences out of this formulation, leading to what seems to be a strict liability standard for bad actions while offering a different standard regarding the consequences of good actions: If one performs a bad action, one is responsible for the bad effects but can receive no credit for whatever good results come from the bad action. If one performs a good action, one should get credit for the good results but should not be blamed for the bad results of a good action. Of course, determining which actions are good and which are bad is still crucial to this distinction, and since she is focused on refuting a consequential analysis of good and bad actions (in which an actions consequences are the scale on which goodness and badness of actions will be measured), she is required to provide an alternate depiction of good and bad actions. She does not do so specifically in "Modern Moral Philosophy," although she hints at a conception that might focus more on character traits in an Aristotelian sense and on the motivations for actions rather than simply the consequences of the actions. She further argues that we lack an adequately rich conception of moral psychology and says, "an explanation is required how an unjust man is a bad man, or an unjust action a bad one; to give such an explanation belongs to ethics; but it cannot even be begun until we are equipped with a sound philosophy of psychology."[54] Moreover, certain conceptions of good and bad action that might accord with religious laws (the "law" conception of ethics) and the requirements of religious reason would be acceptable (this was, of course, the conception of moral acceptability of action as presented by Aquinas, Cajetan, the Salmanticenses, and Gury) only if there was a consensus belief in a divine legislator in society. This is not the case in an overly secular world, and this is a concern for Anscombe's presentation of an ethical standard for actions.[55] She seems to want to replace the divine legislator view with, at minimum, an appropriate philosophy of psychology which can allow us to understand how and why we act in the ways we do that could, at least conceivably, account for secular and religious reasoning together as the general phenomena of human psychology.

2.3 Mr. Truman's Degree. Anscombe privately published and distributed "Mr. Truman's Degree" in 1957 in an effort to protest the honorary degree that Oxford was going to confer on Harry Truman. She did this on the grounds that Truman had perpetrated a horrible crime against the Japanese people (and all of humanity) by ordering the use of atomic bombs on Hiroshima and Nagasaki at the end of World War II. Her aim in this pamphlet, it seems, is to challenge the view that the nuclear terror unleashed upon Japan could be defended by the DDE.

Anscombe addresses, and seems to endorse, the DDE directly in the second part of the paper in her discussion on killing the innocent and murder. While choosing to kill the innocent as a means to your ends (and killing the innocent as an end itself, as the Nazis did) is always murder, killing the innocent as statistical likelihood of your actions is not necessarily murder. For instance,

if you attack a lot of military targets, such as munitions factories and naval dockyards, as carefully as you can, you will be certain to kill a number of innocent people; but that is not murder. On the other hand, unscrupulousness in considering the possibilities turns it into murder.[56]

There is, of course, the lingering question of who the "innocent" are in wartime situations that plays directly into the sorts of calculations that must be made. Anscombe resolves this tension by removing personal responsibility from the equation (so, for example, conscripts who might lack bad feelings or ill will towards their enemies are not morally guilty) and by defining innocence simply as "not harming." But she clings to the proportionality requirement in her discussion of whether or not one can harm sleeping enemy soldiers: while it is permissible to attack a sleeping enemy (because a soldier under arms is harming even if he is asleep), it must be remembered that "the enemy should not be attacked more ferociously than is necessary to put them hors de combat."[57]

She also addresses the first requirement of the principle quite directly: the moral nature of the act itself. While there is something to claims about the horror of killing (and we certainly don't want to lose the horror or else we might become complacent about killing), it would be far too easy to simply write off all killing, or war in general, as evil. This is not something Anscombe wants to do: she needs to show, to defend her claims about the bombing of the munitions plants, that "for one human being deliberately to kill another is not inevitably wrong."[58] For the pacifist, all killings are murder, and this misses the crucial distinction between legitimate acts of killing and those that are intended with ill will. The pacifist would be trapped, should others act in the same fashion, in a world with law but no force to maintain the law, a world where human beings would be miserable, a world where life would be solitary, poor, nasty, brutish, and short.

