On pathology and evolution, or, the Turkish tetrapods

No story yet, in the history of the weblog, has had so many people writing to ask me if I am alive. Even my parents are asking me if it's real.

I am alive, and I have to say it is somewhat heartening that so many people care whether I'm watching. It's been sort of like H.G. Wells' Time Traveller (no Morlock jokes, please), just watching a couple weeks of this story's progress makes a few things about paleoanthropology and the press much clearer.

It is not a hoax. It is a BBC publicity blitz. And to me, it seems like a stunningly cynical PR blitz. In other words, we're being played.

First, the facts, which can be drawn both from a working paper by psychologist Nicholas Humphrey, John Skoyles and Roger Keynes, as well as the published paper in International Journal of Neuroscience by Uner Tan, and the genetic paper by Turkmen et al. (2005) in Journal of Medical Genetics :

  1. There is a family of people in Turkey with 19 children (many adult), 6 of whom have a congenital form of cerebellar ataxia -- a condition in which the cerebellum does not develop to a normal size, with effects on muscle coordination and cognition.
  2. Five of these six affected individuals walk on all fours, with their palms to the ground. One walks bipedally.
  3. The cognitive deficits in the affected individuals include some linguistic limitations, although they communicate in Kurdish with a limited vocabulary and some can speak Turkish as well.
  4. This family was studied by Turkish neuroscientist Uner Tan. Uner Tan is a real, live person (some have speculated that the name might be an anagram for "nature", indicative of a hoax). He has had a long career in neuroscience, and has a substantial publication record extending through the last several years.
  5. Tan claimed in his paper that the connection between cognitive deficits (particularly emphasizing the linguistic) and the quadrupedal gait might indicate a link between these characteristics in the evolution of humans. Tan explicitly raises the idea of one or more macromutations causing bipedal locomotion to emerge suddenly from a quadrupedal ancestor, in association with human cognitive abilities. In his paper, the syndrome is described as a live model for human evolution.
  6. A British group of researchers, including Nicholas Humphrey, were invited to investigate the family. A documentary film production became involved at this point.
  7. A German team of geneticists took samples from the family, and found a region on chromosome 17 that appears to be associated with the ataxia (they term hypoplasia). The condition appears consistent with a recessive Mendelian inheritance; the parents are first cousins (once removed), which raises the likelihood of a recessive disorder being expressed. The German team claims that the gene involved may shed light on the evolution of bipedality.
  8. Humphrey, Skoyles and Keynes deny that the disorder gives any indication of a coevolution of bipedality and language. Indeed, they are skeptical that the genetic basis of the trait says anything about the evolution of bipedality at all. However, they believe that the affected individuals have a unique form of quadrupedal gait, which they call "wrist-walking" because weight is borne on the heel of the palm. They argue that this form of locomotion may be a better model for the ancestral quadrupedalism of hominids than knuckle-walking.
  9. The documentary is scheduled to run on BBC2 Friday, March 17 at 9:00 pm.

What about the science?

Of course, the most important question is whether any of this actually contributes to understanding human origins. Generally, I think the answer is no -- not just an itsy-bitsy no, but a great big honking no.

But I will qualify that for a couple of aspects of the story, since there is the potential of finding some interesting facts. So I will cover the different hypotheses in turn:

Is Uner Tan right? Is the condition really evidence of a rapid coevolution of bipedality and language? No. Human bipedality and human cognition are both highly complex traits involving anatomical, developmental, and behavioral specializations. Each of them involved hundreds, and for cognition I would say thousands, of different genetic changes. There was no small set of macromutations that caused these traits to arise.

Furthermore, the pathology manifested in this family does not indicate any kind of earlier evolutionary stage. There was no time when people with modern human limb proportions walked on all fours. It is doubtful that australopithecines, who were largely apelike in their cognition, ever walked on all fours. A developmental disability that causes mental retardation and inhibits the development of normal bipedal walking does not indicate any kind of earlier form of hominid.

