A much-discussed new law in the state of Nebraska has banned abortion after the 20th week of pregnancy, citing the contested notion of fetal pain. Of course, everyone can agree that we have a duty not to cause pain to others without a just cause. Bioethicists endorse the relieving of pain as an expression of the “principle of beneficence.” And international bodies concur that access to pain relief without discrimination is a fundamental right. As a society we even take efforts to eliminate pain from the process of executing capital offenders whose guilt is manifestly established. But how do we approach the possibility of fetal pain when the science remains uncertain?

I should note that my argument is relative to a community in which abortion is legal. If abortion is wrong by virtue of the kind of act it is, namely an act of unjust killing, which I judge it to be, then aborting a fetus knowing that he or she will or might feel pain makes the act worse. But what about a society like ours where abortion is, sadly, legal?

First, pro-lifers and pro-choicers should be able to agree on the principle of full disclosure. If fetuses feel pain, then where abortion is legal, abortion providers should disclose to the gestational mother the effects of her choice on the fetus. We may disagree about whether abortion is wrong, but we should be able to agree that withholding information relevant to making an informed abortion decision would be unfair to women.

Second, if fetuses feel pain, then where abortion is legal, abortion providers should also take reasonable measures to suppress the pain. We should be able to agree that it would be wrong not to try to remove the pain, just as it is when political authority fails to take reasonable measures to remove the pain from the process of legal execution.

Start your day with Public Discourse

Sign up and get our daily essays sent straight to your inbox.

If, however, evidence demonstrates that fetuses do not feel pain, then, where abortion is legal, abortion providers obviously have neither the duty to relieve that which does not exist, nor to inform women of what’s untrue.

Our moral analysis then waits upon the settling of the empirical question of whether fetuses feel pain. This settling requires convincing data. But when speculative data is used to direct normative considerations of what ought to be done, the measure of certitude justified by the data is decisive for guiding action.

To clarify my meaning, let me use an example. Some abortion opponents have argued that a secondary effect of the most common drug used in emergency contraceptives—levonorgestrel—is to render the uterine lining inhospitable to an implanting embryo. If at commonly prescribed dosages this is the case, then the drug sometimes acts as an abortifacient. Whether this is the case is an empirical question. If it is the case, it has moral implications for the behavior of anyone concerned for embryonic human life and contemplating the legitimacy of taking the drug.

Let us say for the sake of argument that rigorous data is inconclusive. I am then left with a doubt as to whether or not levonorgestrel might render the uterine lining inhospitable. According to my practical knowledge, informed, let’s say for the sake of argument, by the best available evidence, I might kill an embryo if I use this drug in such and such a way. The possibility that my action will cause a death gives rise to the duty, stemming from the requisites of fairness, to refrain from that action. I would need to be reasonably certain that it will not cause death before purposeful action is justifiable. This reasonable certitude can also be called moral certitude. And reasonable doubt and moral certitude about the same fact are mutually excluding.

Let me propose one more example. If reasonable doubt existed as to whether the new device known as the “Mosquito,” which emits a high-pitched noise to disperse loiterers, not only caused minor auditory discomfort but severe pain, the burden of proof would fall upon the manufacturer to give evidence that it does not before the device should be approved for general use. Proof, of course, would be simple to arrive at: ask those exposed to the “Mosquito.” Since fetuses cannot yet provide self-report in language we cannot simply ask them whether they feel pain.

Yet I think the principle still stands: the burden of proof would fall upon defenders of the “Mosquito” to rule out a reasonable doubt that the device causes severe pain before its common use was approved, or to take action to assure that this possibility is mitigated.

The burden falls on the one who might be doing wrongful harm to rule out reasonable doubt that they are. If you were hunting in the woods and saw something moving in the distance, but were unsure of whether it was a deer or another hunter, you would be bound not to shoot until reasonable doubt was dispelled that what was stirring in the distance was not another hunter. When a doubt of fact bears on settling whether an alternative under consideration is immoral (e.g., it would be immoral to shoot in the face of reasonable doubt), one should withhold choosing till the fact has been settled.

So the question to be settled is whether or not reasonable doubt exists concerning a fetus’s capacity to experience pain. Since empirical certitude is not available, I propose, in light of what I said above, the following principle: that the judgment that fetuses do feel pain need only be a reasonable explanatory hypothesis in light of the settled evidence. Whereas the judgment that they do not requires moral certitude before providing a speculative ground for normative judgments about how to act.

Coming from one who is not a scientific expert on the question, but who has read considerably over the past four months on most all dimensions of the question, I conclude strongly that moral certitude that fetuses do not feel pain presently cannot be reached. In other words, fetal pain experience is a reasonable conclusion from the settled evidence. This evidence includes an appeal to fetal anatomical, neurochemical, physiological, and behavioral features, as well as responses to noxious stimuli (behaviors such as facial grimacing, the withdrawing of limbs, clenching of fists, opening of mouth and even crying).

Although we are not warranted in moving from these features and responses to a certain conclusion that fetuses do experience pain—I realize that fetal consciousness is a central factor in the equation—we are justified in concluding from the evidence—in fact, we are rationally required to conclude—that moral certitude does not exist that fetuses do not feel pain.

Unless and until contrary evidence is presented, we have a duty to act with the presumption that they do. In a territory such as our own where abortion is legal, we have a duty: 1) to inform women considering second- and third-trimester abortions that their actions may cause their babies pain; and 2) to guarantee that suitable analgesics and anesthesia be administered to fetuses during second- and third-trimester abortions.

This paper is adopted from remarks given at the conference “Open Hearts, Open Minds and Fair-Minded Words,” held at Princeton University October 15th-16th, 2010.