By extension, the pacifist makes another flawed claim: that war is evil, and that by being involved in any part of war requires that one has to go completely into the darkness and evil that is war. But this is mistaken, argues Anscombe, because while it is bad and unfortunate to be at war, war simpliciter is not evil. If two nations are at war, it is certain that at least one of them is unjust,[59] and this might give justification to at least one of them to commit acts of legitimate killing. Given the falseness of the doctrine of absolute pacifism, in light of the need for force to maintain order in the world, we must avoid falling victim to the conclusion that war is evil in itself. In the midst of her criticism of Truman, we see Anscombe clearly argue for two of the primary requirements of double effect, indifference of the act (killing, and war, are not intrinsically evil acts) and proportionality of the bad effect to the good results. There is also some discussion of the third requirement, that the bad act not be the causal condition for the good results and that one should be responsibly aware of the likely consequences of their actions, in her brusque reiteration of Aristotle's claim "you cannot be or do any good where you are stupid."[60] Simply being ignorant of the causal chain of events is not enough to assuage you of guilt. Anscombe does not claim Truman was ignorant: instead, she takes issue with the nature of the unconditional surrender of the Japanese demanded by the Allies, which seems to move the decision to drop the bombs from one class of action (necessary self-defense) into another camp altogether (a desire for a resounding, unconditional defeat and, possibly, an "exultant itch to use the new weapons"[61]). The DDE would not apply to the bombings of Hiroshima and Nagasaki but it would be imprudent and premature to claim that the DDE could never be applied to the use of atomic weapons or other forms of large-scale military destruction. It might be hard to justify proportionality, but it can't be argued that the use of atomic weapons is, in itself, an evil act.

2.4 War and Murder. "War and Murder," published in 1961, not surprisingly, reiterates some of Anscombe's themes from "Mr. Truman's Degree." What is important here is that we see Anscombe addressing directly the doctrine of double effect and its application to issues in war and morality. The denial of the doctrine of double effect "has been the corruption of non-Catholic thought, and its abuse the corruption of Catholic thought."[62] She points out that Catholics must pay at least lip service to the moral law against killing, yet most Catholics seem to have very little conscience about killing in general. The reason is the pervasive influence (and the destructive tendency to moral thought) of "double-think about double effect."

Despite the fact that the doctrine itself may be defensible, there is also an unfortunate slide towards the misuse of double effect. The reason for this, she claims, can be found in the influence of Cartesian psychology on philosophy and theology, an idea revisited from "Modern Moral Philosophy." Since intention, according to Descartes, is an interior act of the mind, one only needs to "direct your intention" in the appropriate way to avoid blameworthiness. In the case of wartime action the "devout Catholic bomber secures by a 'direction of intention' that any shedding of innocent blood that occurs is 'accidental.'"[63] But it certainly does not seem to be the case that it was merely an unfortunate accident that thousands of people were killed at Hiroshima and Nagsaki, yet it is necessary to claim precisely that if one wants to hold to Christian prohibitions on directly killing the innocent and yet still claim that the decision to drop the atomic bomb was a just one. This seems troubling to Anscombe, despite her continued defense of double effect as a useful principle and despite her adherence to Catholic values, because "it is nonsense to pretend that you do not intend to do what is the means you take to your chosen end."[64] The target here seems to be those who claim to adhere to the doctrine but who are being intellectually and morally dishonest about their intentions. To hold to the Cartesian theory of intention is to lead a double life in one's own mind: to be a "double-thinker" about the doctrine of double effect. This is not to serve as a strong criticism of the doctrine itself, but only of those who claim to use it under false pretenses or under a misunderstanding of the requirements. But she notes that the arguments about double effect, even when offered without Cartesian intention, are being used outside of the scope of Christianity. To claim, for instance, that the obliteration bombing of communists (the scourge of the continued religious existence and liberty of the Church) was justified under command of the Lord is simply to be a dishonest Christian. This might be the ultimate corruption of the doctrine of double effect, one that could lead to its demise and invalidation as a useful practical principle for the daily dilemmas that Christians face. Anscombe does not want to see the DDE go down this road, but she seems at least aware of the troublesome path it seems to have embarked upon.

