Double effect, all over again: The case of Sister Margaret McBride
As media reports have made widely known, in November 2009, the ethics committee of St. Joseph’s Hospital in Phoenix, Arizona, permitted the abortion of an eleven-week-old fetus in order to save the life of its mother. This woman was suffering from acute pulmonary hypertension, which her doctors judged would prove fatal for both her and her previable child. The ethics committee believed abortion to be permitted in this case under the so-called principle of double effect, but Thomas J. Olmsted, the bishop of Phoenix, disagreed with the committee and pronounced its chair, Sister Margaret McBride, excommunicated latae sententiae, “by the very commission of the act.” In this article, I take the much discussed Phoenix case as an occasion to subject the principle of double effect to another round of philosophical scrutiny. In particular, I examine the third condition of the principle in its textbook formulation, namely, that the evil effect in question may not be the means to the good effect. My argument, in brief, is that the textbook formulation of the principle does not withstand philosophical scrutiny. Nevertheless, in the end, I do not claim that we should then “do away” with the principle altogether. Instead, we do well to understand it within the context of casuistry, the tradition of moral reasoning from which it issued.
KeywordsDouble effectCasuistryCatholic hospitalsEthical and religious directives for Catholic healthcare institutionsAbortionPrinciple of choosing the lesser evil
On its face, the highly publicized and much-discussed excommunication latae sententiae, “by the very commission of the act,” of Sister Margaret McBride appears well-founded in the principles of Roman Catholic moral thought. In November 2009, the ethics committee, chaired by Sister McBride, of St. Joseph’s Hospital in Phoenix, Arizona, permitted the abortion of an eleven-week-old fetus in order to save the life of its mother. This woman was suffering from acute pulmonary hypertension, which her doctors judged would prove fatal for both her and her pre-viable child.1 The doctors accordingly advised that she have an abortion; the mother eventually and reluctantly agreed (she was also the mother of four children at home); and so too did the ethics committee, in what the hospital later described as a “tragic case” . Yet, as Thomas J. Olmsted, the bishop of Phoenix, observed in a statement critical of the hospital and of Sister McBride, under the principles of Catholic moral thought, “While medical professionals should certainly try to save a pregnant mother’s life, the means by which they do it can never be by directly killing her unborn child…. The direct killing of an unborn child is always immoral, no matter the circumstances, and it cannot be permitted in any institution that claims to be authentically Catholic” .
In support of this claim, Bishop Olmsted went on to cite both Pope John Paul II’s encyclical Evangelium Vitae, which states that “direct abortion, that is, abortion willed as an end or as a means, always constitutes a grave moral disorder, since it is the deliberate killing of an innocent human being” [3, sec. 62]; and the United States Conference of Catholic Bishops’ Ethical and Religious Directives for Catholic Healthcare Institutions, which states that “[a]bortion (that is, the directly intended termination of pregnancy before viability or the directly intended destruction of a viable fetus) is never permitted. Every procedure whose sole immediate effect is the termination of pregnancy before viability is an abortion” [4, sec. 45]. Bishop Olmsted might also have cited the Catechism of the Catholic Church, which states that “direct abortion, that is to say, abortion willed either as an end or a means, is gravely contrary to the moral law” [5, sec. 2271]. In a statement on the diocese’s web site, as well as in a statement to the media, Father John Ehrlich, the diocese’s medical ethics director, put the point briefly, in language drawn from Paul’s Letter to the Romans (3:8): “no one can do evil that good may come” .
that the action in itself from its very object be good or at least indifferent;
that the good effect and not the evil effect be intended;
that the good effect not be produced by means of the evil effect;
that there be a proportionately grave reason for permitting the evil effect.
