A Duty to Gestate?
Thomson’s claim that a pregnant woman could have no positive duty to gestate a fetus-person has been paid considerable attention both by those who criticize and those who defend the GSTh1 Some critics have challenged her core premise that obligations to assist others must always be explicitly or implicitly assumed, and its implication that positive duties could not arise circumstantially, especially when that assistance is life-saving. They have also challenged the suggestion that such duties could never be owed to strangers.22 If I happen upon a seriously wounded hiker on a deserted stretch of terrain, do I not come under an unquestionable moral duty to lend assistance, given I am the only one who can?
Others have sought to adduce a special responsibility owed by a pregnant woman to the individual fetus-person that inhabits her, grounded in her prior responsibility in becoming pregnant, where sex is consensual, or in responsibilities stemming from biological ties.23 In addition to this, critics of the violinist analogy sometimes point out that the burdens incurred by the person plugged into the violinist are far   
worse than the burdens of typical pregnancy. Pregnancies hardly ever render women bed-ridden and incapacitated for an entire nine months. Perhaps, it is argued, the burdens of sustaining the violinist are not such that any person can be morally required to undertake them in order to save another, but the same is not obviously true of typical pregnancies.
In response, defenders of the GST have rebuffed proposals that consensual sex or biological parenthood could, without more, constitute a source of special responsibility to keep a fetus-person alive.24 Alternatively, they have argued that in any case, the scope of one’s responsibility to assist another, whether that person is a stranger, neighbour, or child, could never encompass the extreme burdens of bodily aid entailed by pregnancy and the violinist scenario.25 Burdens of this magnitude, it is argued—perhaps, even, any burden involving the surrender of one’s very body— are simply outwith the scope of the moral or legal obligations people can have in respect of one another, regardless of the circumstances. As Bertha Manninen has argued, American law would not condone the forcible extraction of a kidney or of bone marrow from one citizen on behalf of another, even if the donor were responsible for the needy person’s situation of imperilment—for instance, if the need arose because of a car crash injury for which the mandated donor was responsible^6 Manninen cites the well-known American case McFall v Shimp, in which a Pennsylvania court ruled that a man could not be required by law to undergo a bone marrow transplant procedure in order to save the life of his cousin, for whom his bone marrow was the only viable match.27
The same is true of English law, which will not compel a parent to deliver up so much as a life-saving vial of blood for his own child, or force a woman to undergo a caesarean section even where it is the only way to preserve the life of her late-term fetus.28 If a pregnant woman were found to be under a legal obligation to offer bodily assistance to a fetus through continued gestation, this would appear to be the only instance of legally imposed duties of bodily aid to others.
In response to this, it might be pointed out that legal and moral obligation are not the same thing, and that we may be morally required to do, or refrain from doing, many things in respect of which we lack legal duties. Under English law, for example, there is no criminal liability for a person who chooses not to assist a child drowning in a shallow pool, although almost anyone would agree that the failure to do so is morally repugnant. Likewise, just because the law does not generally enforce duties of bodily assistance does not mean that weighty moral duties of the kind do not exist. Moreover, some have argued that legal duties of providing adequate parental care do indeed involve the thoroughgoing use of one’s body, if not the actual surrender of flesh and blood. Even those whose parental duties are     
limited to alimony payments must employ their bodies in whatever form of labour they rely on to generate the income with which to pay.29
This last observation about the blurred boundaries between bodily and nonbodily forms of assistance strikes me as a bad argument for the legal enforcement of any duty of bodily assistance that clearly exists in morality. There are good reasons to refrain from using the law to coerce forms of assistance which involve interference with our bodily integrity, and it is surely not a good objection to those reasons to point out that the performance of any duty requires the use of our bodies at least in the restricted sense that all activity is physical.
Be that as it may, Thomson is willing to concede that there may exist some circumstances in which the refusal to provide life-saving bodily assistance is morally deficient, even if not in breach of a clear moral requirement. She considers how our response to the violinist analogy might change if the host only had to remain plugged into him for one hour to save his life.  In that situation, Thomson admits that it may be ‘morally indecent’ for the host to refuse, but that acceding the request still falls short of a moral requirement. Similarly, Thomson imagines a situation in which she is mortally sick and can be saved only by feeling Henry Fonda’s ‘cool hand’ on her brow. Supposing that Fonda is already in the same room, and need only to cross it to provide the life-saving contact, Thomson admits that it would be ghastly of him to refuse, but maintains, all the same, that he is under no moral obligation to comply.
This, however, leads one to wonder what space Thomson believes exists between ‘morally indecent’ refusals and a moral obligation to do otherwise. From one perspective, failing to assist someone when it is morally indecent to do so is just one and the same thing as shirking a moral obligation. Doing the morally decent thing is not acting as a Good Samaritan, since moral decency is not moral supererogation; it is the bare minimum that morality requires. Surely Henry Fonda is morally obligated to provide Thomson with his ‘cool hand’ at such little cost, if we are sure that his refusing to do so is quite monstrous.