Tan's paper has but a single reference in its bibliography -- to himself. And he's an editor of the journal, International Journal of Neuroscience. It happens every year or so that some paleoanthropology story breaks about a paper in some non-anthropological journal. These don't benefit from peer review by paleoanthropologists.

Is the gene underlying the condition important to the evolution of bipedality? This hypothesis is almost impossible to test, because "important" could mean almost anything. If "important" means that the gene underwent at least one substitution as a result of its relationship to bipedality, then the answer may well be yes.

But that doesn't really show that the gene is "critical" or "crucial" or "significant" or any other random signifier. If what we mean by "important" is that bipedality would not be possible without a particular genetic change in that gene, well, I think that is pretty difficult to test. You would have to show a link between gene expression profiles and the development of bipedality in normal people, not pathological ones. You would have to demonstrate a history of substitutions in the gene that match the timing of bipedality. And you would have to show that other functions of the gene either did not interfere with its evolutionary role in bipedality or were deemphasized because of that role.

I think it's really unlikely that a gene that causes cerebellar ataxia was a critical bipedality gene. It may be necessary to walking normally, but it probably (indeed, evidently from the nature of the disorder) is very important to a lot of other things as well. A gene that breaks early brain development is no more likely than other genes to have a specific function role in the development or practice of bipedalism in humans. More on the genetics of broken development below.

Is the specific form of quadrupedalism -- the "wrist-walking" -- of these people a clue about the nature of pre-bipedal hominids? This is the most likely to be scientifically interesting. And it raises the possibility of a test, if you could find bony correlates of the wrist-walking that could be distinguished from knuckle-walking or other forms of locomotion. And the question is well within the tradition of paleoanthropology -- I mean, we even have people studying knuckle-walking anteaters for goodness' sake.

On the other hand, there are several reasons to be skeptical. The working paper by Humphrey, Skoyles and Keynes only includes a couple of paragraphs on the evolutionary question, so it isn't fair to suppose they have presented a summary of all the pluses and minuses of the hypothesis (that early hominids or their ancestors were wrist-walkers in this way). What it has going for it is that early hominids were not chimpanzees or gorillas. Their hands had finger and thumb proportions much like ours, unlike the very short thumbs and long fingers of great apes. In other words, they were not adapted to suspension below branches in the way that apes are. So it is credible to think that if they walked quadrupedally, or evolved from a non-suspensory ancestor, that they might have different quadrupedal gait from living apes.

But although they didn't have the hands of apes, early hominids still had powerful wrist flexor muscles, as indicated by their curved hand bones and large carpal tunnels. The length and strength of these flexors -- necessary for suspension in great apes -- limit the extension of the wrist. This is one of the reasons why chimpanzees and gorillas are knuckle-walkers and orangutans are fist-walkers: their hands do not extend as easily to a palm-down position for weight-bearing. So although monkeys and other primates are palmigrade, apes are not. Humans can extend our wrists more readily, and walk palm-down when we want to, which is usually limited to the first year or two of life. But for early hominids, and presumably their ancestors, this may have been more difficult.

Now the most interesting part about the wrist-walking idea is that it might itself have caused greater hominid wrist mobility, which became more important to hand control and flexibility for toolmaking. But I think that more likely places the effect before the cause: if wrist flexibility and hand morphology evolved for the purpose of toolmaking, then there is no need to invoke an earlier wrist-walking stage. So while the speculation is worthwhile, I think it is unlikely to prove important.

Pathology and evolution

The fundamental question is how much a pathology can tell us about an evolutionary change.

Pathological individuals are by definition to some extent unique: they do not exhibit the characteristic pattern of development or behavior that for most people is normal. Their phenotype did not have to survive and reproduce in the past -- and in the present it generally fails to do one or the other. They may be helpful in uncovering the course of development -- if a critical stage does not happen, what effects does it have?

An analogy mentioned more than once in this story is the alleged relationship between the FoxP2 gene and the evolution of language. But a clear consideration of this link yields some insight about the difficulties linking pathology and evolutionary changes.