2.5 Summary and Conclusion.

To very briefly recapitulate, I should point out that each of the four requirements in the standard view of the DDE make clear appearances in Anscombe's earlier works. Indifference is clearly present in all three papers, most notably in the discussion of killing as an act that is not always impermissible. Mere permission seems to be directly defended when she states flatly "to aim at killing, even when one is defending oneself, is murderous."[65] Simultaneousness, the third requirement, seems to be clearly what she has in mind in "Modern Moral Philosophy" when she argues that a person should get no credit for the good consequences of an evil action, and appears in her criticism of Truman's military decision: "We can now reformulate the principle of doing evil that good may come: every fool can be as much of a knave as suits him."[66] Finally, proportionality seems to lurk in her criticism of the Hiroshima and Nagasaki bombings, particularly considering the amount of innocent casualties in those cases. While she seems to accept that the military casualties were significantly lessened by dropping the bombs, the bad result (civilian casualties) does not seem to be in proportion to the good result (the estimated number of lives to be lost in the ground war on Japan), particularly given the possibility of a negotiated peace settlement between the Allies and the Japanese. In response to the claim that proportionality might not have much to say in 'borderline cases', Anscombe responds clearly and explicitly: "the obliteration bombing of a city is not a borderline case."[67]

All of this is an attempt to show that, despite some of her concerns about the misuse of the doctrine (particularly in "War and Murder"), Anscombe holds true to the standard understanding of double effect and, at various points from 1958 to 1961, endorses all four of the requirements of the standard view. Next I will examine the later analysis of double effect as presented in the 1982 "Medalist's Address" and will connect that analysis not with the standard view but with Aquinas' own view.

§3: The Medalist's Address and Aquinas

3.1 The Principle of Side-Effects. After having years to consider the relative merits and disadvantages of the DDE, Anscombe gives a more explicit evaluation of it in 1982's "Medalist's Address." In this paper, Anscombe explicitly addresses the problem of action and intention as it relates to double effect, a problem she hinted at in her discussion of the prevalence of the Cartesian moral psychology and how one can "choose" one's intention in "War and Murder." She begins by noting the basic problem of double effect: it is held, by most, that one must not kill either as an end or as means to some end. But, not all actions that lead to the deaths of others are murderous or forbidden. The absolute prohibition on murder can cover, according to Anscombe, only certain cases in which one person commits a deliberate action by which another human being is killed unintentionally. What Anscombe wants to defend is the situation in which death is "a side effect which is brought about as well as the effect being aimed at."[68] Dismissing the terminology of double effect as an "unfortunate Latinism," Anscombe changes the terminology into the "principle of side-effects."[69]

The principle of side-effects (PSE) is described in brief by Anscombe as the view that "the prohibition on murder does not cover all bringing about of deaths which are not intended,"[70] although there can be cases in which an unintended death can still be considered murder.[71] Her principle, she says, is more modest than the DDE: it simply says "where you must not aim at someone's death causing it does not necessarily incur guilt."[72] We can see clear examples that offer cases where the risk of future death is remote or uncertain: dangerous surgery, closing of doors to contain water or fire, or having ships and airplanes. While there are risks that these activities will cause death, it is by no means certain that anyone will, in fact, be killed. But what of the cases in which the risk of death seems to be considerably less remote or uncertain? Anscombe offers us the example of the stuck pot-holer to facilitate the distinction between direct and indirect killing in the PSE:

Imagine a pot-holer (cave explorer) stuck with people behind him and water rising up to drown them. And imagine two cases: in one, he can be blown up; in the other, a rock can be moved to open up another escape route, but it will crush his head. He will be killed by it.[73]

It would be impermissible, in these circumstances, to blow up the stuck pot-holer to create an egress to escape from the flooding cave. To do so would be to intend the killing of the pot-holer rather than to merely permit it to occur in the process of performing another action. But to move another rock to open an escape route offers a different way of thinking: "it isn't his being crushed that gives the escape route—and so his being crushed is something you didn't intend."[74] Crushing the head of the stuck pot-holer is merely an unfortunate and unintended side effect of moving a rock to create an escape route, but certainly cannot be considered murderous in the same way that strapping some explosives to him and blasting him out of the way would be.

The DDE is supposed to say, regarding the stuck pot-holer, that people could move the rock but they could not, under any circumstances, blow the pot-holer up. But the problem with double effect, according to Anscombe, is that people find this conclusion (that the two acts are morally different) to be "intolerably artificial and unnatural."[75] This is because, according to Anscombe, double effect tells us more than just what is or is not forbidden, but also what is and what is not allowed. The PSE, on the other hand, does not say when you may foreseeably cause death, [76] but only that the absolute prohibition on murder does not cover all unintended causing of death. The DDE, as Anscombe describes it, is a "package deal" which includes both what is forbidden and what is allowed.