If moral obligations to provide bodily support are not anathema, it will matter for Thomson’s Good Samaritan argument just how onerous pregnancy is compared with being plugged into the violinist for nine months, assuming that we are indeed sure that one cannot be required to sustain the violinist. Whilst it is true that typical pregnancies do not render women bed-ridden for the entire nine months, as in Thomson’s example, the bodily burdens and risks accompanying even ordinary pregnancies are fairly significant. Perhaps pregnancy entails the kinds of burdens which no one can be morally obligated to provide in order to keep another person alive. Still, if we are minded to say, in Thomson’s amended example, that one could be morally obligated to sustain the violinist for an hour, where that is all it takes to save him, the possible duty to gestate (if abortion morality is a matter of positive obligations) might also depend on how much continued pregnancy is required in order to save a fetus. What if a pregnancy became unwanted only an hour before a fetus can be extracted alive? Following our conclusions about the amended violinist case, it would seem that the pregnant woman might well have a duty to gestate for that remaining hour, unless the cost to the woman of extracting the fetus alive, as opposed to aborting it, is itself above and beyond the things we must do to preserve the lives of others.
It warrants saying something confined about claims to the effect that a pregnant woman may fall under a special responsibility to sustain the fetus inhabiting her owing to her responsibility in becoming pregnant, or to relations of biological parenthood. I have so far mostly been treating pregnancy and the violinist case the same in taking both to ask what level of burdens we are morally bound to undertake if and when we find ourselves in circumstances wherein only our bodily assistance can sustain another person’s life. But some have argued that the duties a pregnant woman owes to her fetus may well outstrip those which one person owes to just any other, first and foremost because a woman may owe a special duty of care to a person she brings into existence by foreseeably becoming pregnant.
Boonin labels this the Responsibility Objection. The relevant point of disanalogy with the violinist case is that there you have been plucked at random and forcibly plugged into the violinist. However, where pregnancy results from consensual sex, it could be argued that a woman has at least voluntarily assumed the risk of becoming pregnant. This leads some to say that the violinist analogy only resembles pregnancies that are brought about by rape. In every other case, perhaps a woman’s prior responsibility in becoming pregnant could be the source of a special duty to gestate the fetus when the risk of pregnancy materializes. Moreover, even if Thomson were right that one cannot have responsibility for another unless it is assumed, explicitly or implicitly, perhaps consensual sex constitutes such an assumption.
Thomson in fact pre-emptes this objection at length. There is not enough space here to relay the entire philosophical discussion surrounding the Responsibility Objection. (And if it is true that, in any event, abortion does not equate to a failure to rescue, the issue will be moot.) Most notably, however, David Boonin has forcefully defended GST from the Responsibility Objection, which he usefully divides into two separate forms. One, which he calls the ‘tacit consent version’, claims that by engaging in consensual sexual intercourse, a pregnant woman tacitly consents to providing a fetus with bodily aid should she fall pregnant. The other, which he calls the ‘negligence version’ of the objection, claims that voluntarily engaging in sexual intercourse with the foreseeable risk of becoming pregnant is sufficient to ground a special duty to sustain the fetus whether or not it amounts to an actual assumption of responsibility, in much the same way that we might incur special responsibilities to aid particular individuals whom our negligent behaviour puts in harm’s way.
The ‘tacit consent version’ certainly strikes as the easier to dismiss, since it is far from clear that one can infer actual consent to pregnancy from mere willingness to run the risk of pregnancy. When people knowingly run the risk of contracting sexually transmitted diseases, for instance, we do thereby infer that they consent to fall victim to the disease. The ‘negligence version’ has more allure, for it seems that we might well hold someone responsible to assist another person in peril (especially mortal peril), and responsible in a way that goes above and beyond ordinary duties of assisting strangers, if they had acted in a way which foreseeably brought about that peril. For instance, as Francis Beckwith has pointed out, we hold drink-drivers responsible for the damage their negligence causes to other people, and duty-bound to rectify the damage, even though they were not trying to cause harm.32 Could the same logic be used to impute to many pregnant women the responsibility to assist their fetuses by gestating them to term?
Boonin attempts to refute the ‘negligence version’ of the Responsibility Objection by drawing what he believes is a fundamental distinction between being responsible for a needy person’s existence and being responsible for his neediness given that he exists. He does this with the use of a thought-experiment called the Imperfect Drug story.33 In ‘Imperfect Drug’, you are the violinist’s doctor and have just discovered that he has a rare ailment on the verge of killing him which can only be cured by a drug with an unfortunate side-effect: five to ten years after administration, it causes the kidney condition described by Thomson. Knowing that you alone have the required blood type to save the violinist when his kidneys fail, you prescribe the drug and cure the disease. You are now responsible for the violinist’s existence, in that you have extended his life by offering him the cure. But according to Boonin, this does not now make you responsible for his future neediness, given that he exists. Specifically, you are under no duty to offer him the use of your kidneys later down the line simply because he would not, at that time, be alive but for you. Responsibility in sense one, responsibility for existence, does not entail responsibility in sense two: responsibility for neediness, given that a person exists.
