I Was Once a Fetus:

An Identity-Based Argument Against Abortion


by Alexander R. Pruss

1. Introduction

           First an outline of the argument   Assume that I once was a fetus.  Who will deny this—surely a fetus was what I once was?  Yet, though it is hard to deny, much of this paper will be work to bolster up this portion of the argument.  For now assume this.  But now if the right-to-life (understood as the right not to be deprived of life unless one has deserved such deprivation through a crime that one has been duly convicted of) is an essential property of me, part of what makes me who I am, then since I was a fetus, that fetus was I, and hence it also possessed the right-to-life, as the right to life is an essential property. Alternately, as is reasonable, being a person is an essential property of me.  I could not be myself without being a person.  Whatever is not a person is also not identical with me.  But I was once a fetus.  Hence, that fetus was a person, because I was that fetus and being a person is an essential property of me.  But it is a conceptual truth that all persons have a right-to-life.  Hence, so does the fetus.

2. I was once a fetus

           A pro-abortion interlocutor having heard these arguments may withdraw assent to the claim that I once was a fetus.  She may turn what I said around and say that that fetus was not I, since I am essentially a person and the fetus (he will claim) is not.  But this is untenable.  For suppose, to obtain a contradiction, that the fetus was not I.  Let "F" (rigidly) designate that being which the fetus was.  Then, either F exists now or it does not exist.  (Obviously, even if it exists now, it is no longer a fetus, just as that being that I was 20 years ago, though it still exists, is no longer a child.)  Suppose F exists now.  Ex hypothesi, F is not I.  So what is it? Well, given that F is an organism (that is evident on scientific grounds), and that all the organic parts of the fetus have developed into organic parts of me, it follows that if F exists now, F is a part of me.  But, ex hypothesi, it is not the whole of me, since I was not F and hence still am not F by transitivity of identity. 

           So which part of me is it?  Every part of my body has organically developed out of a part of that fetus. Therefore, I cannot separate out some proper part of my body now and say: "That part of me, that is F."  Therefore, F now contains all of my body. Moreover, it does not contain anything outside of my body, since it is clear that in F's developmental history nothing developed from organic parts of F that remained a living part and did not remain in me.  So, F is an organism whose body is materially identical with my body;  moreover, the parts of F are identical with my parts even qua organic entities.  But there surely cannot be two organisms that have exactly the same organic body parts. Hence, if I am an organism, I am F.  But it was assumed I am not F, and so I am not an organism.  But this is dualism at its worst.  For surely I am a rational animal, and to be an animal is to be a certain kind of organism. If Descartes were right, then of course one could hold that there is an organism, my body, which is distinct from my soul, and hence either the organism will be a proper part of me (if I am defined as an aggregate of body and soul) or will be something apart from me (if I am just the soul). However, Descartes is wrong.  Surely it is evident that we are animals, albeit rational ones, and that therefore we are organisms.  (Note: A Thomistic or ARistotelian dualist account of the soul will not help the abortion supporter here.  For to be an organism or animal is to have an animal soul, and so if I am not F, there are two souls in me: my soul and my body's soul.  But that is absurd and contrary to the subsumption of the lower parts of the soul.)   So we see that we arrive at the absurd conclusion that we are not animals if we assume that F still exists but I was not F.

           A non-Cartesian way of denying that I am an animal is to claim that I am a person and “person” is a functional concept, so that what I am is a functional entity which the fetus is not identical with.  However, the notion of a “functional entity” is multiply ambiguous.  A functional concept is one defined by certain activities.  The first ambiguity is that between the activities and the entity which engages in them.  If we take the claim that I am a functional entity to mean that I am an entity which engages in certain activities, then this claim is fully compatible with the claim that I am an organism and an animal—namely, I am that organism which performs these activities.  But if so, then I am an organism, and my arguments that I was a fetus still goes through. 

           Instead, then, one would need to take the functionalist claim to be not just that I am an entity that engages in certain activities, but that I am these activities.  On the face of it, this is highly implausible.  And it becomes even less plausible when we disambiguate “activities” into type and token.  If am a certain type of activity, then I am an abstraction.  But I am for myself a paradigm of the concrete.  Besides, types do not stand in causal relations, but I certainly do.  Furthermore, it would be logically possible for another person to engage in the same type of activity as I do, which would be absurd were that person also identical with the type of activity she does.  So, I am a token of activity.  But now we have a modal problem.  For surely throughout my life I could have acted and thought differently from the way I in fact did.  But were I to do that, I would either instantiate a different type of activity, which would entail that since the type of activity I instantiate is on the functionalist view essential to who I am I would then not be myself in contradiction to the necessity of identity, or else I would be another token of that same type of activity if the type is defined inclusively enough.  But were I another token of the same type of activity, I would not be myself, since by assumption I am a token of the same type of activity, and so once again a contradiction would ensue.  So the option that I am a token of activity also fails.  The only way to allow for the possibility that I might have acted and thought differently from the way I did is to allow that I am not the activity but that which does the activity, which I have already shown is fully compatible with my argument that I was a fetus.