The Ethical and Religious Directives for Catholic Healthcare Institutions gestures toward the principle of double effect (hereafter, PDE): “Operations, treatments, and medications that have as their direct purpose the cure of a proportionately serious pathological condition of a pregnant woman are permitted when they cannot be safely postponed until the unborn child is viable, even if they will result in the death of the unborn child” [4, sec. 47]. Yet, pace the ethics committee, the case in question does not appear to satisfy all the conditions of the PDE, in particular, the third condition, that the evil effect (here, the death of the fetus) may not be the means to the good effect (here, saving the life of the woman). Instead, in the case in question, the abortion appears to have been “direct” inasmuch as it was “willed…as a means,” to use the language of Evangelium Vitae and the Catechism.
This case, then, appears to be different from operations in which the death of a fetus is an inevitable, secondary consequence. The paradigmatic example of such a case cited by most commentators, and likewise by the USCCB’s Committee on Doctrine in a June 23, 2010, statement , is surgery to remove a cancerous but gravid uterus [11, 12]. In this case, it is claimed that the death of the fetus is not “directly intended”; instead, what is directly intended is the cure of a proportionately serious pathological condition. Such a case thus appears to fall clearly under section 47 of the Ethical and Religious Directives while the Phoenix case does not. For the Phoenix case appears to violate Paul’s rule—named the “Pauline principle” by Alan Donagan —that evil is not to be done that good may come.
On second view
Sometimes, however, a little philosophy is worse than none. For this appearance of clarity—that while some cases clearly satisfy the PDE, the Phoenix case does not—is misleading. It takes more than introductory philosophy to see why.
A first point to note is that I am concerned in this article with the textbook formulation of the PDE. As readers will know, many philosophers and theologians have proposed many revisions to the PDE. These include, for example, no longer calling the PDE a principle, but instead, a doctrine or rule, and (more substantively) collapsing the third condition into the second on the grounds that, if the agent has a right intention (condition 2), he or she does not intend the evil means (condition 3), where a means is not simply a cause but an intended cause of an intended effect. (All means are then causes, but not all causes are means.) I refer to some of these proposed revisions, but my focus is the textbook formulation, because it was some semblance of this formulation that figured in the controversy over the Phoenix case and that is reverberating still . In particular, controversy has centered on the third condition of the PDE, “that the good effect not be produced by means of the evil effect.”
A second point to note is that examples or cases do more than serve to clarify the PDE. Instead, they are needed to specify it, which is to say, to guide its application. Consider the fourth condition: “that there be a proportionately grave reason for permitting the evil effect.” In the contemporary literature, this so-called proportionality condition is interpreted in two quite different ways. On one common interpretation, this condition calls for weighing good and bad effects against one another; the condition is satisfied if there is reason to expect the good to compensate for the bad.2 As Alison McIntyre has pointed out, however, a problem with this interpretation is that it situates the PDE within a consequentialist framework to which it brings merely a small yet, for a consequentialist, likely puzzling restriction, namely, that good effects must not come by way of bad effects . But the agent to whom the PDE is addressed is not a consequentialist seeking permission to maximize good effects. Instead, the addressee is someone wondering, in good faith, whether an action that is normally prohibited on grounds other than consequentialist—for example, killing a human being or exposing someone to the risk of severe injury if not death—is likewise prohibited when it would be brought about in the pursuit of a good end. Such an agent is concerned, not simply with weighing good effects against evil effects, but with the reasons for the action in question, which is to say, with whether the reasons are sufficiently serious to permit going ahead despite the evil that one foresees, or at least risks, doing.
Accordingly, the second common interpretation of the proportionality condition—namely, as Joseph Boyle puts it, that there must be “sufficiently serious moral reasons for doing what brings about such harms” [18, p. 476]—has more to recommend it. What this interpretation has going for it in terms of relevance, however, it lacks in terms of guidance. For it does not provide the criteria according to which one is to determine whether a given set of reasons counts as “sufficiently serious” . The upshot, to reiterate, is that examples or cases do more than serve to clarify the PDE. It is not the case that the PDE can be understood perfectly well on its own and that examples and the like only make it easier to grasp. Instead, without examples and the like indicating some hierarchy of goods and rules of practical reason, the fourth condition of the PDE, and the PDE itself, inasmuch as all its conditions are necessary, leaves one at a loss. This critical role of examples and the like helps explain, I think, why a case like that involving Sister McBride generates such interest and controversy. What is at stake in such discussions may be not only whether the facts of the case do or do not satisfy the conditions of the PDE; what is at stake may also be our understanding of the PDE itself.