Boonin’s retort sounds quite convincing. However, we might feel that some morally salient differences hold between foreseeably bringing a fetus-person into existence and Boonin’s Imperfect Drug story. Francis Beckwith believes it an important difference that in Imperfect Drug, the physician is only responsible for the violinist’s existence in that she extends it.34 She does not bring the violinist into existence. In contrast, the fetus’s progenitors do not extend the life of an already existing human being; they bring into being a brand new one, who is needy by nature, and in need of the pregnant woman’s body to survive. Beckwith thinks this difference important because, by giving him the drug, the physician has still decreased the violinist’s net neediness, given that he was on the brink of death at the time. But the pregnant woman can make no such claim. Her voluntary acts have not decreased the net neediness of the fetus, but rather brought into existence a needy-by-nature being which would otherwise not have existed. Beckwith thinks this sufficient to establish negligence-type responsibility in the pregnancy case.
Finally, some people do find merit in the notion that biological ties to other individuals can ground obligations to them that we do not ordinarily owe to others. The putative point of disanalogy here is that the violinist in Thomson’s case is a stranger, whereas the pregnant woman stands in a relation of biological parenthood to her fetus. Parents are often thought to owe special duties of care to their offspring, whether they have undertaken them or not. Biological fathers, for instance, are under a duty to pay child support whether or not they intended or accepted their biological parenthood. Perhaps Thomson is simply wrong to deny, as she does, that parents have special responsibility for their children unless they have assumed it. The ‘natural claim’ children have over their parents, Beckwith argues, explains why we hold parents to account for the fulfilment of their duties, prosecuting them, for instance, when they neglect to provide care. Some may take the view that the duties of child support which the law poses upon biological parents regardless of whether they intended pregnancy (alimony payments, for example) demonstrate our belief that biological parenthood does ground special moral duties to one’s offspring, regardless of whether those duties are assumed.
The obvious counter to this objection is that it is simply a mistake to think that biological relations per se ground obligations. Like Thomson, Boonin is extremely sceptical about this possibility. He believes it absurd to think that anything about our reactions to the violinist case should change if it were discovered that the ‘stranger’ violinist is in fact your son, whom you never knew existed, created in vitro years ago from gametes you contributed to a medical study. As Boonin argues, biological parents are commonly thought to have strong duties towards their children because they are their guardians. In other words, it is social parenthood which is the true source of obligation. Hence, when children are adopted, it is the adoptive parents, not the biological ones, who are held responsible for their welfare. Thomson claims that biological parents typically assume the social role of parenthood when they choose not to abort pregnancies and take a neonate home. For Thomson as well, it is the accepted role of social parent rather than biological relation which is the source of obligation.
Boonin is surely right to think that social parenthood trumps biological parenthood as far as duties of care are concerned. When a child is adopted, it is clearly the adoptive parents and not the biological ones who are specially responsible for her care. However, even on this ‘social responsibility’ explanation for the typical burdens of biological parenthood, it is not clear that the pregnant woman is off the hook. For Boonin’s argument here contains a further premise: that duties of social guardianship only obtain where they are accepted and cannot be a product of circumstance. One might argue that the first rule of social guardianship of children is simply that one is responsible for a child in need if singled out as the only one who can help, unless and until that responsibility is transferred to someone else. This is certainly, I think, the working moral presumption when it comes to the finding of abandoned or lost children. In post-birth scenarios, being singled out as the only person who can temporarily care for a child only tends to make the responsibility all the more acute. Of course, in most post-birth scenarios that duty is quite quickly transferable. Not so of gestation, though. Pregnancy is special in this regard in that what is required to keep the fetus-person alive is not a transferable duty. No one except the biological ‘mother’ can provide that aid. In this way, the pregnancy scenario resembles the situation where a biological parent is stuck on a desert island with her child, with no option for transferring guardianship. Presumably, Boonin would not agree that post-birth child abandonment is permissible, even in circumstances where the parent (or adult in loco parentis) had no option for transferring social responsibility. It will not be an adequate reply here for Boonin to reiterate Thomson’s claim that the parent’s assumption of responsibility was originally constituted by not aborting the child and taking her home after birth. The claim thrown into question here is simply the one which holds that responsibilities to children must always be voluntarily assumed.
