As promised below, since I wouldn’t want Ramesh Ponnuru to come away from the aftermath of his first book with the impression that supporters of abortion rights can’t do any better than to mimic Leon Kass, it’s worth trying to take a look at the sort of arguments he makes and seeing what kind of responses are available. (Apologies to longtime readers, who will probably have seen a substantial number of these arguments here before.) Now, I haven’t read Ramesh’s book yet, but I have debated him on the topic and read plenty of his shorter articles on the same theme, so I’m guessing (and the reviews I’ve seen seem to confirm) that his argument tracks pretty closely the one he develops at the end of this Hoover Digest article. With luck I’ll have time to get around to the book itself one of these days, and see if there are major threads I’ve missed. In the meantime, it’s worth a look at the parts of his usual argument that are familiar from these other sources. Advance warning that there will be some very weird thought experiments involved, which I think are sometimes necessary to isolate a particular issue from other confounding considerations.
One thing that’s worth noting is that for all his eagerness to cast his opponents on the side of sentiment and intuition and give his own position the mantle of a reasoned system of universal rights, Ramesh typically offers an argument that relies absolutely centrally on intuitive reactions in particular kinds of cases to the principles he believes alternative moral views yield. He assumes that what’s most appealing on face is a view that makes moral status depend on some kind of mental features: A capacity to feel pain, or self-consciousness, or some kind of higher rational faculties. (Let’s just say “Property M” to refer to any of a range of mental characteristics one might think make that moral difference, and any moral view that makes some Property M the determinant of moral status an M-theory.) For a lot of those candidate properties, I think you can say something interesting and direct about why we might regard that feature as morally salient. But while he might ultimately do it in the book, I’ve never seen Ramesh offer any direct argument for why we should think his preferred core feature, species membership or genetic structure, is in itself something that it makes sense to believe should give us reasons to treat some organisms as morally special. Rather, the arguments we usually see are indirect, working essentially by process of elimination from the purported repugnant results of the alternative rights-conferring traits. (For example, it is just supposed to be obviously wrong if a view yields the conclusion that very young infants might not have rights.) And that seems like a disadvantage, insofar as you presumably want a view on which your criterion of moral personhood doesn’t just avoid having (any? too many?) intuitively repugnant implications, but actually has something to recommend it independently. This kind of direct account might also give us some kind of guidance if we want to know how to treat other species. If I want to know, for instance, whether it is OK to eat meat, whether there might be something to say against torturing a chimp for fun (beyond the idea that we don’t want to encourage a disposition to take pleasure in the infliction of pain.) A clearer sense of what was so special about species membership, beyond the brute fact of a human/non-human divide, might give us some insight here. Instead, it looks as though species membership is just what’s left when we’ve (supposedly) seen why no more prima facie appealing Property M will do the trick. So let’s review why Ramesh thinks it won’t.
One central part of the case, perhaps the most direct, runs something like this: We take it as a premise that, say, a young child has rights. You can’t kill a toddler just because he’s tiny. But as we run back in time, we find no clear moment of sharp transition that seems like it would correspond to a shift from an entity without rights to an entity with them. A newborn is physically pretty much the same as a fetus one day before delivery. And at any time in the gradual development (both physiological and mental) of the fetus, it seems weird to suggest that just at time T, the fetus has rights, whereas at T-minus-five-seconds, it was just a blob of cells with no special moral status. This same argument sometimes shows up in two other forms: As a sociological slippery slope claim (“If we don’t say all human organisms have rights, we’ll be offing unwanted children someday.”) and as a reductio (“If it’s permissible to kill a fetus, then we can keep stepping forward in five second increments until we’ve inferred it’s permissible to kill a child, which we know to be false, so the premise must be false.”), the latter form of which Ramesh also occasionally invokes. Since conception seems to provide the only such bright-line point, it must be this point at which rights inhere, and biological membership in the human species that’s the key to our moral status.
I think we can dub the faulty rule of inference Ramesh employs here the Sorites Fallacy, in honor of the Sorites Paradox, usually attributed to the Greek philosopher Eubulides, and which I’ve written about elsewhere. Two or three grains of sand are plainly not a heap (runs the argument), and it surely can’t be that a single grain of sand makes the difference between something that is and something that isn’t a heap. By iteration, we conclude that ten thousand grains of hand are not a heap either. Most philosophers regard this little puzzle as an interesting way to start thinking about the logic of certain of our ordinary-language terms and the vagueness of certain properties. Ramesh, presumably, regards it as a sound argument.
