The mistake I made was claiming that Don Marquis "moved the goal posts" in his justification for why the future-of-value criterion (FOVC) does not imply the immorality of birth control. That part wasn't wrong; he does move the goal posts, just not where I said he did. What I said was that his justification required that a future-of-value be bound to a particular thing, and that this was not part of his original criterion. That was wrong. It was part of his original criterion, as commenter Publius kindly pointed out.
Here is Marquis original presentation. I've added a highlight to the part that I missed (or at least forgot about):
What primarily makes killing wrong is neither its effect on the murderer nor its effect on the victim’s friends and relatives, but its effect on the victim. The loss of one’s life is one of the greatest losses one can suffer. The loss of one’s life deprives one of all the experiences, activities, projects, and enjoyments which would otherwise have constituted one’s future. Therefore, killing someone is wrong, primarily because the killing inflicts (one of) the greatest possible losses on the victim.He continues, but notice the subtle shift from the third to the first person:
To describe this as the loss of life can be misleading, however. The change in my biological state does not by itself make killing me wrong. The effect of the loss of my biological life is the loss to me of all those activities, projects, experiences, and enjoyments which would otherwise have constituted my future personal life. These activities, projects, experiences, and enjoyments are either valuable for their own sakes or are means to something else that is valuable for its own sake. Some parts of my future are not valued by me now, but will come to be valued by me as I grow older and as my values and capacities change. When I am killed, I am deprived both of what I now value which would have been part of my future personal life, but also what I would come to value. Therefore, when I die, I am deprived of all of the value of my future. Inflicting this loss on me is ultimately what makes killing me wrong.He does this because he wants one of the consequences of his theory to be that killing hermits is morally wrong. The only way to do that is to measure the future value of a human life by the quality metric of the person living it. We don't want to have to find someone else to vouch for us in order to establish our own value.
Marquis continues by concluding:
This being the case, it would seem that what makes killing any adult human being prima facie seriously wrong is the loss of his other future.This of course does not mean that the FOVC only applies to adult human beings. The form of the argument is, "what makes the killing of adult human beings wrong is FOVC, therefore FOVC is a valid criterion by which to judge the wrongness of killing, and hence it is wrong to kill anything that values its own future."
Marquis then goes on to list four redeeming qualities of FOVC which I listed in the original post. The fourth of these is:
In the fourth place, the account of the wrongness of killing defended in this essay does straightforwardly entail that it is prima facie seriously wrong to kill children and infants, for we do presume that they have futures of value.Note the highlighted words. This are where he actually moves the goal posts. It's a subtle but crucial shift, and I think that may be why I missed it the first time around: "WE presume that THEY have futures of value." Indeed fetuses do have futures of value relative to other people's quality metrics. But Marquis has explicitly disclaimed this mode of reasoning! It is not the effect of killing on friends, family, or concerned bystanders that makes killing wrong, it's the negative impact on the victim as assessed by the victim. This is not an accident; it's the only way to save the hermits. It's also the only way to not arrive at the conclusion that euthanasia is wrong.
The problem for Marquis is that fetuses do not and cannot possibly value their own lives. To value anything you have to have a brain, and fetuses don't. And it's even worse than that: the essential ingredient for valuing things is not a brain but a mind. (This is why it's generally considered OK to kill brain-dead people and harvest their organs: they have brains, but not minds.) Newborn babies have brains, but whether nor not they have minds is debatable. In particular it's debatable whether a newborn human has more of a mind than, say, an adult chicken. In fact, if you present the question to a chicken in a form that it can understand (e.g. standing over it with butcher's knife in your hand) I'll wager it will give you some pretty definitive indications that it does indeed value its own future.
So not only does FOVC fail to save fetuses, it even fails to save newborns, at least as long as we find it acceptable to kill chickens for food. Oh well, at least the hermits can breathe a sigh of relief. (And maybe the chickens if people really start to take Marquis seriously.)
I suppose the reason I missed this is that I was trying to give Marquis the benefit of the doubt, because the theory as he actually presents it is just hopeless. The only way I can see to salvage it is to accept the moving of the goal posts, accept the premise that babies have futures-of-value because we adult humans say they do, and reason from there, at which point you run into the problem I described in the previous post, namely, that it's hard to decide where to stop the extrapolation backwards in time. If you're going to impute value all the way back to the zygote, why stop there?
Enter Peter Donis with an unusually innovative (by the standards of the abortion debate) proposal to draw the bright line at implantation rather than fertilization. Note that it is not even worth considering this unless we have already abandoned Marquis's FOVC-AABTV (As Assessed By The Victim). We have to accept, either as an axiom or as a consequence of some other criterion, that the value of a newborn infant has already crossed the threshold beyond which it is morally wrong to kill it. Then -- but only then -- we can ask: where was this threshold crossed?
The overwhelmingly most popular answer to this question (by those who accept its premises) is: at conception. But this has problems with regards to the moral status of frozen embryos, the destruction of which most people do not regard as a moral transgression on a par with murder. Peter's suggestion of drawing the line at implantation rather than conception is designed to solve that problem, along with several others that depend on events that are common before implantation but rare afterwards.
But this is only a temporary solution. The problem of the moral status of frozen embryos only exists because we actually have the technology to freeze embryos. Implantation is only a bright line because we don't yet have the technology to incubate an embryo outside of a womb. But that constraint is probably only temporary, and it would be nice to have a moral framework that was AW-ready (Artificial Womb) as well as IA- and AI-ready.
The straightforward extrapolation of Peter's implantation criterion to artificial wombs is that an embryo crosses the moral threshold when it is taken out of the freezer and implanted into an artificial womb. So let's do a thought experiment: a couple decides to have a kid, takes an embryo out of the freezer and puts it (literally!) in the oven. Let's suppose that this is early days and the technology has not yet advanced to the point where you can order a Mr-Womb machine from Amazon. You have to pay a company to rent and operate their machine.
Three months in to the process, both parents lose their jobs and are no longer able to pay the bills. What should happen?
Or suppose that the technology has advanced to the point where you can buy a Mr. Womb for $199 and conduct this entire process in the comfort and privacy of your own home. Now one day the couple's six-year-old daughter decides she wants a little sister, takes an embryo out of the freezer, pops it in and pushes the button. Some hours later, the parents wake up and are horrified to discover what little Suzie has done. They can't afford another child. If they pull the plug at this point, have they committed murder?
It's also interesting to construct similar thought experiments based on hypothetical "gestational" processes for AIs. I started writing one of those up and it turned into a very long passage (I think it could actually make a good premise for a science-fiction novel!) so I'm going to set that aside for now.