As I have already noted, a commonly cited case is surgery to remove a cancerous but gravid uterus: more fully, hysterectomy against uterine cancer even when a woman is pregnant. (Whether it was a mistake to apply the PDE to this case is a disputed point in the literature,3 but as I have also noted, the USCCB’s Committee on Doctrine cites just this example in its defense of Bishop Olmsted.) This procedure is commonly thought to be permitted under the PDE since the object of the action—namely, that to which it is immediately directed, removing the cancerous uterus—is good or at least indifferent (condition 1); the action is done with the right intention of preserving the woman’s life (condition 2); the death of the fetus (which is understood to be not yet viable ex utero) is only foreseen and does not figure as the means toward preserving the woman’s life (condition 3); and preserving the woman’s life is a sufficiently serious reason to go ahead with the action despite the harm that it will also bring about and for which the agent is then responsible, though in this case, not culpable (condition 4).
As for why saving the woman’s life qualifies as a sufficiently serious reason, if one thinks of the doctors performing the hysterectomy as the woman’s agents, one can perhaps invoke Aquinas’s natural-law justification of self-defense (though it must be acknowledged that describing the hysterectomy as an act of self-defense is something of a stretch, as what one is “defending” oneself against is not another person but a pathology). In Aquinas’s words, “an action of this kind, from which is intended the conservation of one’s own life, does not have the character of being unlawful, since it is natural that one keep oneself in being as far as possible” [20, p. 611]. On this account, it is then the prerogative of self-defense arising from the natural inclination to self-preservation that does the work of justifying the action in question . Aquinas immediately adds, however, an important proviso: “nevertheless, a particular action coming from a good intention can be rendered unlawful if it should not be proportionate to the end. And so if one uses greater violence than is necessary for defending one’s life, it will be unlawful” [20, p. 611]. For Aquinas, in other words, the proportionality condition does not concern weighing effects against one another. Instead, what needs to be rightly proportioned is the act to the end; hence his statement that it is unlawful to use greater violence than is necessary to defend one’s life. But he further writes, lest he be thought to oppose all killing in self-defense, that it is not “necessary for salvation that a man omit an action of moderate self-defense [that is, an action proportioned to this end] in order to avoid killing another, because a man is more obliged [plus tenetur] to provide for his own life than the life of another” [20, p. 611], a claim again resting on natural-law grounds.
Hysterectomy against uterine cancer even when a woman is pregnant is commonly contrasted with fetal craniotomy in the course of labor gone tragically wrong.4 This latter case is commonly thought not to be permitted under the PDE, as it appears not to satisfy the third condition, namely, that the evil effect (here, the death of the fetus) may not be the means to the good effect (here, saving the life of the woman). From this point of view, fetal craniotomy to save a woman’s life, because it is done as a means, is the direct killing of an unborn child, and as such, gravely wrong. The customary, textbook interpretation of the PDE affirms the dictum that “he who intends the end intends the means.” In other words, it regards as intended not only the good effect that one seeks, but the means to that effect, which is then independently evaluated as good or evil . So a hysterectomy against uterine cancer is permitted even when a woman is pregnant, but an abortion when a woman’s life is endangered by complications of labor, or for that matter by pulmonary hypertension, is forbidden. For, in the first case, the death of the fetus is, strictly speaking, not intended but only foreseen—it is a so-called indirect rather than direct abortion—whereas in the second case the death of the fetus is intended, since he who intends the end intends the means.