           Hence, all roads to denying that I was F on the assumption that F still exists lead to absurdity.  What if instead we say that F no longer exists?  The one wants to know: What happened to it?  When did it cease to exist?  After all, every major organic part of F, except for umbilical cord, is still in organic existence, albeit perhaps transformed and developed.  Moreover, these parts have not radically changed their interrelation so as to dissolve F.  The only way that the loss of the one part that was destroyed, namely the umbilical cord, could have annihilated F is if F was a part of the woman prior to birth. Admittedly, when an organic part becomes cut-off from that of which it was a part, it is thereby destroyed.  The cut-off finger is, according to Aristotle, a finger in name only, and is no longer the same entity as it was before, since its essence was to be a part of the body.

           Therefore, the most promising way of holding F to have ceased to exist is to hold that F was a part of the woman's body.  (We cannot say a part of the "mother's" body, then, properly speaking since one is a mother of a separate entity and not of a part of oneself.)  But there is good scientific reason to deny this.  First of all, F has a different genetic composition from that of all of the other parts of the woman.  But, perhaps, it might be argued that through transplantation one might gain a part that has different genetic make-up from the rest of oneself.  This is not a decisive reply, because transplantation is a non-natural process;  but the fact that something that comes about naturally has a disparate genetic identity from oneself is good reason to believe it is not a part of oneself.  Moreover, if a part of one's body acquires a different telos from that of benefiting the whole body and moreover starts acting for that telos on its own, apart from the direction it receives or has received from the rest of the self, it is reasonable to say it is no longer a part of the body.  The cell that has turned cancerous and whose telos is now simply to reproduce madly independently of the direction of the rest of the body is no longer a part of the body.  On the other hand, the transplanted body part starts to benefit the rest of the body and interacts with it in a way that allows the rest of the body to exercise some control of it.  However, the fetus is not controlled by the woman's body and its telos is its own development, and not the benefit of the woman.  Therefore, we have good reason to think it is not a part of the woman's body.  The umbilical cord cannot be definitive of the fetus' being a part of the woman because the causal interaction through the umbilical cord neither has the woman's body exercising control over the fetus nor does the fetus interact back through the cord for the woman's benefit.

           Moreover, the fetus comes to exist from the union of an ovum from the woman and a sperm from a man.  The informational input of both is equal, and so the formation of the embryo from ovum and sperm is being equally guided by the man's body and the woman's body, despite the fact that this formation occurs within the woman's body.  Hence, the embryo that comes to exist can neither be said to have been produced by the woman nor to have been produced by the man.  It is thus an entity that is organically distinct from them both.  In the first few days of its existence, the embryo freely floats around taking in nourishment.  The fact that its informational guidedness comes equally from the man and from the woman, together with its free floating around, shows that at that time it is a separate entity.  Designate (rigidly) this entity by "E".  Then, E is not an organic part of the woman, but under the hypothesis we are exploring, F is.  If being-an-organic-part-of is an essential property, then when the embryo becomes a part of the woman's body, it is thereby destroyed.  But there is no such moment of destruction.  Certainly, it does not become destroyed by eventually implanting in the endometrial lining and eventually growing an umbilical cord, just as Judith Thompson's sick violinist is not killed by being hooked up with tubes to his host.  Rather, E continuously organically develops into F.

           But perhaps being-an-organic-part-of is not an essential property and so E does not cease to exist when it becomes F, and so in fact E is identical with F.  But if E does not cease to exist upon becoming F, presumably because of the organic continuity of the process, and despite ex hypothesi becoming a part of the woman's body, then likewise F should not cease to exist when it ceases to be a part of the woman's body, and so the fact that F was a part of the woman's body does not contradict its being identical with me.  However, it is highly plausible that being-an-organic-part-of is an essential property.  We individuate organic parts by their organic functions and their continuation in the same organic configuration.  It is thus highly plausible that to be an organic part of something is an essential property, and hence if E is not a part of the woman, neither is F.

