On Not De-Extincting Homo neanderthalensis

Making “to de-extinct” into a verb is probably a bad idea.  Trying to de-extinct Neanderthals is probably a worse one.

The idea of bringing species back from extinction has been getting increasing attention recently and will get even more attention soon.  I generally support the idea, though with some reservations.  One of them is that this shouldn’t be tried with extinct hominids any time soon (if ever).

The Neanderthal possibility kicked off a worldwide teapot tempest in January 2013 when Der Spiegel, a major German magazine, published an interview with George Church, a prominent (and wildly creative) Harvard bioengineer.  In a question, the interviewer noted that, in his recent book, Church says “an extremely adventurous human female” could serve as a surrogate mother for a Neanderthal.  George has since said that his discussion of the topic was a thought experiment that escaped containment, which, on reading the whole Spiegel interview (here) seems mainly right (qualified by the fact that Church, whom I know, is not always averse to stirring up controversy).  I suspect the controversy was largely the result of the British tabloid press, notably a headline (between quotation marks!) in a story in the Telegraph about the Spiegel interview that read “I can create Neanderthal baby, I just need willing woman.”  There’s no evidence Church ever said that.

But the controversy does raise the question – if this were possible, should it be tried.  I think clearly not – at least for now and a long, long time to come.  My reasons are both medical and social.

Some Science – or Technology

But let me start with a moment’s explanation.  The stories often talked about “cloning” a Neanderthal.  That, in the sense in which we have, since Dolly, talked about “cloning” animals seems to me highly unlikely.  The Dolly process, somatic cell nuclear transfer, required taking a cell from one creature (in Dolly’s case a cell line derived from mammary tissue of a Scottish Blackface sheep) and fusing it (and, most crucially, its DNA-containing nucleus) into another creatures (a Finn Dorset sheep) from which nucleus (and hence almost all the DNA) had been removed.

The problem with doing this for a Neanderthal is finding a useful Neanderthal cell.  If cells have been carefully prepared and then frozen at liquid nitrogen temperatures, they can be used in cloning – as Dolly showed. But even if a well-preserved Neanderthal were found in permafrost or, like Ötzi (a modern Homo sapiens), in a glacier, the odds of finding a viable cell or cell nucleus seem close to zero.  It’s an empirical question and we don’t actually know the answer yet – but no one has announced finding viable animal cells or nuclei from ancient frozen animals yet.

As the Spiegel interview explains, Church doesn’t expect anyone to find a Neanderthal cell with a viable nucleus, but instead thinks someone could make one. Starting with a modern human stem cell line, he could slowly replace the modern human DNA with Neanderthal DNA where the sequence from the two species differed.  We know “the” whole genome sequence for modern humans (or, at least, the most important 90 percent of that sequence for several thousand humans) and we know the sequence for (several) Neanderthals.  When these cumulative changes made the stem cell line “sufficiently Neanderthal” in its DNA sequence, a cell from that line could be used for somatic cell nuclear transfer (the Dolly procedure) or potentially more directly to replace the inner cell mass cells in an existing human embryo.

This is, basically, the technique Church and others want to use to revive some extinct species for which we have good genome sequence data (and a closely related existing species), such as passenger pigeons.  It remains to be seen whether and how well this would work, though I suspect it will in the next decade or two.  If it works for extinct pigeons, why not Neanderthals?

Medical Concerns

The first problem is that we don’t know whether it would work or how well it would work, in any species, let alone in humans.  The potential health risks, for baby and mother, are enormous. That alone would make such an attempt in humans beyond reckless.

We do have some experience with making animals with genomes that have been tinkered with, either to knock out some genes or to “knock” some in.  This has mainly been done with rodents.  It works, though it is generally terribly inefficient.  We also have some experience with somatic cell nuclear transfer with animals.  It, too, works – for many species although not yet for any primate (in spite of tens of thousands of efforts with monkeys).  It is neither very efficient nor, more importantly, very safe.  Most species, when cloned, produce an unusually high number of stillbirths or deformed infants.  The only extinct subspecies ever revived, the Pyrenean ibex, was cloned from a  tissue sample  carefully taken and prepared from the last survivor of the subspecies. The clone was born alive and died within ten minutes – it had an extra, non-functional, “sort of lung” in its chest, which appears to have caused its death.

We are willing to accept low efficiency, high miscarriage and stillbirth rates, high early death rates, and high deformity and disability rates when dealing with laboratory animals.  Even there, we have (appropriately) some limits in terms of the amount of suffering an experiment can be allowed to cause.  But as close as Neanderthals appear to be to us, it seems to me that the only appropriate measure for the “right” amount of risk to take is to ask whether we would allow this with modern humans.  We might (or might not) eventually decide that they were not “fully human” but they certainly would be much closer to us than mice, rats, or even chimpanzees are to us.  I think the only careful think to do is to assume that they would be, for any morally relevant purposes, “human.”  A disabled mouse pup is one thing; a disabled Neanderthal infant is, in effect, a disabled human baby created by research.

Unless and until these techniques were shown safe and effective in modern humans, their use to make Neanderthals cannot be considered. (Even if they were safe and effective for modern humans we still could not know that they would work for a reconstructed Neanderthal genome, but, depending on the context, we might have enough confidence to go forward.)  The FDA might (should) insist on strong proof of safety and efficacy before allowing this kind of research in humans (or extinct hominids).  It has asserted jurisdiction over human cloning, ooplasm transfer, and similar techniques.  That kind of proof might be forthcoming someday, but no time soon.