The PSE, however, does not offer to us any assessment of the permissibility of moving the rock "except perhaps that it is not excluded on the score of being intentional killing."[77] Anscombe admits that, depending on the facts of the rock-moving cases, we might have to exclude it as a possibility: if the rock is balanced precariously over the head of the pot-holer and the result of moving the rock is "so immediate that the action could not be called taking the risk that that would happen,"[78] we could not feasibly argue that we did not intend that result. I will return to the facts of the pot-holer case shortly.

The DDE gives us access to what Anscombe refers to as an absurd device: it allows us to choose the description under which the action is intentional.[79] Simply by saying "I am moving what blocks the egress" or "I am moving the rock that blocks our escape," I imply that this is all that I am doing as a means to my ends. But this is, of course, absurd. One cannot claim "I am merely pulling the trigger of this .44 Magnum" or "merely swinging this running chainsaw through this region of space" without noting that the .44 Magnum is pointed towards a crowd of people or that the region of space is occupied with human beings. There is a certain threshold of reasonableness that is violated by making these claims about one's simple intentions. An act "does not merely have many descriptions, under some of which it is indeed not intentional: it has several under which it is intentional" and, because of this, "you cannot choose just one of these and claim to have excluded others by that."[80] This also prevents the inappropriate psychological phenomenon of "directing your intention" because the circumstances, the undeniable facts of the situation, will dictate the descriptions of intentions that you can admit to. Swinging a chainsaw through a space occupied by another person cannot, on the basis of the situation, be described as only swinging the chainsaw through said space. It must also be considered that a running chainsaw will cause great bodily harm and quite possibly death, to those present in that space.[81]

This brings us back to our unfortunate friend, the pot-holer. The circumstances of his confinement and the path and precariousness of the rock that must be moved to open the alternate egress must be taken into careful consideration before the action is performed. Deliberate ignorance ("I won't look through the crevice to see what the path of this stone might be just in case it is something I won't want to see") does not free one from moral responsibility and reprobation once the act is performed. Legitimate ignorance (where one can't tell, given one's situation, where the rock will fall) can be a grounds for saying you did not intend the killing: "you might not know that in moving the rock you would crush his head."[82] On the other hand, as was mentioned before, if the situation was that the rock to be moved was in a precarious position over the head of the pot-holer, with no place to go but onto the poor man's skull, and those trapped below in the water-logged cave were aware of all of these facts, then the situation is different. Given these sorts of facts, "if you do know, then where the crushing is immediate you cannot pretend not to intend it if you are willing to move the rock."[83] If the side-effect is clearly what will follow from the action, given the relevant facts of the case, it cannot be argued that simply saying "but I did not intend the side-effect, but only the primary effect" will assuage you of moral responsibility of your action. One must be responsible enough, as an action-performer, to be aware of the likely (and the certain) side-effects of one's actions.

3.2 The PSE and the denial of the "package deal." The case is slightly different when the result is not immediate: if the rock must roll down a slope towards the head of the pot-holer. In this case, Anscombe argues, "there is indeed room for saying that you did not intend that result, even though you could foresee it."[84] Where the DDE provides allowances, if the balance of good over evil is favorable, the PSE says only that moving the rock is not excluded by the prohibition on intentional killing. The PSE is not a "package deal" in the same way that the DDE, in the standard view, is. In her previous explications of the DDE, she claims, she was only trying to defend the thesis of the PSE and not the whole package of what is both permitted and forbidden, but had not realized that the DDE could be split into its component parts, including the PSE.