           We have thus seen that there is no reply to the claim that I was F based on any claim that F was a part of the woman's body.  Remember that we are now considering the horn of the dilemma according to which F no longer exists.  Once the option that F was essentially a part of the woman's body is removed, there is no reasonable way of holding F to have ceased to exist. On the contrary, its parts have flourished and developed further.  The functions that the body of F has subserved continue to be subserved by a body that has developed continuously from that body.  One cannot point out any event which destroyed F.  Therefore, F continues to exist, and we are back to the first horn of the dilemma which has already been discussed.

           There is another intuitive argument for why I was a fetus or even an embryo on the assumption that I am an organism.  It is evident that there is a continuous development from an embryo, to a fetus proper, to me.  At any fixed time t after fertilization there exists a single organism as a stage in this development process.  We can individuate the organism as a single organism at all times after fertilization.  Denote by Ot that organism which exists at the stage that is at time t.  Then, Onow is I, since I am a part of this continuous development sequence.  If I can show that for all t, Ot is the same entity, then I will have shown that I was a fetus and an embryo.  For a reductio, assume that Ot does not designate the same entity at every time t within the continuous development sequence.  At any time t at which an organism Ot within this sequence can be individuated, that organism is either identical with me or not identical with me.  Evidently, if s<t<now, and Os is identical with me, i.e., with Onow (or with Ofive years ago for that matter), then likewise Ot&shy; is identical with me.  Since we have assumed Ot&shy; is not always the same entity, it follows that there is a time t with the properties: (a) if a<t is any time within the continuous development sequence during which an organism Oa&shy; can be individuated, then Oa is not identical with Onow, and (b) if b>t and b is not after my death, then Ob is identical with Onow.[1]  Now, let b be t plus one femtosecond and let a be t minus one femtosecond.  Since nothing organically significant has happened over the continuous development of the organism in two femtoseconds, it follows that Oa and Ob are the same organism.  Since Ob=Onow, it follows by transitivity of identity that Oa=Onow.  But by our assumptions, Oa was different from Onow, and hence we have an absurdity.  Therefore, Ot designates the same entity at all times t, and since it once was a fetus and an embryo and now is I, I was once a fetus.

           A natural response is that this is a Sorites argument, and as such is subject to the usual response to Sorites arguments’, namely that some crucial property (e.g., is a heap) is vague.  But this will not work here, because identity is not vague.  Surely, any individuated entity either is identical with me or is distinct from me.  To specify an entity is, inter alia, to specify implicitly what the criteria of identity for that identity are and hence what it is identical with.  If we can specify a single entity Ot, then we can meaningfully ask whether that entity is identical with me.  Now, Sorites-type arguments are often applied against abortion.  A natural response that is sometimes made against them is that the process of fertilization itself is a continuous process and hence a sharp line cannot be drawn between what has rights, the child, and what does not, the ovum and the sperm.  However, the present argument avoids any such response.  For I am not claiming here that at any given time t there is a sharp fact of the matter as to whether at t there exists an entity identical with me.  Rather, I am claiming that for any given individuated single entity there is a sharp fact of the matter whether that entity is I.  Continuity in the process of fertilization may leave vague the question of when exactly there cease to be two organisms, the ovum and the sperm, and there comes to be one individual organism, the embryo.  But nonetheless, evidently, at least by the time of the completion of the first cell division it is clear that there is a single individual organism there which we can individuate, and we can ask about that organism whether it is identical with me.  The question of when exactly after sperm and ovum start to overlap there is a single organism can be left vague.  Nor is it of practical importance, since all abortion methods (and logical rigor forces one to include drugs which prevent implantation) act after the first cell division when it is completely unambiguous that there is a single individual organism there.

3. Conclusions

           Therefore, we have seen in more than one way that one cannot maintain that I and the fetus F that has developed into me are not identical.  Hence, I and F are identical, and if my right-to-life is an essential feature of what I am, as seems highly plausible, then F has a right-to-life as well. 

