Following on from Is Abortion Liberal? Part 1 and Is Abortion Liberal? Part 2, discussion in the comments section turned to the issue of sentience. Commenters asked whether perhaps sentience is the property that a newborn possesses and a fetus does not that warrants such unequal application of the non-initiation of force principle by liberals. Is sentience the property that enables the liberal to be consistent in otherwise violating the non-initiation of force principle when it comes to abortion, by permitting feticide, and yet not when it comes to liberal opposition of infanticide?

Some commentators claim that feticide is not homicide until the fetus is sentient. Bonnie Steinbock made perhaps the most sophisticated defence of this position. In an article summarising her position Steinbock states, “my thesis is that killing fetuses is morally different from killing babies because fetuses are not, and babies are, sentient”.[i] She states that she finds the notion that feticide is on a par with infanticide and hence homicide, “completely implausible” because:

A newborn can feel, react, and perceive. It cries when it is hungry or stuck with needles. Very soon after birth it cries from boredom or loneliness as well and can be soothed by being rocked and held. By contrast, the first-semester fetus cannot, think, feel, or perceive anything.[ii]

I propose two responses. Firstly, I will analyse Steinbock’s argument in favour of this thesis, then I will criticise her conclusion in a subsequent post. In both cases, I will argue that Steinbock does not provide a reason for thinking that feticide is different from infanticide. On the contrary, her arguments in favour of feticide are also arguments in favour of infanticide.

Steinbock’s Argument
Steinbock argues that being sentient is essential for possession of interests.

[S]entience is important because nonsentient beings, whether mere things (e.g., cars and rocks and works of art) or living things without nervous systems (e.g., plants), lack interests of their own. Therefore, nonsentient beings are not among those beings whose interests we are required to consider.[iii]

Why is sentience necessary for possession of interests? Steinbock says “[I]t is only sentient beings to whom anything matters... since non-sentient beings cannot be hurt or made to suffer it does not matter to them what is done to them”.[Emphasis original][iv] Elsewhere Steinbock states that what is necessary for a being to have interests in something is that it desires or wants the thing in question.

Steinbock’s argument then rests on these claims. Firstly, that killing an individual is homicide only if continued life is in the interests of that individual. Secondly, an individual cannot have an interest in living unless it cares about living. Thirdly, that sentience is necessary to be able to care about anything.

This argument equivocates on two separate senses of the word ‘interests’. Interests can refer to things a person is interested in, what a person likes or cares about; in this sense, tramping is one of my interests. On the other hand, to talk of someone’s interests can mean to talk about what is best for him or her, what enhances and promotes his or her welfare.

When we examine the second premise of this argument it is clear that it is true only if interests are defined in the first of these senses. In order for something to be in someone’s interests in the first sense, then it must matter to him or her. He or she must care about the thing in question. It is not necessary for a person to have interests in the second sense. A thing can be in someone’s best interests even if he or she is not interested in it. Matters in the best interests of children such as nutritious food, clothing, shelter and education are often things they have never taken an interest in. People who suffer from mental illness can fail to take care of their interests correctly. Consequently, the second premise of the argument is true only if the word ‘interest’ is used in the first sense.

However, if the first premise is to have any plausibility at all then interests must be understood in the second sense. If one has a duty to refrain from taking something from an individual only if the individual cares about the thing in question, then depriving the insane or children of what is in their best interests will not be wrong as long as they are not interested in it. Infants lack the mental ability to conceive, and hence, care about their future existence so infanticide would not be homicide by this line of reasoning. The argument is unsound.

This problem with Steinbock’s position leads naturally to a second one. From the quotes above Steinbock seems to think that sentience is not only a necessary but also a sufficient condition. She concedes that it would be wrong to kill unwanted infants but adds that this is different from killing a fetus “because fetuses are not, and babies are, sentient”.[v]

The problem is that her above argument does not justify this conclusion at all. What the argument says is that sentience is necessary for possessing interests. Moreover, it explicitly states that in order to have an interest in continued existence, an individual’s existence must matter to it; it must care about it in some way. While sentience is clearly necessary for caring about one’s continued existence various other factors are also necessary.