Accepting the PSE, but without delving into the permissions granted by the DDE, we must still provide some further principles on which to judge the unintended causing of death. The certainty of the death of the victim would be of primary concern here: if it were certain, or highly likely, given the facts of the case that moving the rock would cause the death of the pot-holer, we would have to add the moving of the rock from the list of prohibited actions. These prohibitions on action that the PSE gives us are there to tell us not what we can do, but only what we cannot do, and the PSE is best suited to cover only the causing of death, and not other sorts of harms. When examining a particular action in light of the PSE, what is most important is that one moderates the force one uses in achieving the end desired. This, according to Anscombe, is all that Aquinas means when he speaks of proportionality. A glance back at the discussion of proportionality seems to show just this:

An act that is properly motivated may, nevertheless, become vitiated if it is not proportionate to the end intended. And this is why somebody who uses more violence than necessary to defend himself will be doing something wrong.[85]

There is not any explicit calculation of good and evil consequences going on here, only a discussion of the nature of a particular action type (self-defense) and an attempt to keep the amount of violence going on in a self-defense action to the minimum required to be successful. If one strikes at one's assailant (who, unbeknownst to the victim, happens to have a particularly soft skull) and lands a killing blow to the head, but one was using what would have been reasonable force for the "average" assailant, this would be an accident and not something planned. Once again, the relevant facts, and the facts one should be aware of, come into play here. If the soft-skulled assailant was known by the victim to be soft-skulled, the PSE would kick in and warn the victim that a blow to the head would likely to be fatal, and the blow would therefore be morally prohibited

Those (including Gury and the Salmanticenses) who wished to defend the "package" view of the DDE were introducing their own doctrine on the proportion of good over evil consequences, a view of the consequences in the long run. For Anscombe, and for Aquinas, what is more important is whether the act is pre-conceived or accidental. She quotes Aquinas:

If it (the consequent event) is pre-conceived, it manifestly adds to the goodness or badness of the action. For when someone considers that much that is bad can follow from what he does, and does not give it up on that account, this shows that his will is more inordinate. But if the consequent act is not pre-conceived, then it is necessary to distinguish. For if it follows from that kind of action per se and in most cases, then the consequent event does accordingly add to the goodness or badness of the action; for it is clear that action is better in kind, from which more goods can follow, and worse, from which more evils are laible to follow. But if it is per accidens, and in rather few cases, then the consequent event does not add to the goodness or to the badness of the action: for there isn't judgment on any matter according to what is per accidens but only what is per se.[86]

The PSE, which Anscombe now finds the only part of the DDE that she is willing to offer a clear defense, offers only prohibitions on what one cannot intend to do. It offers no advice on which of the actions that are not ruled out by the PSE one is permitted to perform. The DDE, on the other hand, provides both the PSE (in prohibitions on certain intentional actions) and a scale upon which one can balance the consequences of good over evil (via proportionality) to determine which unforbidden actions can be performed.

3.3 Conclusion: A Shared View of Indifferent Acts. I have previously argued that Aquinas himself only offers the requirements of indifference and proportionality in his own presentation of the DDE, but that he does not distinctly offer mere permission or simultaneousness. Given Anscombe's new understanding of Aquinas' presentation of proportionality (that it must be considered only in the context of the proportion of force needed to successfully complete the desired action and not to achieve some balance of good over evil), it seems that only the requirement of indifference is left from Aquinas' understanding of double effect. Here he and Anscombe seem to share an understanding that certain intentional actions, despite the differing descriptions that might be offered for them ("I was merely swinging the running chainsaw through this {occupied} area of space"), as evil acts simpliciter. Blowing up the pot-holer is ruled out despite the good consequences that might follow (if those in the cave below included the Pope or Aquinas himself, for instance) because the act of intentionally killing the pot-holer is forbidden. Moving the rock is not forbidden by the PSE, but it says no more than that. Those trapped below the pot-holer have one option ruled out and one option that is not; but they will have to appeal to some other principle or set of principles to determine whether or not they can actually attempt to move the rock, knowing it is likely to crush the head of their trapped colleague. Anscombe does not give us any of those principles in the "Medalist's Address," but she certainly leaves us with many possibilities for them in her other writings on moral philosophy.

