The claim that there are certain actions that are intrinsically bad and ought never to be chosen was once part of the common pagan, Jewish, and Christian moral inheritance of the West. Today, however, it is severely challenged. The reasons for this, of course, are many. Vindicating this once common view depends upon justifying the distinction between intention and foresight: that is, the distinction between intending and choosing an action and merely foreseeing the consequences and side-effects of an action. Moral absolutes apply to intention and choice. In our previous article, we introduced a natural law account of social reality and showed its relevance to the moral evaluation of actions. Now we put that theory to work and address particular moral dilemmas.
Take, for example, the nun who, in her capacity as ethicist at a Catholic hospital in Phoenix, recently approved an abortion. She made this difficult decision in order to reduce a severe risk to a pregnant mother who had developed pulmonary hypertension. By approving the procedure, did the nun share in an intention to kill the unborn child and thereby violate an exceptionless norm against intending the death of an innocent human being? Or did she blamelessly intend to reduce grave risk to the mother, merely foreseeing the unborn child’s death as a certain but tragic side-effect?
The answer, barring further revelations about the specifics of the case, is that the nun shares responsibility for intentionally killing the unborn child and violating the exceptionless norm. She cannot be exonerated from intending the child’s death by appealing to the principle of double effect, because although she had an admirable goal in saving the mother, the means she approved of choosing for the sake of that goal was the death of the innocent child. The principle of double effect permits sometimes bringing about bad effects foreseeably but prohibits choosing them, and she chose the child’s death as a means to reduce risk to the mother.
Why shouldn’t we say that the nun chose merely to “end the pregnancy” and not to “kill the child”? No doubt she regretted the child’s death, and if the child had miraculously survived the abortion, then she would have rejoiced; surely there is nothing wrong with “ending a pregnancy,” so described, since after all, a birth ends a pregnancy as well. This objection rests on a faulty Cartesian model of act analysis, which misunderstands what is included in a means to an end, and opposes the natural law account we sketched in our previous essay. Recall that the natural law account emphasizes that human acts—intended ends and chosen means—have a dual dependence upon both (a) the nature of the individual people who perform them and (b) the social practices in which they are performed.
The Cartesian, by contrast, analyzes intentions solely in reference to the choosing self’s professed plan in acting. On this view, a means to an end is a conceptual construct, which the agent creates through introspection about his goals, and is not predisposed by any natural or social facts about human behavior. In so doing, the Cartesian misunderstands intention as a private phenomenon, unmoored from both the particularities of human animal and social life. So long as the agent is sincere about his professed plan, the Cartesian holds that his intention in acting just is what he says it is—neither more nor less. In neglecting social practice, the Cartesian commits what might be called the “Humpty Dumpty fallacy,” after a famous passage from Lewis Carroll’s Through the Looking Glass:
“I don't know what you mean by ‘glory’,” Alice said.
Humpty Dumpty smiled contemptuously. “Of course you don’t—till I tell you. I meant ‘there’s a nice knock-down argument for you’!”
“But ‘glory’ doesn’t mean ‘a nice knock-down argument’,” Alice objected.
“When I use a word,” Humpty Dumpty said, in rather a scornful tone, “it means just what I choose it to mean—neither more nor less.”
“The question is,” said Alice, “whether you can make words mean so many different things.”
“The question is,” said Humpty Dumpty, “which is to be master—that’s all.”
Carroll’s ironic dialogue is a nice illustration of how, pace Cartesianism, actions get constituted by social practices. Words are uttered in speech acts that are parts of the social practice of language, so when you use a word to express a meaning, that meaning is a function of the linguistic practice, not of your subjective wishes. Your wishes determine which words you choose, not the meanings of the words you choose.
Suppose that you utter the sentence, “God lacks all glory.” This is an act of blasphemy (unless, say, you’re speaking as a character in a play). Suppose, further, that you claim to be innocent of intentional blaspheming, since although you know what the standard of meaning of “glory” is in English, you intend your use of the word to mean, on this occasion, “fault.” After all, we can certainly conceive of a possible world where “glory” did mean “fault,” even though, in the actual world, it means magnificent splendor. This defense would be sophistical, because you would have supposed, like Humpty Dumpty, that word meanings are products of private willing, not public linguistic convention. But contra Humpty Dumpty, our words come with ready-made public meanings, and private willing comes in when we deliberate over which word meanings to use. It is possible to commit the Humpty Dumpty fallacy not just in choosing a word in order to mean something, but also in choosing an action for the sake of an end. Just as words get their meanings conferred by the practice of language, so non-linguistic acts get their significance as choice-worthy pieces of behavior conferred in part by medicine, warfare, parenting, voting, playing, and so on.
In the case of the Phoenix nun we were examining above, the relevant social practice is medicine. It is impossible for her to have intended to “end the pregnancy” without thereby intending to “kill the unborn child,” because the nature of the procedure that the nun approved was fixed by the practice of medicine, regardless of any subjective wish to pry apart descriptions by conceptual fiat. She approved an abortion procedure that is performed upon the child and whose success criteria include the destruction and removal of the pre-viable child from the womb; evacuating a pre-viable child from the womb is a way of killing the child. Therefore, to intend the former is to intend the latter, differently described. What somebody chooses is not determined by the miraculous possibilities she is capable of imagining in view of the benefits she seeks, but by the practices in which she is a participant. Your intentions determine which means you choose, not the significance of the means you choose.