FoxP2 was singled out because of a mutant allele that causes specific language impairment in a well-studied family. The disorder has some other effects, but it targets language fairly narrowly. That argues that the normal form of the gene is necessary for normal language development. Moreover, the gene underwent a recent substitution in humans, with a new allele arising and becoming universal within the last 200,000 years. This change suggests the hypothesis that the rise of language required this substitution.

But as yet, that is only a hypothesis. And is quite a bit of evidence that detracts from it. For one thing, only two amino acid substitutions separate the human FoxP2 gene from the chimpanzee form. If this gene was critical for the appearance of language, that criticality did not involve repeated changes. Considering that language is a complex trait, that means that many other genes are probably also involved. Which are the critical ones? And does FoxP2 count as critical with only two possible changes? And how can we establish that the most recent sweep in FoxP2 was related to language at all? After all, evolutionary changes in other genes could have left language dependent on FoxP2 function, without necessitating any changes in FoxP2 function. So the link between FoxP2 and language is very suggestive, but to say that the gene is necessarily critical in the evolution of language is just not yet demonstrated.

The main point is this: the fact that a gene breaks something doesn't mean that it was the key gene necessary to create something. Suppose that you want to figure out how a car works. So you look at cars that aren't working right, and you see what is broken. Now, you will notice that cars run sort of poorly with flat tires, they run with depleted batteries but won't start, they will run for a bit without motor oil, but then seize up, and so on.

Cerebellar ataxia breaks a whole lot of things. It's like breaking the crankshaft -- the engine might run, but it is going to make a whole lot of noise, and the car isn't going to move. We may conclude that the crankshaft is necessary for the wheels to move. But does that mean that the crankshaft is the key component of the wheel? Clearly not.

The analogy between cars and organismal development is useful because both systems depend on hierarchical functions. Early things must all work right for later things to develop. When an upstream gene (or part) breaks, it doesn't mean that downstream things affected by the broken gene were caused by the broken gene.

But in biological systems these hierarchical functions sometimes involve the same genes -- regulation can make early-acting things also have later effects. So we may not prove out a critical relationship between a gene and a function on the basis of pathology, but we may find it equally difficult to rule it out. Here it is clearly useful to keep in mind the null hypothesis -- the gene isn't the unique, specific, or important cause until proven otherwise.

In this context, it surprises me that nobody has commented on the other prominent connection between Nicholas Humphrey, pathology, and human evolution. One of his best-known essays compares the cave art of France and Spain to the figure drawings of autistic children. The point is that understanding certain pathologies in the present may give hints about the evolution of humans in the past. In the case of autism, the idea is that ancient human minds may not have had a very different ontogeny as language and the ability to use and understand symbols emerged.

The paintings and engravings must surely strike anyone as wondrous. Still, I draw attention here to evidence that suggests that the miracle they represent may not be at all of the kind most people think. Indeed this evidence suggests the very opposite: that the makers of these works of art may actually have had distinctly pre-modern minds, have been little given to symbolic thought, have had no great interest in communication and have been essentially self-taught and untrained. Cave art, so far from being the sign of a new order of mentality, may perhaps better be thought the swan-song of the old.

It is a more sophisticated idea than that autism is an atavistic trait -- a throwback to an earlier stage. It is an analogy -- the cognitive traits that autism affects may not have been present in some earlier human. I don't believe that story either -- although it may be worthwhile to explore what it would imply for certain human traits to be absent, the fact remains that ancient people had to live somehow. They were social primates, and traits underlying their (and our) sociality are deeply embedded in our development. Autism may affect some of those, for genetic or nongenetic reasons, but there is no reason to think that the trait associations found in that condition are relevant to the evolution of an earlier hominid.

With the quadrupedal story, they are on safer ground. But in that case, if the claims are not so bold, neither are the possible conclusions if they are right.

The PR

That some kind of PR machinery started this story is obvious. After all, a single-reference paper in an obscure neurological journal by a respected but obscure Turkish scientist would not ordinarily make international headlines. Nor would a story about a single family carrying a very rare mutation, regardless of what it did (indeed, the genetic paper by the German team in December did not create any press). And the fact that the documentary has clearly been timed to coincide with the release of Tan's paper is pretty blatant.