It has been suggested that the death of the fetus in the Phoenix case might likewise be categorized as an indirect abortion, or in other words, as only indirectly intended. (By the way, the point of this terminology of direct and indirect intention, as Warren Quinn has observed, appears to be to acknowledge the “linguistic impropriety in an agent’s asserting, with a completely straight face, that a clearly foreseen harm or harming is quite unintended” [23, p. 335] (more on this shortly).) Writing in America, Kevin O’Rourke noted that “the hospital’s ethics committee identified the pathological organ as the placenta,” which “produces the hormones necessary to increase the blood volume in pregnant women,” and in this case, apparently “put an intolerable strain on the woman’s already weak heart” [24, p. 16]. What the committee might have recommended, then, was removal of the organ in which the placenta is located, namely, the uterus, thereby making this case analogous to hysterectomy against uterine cancer even when a woman is pregnant. As a physician observed in a letter on O’Rourke’s article, the committee might also have authorized, yet more precisely, removal of the placenta itself, “even though the procedure would indirectly result in the loss of the pregnancy” . Inasmuch as the death of the fetus would not serve as the means to saving the life of the woman—inasmuch as removing the placenta would do so—this procedure, too, might be cast as analogous to hysterectomy against uterine cancer, and so thought to be permitted under the PDE.
Let us put aside, however, the medical question of the identity of the pathological organ. I think one lesson that we can draw from the Phoenix case is that cases that appear to fall clearly under this or that precedent might appear quite different after casuists have had a look. Another lesson, perhaps, is that casuistry’s bad name is not altogether undeserved—at least insofar as it proceeds according to what has been called the geometric method, judging actions in view of whether they satisfy idealized, atemporal, and necessary conditions, and seeking to justify a controversial action by re-describing it in terms that satisfy the conditions in question [15, 26]. Whether O’Rourke has proceeded in this way is a separate question, but is it really the case that abortion in the Phoenix case would be morally justified if the doctors had intended to save the woman’s life by removing the placenta, thereby indirectly aborting the fetus, but not morally justified if the doctors had intended to save the woman’s life by directly aborting the fetus? Pressing this question further, Is there really a morally significant difference between these two cases, or, to use more often cited examples, the cases of hysterectomy and fetal craniotomy? Two philosophical questions are relevant here. One hard question is, In a paradigmatically permitted case like that of a hysterectomy, what grounds the claim that the evil effect, in this case, the death of the fetus, is not part of the intended means but instead only a foreseen side effect? Another hard question is, What grounds the claim that so-called instrumental intending—intending a harm only as a means to an end—is in all cases objectionable?
Consider three criteria for determining what is to be counted as part of the means: a causal connection criterion (something is part of the means if it is causally connected to what is done toward an end); an inevitable connection criterion (something is part of the means if it is inevitably connected to what is done); and a “same action” criterion (something is part of the means if it is the same action as what is done) . The problem for the PDE is that, in a hysterectomy, as in an abortion when the woman’s life is in danger, the fetus’s death is causally connected to what is done toward the end of saving her; the fetus’s death is inevitably connected to what is done; and what is done is the same action as killing the fetus, inasmuch as an action is not just a bodily movement, but is identical “with the causing of each and every consequence to which the doer’s agency in doing it extends” [13, p. 160].5 The upshot is that, if, on the basis of these three criteria, the fetus’s death is an intended part of the means in an abortion when the woman’s life is in danger, the fetus’s death in a hysterectomy is likewise an intended part of the means.