The unjust killing in the Phoenix case can be contrasted with surgical procedures that also cause deaths but are permissible because they do not involve intentional killing: the removal of a cancerous uterus from a woman who happens to be pregnant and the (admittedly trickier) case of treating an ectopic pregnancy. Both of these cases highlight how human animality—and in particular, proper biological functioning—determines the nature of human actions in addition to social practice. In the cancerous uterus case, the object of the surgical procedure is the uterus and not the child, whose presence is unrelated to the lethal risk that the diseased organ presents to the mother—a risk that could arise apart from the child. Therefore, it is possible to treat the cancer by removing the womb, foreseeing the death of the child who happens to be present and whose development is incidental to the cancer, but not intending the death.
Evaluating the treatment of an ectopic pregnancy is more difficult, because in an ectopic pregnancy, it appears to be the unborn child’s very presence that threatens the mother. In an ectopic pregnancy, the embryo implants outside the uterus, e.g., in a fallopian tube. And although outside the uterus, the embryo’s maturation is impossible and death certain; even its partial continued growth can kill the mother by causing the fallopian tube to rupture. We would argue that it is possible to prevent rupture by removing the tube (either surgically or chemically) without intending the embryo’s death, even though this is foreseen with certainty.
Why? The reason is that the object of the procedure, fundamentally, is the rupture-prone tube and not the embryo itself. Although the embryo’s growth in the tube is what causes the threat of lethal rupture, the embryo’s causality is only per accidens and not per se. This fact is the crucial difference from the Phoenix case and is a principled ground for individuating act descriptions and applying double effect. Lethal rupture of fallopian tubes is not a per se effect of human embryological development. In preventing such a rupture by removing a tube, a surgeon need not therefore intend to disrupt embryological development, which would amount to wronging the embryo, because in any case, development has already been disrupted by defective and non-viable implantation of the embryo outside the uterus. In removing the tube, the surgeon is not depriving the embryo of a condition that is sufficient for its survival, nor of a condition that is proper for the embryo at that point in its life history. The object of the procedure is the rupture-prone tube, not the embryo itself.
In both the ectopic and pulmonary hypertension cases, the continued presence of the embryo/fetus is a serious risk to the mother. It is our claim that there is a crucial further difference between the cases in the causality of the risk involved. In the ectopic case, malfunctioning in embryonic development causes the risk to the mother, which the surgeon remedies by removing a fallopian tube. In the hypertension case, proper functioning in fetal development causes the risk to the mother, which the surgeon reduces by disrupting the normal development of the fetus and depriving it of the environment necessary and proper to its survival—in short, the surgeon pits the mother against her unborn child, sacrificing the life of the latter for the health of the former. The mother’s own illness is tragic, but because her unborn child is not an unjust aggressor, her child may not be killed for the sake of her health.
It’s worth noting that the distinction between per se and per accidens causality is operative everywhere in biology. The heart causes blood to circulate per se, but causes a certain audible rhythm only per accidens. The relevance of the distinction is also clear in moral dilemmas analogous to the pregnancy cases. Suppose, for example, that after exposure to a nuclear disaster, a radioactive victim is mistakenly placed in a hospital room with another patient in critical condition who hasn’t been exposed to radiation and is expected, eventually, to recover. Both patients are in an extremely fragile condition, and if they are moved, they will almost surely die. The first victim will soon succumb to radioactive poisoning and is receiving palliative care, but if he remains in the room, he will expose the other patient to a lethal dose of radiation, which will be sufficient to kill him, too. In this case, it is possible to move the radioactive patient out of the room, foreseeably but not intentionally bringing about his death by stress, in order to save the recovering patient from lethal exposure. Why? The terminal patient may be the cause of the threat to his roommate, but only per accidens, because radioactivity is not a per se effect of his properly human causal powers.
If human nature were radically different, e.g., if we didn’t reproduce sexually, then what it would be possible to choose and intend would be radically different. But given our proper biological functioning as it is, our possible intentional actions are circumscribed and predisposed in certain ways. Intention stands to behavior as form to matter, and therefore, any bit of behavioral matter eligible to be chosen, as it were, is patterned to be informed by a fixed range of intentions.
As we argued in the previous essay, one of the ways in which human nature is naturally inclined is towards forming conventions. This natural fact is why conventional social practices, too, can structure the nature of our intentional actions by structuring what is included in a means to an end. We have focused here on the implications of this account for hard cases in medical ethics involving the exceptionless norm against killing the innocent. There are, of course, other exceptionless norms, the application of which also requires appeal to social practice and human animality. Adultery, treachery, and torture are intrinsically evil kinds of action, and if somebody wants to act virtuously, he can never choose any action that satisfies one of these descriptions.
Although there are important arguments to be had over which types of action really are intrinsically bad, there is a deeper issue that must be confronted. This issue is the possibility of nature as a moral criterion. For according to an oft-repeated tale, science has disenchanted the world and humanity along with it. Nature now stands inert and silent in the face of questions about how we ought to live. So-called natural laws are just observed regularities that have no claim upon us as precepts. In the next and final installment of this series, we will take up this basic problem about the naturalness of morality.
Matthew B. O’Brien holds a Ph.D. in philosophy and Robert C. Koons is a professor of philosophy, both at the University of Texas at Austin. Koons is a senior fellow of the Witherspoon Institute and sits on the editorial board of Public Discourse.