monstrosities, every member of our species is indubitably a person, a human being, at the very latest at birth.”10 Is it really indubitable that newborn babies are persons? Surely this is a wild contention. Wertheimer is falling prey to the confusion naturally engendered by the practice of using “person” and “human being” interchangeably. Another example of this is provided by Thomson: “I am inclined to think also that we shall probably have to agree that the fetus has already become a human person well before birth. Indeed, it comes as a surprise when one first learns how early in its life it begins to acquire human characteristics. By the tenth week, for example, it already has a face, arms and legs, fingers and toes; it has internal organs, and brain activity is detectable.”11 But what do such physiological characteristics have to do with the question of whether the organism is a person? Thomson, partly, I think, because of the unfortunate use of terminology, does not even raise this question. As a result she virtually takes it for granted that there are some cases in which abortion is “positively indecent.”12
There is a second reason why using “person” and “human being” interchangeably is unhappy philosophically. If one says that the dispute between pro‐ and anti‐abortionists centers on whether the fetus is a human, it is natural to conclude that it is essentially a disagreement about certain facts, a disagreement about what properties a fetus possesses. Thus Wertheimer says that “if one insists on using the raggy fact–value distinction, then one ought to say that the dispute is over a matter of fact in the sense in which it is a fact that the Negro slaves were human beings.”13 I shall argue that the two cases are not parallel, and that in the case of abortion what is primarily at stake is what moral principles one should accept. If one says that the central issue between conservatives and liberals in the abortion question is whether the fetus is a person, it is clear that the dispute may be either about what properties a thing must have in order to be a person, in order to have a right to life – a moral question – or about whether a fetus at a given stage of development as a matter of fact possesses the properties in question. The temptation to suppose that the disagreement must be a factual one is removed.
It should now be clear why the common practice of using expressions such as “person” and “human being” interchangeably in discussions of abortion is unfortunate. It would perhaps be best to avoid the term “human” altogether, employing instead some expression that is more naturally interpreted as referring to a certain type of biological organism characterized in physiological terms, such as “member of the species Homo sapiens.” My own approach will be to use the term “human” only in contexts where it is not philosophically dangerous.
III The Basic Issue: When is a Member of the Species Homo sapiens a Person?
Settling the issue of the morality of abortion and infanticide will involve answering the following questions: What properties must something have to be a person, i.e., to have a serious right to life? At what point in the development of a member of the species Homo sapiens does the organism possess the properties that make it a person? The first question raises a moral issue. To answer it is to decide what basic14 moral principles involving the ascription of a right to life one ought to accept. The second question raises a purely factual issue, since the properties in question are properties of a purely descriptive sort.
Some writers seem quite pessimistic about the possibility of resolving the question of the morality of abortion. Indeed, some have gone so far as to suggest that the question of whether the fetus is a person is in principle unanswerable: “we seem to be stuck with the indeterminateness of the fetus’ humanity.”15 An understanding of some of the sources of this pessimism will, I think, help us to tackle the problem. Let us begin by considering the similarity a number of people have noted between the issue of abortion and the issue of Negro slavery. The question here is why it should be more difficult to decide whether abortion and infanticide are acceptable than it was to decide whether slavery was acceptable. The answer seems to be that in the case of slavery there are moral principles of a quite uncontroversial sort that settle the issue. Thus most people would agree to some such principle as the following: No organism that has experiences, that is capable of thought and of using language, and that has harmed no one, should be made a slave. In the case of abortion, on the other hand, conditions that are generally agreed to be sufficient grounds for ascribing a right to life to something do not suffice to settle the issue. It is easy to specify other, purportedly sufficient conditions that will settle the issue, but no one has been successful in putting forward considerations that will convince others to accept those additional moral principles.
I do not share the general pessimism about the possibility of resolving the issue of abortion and infanticide because I believe it is possible to point to a very plausible moral principle dealing with the question of necessary conditions for something’s having a right to life, where the conditions in question will provide an answer to the question of the permissibility of abortion and infanticide.
There is a second cause of pessimism that should be noted before proceeding. It is tied up with the fact that the development of an organism is one of gradual and continuous change. Given this continuity, how is one to draw a line at one point and declare it permissible to destroy a member of Homo sapiens up to, but not beyond, that point? Won’t there be an arbitrariness about any point that is chosen? I will return to this worry shortly. It does not present a serious difficulty once the basic moral principles relevant to the ascription of a right to life to an individual are established.
Let us turn now to the first and most fundamental question: What properties must something have in order to be a person, i.e., to have a serious right to life? The claim I wish to defend is this: An organism possesses a serious right to life only if it possesses the concept of a self as a continuing subject of experiences and other mental states, and believes that it is itself such a continuing entity.
My basic argument in support of this claim, which I will call the self‐consciousness requirement, will be clearest, I think, if I first offer a simplified version of the argument, and then consider a modification that seems desirable. The simplified version of my argument is this. To ascribe a right to an individual is to assert something about the prima facie obligations of other individuals to act, or to refrain from acting, in certain ways. However, the obligations in question are conditional ones, being dependent upon the existence of certain desires of the individual to whom the right is ascribed. Thus if an individual asks one to destroy something to which he has a right, one does not violate his right to that thing if one proceeds to destroy it. This suggests the following analysis: “A has a right to X” is roughly synonymous with “If A desires X, then others are under a prima facie obligation to refrain from actions that would deprive him of it.”16
Although this analysis is initially plausible, there are reasons for thinking it not entirely correct. I will consider these later. Even here, however, some expansion is necessary, since there are features of the concept of a right that are important in the present context, and that ought to be dealt with more explicitly. In particular, it seems to be a conceptual truth that things that lack consciousness, such as ordinary machines, cannot have rights. Does this conceptual truth follow from the above analysis of the concept of a right? The answer depends on how the term “desire” is interpreted. If one adopts a completely behavioristic interpretation of “desire,” so that a machine that searches for an electrical outlet in order to get its batteries recharged is described as having a desire to be recharged, then it will not follow from this analysis that objects that lack consciousness cannot have rights. On the other hand, if “desire” is interpreted in such a way that desires are states necessarily standing