Human Nature

The Machine of a New Soul

The messy biology of human embryos.

Linda Hamilton, former half-embryo

Are embryos morally equal to people? I say no. Robert George, a member of President Bush’s bioethics council, and his colleague Christopher Tollefsen say yes. In their new book, Embryo: A Defense of Human Life, George and Tollefsen conclude not only that embryo-destructive stem-cell research should be defunded but that any research involving embryos should be banned if it even slightly risks an embryo’s health. They propose to halt the common practice of producing extra embryos during in vitro fertilization and to require that every IVF embryo be transferred to a womb.

In Sunday’s New York Times, I reviewed the book’s arguments. A day later, the authors replied on National Review Online. This is a conversation worth pursuing. George and Tollefsen are pushing the discussion into an area—embryology—where, in contrast to the usual shrieking about abortion, real progress can be made. They’re civil, logical, and smart. I’ve seen George pick apart fuzzy-thinking adversaries at meetings of the bioethics council. It’s like watching a cat with mice. Today, unfortunately, I’ll be the mouse.

The virtue of Embryo is that the authors stake their case on science and logic, not religion. What makes you a human being, they argue, isn’t a soul, but “a developmental program (including both its DNA and epigenetic factors) oriented toward developing a brain and central nervous system.” They believe that this program starts at conception and therefore, so does personhood.

I like this bet on science. It’s scrupulous, brave, and constructive. Let’s toss in our chips and call the bet. We’ll have to accept what science shows: Conception is, as George and Tollefsen argue, the sharpest line we could draw to mark the onset of moral worth. But they, in turn, will have to accept the other side of what science shows: The lines of embryology are dotted, not solid. Such lines don’t warrant severe categorical restrictions on stem-cell research or assisted reproduction.

Start with the line between embryo and mother. They send signals back and forth to facilitate the embryo’s migration, implantation, and nutrition. The embryo carries the mother’s RNA, which directs its growth. What’s more, the embryo is already on her way toward motherhood, with primordial germ cells up and running in her second week of development. The same program that created her is creating her children. It will kill her, and later them, revising not just individuals but families and species. You can’t isolate life’s program in one body, any more than you can isolate the Internet on one computer. Indeed, life is far more fluid than the Internet, with shared software that remakes its hardware as well as itself.

George and Tollefsen assume a clear distinction between wholes and parts. Eggs and sperm are parts, they reason, while an embryo is a whole. At conception, the parts become a whole, the program launches, and personhood begins. But it isn’t that simple. Some embryos divide after conception to become two or more people. Are those embryos, prior to twinning, an individual? Furthermore, all of us came from embryos that were part “embryoblast” (the segment that became a person) and part “trophoblast” (the segment that became placenta). The placental lineage grew you to birth, separated, and died. In computer terms, it’s like a Zip file. In human terms, it’s a bit like a mother. In these ways, the early embryo is simultaneously a whole, a part (of the mother-child system), and a dyad (of potential twins or of embryo and placenta).

The egg-embryo distinction, too, is permeable. George and Tollefsen write that eggs must combine with sperm or die. They say an organism “was never itself a sperm cell or an ovum.” But look what just happened at a zoo in Kansas: another case of parthenogenesis—eggs becoming offspring without fertilization. This process has produced adults in dozens of vertebrate species, including sharks and turkeys.

And those are just nature’s tricks. With biotechnology, we’re adding our own. Through IVF, we’ve separated, for the first time, internal and external elements of the embryonic program. Through cloning, we’ve turned adult cells into embryos. Through viral injections, we’ve turned adult cells into embryonic stem cells. Through aggregation, we’ve made embryonic mouse stem cells grow into mice. By tweaking a single gene, we’re learning to alter embryogenesis so that what would otherwise become an embryo becomes instead a disorganized bunch of stem cells.

In their rebuttal, George and Tollefsen try to sharpen these blurred lines. “The embryo is not a maternal body part,” they observe. That’s true, but it misses the point. The problem isn’t that they put the embryo in the wrong category. It’s that the embryo defies such neat categories. It’s a part, a whole, and a dyad.

Maternal factors don’t alter the embryo’s genetic humanity, the authors write; they “merely enable it to continue to grow and develop.” True again. But the embryo’s dependence on these factors for its very life makes them more central to the embryonic program, not less. Indeed, this is the logic behind viability as a standard of abortion jurisprudence: The less the unborn human relies on its mother, the more it encompasses its own developmental program, and the more we should treat it like a born child.