And even if the “genome editing” and “baby producing” steps were both proven safe and effective in humans, there is another risk, the one the Telegraph hinted at – what about the mother?  Is it safe for a Homo sapiens woman to bear a Homo neanderthalensis fetus?  We don’t know.  It’s not even clear how we could find out, short of trying.  I suppose it is possible that inter-species pregnancies in species similar to us – seeing that, say, gorilla females could carry orangutan fetuses or chimp females could carry a bonobo fetus – might provide some confidence.  Unless an appropriate level of confidence could be reached, we could not ethically – and regulators should not allow legally – that kind of experiment to be done, no matter how brave and willing the women.  (Alex Capron, at his Law and Biosciences Workshop talk in February at Stanford Law School, said that lots of people had volunteered to be the first person to go to the moon, even though we could not get them back – volunteering for a one way trip.  Their offers, though made seriously, were not taken seriously.)

Alternatively, one might get to the point of having an “artificial womb.” I think this is plausible – probably an actual human uterus constructed, ex vivo, from stem cells and attached to machines that provided appropriate life support (blood with oxygen, glucose, hormones, etc.)  I can imagine that, sometime in this century . . . but not that it would produced and been proven safe and effective during my lifetime.

So the medical risks to infant are unknown but quite likely high. In the case of the surrogate mother, they are quite possibly high.  We might be able to get enough confidence in sufficient safety sometime this century, though, I would argue, not for many decades.  If so, could we ethically bring back Neanderthals?

Social Issues


Again, I think not – at least, not anytime soon.

Part of me wants to just make a Kantian argument that these Neanderthal babies would be humans and humans must be viewed as ends in themselves, not as a means to something else. Creating Neanderthal babies for research would, in this sense, be as bad as creating modern human babies strictly for research.

But I don’t generally use Kantian arguments, including this one, even though it seems to me that it might be particularly appropriate here.  Instead, I rest my opposition on the likely consequences to the Neanderthal (or Neanderthals) we create.

Frankly, I do not trust us to treat them well. By well, I mean as persons, entitled to human rights.  Perhaps they would be “poor quality” persons – below the modern human medians for intelligence or socializabilty or emotion or other behavioral traits. But so are plenty of existing modern humans (after all, half are at or below the median levels of any trait).

Rob Streiffer, a philosopher at the University of Wisconsin – Madison, has made this argument in his work on human/non-human chimeras.  I respected and largely agreed with his arguments, but thought they were unlikely to be relevant – the chance that anyone could or would create a human/nonhuman chimera that was “human enough” (whatever that means) to be entitled to treatment as a “person” seems to me quite remote.  On the other hand, that a Neanderthal baby should be treated as a person seems to me the presumption we should apply, so Streiffer’s arguments are very salient.  Perhaps, unlike the chimeras he worries about, we would not slice, dice, sacrifice, and autopsy a Neanderthal baby the way we might a monkey with a human-like brain, but human history gives me no confidence that we would treat it – make that “him” or “her” – well.

Could that fear ever be overcome?  I’m not sure.  Substantial and successful experience with treating other non-modern human but intelligent entities – little green men, androids, truly intelligent computers, communicative porpoises – well might give me some hope, but we cannot count on that happening anytime soon, if at all.  Would it be enough if humanity, in general, became vastly warmer and kinder, singing Kumbaya all day and night?  Maybe, but I suspect we’ll see little green men in flying saucers first.  No, I think the risks that we would mistreat Neanderthals are, and will remain for the foreseeable future, too great to make their creation ethically justifiable even if the major medical risk issues were overcome.


I am not entirely happy about my conclusions. I would dearly love to know what a Neanderthal was like, in ways that bones and artifacts cannot reveal. (Leaving aside, for the moment, the question whether an even 100 percent recreated Neanderthal genome would “actually” be a Neanderthal in the face of different epigenetics, microbiomes, environments, and cultures.)

It is conceivable, barely, that we will learn enough about genetic variations that we could “recreate” the Neanderthal in silico from its genome, predicting (accurately) what it was like from its DNA sequence even without that sequence being embodied in a biological organism.  Maybe – but I’m not holding my breath.  And, even so, in silico wouldn’t, at least to me, seem the same as a living, breathing organism.

Plus, I think it would be fascinating to see our societies grapple meaningfully with the question of what entities, other than biological organisms born alive as the result of a merger of human’s egg and sperm, deserve to be treated as “persons,” ethically and legally.  But I’m not prepared to countenance steps to test that, at least until I thought we would be likely to answer that question well.

So, I personally have no deep ethical objections to the de-extinction of passenger pigeons, Tasmanian tigers (thylacines), wooly mammoths..  But Neanderthals, Denisovans, Hobbits, Homo erectus – no.  Or, at least, not in my lifetime.  Alas.

Hank Greely

PS  An organization called Revive and Restore, with help from the National Geographic Society, has been largely responsible for any general discussion of de-extinction. I attended a workshop the group held in February 2012 and served on the planning committee (and spoke at) another workshop in October 2012.  It is holding a TEDx event on March 15, 2013 in Washington, D.C. at which I am speaking.   I am not, however, currently affiliated with Revive and Restore.  (Oh, and they aren’t advocating de-extincting Neanderthals, either!)