           Peter Singer has argued that the fetus is not harmed by an abortion at an early time because the fetus is not conscious then.  This is completely besides the point.  Certainly, I will be harmed if someone should murder me when I am asleep.  And, likewise, I would have been harmed had someone in the past murdered me when I was unconscious.  This is true regardless of which point in my existence we look at, as long as we actually look at a point in time when I existed.  Killing me at time t0 would deprive me of the life that I would have otherwise led after t0.  If t0 were earlier, the harm is, if anything, greater since I would have been robbed of more.  The harm in the killing is the loss of the future life.  Certainly, one need not be aware of having been harmed to have in fact been harmed.  If a surgeon at night removes much of my brain by stimulates the pleasure center of my brain for the rest of my life, I may never be aware of the harm that has come upon me, the loss of a meaningful intellectual life.  If killing me earlier in my life is, if anything, a greater harm to me, and if the fetus was I, and if no justification in terms of the fetus's capital guilt is possible, then to kill the fetus is, if anything, a greater evil than to kill me now.  If it would be unjustifiable to kill me now in order to, say, save a woman from great distress, it would likewise have been unjustifiable to kill me when I was a fetus for the same purpose.  Indeed, it would be unjustifiable to kill me now in order to even save another person's life, and so likewise killing a fetus for that purpose is also unjustifiable.

           Note that this paper does not attempt to establish conclusively that the fetus is a person, except in its remark that if personhood is an essential property then the fetus that I was was a person.  All that the main arguments need is that the fetus was I, not that the fetus was a person.  Observe that there is a sense in which abortion is often an even greater evil if the fetus were not a person (though in fact, I think being a person is an essential property and hence a fetus is a person).  For, if one kills a person, one has not robbed her of all of the worldly personal life that she would otherwise have had.  But if one kills the fetus, if a fetus is not a person and unless we are talking of a special case where the fetus could not have lived until attaining personhood (a case where I take it abortion would still be wrong by my previous argument, though perhaps to a lesser extent), one has robbed the fetus of all of her worldly personal life, and this is a greater theft, with numerically the same victim.  It is crucial for this argument that there is a subject who is robbed of something that otherwise she herself would have.  If the fetus were not the same being as the person she could develop into, then it would be fallacious to say that the fetus was robbed of anything.  But the fetus is the same being, and hence is robbed of something she would have had otherwise.

           A different way to the same conclusion that because I was a fetus therefore the fetus has, roughly, the right-to-life is the Rawlsian way.  On the best interpretation of the Rawlsian veil of ignorance, one needs to be ignorant precisely of which worldly role one fills if and only if this is a role that one might not have filled.  I need not be ignorant of my being a non-mosquito, because I could not have lacked that property.  But I could have been, say, in the role of a slave, or of a pregnant person (there is no contradiction between having the role of pregnancy and being myself, even if being male is an essential property, because it would be logically possible for a man to fulfill the role of a pregnant woman with some science-fictional surgery).  In the Rawlsian negotiation, I would assign something rather like a right-to-life[2] to those persons who would fill any role that I am ignorant of whether I fill it or not, and these would be those roles which I could both fill and could not fill.  Since I was a fetus, it is metaphysically possible for me to be a fetus, and hence the knowledge that I am not a fetus must fall under the veil of ignorance.  But under such a veil, I would assign the kind of right-to-life to the fetus as I would to the adult, since I would not want to take the risk that I might end up being a fetus.

           All the above arguments end up claiming that my right-to-life is an essential property of me, from which it immediately follows that the fetus that I was had a right-to-life.  We can also support the claim that my right-to-life is an essential property as follows.  For a reductio, suppose that the right-to-life is not an essential property of me.  Then, if ethics is not to be arbitrary or merely a matter of societal fiat, there will be some non-essential, valuable quality in me in virtue of which I have the right-to-life or not.   For instance, it might be held that rationality is not an essential property of me, but it is in virtue of my having rationality that I am have a right-to-life.  However, any non-essential, valuable property that could plausibly be proposed as that property in virtue of having which I have a right-to-life will be a property that normal innocent adult human beings have to differing degrees.  This is clear in the case of rationality: some people are obviously more rational than others.  Likewise, if the property is the capacity for reasoned moral decisions, some have this capacity to a greater degree than others, again because some have a greater share in reason than others. 

           But isn’t personhood such a valuable property that grounds the right-to-life but does not come in degrees?  This, however, the defender of abortion cannot argue.  For if personhood does not come in degrees and I did not have it when I was a fetus, then at point in my life I acquired it in toto, not having had it at all earlier.  But this is absurd in view of the continuity of my development, and so one who seeks to deny that the fetus has personhood must claim that there are degrees of personhood.  However perhaps, even though personhood in general comes in degrees, all normal adult human beings have it to the same degree?  Perhaps, for instance, there is a threshold that each human being who becomes a normal adult crosses, and once the threshold is crossed, one has the fullest degree of personhood and cannot have more.  But this is not a plausible suggestion.  First, all those capacities that admit of degrees at all wax and wane at different times in the life of an adult human being, with exercise, disease, drugs, education, etc.  If personhood is a capacity for certain kinds of activities, then it too waxes and wanes.  This is clear on all accounts of personhood that make personhood non-essential to being a human being.  (And I have already argued that if personhood is essential to a human being, then fetuses have the right to life.)