My primary focus in this paper was to provide some clarity to the historical intricacies of the DDE. My core claim, that the DDE as presented by the (textual) Aquinas is in fact closer to the PSE presented by Anscombe than to the standard view as presented by the Salmanticenses and Gury, does not offer any theological or philosophical challenge to the doctrine itself. The fact that the standard view is attributed to Aquinas is a mistake: his view is considerably more bare-boned that the versions attributed to him. My project is not explicitly normative, but instead historical: I wish to clarify the often confused doctrine of double effect, and to acknowledge with Anscombe that the ways in which many defenders of the doctrine misuse the concept of intentionality (falling prey to the Cartesian error of 'directing one's intention') seem to lead to an abuse of the doctrine. If there is any suppressed normative claim lurking under the surface of this paper, it might be that we should take seriously Anscombe's claims in "Modern Moral Philosophy": we need to examine the DDE (or the PSE) in light of recent work in empirical ethics and moral psychology.[87] It might end up that the doctrine, despite its long history in both Catholic and secular ethics and its deep intellectual roots, simply cannot hold up to critical evaluation.


[1] I am grateful to Larry May, Jose Luis Bermudez, Emily Austin, Steve Viner, and the editors of Studies in the History of Ethics for comments on earlier drafts of this paper. All errors, of course, are my own.

[2] Also known as the principle of double effect and abbreviated for the purposes of this essay as DDE and sometimes referred to as "the doctrine." Some sources note explicitly that it cannot be considered a doctrine per se because it is not taught as one according to the Church. I will avoid those murky waters here by simply referring to it in the most common form as a doctrine often used in moral philosophy and not necessary accepted as doctrine by all members of the Church.

[3] There may a distinction between the use of "ends" compared to "results" or "consequences." I offer this rough and ready analysis in part to explain how the doctrine is commonly expressed by philosophers and theologians alike.

[4] One underlying theme to this discussion, which is not the primary focus of this paper, is the difference between the "God's eye view" of intention that Aquinas holds (following Augustine) and Anscombe at least hints at and the far more subjective understandings of intention present elsewhere. Whether or not there is an all-seeing moral arbiter seems to provide a great deal of impetus for moral agents to develop the proper intentions. In situations where one's intentions are completely private, it would seem to be considerably easier to convince oneself that one intended to A when one really intended to do the more insidious B.

[5] See, for examples of the application to secular moral philosophy, several of the essays collected in The Doctrine of Double Effect, ed. P.A. Woodward (Notre Dame, IN: University of Notre Dame Press, 2001). These include Joseph M. Boyle, Jr., "Toward Understanding the Principle of Double Effect," G.E.M. Anscombe, "Medalist's Address: Action, Intention, and 'Double Effect,'" Nancy Davis, "The Doctrine of Double Effect: Problems of Interpretation" and Philippa Foot, "The Problem of Abortion and the Doctrine of Double Effect." These particular essays are reprinted and date from 1978 to 1984. Since then, of course, the philosophical floodgates about double effect have opened full stop. A good survey of the theoretical literature can be found in R.G. Frey's essay "The Doctrine of Double Effect" in R.G. Frey and C.H. Wellman, A Companion to Applied Ethics (Malden, MA: Blackwell, 2003).

[6] Note here that there is a distinction made between the ACT and the CONSEQUENCES almost immediately. When considering the morality of a act-token, what is being evaluated in the first prong of the test is the acceptability of the act-type: whether or not there are any act-types which are going to be absolutely forbidden is the devil in the details, although I could imagine, for instance, that rape or torture could easily be cited as impermissible act-types. Murder might be another, although the difference between justified killing and murder may be hard to clarify in many cases. Note the move from the action to the consequences of the action in the remainder of test.

[7] This is, for the most part, my own shorthand although the fourth requirement is frequently discussed explicitly as the proportionality requirement in current and historical work on DDE.

[8] For medieval writers, ensoulment was said to occur after 40 days (for the male fetus) and 80 days (for the female fetus). Aquinas himself argues "someone who strikes a pregnant woman performs an unlawful act. Therefore, if it results in the death of the woman or of an animated fetus, he will not escape the crime of murder, especially if death follows as the immediate result of the blow." See Summa Theologiae II-II, qu. 64, a. 8. Citations from Aquinas come from the following edition: Summa Theologiae (Blackfriars/ McGraw-Hill, London: 1970). Hereafter, citations from the Summa will be abbreviated ST.

[9] There is debate as to whether or not the "direct" intention is of any use here, or whether this should be discussed only in terms of intention simpliciter. I use the term "directly intended" following Aquinas' own discussion of Augustine: "Augustine must be understood to be referring to a man's direct intention to kill another in order to save himself from death." ST II-II, qu. 64, a. 7, but also acknowledge the possibility of the overly convoluted utilization of "direct" here.