It is not, of course, a sound argument, and I think by extension neither is Ramesh’s. What it does point up is the problem of thinking about moral status in strictly binary terms. What I mean is, thinking either that there’s some one instant of development where you go from being morally equivalent to a rock to being morally equivalent to Ramesh Ponnuru, or even that the lines between discrete moral classes are going to be brightly marked. (It might, one supposes, be that there actually is some precise moment at which a brain begins to support an experiencing mind—a binary line between consciousness and its absence—though we can’t be certain exactly where this happens. My old philosophy advisor Peter Unger thought that a case for dualism could be made on the basis of this supposed either-or nature of consciousness. Though I think a position that specified this as Property M would be more plausible than Ramesh’s, I don’t find it satisfactory either—for if it’s just consciousness, rather than consciousness of a certain sort, we get no moral distinction between persons and, say, squirrels. )
What gives the binary way of thinking about it an appeal, of course, is that the question “is it OK to kill this organism?” seems like a pretty either-or sort of thing, if you want to invoke the language of rights, at least. If we abjure rights-talk, the story is a bit different. You might, for instance, have a view where you think certain animals can be harmed only in the interest of producing some great benefit: You might think it’s wrong to do chemical experiments on a chimp to test a new cosmetic, whereas it would be acceptable if the tests were likely to result in a cure for AIDS. But (many of us are inclined to think) that is not how it is with persons, who have a dignity and an inviolability that precludes this kind of instrumental calculus: Humanity, to borrow Kant’s formula, should never be used merely as a means. A consequentialist has an easy go of it here: He can just say the mental level of an organism is one more factor in determining how bad it would be to kill it when weighing that against the benefits. But what if we do want to take a more Kantian position? Aren’t we then committed to there being some kind of “bright line” of the sort Ramesh thinks only conception provides?
I don’t think we are. For it is one thing to say that there is a difference in kind between rational or reflectively self-aware persons (who are owed robust rights) and other merely sentient creatures (who might merely be due some kind of moral consideration), and it is quite another to insist that the threshold actually be precisely specifiable. It had better be, for Ramesh’s sake, since this is not, in fact, the case for biological species, which are themselves a topic of rich philosophical debate. A Sorites style argument can simply be repeated at that boundary—if not as regards actual organisms currently living, then certainly in principle. Being genetically human is, as it were, a statistical question: At the boundaries, there aren’t going to be any brighter lines than with an M Property. If that is supposed to be fatal to a mental-criteria theory of moral personhood in principle, it will be no less damaging to one based on species membership.
But there is no reason to think that it should be: There is no guarantee that the world is set up in such a way as to make our moral decisions easy. (Are we guaranteed these bright lines when we face other moral dilemmas? Most people, presumably Ramesh included, think it’s wrong to kill an innocent person to save three other lives. Is anyone so hardcore as to say we can’t break an innocent person’s arm to save ten-thousand lives? Is it so terrifying to admit there aren’t always going to be bright lines that the appalling extremes here, a crude utilitarianism or a fiat justitia, ruat caelum rights absolutism, are preferable?) Probably there is a moral twilight area where there is no very good way to say whether some organism has reached the level of mental complexity at which we should regard it as having rights. But the worst possible way to deal with this problem is to simply insist there must be such a clean and discoverable line and adopt as your criteria of personhood whatever property holds out the hope of one, like the drunk in the old joke searching for his lost keys only under a streetlight, despite having lost them elsewhere. It is unfortunate if our nature is such that the question “is this a person with rights?” will sometimes be genuinely indeterminate; that does not make it false. Nor do such indeterminacies plunge us into relativism or nihilism. As an old friend of mine used to say, it does not follow from the existence of twilight that there is no such thing as night and day.
A related argument we sometimes get from Ramesh is that since any Property M will vary continuously within the human population, it will be an implication of any M Theory that rights vary from person to person, so that perhaps someone who was as smart as Eugene Volokh and as self-reflective as Marcel Proust would get to treat the rest of us like gerbils. But that just doesn’t follow. (And, in fact, Robert Nozick gave a nice account of why in The London Review of Books when Ian Hacking charged him with holding such a view—though their exchange, alas, is not online.) We can have rights in virtue of a Property M without getting more rights when we exhibit the components of that property in greater degree. (Past some threshold, at least: We do typically countenance paternalism toward persons of very limited mental capacity&mash;children, the insane.) Temperature is what makes the difference between ice and water, but a bucket of H20 at 50 Celsius is obviously not more waterish, or imbued with water-ness to a greater extent, than the same bucket at 15 Celsius.
Obviously, in actually developing an M Theory, you’d have to say something about what the approximate thresholds might be, and why a threshold structure rather than a continuous-increase structure makes sense. But there’s no ex ante reason to suppose you could only have the latter sort of view. Ramesh asserts that it will be “impossible to say” why the former might be right, but doesn’t seem to try very hard to imagine why it might be possible. The conclusive thing to do, of course, would just be to develop one such M Theory that has this structure and see how it fares. This post is too long already, and even I lack the hubris required to imagine I could do an adequate job of that if I had the space, but maybe I can make some quick schematic remarks on that score.