Another way to come at this problem is to consider H.L.A. Hart’s notorious objection to the PDE’s prohibition of fetal craniotomy. “[I]n such a case,” Hart claimed, “it could be argued that it is not the death of the foetus but its removal from the body of the mother which is required to save her life,” with the upshot that the death of the fetus in the course of removing it ought to be thought of “as a ‘second effect,’ foreseen but not used as a means to an end” [27, p. 123]. Phillipa Foot countered that this claim “makes nonsense” of the PDE, as there is no room, so to speak, between crushing the fetus’s skull and killing it, and thus, no conceptual space, again so to speak, to apply the intend/foresee distinction (that is, to claim that the crushing of the fetus’s skull was intended, but the fetus’s death was strictly unintended and only foreseen) . I think Foot is perfectly right that there is “no room” between crushing the fetus’s skull and killing it—crushing the fetus’s skull constitutes killing it, such that one cannot conceivably aim at the former without aiming at the latter—but is there any more room between removing a woman’s uterus and killing the previable fetus within it, such that one could conceivably aim at the former without also aiming at the latter?6 It seems not.
Perhaps it could be countered, in turn, that the death of the fetus does not form part of the plan in the case of a hysterectomy, in the sense that the doctor performing the hysterectomy is not guided in his or her choice of instruments by the aim of killing the fetus . But the doctor performing the fetal craniotomy could reply that he or she also is not guided by the aim of killing the fetus—that he or she is guided by the aim of removing the fetus from the woman. So again, the hysterectomy case and the fetal craniotomy case collapse into one another . A final, oft-invoked way to distinguish the cases, the so-called test of failure, likewise fails. In Donagan’s words, “A good test of whether or not you intend a particular foreseen effect of an action is to suppose that, by some fluke or miracle, the action does not have the effect you foresee, and to ask whether you then consider your plan carried out and your purpose accomplished” [31, p. 496]. But, of course, both the doctor in the hysterectomy case and the doctor in the fetal craniotomy case could answer, “yes, plan carried out and purpose accomplished,” even should both the woman (as intended) and the fetus (contrary to all expectations) live.
It must be admitted that there is reason to be uneasy with what might be called the picture of intention that Hart’s objection operates with. To intend is to will an action for the sake of an end. As J. L. Austin has put it (roughly but as well as anybody), an intention is then “as it were a plan, an operation order…on which I am acting, which I am seeking to put into effect, carry out in action: only of course nothing necessarily…so full-blooded as a plan proper” [32, p. 283]. Now, an action counts as intentional only under the descriptions that an agent knows it falls under—so, for example, it is mistaken to claim that Oedipus intentionally killed his father—but an agent may not pick and choose which description, of those that she knows the action falls under, is to count as the one and only description under which her action is intentional. Instead, as Elizabeth Anscombe has remarked, “Circumstances, and the immediate facts about the means you are choosing to your ends, dictate what descriptions of your intention you must admit” [33, p. 23]. To claim otherwise is practically to invite the ridicule that Pascal, in the seventh of his Provincial Letters, heaped on his Jesuit’s grande méthode de diriger l’intention, permitting one to do prohibited actions on the condition that one is able to find for these same actions a lawful object. The picture of intention that Hart’s objection operates with does not, then, have much to recommend it—yet this picture appears to be likewise operative in the claim that the death of the fetus in the hysterectomy case is, strictly speaking, not intended but only foreseen, as if one could pick and choose the descriptions under which one’s action is to count as intentional. (Recall here Quinn’s observation that the terminology of direct and indirect intention acknowledges the “impropriety in an agent’s asserting, with a completely straight face, that a clearly foreseen harm or harming is quite unintended.”) The upshot is that, if it is nonsense and bad faith for a doctor performing the fetal craniotomy to claim that he or she intends only to remove the fetus, which it certainly does appear to be, it appears to be equally nonsense and bad faith for a doctor performing the hysterectomy to claim that he or she intends only to remove the uterus.