Of course, some discussants are reluctant to relinquish the notion that biological parenthood alone is a meaningful moral relation, and not entirely without reason. Boonin’s suggestion that in the amended violinist scenario it would be absurd for you to feel differently upon discovery that the violinist is in fact your son seems to me to be misplaced. Surely most people would feel differently upon such a discovery, even more so if the violinist were not a grown violinist at all, but a small child. Moreover, this change of feeling is not one that it seems right to view as inappropriate, pathological, or surprising. It might be replied that even though explicable, the feeling of special responsibility towards one’s biological progeny is not a genuine reflection of moral obligation, but only a by-product of the huge cultural significance attaching to biological relations. This cultural significance surely owes in large part to the overwhelming correlation between biological parenthood and social guardianship. Far from showing the pull to care especially for one’s biological relations to be irrational or misplaced, however, explanations such as this—as well as theories which carve out an important place for genetic relation in our self-understanding and personal identity5—might provide a rational underpinning for those attitudes.
Whatever one’s conclusions about the importance of biological relations, though, it is still open for critics of GST to argue that either bodily burdens in general or the typical burdens of pregnancy in particular are simply outwith the scope of any moral obligation one could ever owe to any other, regardless of any special duties of care grounded in either biological relation or negligence responsibility. If this is correct, then arguing successfully for a source of special responsibility will not make a difference to the overall conclusion: that there can be no positive duty on a pregnant woman to gestate a fetus. Reiterating Manninen’s argument, someone might also point out that parental and negligence-based responsibilities still have very clear limits in law, and do not ever extend to bodily burdens.
Whatever the legal situation, however, it is far less clear (I think) that bodily assistance could never form the content of a moral duty to assist another. And if moral duties of bodily assistance could possibly exist, we can again pose the question whether ordinary pregnancy exceeds the moral obligations of physical help one could ever owe to one’s offspring or to those whose needy situation (in the case
of pregnancy, needy existence) one is responsible for creating. Most, I think, would find it morally deficient to refuse one’s child, or the victim of one’s negligence, a life-saving bone marrow or blood donation. Whilst I think it clear that even the most straightforward pregnancies go beyond this level of burdensomeness, they are not so burdensome as being plugged into Thomson’s violinist for nine months. This makes it a live question whether those burdens go beyond the scope of a possible moral obligation to assist a fetus, and not one that can be answered using the analogy of the violinist.
-  Boonins treatment of this hefty sub-issue can be found at pages 148—88, 227—54, and 266—74(Boonin, n 11).
-  See, John Finnis, ‘The Rights and Wrongs ofAbortion: A Reply to Judith Thomson’ (1973) [Wiley]2 Philosophy and Public Affairs 117, 122—3; Christopher Robert Kaczor, The Ethics ofAbortion: Women'sRights, Human Life, and the Question of Justice (Routledge 2011), 150—1; and Joel Feinberg, ‘Abortion’,in Joel Feinberg, Freedom and Fulfillment: Philosophical Essays (Princeton University Press 1994).
-  See, Mary Anne Warren, ‘On the Moral and Legal Status of Abortion’ (1973) 57 The Monist 43(Warren objects to Thomson’s particular argument on the ground that it does not account for an obligation a woman may be under to sustain the fetus on account of her engaging in consensual sex, although,like Thomson, she rejects the ‘personhood proviso’ in any event); Kaczor (n 22) 162—8; Patrick Lee,Abortion and Unborn Human Life (2nd edn, Catholic University of America Press 2010) 108—10 and117—20; and Francis J. Beckwith, ‘Defending Abortion Philosophically: A Review of David Boonins‘A Defense of Abortion’’ (2006) 31 Journal of Medicine and Philosophy 177. Boonin labels the first ofthese arguments the ‘Responsibility Objection’, and discusses it at pages 167—88 (Boonin, n 11).
-  Boonin (n 11) 148-88, 227-34, and 246-54.
-  See especially, Bertha Alvarez Manninen, ‘Rethinking Roe v. Wade: Defending the AbortionRight in the Face of Contemporary Opposition’ (2010) 10 American Journal of Bioethics 33.
-  26 ibid 40. See also Bertha Alvarez Manninen, Pro-Life, Pro-Choice: Shared Values in the AbortionDebate (Vanderbilt University Press 2014) chapter 2.
-  10 Pa D & C 3d 90 (26 July 1978).
-  See S v St George’s NHS Trust  2 WLR 936.
-  Michael Levin, for instance, has argued that ‘all child-support laws make the parental body anindirect resource for the child. If the father is a construction worker, the state will intervene unless someof his calories he extends lifting equipment go to providing food for his children’ (see Michael Levin,‘Review of ‘Life in the Balance’ by Robert Wennberg’ (1986) 3 Constitutional Commentary 500, 511).
-  Thomson (n 8) 59.
-  Boonin (n 11) 118.
-  2 See J David Velleman, ‘Family History’ (2005) 34 Philosophical Papers 357.