If we want to say that while other animals might be due some kind of defeasible moral consideration, people have these strong rights claims, I think what’s motivating the thought is something like this: Whereas other creatures can experience suffering or pleasure, and maybe all we can say there is that it’s better to have less of the first and more of the second, human ends are more complex, even for the simplest of us. We don’t just have an experiential welfare that can be augmented or diminished, but a set of abstract second order values, many of which have nothing directly to do with our own experience. We have values, what Rawls liked to call a “comprehensive conception of the good.” I’m not talking here about anything so grandiose as an overarching, coherent Philosophy of Life; just the ordinary kinds of diverse goals all people have.
A lot of implications might be drawn from that fact, but here are two to start us off. One is that instead of just acting on our strongest urge of the moment, we represent reasons to ourselves by reference to that system of ends. (A gerbil might dislike pain, or act on an impulse to escape pain; it is a little queer to say it has a reason to do so.) Now, when we ask what we’re obligated to do morally, we’re asking for a certain species of reason for action (or inaction). And it might be that being open to moral reasons involves being open to other people’s reasons—that is, letting a combination of facts and values that constitute a reason for someone else (their reasons to attempt to perpetuate their own lives, for instance) count as a reason for you. Animals might have a welfare that can be translated into a reason from one’s own perspective; people will have values that yield reasons from their own perspectives.
The other implication is that once you’re talking about abstract systems of ends rather than raw feelings of like or dislike, there’s going to be a problem of incommensurability. A monk has a set of goals involving living a quiet life of ascetic contemplation; an artist has a series of creative projects he wishes to complete; a bon vivant wants to enjoy good food, conversation, music, and sex. You might have your own opinion about which set of projects is more worthy, but the point is that when we look at human ends, we’re not looking (as we might be if we were just thinking about pleasure and pain, say) at some kind of more-or-less uniform quality we can gauge with a common metric. Now if you’re looking at a group of gerbils, and it seems like they’re pretty much all about seeking pleasure and avoiding pain, a moral approach where you act in a way that maximizes aggregate pleasure and minimizes pain might work out: Maybe one gerbil has to die quickly so that a larger number don’t suffer at length or something. But that’s not going to work—or anyway, it’s going to be a lot more complicated—for human ends and values. What are you “maximizing”? Not people’s feeling of subjective satisfaction in the achievement of their values, because the whole point is that they (often) care about the value or the end, not just the warm fuzzies they get from having them realized. That incommensurability might be one motivation for thinking about how to express respect for their reasons in terms of boundaries moral boundaries (i.e. rights) within which they can pursue their diverse ends rather than some kind of general good we aggregate and promote.
Take those two together, and you have at least a gesture in the direction, first, of why entities having the kinds of M properties that go with representing reasons to oneself might have a special kind of place in our own moral reasoning, and why they might also fit better into some kind of deontic rights structure than an aggregative one. In both cases, as I’ve conceded, you might run into some fuzziness about just what properties and in what degree combine to push someone over the line into this category. But in both cases, I hope equally clearly, the way the logic works makes it matter whether the others are thinking and representing this way at all, not how well or to what extent they’re doing it. I’m acutely aware of how thin and inadequate what I’ve just said is in terms of actually getting to a satisfactory M Theory. There are all sorts of obvious objections and clarifications and whatnot you’d have to deal with. But it seems far too hasty to just say at the outset that an account with something like this shape is “impossible.”
Note, incidentally, that a view that makes some Property M central to rights, but not necessarily a requisite for any kind of moral consideration might allow us to draw a legal line at birth even if we don’t think the newborn has Property M and the rights that go with it. For even if the newborn lacks rights, it might be that, in virtue of being able to experience discomfort and so on, it shares a category with many other animals, such that it would be wrong to kill it except to realize some other substantial good. (As in the case imagined above where someone will accept medical experimentation on primates to cure AIDS, but objects if the point is only to test some new cosmetic.) Even if Property M and its corollary rights accrue only at some point (hard to pinpoint precisely) some months after birth, say, it might at any rate be due some kind of consideration from the point it develops a nervous system sufficiently advanced to support just raw, unreflective experience. And at that stage, the burden of pregnancy on the host mother might be the decisive factor that makes it reasonable to permit abortion at any stage, but begin recognition of personhood and rights at birth rather than picking some later time, closer to when they’re probably actually arrived at.
