My case

I will affirm my argument with the following syllogism:

1. P1: An unborn child is a person from the point of conception
2. P2: If an unborn child is a person, abortion is morally impermissible in the majority of cases
3. C1: Therefore, abortion is morally impermissible

P1: An unborn child is a person from the point of conception

Determining the best Standard for Moral Consideration

Per Solomonoff's theory of inductive inference, we should generally assume that the most likely model is the simplest one that accounts for our empirical observations. In the case of morality, the closest thing we have to empirical observations are foundational moral assumptions, or strong, widely-shared moral intuitions. We should accept the simplest definition or standard that accounts for our strong, widely-shared moral intuitions.

“Ability to be Harmed” as a Standard for Moral Consideration

When an entity is considered to be a person, there are certain rights that seem straightforward to grant that entity. For example, if a person has an internal biological process that would grant them utility down the road, disrupting that process (such as by poisoning them and stunting their growth) is a violation of their rights. Furthermore, if a dependent, such as a child, is suffering from a deficiency that is harming/disrupting an internal biological process (such as from lack of oxygen, a hormone deficiency, or starvation, which can cause direct adverse effects on the body), then their parents have a responsibility to mitigate this harm or seek treatment for it, given they have the capability to do so. Note that a child has the right to the continued internal process even if the process must be aided by nutrients or some amount of external support. However, it is not an automatic moral obligation to help someone generate a new internal process, even if they strongly desire it. I have no obligation to help someone create an immortality potion, for example, or to perform an operation that would enable someone to fly, even if those things were somehow feasible. These differences are essentially between a process that is in progress and must be aided vs. a process that has not begun at all.

Since these moral intuitions seem pretty foundational, we should be seeking the simplest set of standards for moral consideration that incorporate them and other foundational intuitions. The model I propose is that we maintain the criterion of “internal biological process that would grant an entity utility down the road” as the only criterion for moral consideration. Essentially, if an entity possesses said internal process at all, they are a person and warrant moral consideration in proportion to the utility they stand to lose. This model is inherently simpler than other models for personhood, because in order to maintain the foundational moral intuitions described above, these other standards would have this foundational standard for moral consideration, and then an additional standard or standards for an entity to clear the bar of “personhood.” But if this criterion is sufficient to establish personhood, we have a simpler, more elegant model. To prove this model infeasible, an opponent would need to show that granting moral consideration from the point of conception violates some additional, almost universally shared moral intuition.

Abortion vs. Contraception

I bring this up to distinguish between abortion and contraception. An unborn child will keep growing and develop the capacity for consciousness unless directly harmed (if their bodily functions are impeded through starvation or losing oxygen or hormonal deficiency). The zygote is already growing and will become a newborn unless this process of growth is impeded, much like a newborn is in the process of growing into later stages of development. Hence, an unborn child meets this standard for personhood, while a gamete does not.

But sperm will not develop into a person unless combined with an egg (if bodily functions are added). Growth into a later stage of human development requires it to become an entirely new organism with entirely new bodily functions. Conception is categorically different from pregnancy in that an entire gamete is being added to merge with, rather than nutrients which prevent nutritional deficiency, or direct harm, in an existing organism. Again, this is not a comprehensive description of pregnancy, but this should establish that abortion causes direct harm that contraception does not cause.

Further Support from Mathematical Limit

The burden of a disease is often measured in quality-adjusted life years. In simplified terms, this means that generally doctors are trying to maximize years of life saved rather than the number of lives saved. If we are distributing scarce resources and a treatment would save Patient A twenty years of additional life, all things being equal, then their access to the treatment is more pressing than for Patient B, for whom it would only save one day of additional life.

This means that if we select a life expectancy of, say, 70 years, the amount of harm from the death of a patient would be relevant to their age and how long they have left to live. The death of a 50 year old expected to live to 70 otherwise would represent twice as much harm as the death of a 60 year old expected to live to 70 otherwise. (Per rational decision theory, a half chance of losing 20 years of life is equivalent to losing 10 years of life.) We might skew things a bit to account for quality of life, but in general terms, the point stands.

For entities with a similar life expectancy, all things being equal, this trend is basically linear, with moral consideration , increasing as past sentience decreases and future expected lifespan increases. A one-year old with 69 years remaining would have the most to lose and warrant more moral consideration than someone with one day left to live. As the age of the entity approaches zero, the expected life years approach 70, representing a significant, high amount of moral consideration.

Presuming we are seeking the simplest standard for personhood that matches these observations, a model that grants an unborn child significant moral consideration is simpler and more elegant than one granting it no moral consideration. Granting the unborn entity no moral consideration would seem to imply that the linear trend breaks and somehow drops to zero when past sentience hits zero, but this seems rather arbitrary. As age decreases, past sentience has been continually decreasing, yet moral consideration has been going up. If we grant a five year old more moral consideration than someone expected to die tomorrow, but no moral consideration to a fetus, then this is a bit like me having $1 and saying “I’m better than the guy with $0 because I have $1. But I’m just as good as the guy with $10,000, because after one cent, money doesn’t matter.” Once we’ve attributed moral consideration to a continuous variable, it doesn’t make sense to set an arbitrary cutoff, particularly one that goes in the reverse direction of the trend. It only makes sense to have a hard cutoff with a non-continuous variable, such as “ability to be harmed” as outlined earlier (i.e. a zygote has an internal bodily process leading to sentience while a gamete has no such internal process). Even if we were granting equal moral consideration to the five year old and someone expected to die the next day, a hard cutoff based on lack of past sentience would still be arbitrary, as it has been continually decreasing with age yet not causing a reduction in moral consideration.

P2: If an unborn child is a person, abortion is morally impermissible in the majority of cases

Direct harms vs. Indirect harms

It is sometimes argued that abortion is simply “letting someone die” in early stages of pregnancy when the unborn child depends on the mother. However, refusal of bodily support can still equate to direct killing in some circumstances. To understand why, consider the following scenario:

A woman carries a child deep into a forest on a camping trip. For whatever reason, she leaves the child there, where it is eaten by wild beasts. Clearly, this is immoral.

I argue that abortion is morally comparable to this. Despite the fact that carrying the child into the forest puts the child in a state of physical dependency, carrying the child into the forest alone is not a direct harm. Parents bring their children to all sorts of places, but harm does not occur until the child is abandoned. This isn’t just “failing to save” or even “failing to save after accidentally causing harm,” but directly harming via abandonment. Similarly, when a woman aborts a child she conceived, removing bodily support is a direct harm to the child.

Modus Ponens

1. P1: Removing bodily support after making someone dependent on oneself is a direct harm.
2. P2: Abortion is removing bodily support after making someone dependent on oneself.
3. C1: Abortion is a direct harm

The forest analogy establishes P1, and P2 is true under usual circumstances, as the vast majority of abortions arise from consensual sex. P1 follows even if the pregnancy is not the intended consequence of sex—the woman from our earlier analogy might have gotten lost in the forest by accident, yet bringing the child there is still a result of her actions. Hence, abandoning the child would still be a direct killing.

Insufficient Justifications

Hence, the question that matters is whether a justification exists in the case of abortion for directly killing one’s child. For the ends to justify the means, the good achieved from an action would have to at least outweigh the bad. But being killed as a child prevents one from experiencing their entire life, while pregnancy burdens someone for only nine months. Therefore, the former is a greater harm, and the ends cannot justify the means—it would be akin to killing ten people to save a single person.