[10] "Moreover an organ that is endangering an individual's whole body may legitimately be removed by his own consent for the sake of the well-being of his body as a whole since each individual is responsible for the well-being of his body as a whole." ST II-II, qu. 65, a. 1.

[11] See the discussion on pp 44-46 of Joseph T. Mangan, "An Historical Analysis of the Doctrine of Double Effect," Theological Studies (Volume X, No. 1, 1949: pp 41-61). See also F.J. Connell, "The Principle of Double Effect," New Catholic Encyclopedia, Volume IV (Washington D.C., Catholic University of America Press: 1967), pp. 1020-1022.

[12] ST II-II, qu. 64, a. 7. (pp.41-43)

[13] Ibid.

[14] Ibid 43. Italics in the original.

[15] The description of the attacker as "unjust" is common in the contemporary literature, but it is not explicitly stated in Aquinas' depiction of the doctrine. In the discussion, he mentions explicitly situations in which thieves are killed while breaking in, when enemy soldiers are killed in battle, and when an official of the court fights against robbers, all of which seem to indicate the permission of killing an unjust aggressor. But this is not an explicit condition of Aquinas' presentation. It does seem to be important to note that all of the acceptable cases mentioned involve a clearly unjust attacker, but it is not clear to me that simply being attacked means that the attacker is an unjust aggressor.

[16] ST, ibid.

[17] ST I-II, qu. 6, a. 3.

[18] Although this falls slightly outside of the scope of Aquinas' discussion of homicide, it does not seem at all clear that certain forms of incapacitation would be any better than killing. Shooting an assailant in both kneecaps, for instance, would likely stop any sort of attack, but might not be proportionally acceptable.

[19] I use Anscombe's terminology on intentions here, but it certainly can also be found in a "God's eye view" understanding of intention as seen in Augustine.

[20] ST II-II, qu. 64, a. 7.

[21] ST II-II, qu. 64, a. 3.

[22] Recent psychological literature on self-assessment seems to indicate that agents are, at the very least, not always accurate in their self-assessments of ability; I assume self-assessment of reasons is at least as tricky a proposition.

[23] For an interesting discussion on situations where the intention is quite clear but the action itself is less so, see Anthony Kenny's analysis of the African case Nyuzi and Kudemara v Republic in his Freewill and Responsibility (London: Routledge, 1978).

[24] ST II-II, qu. 64, a. 6.

[25] Ibid.

[26] ST I-II, qu. 12, a. 2.

[27] The translation and analysis of Cajetan's commentaries come from Mangan, 52.

[28] ST II-II, qu. 64, a. 6.

[29] ST II-II, qu. 64, a. 5. A prominent Biblical example of this situation can be found in 1 Machabees Chapter 6. Eleazar, fighting in the army of the Machabees, sighted an elephant upon which he believed the enemy leader was riding. Eleazar ran under the beast and slew it from below, knowing fully that he would be crushed by the falling elephant, but also intending the death or disablement of an enemy leader. What might be considered suicidal under most descriptions would be permissible, applying double effect, in this case.

[30] Passive scandal is the effect that performing one action has on another, proximate, agent. The standard example of this in theological literature is the permissibility of a priest visiting the sick in a hospital even though one of the sick, upon seeing the priest, becomes angry and blasphemes.

[31] This name is the designation of the authors of theological courses at the theological college of Salamanca, beginning around the middle of the Seventeenth Century. The Salmanticenses were strong adherents to Thomistic scholasticism and their work was held in high regard by the Catholic Church.

[32] Salmanticenses, Cursus Theologicus, t. 7, tr. 13, disp. 10, dub. 6, n. 211-213.

[33] Ibid.

[34] Ibid.

[35] Contrast this with a more deontological theistic view that absolutely forbids certain act-types regardless of the consequences of the disallowance of those acts. Whether theistic (Catholic) ethics is allegedly deontological or consequentialist is an interesting underlying question, but not my primary concern here.

[36] My citations from the Compendium here come from the edited version of Gury's Doctrine of the Jesuits, trans. Paul Bert (Boston: F. Bradbury and Company, 1880) that edits and translates selections from Gury's works from French to English. Where these translations were missing some of the fell textual references, I rely on the translation of the full text by Mangan.