Enough about M Theories. How about species membership, then? As Derbyshire’s rather blunt approach brings out, the most obvious problem is that this is just a totally implausible candidate for what’s morally important about people. (Though there is also the problem mentioned earlier, that if the Sorites argument goes through, it would cut against this view too.) This isn’t, as Ramesh would like to have it, some kind of errant sentimental reaction that needs to be regulated by moral abstraction: To put having a certain kind of genetic structure at the core of morality requires tossing out not just a few intuitive responses to a few particular kinds of cases, but the entire basis of our ordinary moral reactions—which, it’s worth noting, predate our knowing that there was any such thing as a “genetic structure.” And it entails conclusion that are at least as repugnant—I think clearly more so—than any of the problematic inferences Ramesh thinks follow from alternative positions.
It entails, for instance, that if we were to learn that some sub-population of what we’d previously regarded as human beings actually had a very different genetic structure, their moral status would be an open question. The same would hold if we discovered some species of intelligent mer-creatures who, despite having DNA much more similar to fish (say) than humans, created complex societies, composed music, and wrote polemical books with inflammatory titles.
Now, presumably, Ramesh wouldn’t actually want to say there was any such question: He might revise his position to claim that it’s not human species membership that confers moral status, but rather membership in or shared genetic structure with any class of creatures that typically have those sorts of features. Even this won’t be quite enough, for presumably he’ll want the moral embrace of species membership to run in only one direction. That is, imagine that we have a creature that is unambiguously a genetic horse, a clear member of genus Equus, which like Mr. Ed has somehow become just as intelligent and self-conscious as any human. Maybe some freak mutation, maybe some mad neurologist has rewired its brain, whatever. It’s a horse, but mentally has all the typical characteristics of a human. (I hope it’s obvious that it won’t do to try and insist that the horse is now ipso facto a member of a different species.) It would clearly be monstrous to think this creature could be killed with impunity—or, if you think animals oughtn’t be killed anyway, generally treated in ways it would be OK to treat a horse.
If Ramesh doesn’t want to say that, he has to allow that M Properties might give grounds for affording an organism more rights than the modal adult member of its species, but maintain that species membership provides some kind of moral floor such that lacking M Properties will never be a reason to treat an organism as having fewer rights (or at least not as lacking certain core rights?) than that modal adult member. Maybe such a view can be made to work, but I hope it’s starting to become clear that there are going to be some serious difficulties making species the fundamental category.
Say Ramesh does make the move I suggest, granting that it’s not just human species membership that confers moral status, but rather membership in any species that typically has certain M Properties. If he does not make this move, he is in what looks like the very ugly position of suggesting that it might be OK to kill someone just like you—with all your desires and goals and thoughts and so on—if your mind were housed in the wrong sort of body. But if he does want to make that move, he’s got to concede what he seems to generally want to deny: That these kinds of mental traits really are at the core of interpersonal morality after all, and species membership is, at most, a way for certain kinds of similar creatures who lack these morally central traits to “hitch a ride” on the moral status of their relatives. That would be an importantly distinct position—albeit a less weird one on face—from the one he seems to take. His arguments are not for why individual creatures lacking the morally central Property M should be able to hitch a ride on the status of similar organisms who lack it; they are arguments for why a Property M is not central at all.
If Ramesh is willing to move to this kind of hitch-a-ride position, we actually have a much narrower debate. We no longer have something that looks like a fundamental dispute about what is ultimately morally valuable, but a scope dispute about the extent of, if you’ll forgive the phrase, the penumbras and emanations from an agreed upon source of value. This would involve different kinds of arguments. Some of them would be empirical arguments about neurological development. Some would be moral arguments about the specific components of Property M. Some of them would be hybrid epistemic/moral arguments about how to deal with uncertainty in face of the inevitable fuzziness surrounding those properties. The problems here would be thorny and difficult, and grappling with them much less fun than declaiming in ringing terms about “parties of death” and the intrinsic value of life. But we would at least be started off on the shared project of approximating a correct view.
Addendum: Ross Douthat gamely e-mails an offering in this vein. While I doubt that the sort of argument he’s proposing will ultimately work, I do think it’s the right kind of argument to be making, so I’ll just offer it without rebuttal for now as an instance of what seems to me like a much more sensible approach to the pro-life case:
I think a pro-lifer might accept that some “Property M” other than species membership is at stake in deciding which organisms qualify for human rights, but counter that rights accrue to organisms that have “Property M” even when they aren’t displaying/experiencing that property, because an organism doesn’t exist in a moment, but over a lifespan, and you can’t judge a creature’s rights-worthiness on a moment-to-moment basis, since creatures pass in and out of different states. So let’s say that “Property M” is consciousness: it wouldn’t be okay to kill me while I’m in a dreamless asleep, or if I went into a coma for a few weeks, because you know that eventually, if left alone, I will begin experiencing “Property M” again. Similarly, an embryo is an organism that will experience “Property M” eventually, so it ought not to be killed (whereas a sperm would have to undergo an ontological change, and become a different organism, in order to experience “Property M”, and so it doesn’t qualify for legal protection).