I turn now to my second philosophical question: What grounds the claim that so-called instrumental intending—intending a harm only as a means to an end—is in all cases objectionable? Some counterexamples might lead one to think twice about this claim. Imagine a dentist who intends to inflict not insignificant, though temporary, pain, not as an end in itself, but as the only effective means to discover a patient’s areas of sensitivity (in other words, as a heuristic or as a diagnostic tool) . Or imagine a censor of obscene literature who, in order to detect the obscenities from which he intends to protect the reading public, must allow and even intend for illicit thoughts to arise in his mind, not for the prospect of the pleasure they provide, but as the only effective means to his end . Is the dentist morally in the wrong on the grounds that evil effects may not be means to good effects? Is the censor? I assume that it is unlikely that one would want to say that the dentist is morally in the wrong; one might have qualms about the censor because of one’s qualms about the end of censorship, but not, I think, because the censor is allowing himself pleasure as a means to his end. Of course, one would be highly critical, to say the least, of a dentist who sought to inflict pain as an end in itself, and one might also be critical of a censor—in any event, other censors likely would be—who did the job precisely for what might be called its side benefits. (If nothing else, he would be a hypocrite.) But here is the point: the third condition of the PDE does not, on its own, give one reason to think that instrumental intending—intending a harm only as a means to an end—is in all cases objectionable. Sometimes doing a lesser evil appears justifiable as a means of preventing a greater evil. One needs to call on other moral principles to judge the actions in question.
Thomas Nagel has claimed, in an oft-cited discussion, that “to aim at evil, even as a means, is to have one’s action guided by evil.” Further, to be guided by evil is wrong, since “the essence of evil is that it should repel us” and so move us to work “toward its elimination rather than toward its maintenance” [34, pp. 181–182]. If the question, however, is why it is objectionable to aim at evil only as a means, then Nagel has not answered this question, but begged it, by simply stating without explanation that even aiming at evil as a means is “to have one’s action guided by [it]” [17, 30]. Moreover, as my counterexamples have shown, there is reason to doubt this claim: surely it is nonsensical to say that, for example, the dentist who inflicts pain only as a means is guided by evil in what he does, though this claim would be perfectly right for a dentist who inflicted the same pain as an end.7
Another way to defend the third condition of the PDE is to reject the counterexamples that I have given on the grounds that the instrumental harms in question—temporary pain and illicit pleasures—do not constitute sufficiently evil effects, and that it is instrumental attacks on so-called basic goods, like life and knowledge, that the third condition should be understood to rule out. So, for example, killing as a means to self-defense would be ruled out, as might lying as a means to saving someone’s life. I am skeptical that not insignificant, though temporary, pain, to consider only one of the examples, does not constitute an evil effect, but in any event, what needs to be asked here is, What is it to attack a basic good? Does one attack the basic good of life in aiming to kill an attacker where the only way to stay alive is to aim to kill him? The answer to this question appears to depend on whether one’s action is guided by evil when one aims to kill only as a means, so this defense takes one no further than Nagel’s did.
Pursuing the example a little further, however, imagine a case of self-defense where killing the attacker is “moderate” or rightly-proportioned to the end of saving one’s life. In such a case, assuming, of course, that one has not already been mortally wounded, killing the attacker is admittedly a means toward saving one’s life, but killing the attacker is not a first, intermediate step in a succession of actions. One needs to be careful here not to assimilate unlike cases. Unlike, say, the destruction of human embryos for the purpose of curing disease sometime in the future, killing the attacker with the intention of saving one’s life is not a means prior to the achievement of this end. Instead, killing the attacker and saving one’s life are the same action, only differently described. (Otherwise we are not talking about a case of double effect.) For one does not kill the attacker (call this action 1) and then, thereafter, save one’s life (call this action 2). Instead, killing the attacker is immediately saving one’s life. It might be countered—rightly—that the killing is initiated before the saving of one’s life is realized; but it is also right that the initiation of the killing is the initiation of the saving of one’s life. From the point of view of the agent acting with the intention to save his or her life, there is not first one action (killing) and then another (self-defense). Instead, there is one, indivisible action guided by the intention of self-defense.8 To claim, to the contrary, that the agent is guided here by evil seems simply false. Instead, it is the good end of saving one’s life that does the guiding. Accordingly, it is misleading to say that one is doing evil here that good may come. What one is doing is seeking to save one’s life—full stop. There is no question, then, of contravening the authority of Saint Paul.