But this is the interesting part: Even the hoax-like appearance of it is part of the PR. The first news stories could not have broken through internationally without the sensationalist claims -- the "evolution running backward". To be sure, this story has had unusually long legs because newspaper editors are drawn to its freakshow quality (generally showing little apparent regard for its unfortunate subjects). And there is the video showing the people lurching around, which has become a popular e-mail chain mail. It is the same force that makes people tune into the Discovery Health Channel to watch "Trash Can Full of Skin", and "200 Pound Tumor".

Let's face it: World Science (slogan: "Long before it's in the papers") is not the New York Times. It seems like a great place to float a strange story. And after the first story, they followed up several times, first with news about the genetics.

By this time, blogs were on the story, and it might have been expected that I might write about it. I certainly got several e-mails from colleagues passing the story around. Thanks to all those who sent references and pictures!

It was not clear at this point that there was a documentary. That story emerged during the first week of March, with the BBC2 release, the producers at Passionate Productions themselves, and another World Science story about the feud between Tan and the British researchers. This story raised the question of payments to the family associated with the research and documentary, quoting bioethicist Arthur Caplan.

At this point, mainstream science writers picked up the story, notably including The Times online and a skeptical Carl Zimmer. And in what I think may be the first case of active blog-publicity-managing by a key player in an emerging paleoanthropology story, Zimmer retracted his post after a correspondence with Nicholas Humphrey [see update below]. And National Geographic News covered the story with perhaps the most unintentionally hilarious quote I've seen in a long time:

This bizarre case is not a hoax, according to experts who have studied the family.

The cynical part is that the entire progression of the story was predictable without any necessity of orchestrating it. You see, it has all happened before, with almost exactly the same sequence of events. It is the Flores story reborn.

Consider: a rare, unique, and strange discovery comes to light. Immediately many experts are skeptical, but it takes time for them to organize a response. In the interim, a few strong backers explain that unique evolutionary principles can explain the find, and some experts cautiously agree. Then, two things happen: first, there is a conflict between the Western backers of the find and a senior local scientist; second, a high-profile documentary appears in which the discoverers can take their case directly to the public.

It's an autocircus.

Only the documentary itself can save us from this madness. At least that seems to be the import of this quote from one of the producers, on Gene Expression:

Our film will, I hope, redress the balance and be seen as a sensitive, thorough and thought-provoking record of the family/phenomenon. We made it not for voyeuristic reasons (although of course the quadrupedal siblings are visually arresting) but because we thought it raised all kinds of fascinating scientific, and many other, questions. I think the reason it's created such a fuss because bipedality is something that defines us as human beings - separate and distinct from beasts - and their existence is challenging philosophically. You only have to look at the Bible, for instance, to see how the word "upright" is loaded with meaning about purity, morality etc. That's ingrained very deeply in us.

Yeah. Um-hmm. "Upright" is loaded with meaning about purity, morality, etc. Ingrained deeply. Yep.

UPDATE (3/14/2006): Carl Zimmer wrote me about his earlier comments, and wanted to make sure his post and retraction weren't taken out of context by either me or by readers who might not follow the link to his explanation. He writes:

As I explained, I retracted my post specifically because I realized I shouldn't be making speculations about the relationship between the research and the show if I hadn't done the reporting to back them up. In retrospect, I'd now says that if the post had been limited to my skepticism about their scientific claims, it would be still up now. But I screwed up, and I took responsibility for it. I don't appreciate your implication that I was being managed, which is not supported by the fact.

Definitely nothing sinister implied on my part, just that it is novel that somebody is watching blogs!

References:

Tan U. 2006. A new syndrome with quadrupedal gait, primitive speech, and severe mental retardation as a live model for human evolution. Int J Neurosci 116:361-369. DOI link

Turkmen S, Demirhan O, Hoffmann K, Diers A, Zimmer C, Sperling K, Mundlos S. 2005. Cerebellar hypoplasia and quadrupedal locomotion in humans as a recessive trait mapping to chromosome 17p. J Med Genet DOI link