Michael Tooley has argued one must also be able to conceptualise one’s continued existence and this requires awareness of one’s self as a subject of future experiences. An infant does not acquire the neurological capacity for this until some time after birth. In fact, an infant’s awareness of itself and future-orientated preferences to live are less developed than those of a mature cow. Consequently, if the premise of the argument offered is correct then infants do not have an interest in continued existence and consequently infanticide is not homicide.[vi]

Steinbock offers a second argument in favour of sentience as a significant threshold.

“[G]olden rule”- type reasons do not apply to nonsentient beings. That is, no one would explain opposition to burning the flag of the United States of America by saying, “How would you like it if you were a flag and someone burned you?[vii]

Here, Steinbock appeals to a version of the Golden Rule. This rule prescribes consistency in one’s ethical judgments so that if one holds that it is permissible for you to do a particular action X to another person in certain circumstances then one must also hold that it is permissible for someone to do X to you in the same circumstances. Steinbock’s point is that such consistency is impossible with non-sentient beings. It does not make sense for one to object to destroying a flag because if one were a flag I one would not object to being burned. Flags after all cannot object to anything.

Again, Steinbock’s argument relies on concealed ambiguities. Harry Gensler has distinguished two different interpretations of the Golden Rule. In one, the rule asks for our reaction to a hypothetical case involving ourselves. On the other it asks how we would react if faced with the actual case itself.[viii]

Two examples demonstrate the difference between these two interpretations; a temporarily-unconscious individual and a case of infanticide. Taking the unconscious individual first, the first interpretation of the Golden Rule asks, would I object to the idea of someone killing me if I ever fell into a temporary coma? The second interpretation asks, would I object to being killed if I were, in fact, already in a coma? The answer to the first question is yes, I do object to the idea of someone killing me in the event I fell temporarily unconscious. I do not think it permissible for a surgeon to kill me while I was under general anaesthetic, for example. However, the answer to the second question must be no. If I were already unconscious, I could not object to being killed. While unconscious, I would not be capable of objecting to anything at all.

The infanticide example demonstrates these differences further. Applying the first interpretation I would ask, would I object to the idea of being killed when I was an infant? The second interpretation asks, would I object to being killed if I was an infant? Again, I would answer yes to the first question but no to the second. I do object to the idea of being killed when I was an infant. However, I clearly could not object if I were an infant because infants lack the cognitive development to be able to object or even understand what it means to object to being killed.

Returning now to Steinbock, it is evident from her discussion that she is interpreting the Golden-Rule in the second way. Her question, “how would you like it if you were a flag and someone burned you?” asks us how we would react if we were a flag. Such a question is absurd because as a flag we could not have any cognitive reactions at all. However, interpreted this way it is equally absurd to apply the Golden Rule to infants or the temporarily unconscious. Consequently, if feticide is not homicide because the Golden Rule questions interpreted in this manner do not apply to fetuses, then killing infants and killing the temporarily unconscious are also not homicide.

On the other hand, if the golden rule is interpreted the first way a different result emerges. It remains absurd to ask whether I would object to the idea of being killed in the event that I became a flag. It is dubious that I could ever be a flag, that a human adult and a flag could ever be the same individual. However, it does make sense to ask whether I object to the idea of being killed when I was an infant or when I was a fetus. This is because rational agents like myself once were fetuses and infants. The fact that a being is a potential rational agent makes it possible for rational agents to ask Golden Rule questions about them.

Steinbock’s arguments for the sentience threshold are unsound. Further, when conjoined with certain facts of neurology they entail that infanticide is not homicide. The claim that feticide is not homicide while infanticide is appears ad hoc and arbitrary.

I will address Steinbock’s conclusion in my next post, Sentience Part 2.

[i] Bonnie Steinbock, “Why Most Abortions are Not Wrong,” in Advances in Bioethics: Bioethics for Medical Education, Vol. 5, ed. Rem B. Edwards & E. Edwards Bittar (Stanford, CT: JAI Press, 1999), 248.
[ii] Ibid.
[iii] Ibid.
[iv] Ibid.
[v] Ibid.
[vi] Michael Tooley, Abortion and Infanticide, (New York: Oxford University Press, 1983) 47.
[vii] Steinbock “Why Most Abortions are Not Wrong,” 248-49.
[viii] Harry Gensler, “Abortion and the Golden Rule,” in The Abortion Controversy 25 Years after Roe v Wade: A Reader, ed. Francis Beckwith & Louis Pojman (Belmont, CA: Wadsworth, 1998), 323.