There are, of course, a number of further questions to consider. One is whether, seeing that it apparently does not withstand philosophical scrutiny, the textbook formulation of the PDE should be done away with.9 To be clear, I am not proposing in this article that the PDE be kept but that the third condition be thrown out. Such a proposal would effectively reduce the PDE to the fourth proportionality condition, inasmuch as the third condition (evil effect not a means to good effect) may be understood as an elaboration of the second condition (rightful intention) and even the first (lawful object). In other words, I am not at all advocating that the end always justifies the means—though I have proposed, in my counterexamples to the claim that instrumental harming is in all cases objectionable, that an end may in fact sometimes justify a means. In any event, I am not at all suggesting that there is no moral difference between terror and obliteration bombing, on the one hand, and so-called strategic bombing, on the other; or between preparing a narcotic in order to kill a suffering man and preparing a narcotic in order to relieve a man’s suffering, with the foreseen effect of hastening his death. But the moral difference is not even partly explained by claiming that, in the second case of each of these pairs, the evil effect is not part of the intended means. For no one has yet made good on this claim. And the moral difference is also not even partly explained by claiming that, in the first case of each of these pairs, the evil effect is intended instrumentally. For there are no answers to why instrumental harming is in all cases objectionable.
A second question is how to judge the Phoenix case involving Sister McBride. Unless there is merit to the claim that the pathological organ in this case was the placenta, it seems that the ethics committee’s decision to permit an abortion cannot be justified under the PDE in its textbook formulation. The Phoenix diocese, of course, drew just this conclusion. Given the problems with the PDE, however, one might well wonder why it matters whether the decision could or could not be justified in its terms.
What is needed to answer both of these questions, I think, is some insight into the practice and development of casuistry, the tradition of moral reasoning from which the PDE issued. As its name suggests, casuistry is case-based: it proceeds not “geometrically,” by subsuming cases under timeless principles, but “taxonomically,” by comparing, contrasting, and evaluating “the contours of one case … against other cases” considered to be already successfully resolved [15, p. 296].10 Over time, principles like the PDE may then arise, but the important point, as James Keenan has put it, is that “the principle is a shorthand expression of the taxonomic relationship among a number of paradigm cases,” and does not itself justify or give authority to these cases, which have an independent, internal certitude that the principle rather reflects . John C. Ford made much the same point some years ago: the PDE is “not … a mathematical formula, nor an analytical principle. It is a practical formula which synthesizes an immense amount of moral experience, and serves as an efficient guide in countless perplexing cases”—so long as it is “applied by a hand well practiced in moral principles and moral solutions” [37, pp. 289–290]. Casuistry, however, is vulnerable to degeneration into moral legalism, with prudence or practical wisdom giving way to rote application of rules under the guidance of “ethics experts.”11 Whereas the PDE should have what Keenan calls “a heuristic and confirming function” —suggesting that, if a new case satisfies its conditions, there is reason to think that one can be as certain about it as earlier thinkers were about the paradigm cases—what happens instead, when moral legalism prevails, is that the PDE is invested with authority of its own and made into the up-or-down test of a case’s morality. In other words, it is enshrined into, precisely, a principle. To put the point somewhat dramatically, the moral wisdom reflected in the PDE is replaced by an image of itself, the principle so derived, which poses the danger of distorting evaluations of new cases that do not satisfy this principle and that might better be resolved by looking to other cases and principles altogether.