[37] Gury, 46.

[38] Gury, 47.

[39] Ibid.

[40] Ibid.

[41] Mangan, 60.

[42] Ibid.

[43] It is not clear here whether Gury refers to laws of man or laws of God or Nature. I leave this question open, although I assume he means the latter and that those will be relatively similar to the former.

[44] Mangan 60-61.

[45] In the interest of space, I have edited this section considerably from earlier versions. What remains are the highlights of the early Anscombian views on the DDE, but many of the details are left out.

[46] See footnote 2 for citations on Foot and Boyle. See also J. Bennett, "Morality and Consequences" in S. McMurrin (ed.), The Tanner Lectures on Human Values, vol. 2, (Salt Lake City: University of Utah Press, 1981) pp 47-116 and Judith Jarvis Thomson, "Killing, Letting Die and the Trolley Problem," The Monist 59 (1976), pp. 204-217.

[47] See G.E.M. Anscombe, "On Transubstantiation" in The Collected Works of G.E.M Anscombe, Volume III: Ethics, Religion and Politics (Oxford: Basil Blackwell, 1981).

[48] "Modern Moral Philosophy" in 20th Century Ethical Theory, eds. Steven M. Cahn and Joram G. Haber (Upper Saddle River, NJ: Prentice-Hall, 1995), pp. 351-364, and "Medalist's Address: Action, Intention, and Double Effect," in The Doctrine of Double Effect, ed. P.A.Woodward (Notre Dame: University of Notre Dame Press, 2001), pp. 50-66. It can also be found in Human Life, Action and Ethics, eds. Luke Gormally and Mary Geach (St Andrews: St Andrews Press, 2005), a recent collection of mostly unanthologized articles and lectures given by Anscombe.

[49] "Mr. Truman's Degree" in The Collected Works of G.E.M .Anscombe, Volume III: Ethics, Religion and Politics (Oxford: Basil Blackwell, 1981), pp. 62-71 and "War and Murder" in Woodward, pp. 247-260.

[50] MMP, 358.

[51] Ibid.

[52] MMP, 359.

[53] Ibid.

[54] MMP, 354.

[55] MMP, 360.

[56] MTD, 66.

[57] Ibid, 67.

[58] Ibid, 68.

[59] This is Anscombe's claim: apparently for her, there is no possibility that both sides are waging a just war. I am skeptical of this claim in light of arguments made by Gentili and Suarez about the possibility of war being just on both sides. See Alberico Gentili, The Law of War and Peace, ed. John Rolfe (Oxford, At The Clarendon Press, 1933), Book I, Chapter VI and Francisco Suarez, Selections from Three Works, ed. Williams, Brown and Waldron (Oxford: At the Clarendon Press, 1944), Disputation XIII, Section VI.

[60] MTD, 65.

[61] Ibid.

[62] Ibid.

[63] Ibid, 257.

[64] Ibid.

[65] MTD, 68.

[66] MTD, 65.

[67] WAM, 265.

[68] "Medalist's Address," 60.

[69] Ibid, 60-61.

[70] Ibid, 61.

[71] Anscombe does not offer any clarificatory examples here, but she might be referring to situations in which reckless actions lead to death or if the action that led to the unintended death was illicit nature (killing in the middle of a robbery).

[72] Ibid.

[73] Ibid.

[74] Ibid.

[75] Ibid, 62.

[76] Ibid.

[77] Ibid.

[78] ibid, 63.

[79] Ibid. Also see "War and Murder, " p.257.

[80] Ibid.

[81] Robert Nozick makes a similar argument involving swinging a baseball bat for fun in space occupied by a cow's head. See his Anarchy, State and Utopia (New York: Basic Books, 1975), p.37.

[82] Ibid.

[83] Ibid.

[84] Ibid, 64.

[85] ST II-II, Qu. 64, A. 7.

[86] ST II-II, Qu. 20, A. 5.

[87] In a paper forthcoming in Cognition, Shaun Nichols and Ron Mallon analyze psychological responses to double effect cases and 'trolley cases' to determine whether they are as morally accepted as their defenders would claim them to be.