           Thus, whatever the valuable non-essential property that grounds the right-to-life is, it will have to be something that is exhibited to different degrees by different normal, innocent, adult human beings.  Since the right-to-life is grounded in the value of this property, it is difficult to deny that those who exhibit this property to a greater degree have the right-to-life to a greater degree.  E.g., if rationality is this property, then those who are more rational have a greater right-to-life.  But this conclusion contradicts the basic democratic idea that all of us normal, innocent, adult human beings equally have the right-to-life, and hence is morally absurd.  Therefore, we have reduced the assumption that the right-to-life is had by me in virtue of a non-essential property to absurdity.

           One might challenge the claim that all normal, innocent, adult human beings equally have the right-to-life.  Perhaps, it might be claimed, if one had the choice between saving a famous doctor who is on the verge of discovering a cure for cancer or two less intelligent people from drowning, one should save the doctor.  However, this judgment does not support the claim that the doctor’s life is more valuable than that of the two other people;  rather, what is more valuable than their life is the life of the doctor together with the lives the cure for cancer is expected to save.  Thus it is not the doctor’s right-to-life that outweighs the rights-to-life of the other two drowning people, but the right-to-life of the doctor and of the large number of cancer sufferers.  Other cases where one’s intuitions might grant a greater right-to-life to one innocent, normal, adult human being than to another are, I suggest, either cases of mistaken intuitions or cases where one is confusing the value of life itself with the sum of the value of a life and that of the expected societal contributions of the person living that life.

           Consequently, indeed, we should uphold the claim that all normal, innocent, adult human beings have the right-to-life to an equal degree, and hence by the above arguments cannot hold that this right-to-life is a function of a non-essential property.[3]  But we have already seen that since I was a fetus, that fetus had the right-to-life if the right-to-life is a function of an essential property.  Since I am in no way special, the same must hold for every other fetus.

[1] Just define t as the supremum of all times a at which Oa is not identical with Onow.

[2] Not exactly a right-to-life as defined above.  On Rawlsian principles, it can be argued that it is acceptable to kill an innocent human being (of whatever age) if she would die soon anyway and if one would die were one not to kill her (see my manuscript “De re modality, Rawls and abortion”).  I find this conclusion repulsive and a reductio of Rawlsianism, but this is essentially the only kind of a case under which a Rawlsian “right-to-life” differs from the “right-to-life” as I have defined it.  In the case of abortion, this would correspond to the case where both the mother and the fetus would die should the fetus not be aborted but the mother would live if the fetus were aborted.  I hold that abortion is still wrong in such cases on the non-Rawlsian deontological approach outlined above, but the present argument is being made on the assumption of Rawlsianism.

[3] One might try to make the following counter-argument.  It is a human being’s character’s moral value that grounds her right-to-life.  A human being’s character prior to “the age of reason” does not have moral value, but that of a normal adult may.  The problem with this account is that moral value does differ even between innocent, normal, adult human beings.  Of two innocent persons, one might exhibit heroic virtue while the other might exhibit only run-of-the-mill virtue.  But there is something repugnant about considering the morally superior person as per se having more of a right-to-life than another person who, though innocent, has not scaled the same heights of virtue.  Moreover, consider those people who neither exhibit significant virtue nor exhibit significant vice.  Why would such a person’s life have a greater value than that of a human being prior to the capability of making morally significant choices?  Or consider a person who exhibits slightly more vice than virtue, but not enough vice to render her worthy of capital punishment.  Such a person has a right-to-life, and if she has a right-to-life, despite a negative virtue-to-vice balance, then a fortiori a fetus should have a right-to-life having a zero virtue-to-vice balance.  If it is said that it is the very capacity for making morally significant choices that grounds the value of life, and the fetus lacks this capacity, then the same abhorrent conclusion follows as considered in the body of the paper.  For that capacity is found in different degrees even if we only consider people who neither exhibit significant amounts of vice nor of virtue.  (There is an ad hominem argument for this last point: according to the pro-abortion position currently under consideration, this capacity develops from zero to a full amount as a human being grows up;  but all such capacities are had by different people to different degrees.)