Another principle within Catholic moral thought is the principle of choosing the lesser evil in circumstances where, however one acts, an evil is bound to occur . There is no question here, it must be noted, of doing evil that good may come: in the circumstances appropriate to this principle, one is not free to do good, but only to choose between evils. It seems that the Phoenix case involving Sister McBride is more congruent to the paradigm cases that gave rise to this principle than to the paradigm cases that gave rise to the PDE. That the principle of choosing the lesser evil has lately been obscured, for whatever reason,12 is only more reason to look at it again. I think that the Phoenix case also indicates that there is reason for the United States Conference of Catholic Bishops to consider the incorporation of this principle, duly qualified, into the Ethical and Religious Directives. For purposes of helping prudence prevail, there is more work to be done than the PDE can do.
For citations, see Keenan . Perhaps the most important development of this interpretation in the twentieth century was that of Peter Knauer, S.J. On his account of proportionate reason, “on doit admettre un mal si c’est la seule manière de ne pas contredire exactement le maximum du valeur qui s’y oppose” [16, p. 371].
See Sulmasy . Sulmasy claims that it was a mistake, originating in the early twentieth century, to apply the principle (which he prefers to call a rule) to the cases of the cancerous gravid uterus and tubal ectopic pregnancy. Yet he acknowledges that “the traditional view” is that “these are typical applications of the RDE (even paradigmatic instances of its use).”
For an instructive, though doleful, discussion of fetal craniotomy from a medical perspective, see Parikh .
Cf. Germain Grisez : “a performance considered as a process of causation in the order of nature includes not only the bodily movements of the agent but also the inevitable physical effects which naturally follow from those movements. For example, the performance of lighting a match includes the match igniting; the performance of eating includes eliminating hunger…” [11, p. 88]. The claim that one’s agency is identical “with the causing of each and every consequence to which the doer’s agency in doing it extends” requires, however, two qualifications. First, one’s agency is cut short, so to speak, by the intervention of a new agent. Second, one’s agency does not extend to abnormal events, which is to say, to events not reasonably expected to follow. For a discussion of these qualifications, see Donagan .
Cf. Peter Clark on the use of methotrexate for tubal pregnancies: “A foreseen consequence of this action is the death of the embryo but this is not intended as a means or an end. What is directly intended is to stop the DNA synthesis so that the life-threatening condition to the mother is avoided” [29, p. 16].
By the way, a dentist who inflicted pain in all instances both as a rightly proportioned means and as an end likely would not be as dangerous as a dentist who inflicted pain as an end only. But still, there would be reason to criticize both dentists morally. Moral evaluation concerns, after all, not only what is done, but the reasons for actions.
Cf. Knauer who is especially clear on the mistake of “le morcelage des actes selon leur structure physique extérieure, alors qu’il faudrait les considérer dans l’unité de l’objet moral” [16, p. 360].
The claim that casuistry proceeds not “geometrically” but “taxonomically” needs, however, to be qualified: sixteenth century “high casuistry” proceeded “taxonomically,” but what Keenan calls the seventeenth century “casuistry of accommodation” became deductive in its method. For more on this change, see Keenan [35, 36].
Cf., e.g., R.M. Hare, who speaks witheringly of “some ecclesiastics and lawyers, who simply make the old casuistical moves without any attempt to justify them” [38, p. 70].
The explanation, to speculate a bit, may have to do with the more or less consequentialist interpretation of the fourth condition of the PDE, in terms of weighing good effects against bad effects. This interpretation was advanced in the twentieth century, in much qualified form, by so-called proportionalists. As the tradition of Catholic moral thought is antipathetic to consequentialism, more conservative representatives of the tradition reacted by magnifying the third condition of the PDE, with the result that any doing of evil became nearly unthinkable. For precision on the “consequentialism” of proportionalists, see Lisa Sowle Cahill . For an interesting discussion of the reaction of proponents of the so-called new natural law, see Sulmasy, who notes “a revisionist tendency to interpret the entire Catholic moral tradition through the PDE, even if it demands more of the faithful than the Tradition would demand” [41, p. 124], in particular, on the question of withholding and withdrawing an extraordinary means of treatment.
I thank Lydia Moland and Peter Clark for discussions on this article and two anonymous reviewers for constructive criticisms of an earlier draft.