As for that pesky maternal RNA, George and Tollefsen quote a scientist who says, “Once an embryo has come into existence, the maternal RNA, like the embryo’s genome, belong to the embryo itself.” It’s a handy quote, but it’s not scientific. The RNA is in the egg, and, as the authors note, the egg is part of the mother. To claim that the RNA suddenly belongs to the embryo, and not the mother, simply because the “embryo has come into existence” is to assume what was to be proved: that the embryo is fully distinct.

Same with the placenta. “The embryo generates through its own activities that organ,” the authors write. “Its reliance on the placenta is bound by narrow temporal conditions and the placenta will (like baby teeth) be discarded in time.” But the placenta is nothing like baby teeth. Research in mice indicates that the embryoblast-trophoblast distinction may begin at the two-cell stage. *  If this theory is confirmed in humans, then fully half of the two-celled embryo that became you wasn’t exactly you. It was your support system, and the zygote that spun it off was your software installation package. To dismiss that half as an embryonic “organ” again assumes what was to be proved: the embryo’s continuous identity. The initial, total embryo that “generates” the placenta and the later, specialized embryo that “relies” on it aren’t quite the same thing.

On the other hand, when continuous identity confounds the authors’ distinctions, they reject it. Take parthenogenesis. “In all such cases—just as with the typical generation of a human being—there is a critical change from an entity that acts only as a part of a larger biological whole to an entity that acts in the way characteristic of an embryo,” they write. True again. But in many of these cases, the initial entity that acts like a part and the later entity that acts like an embryo are, genetically, the same thing. No scientist swaps out its nucleus or zaps it with electricity. It simply crosses the egg-embryo barrier.

Continuous identity also confounds George and Tollefsen’s position on twinning. In an ordinary pregnancy—specifically, mine—they argue that “it would be a howler of a scientific mistake to say that once upon a time there was an embryo that was something distinct from the living human organism that is now Will Saletan, but that got transformed from whatever it was into the organism that is Will Saletan at some point after the embryo came into existence.”

But in the case of twinning, they walk right into this alleged mistake. “The original embryo A lives until twinning occurs,” they propose, “and at that point, either A continues to exist and a new embryo comes to be by ‘budding’ from the original one, or (less likely, given recent findings) A ceases to be and two new embryos, B and C, come to be.” In other words, if I’m a budded twin, once upon a time there was an embryo that was something distinct from the living human organism that is now me, but that got transformed into the organism that is me at some point after the embryo came into existence. To escape this conclusion, George and Tollefsen have to accept either of two alternatives: that my twin and I are, through continuous identity, the same person; or that the original embryo was not, as they had posited, a “determinate individual.”

I envy the authors’ philosophy. It’s wonderfully clean and rational. It just doesn’t match the messiness of biology. Each time the biology gets complicated, they have to simplify it or brush it aside and retreat back into philosophy. Take the observation that life’s program runs on a network of evolving species. George and Tollefsen dismiss this critique as “veering away from straightforward biology toward metaphysical speculation of considerable abstraction. Human embryologists focus strictly, and rightly, on the life of a developing human, and his or her developmental program. Their business is not with the biological program for humanity.”

The objection is backward. “Straightforward” science, like metaphysical speculation, is what Aristotle did. You focus on one thing at a time, classifying it based on what you see. Modern biology goes beyond this. It penetrates superficial distinctions and tries to decode the underlying programs. The communality, transmission, and evolution of these programs aren’t abstract; they’re real. They just happen to be inconvenient if you’re trying to validate an equation between embryos and people. The suggestion that we should stick to the “business” of looking within an isolated embryo isn’t an argument. It’s a confession that if we broaden our investigation, the program will get too complicated.

The complications don’t ruin George and Tollefsen’s main point. The embryo does have a program, and its launch does mark the sharpest line in human development. But even that line is dotted. Transitions that are supposed to happen at fertilization happen after or without it. As the embryo grows toward maturity, it becomes more like a person. Its individuality solidifies. Its body plan and nervous system develop. Its boundary with the mother closes. Its placenta passes away. These, too, are lines in human development. It’s reasonable to build moderate IVF, stem-cell, and abortion policies along such lines, even if, like the rest of biology, they’re not absolutely clear.

Correction, Feb. 15, 2008: As evidence for trophoblast lineage distinction at the two-cell stage, the article originally linked to a 2006 paper in Science. That paper has since been retracted due to “falsified or fabricated images” by one co-author. However, according to subsequent communication with a different co-author who was not implicated in the fabrication, the essential findings “ have been replicated and appear to be as reported.” I have rephrased the “finding” as a “theory,” pending publication of the replicating study. (Return  to the corrected sentence.)