The problem of non-ancestral "ancestors"

by
Jim Moore

Jim Moore is a professor of anthropology at the University of California at San Diego.

One creationist criticism of the hominid fossil record as evidence for human evolution is that (recent) fossils that look a lot like us (e.g., Homo erectus) are us--fundamentally already modern human--and (older) fossils that look quite distinct (e.g., australopithecines) are distinct because they are not ancestors at all, but fossil apes. Thus, there is no connection between fossil humans and fossil apes, no transitional form. This argument is, for example, made at the Institute for Creation Research (ICR) Museum in Santee (CA) in a series of wall displays illustrating gracile and robust australopithecines, H. erectus, neandertals, and anatomically modern humans (click here for a view; the fossils are in the background (black blobs are replicas of the Laetoli footprints).

Those familiar with the fossil record will note the absence of Homo habilis from that list; indeed, H. habilis is not mentioned in the exhibit. However, this is not because creation scientists lack a story for habilis. Homo habilis, they say, is not a "real species" -- it is a "wastebasket" taxon, made up of specimens representing an unknown number of actual species, and therefore cannot be a true ancestor (after all, how can you have an ancestral species without having a species?). See, for example, the 1997 ICR video "The Image of God - Human Origins: Creation or evolution?" which features paleoanthropologist Chris Stringer acknowledging that 'habilis is a taxonomic mess'. Also see Lubenow, M. L. (1992). Bones of Contention: A Creationist Assessment of Human Fossils. Grand Rapids (MI): Baker Books. Both sound plausible as long as you aren't familiar with the actual evidence and scientific arguments.

They are probably right about the wastebasket taxon, but mistaken in the conclusion drawn from that. It is a particularly informative sort of mistake, worth examining a little, because it speaks to the more general point often made by creationists about the supposed absence of "transitional species" in the fossil record.

First, a little about the dramatis personae:

Gracile and robust australopithecines had brains very slightly larger (on average) than modern chimpanzees and (relative to body size) gorillas--very roughly,

: graciles: 430-520cc; robusts: 500-550cc; chimpanzees: 350-450cc; gorillas: 450-550cc

(Note that gorillas are much larger-bodied than either Australopithecus!). Unlike any modern ape, both were habitually bipedal (though still with some adaptations for climbing, like very slightly curved finger-bones), and their teeth resemble humans' in having relatively thick enamel on the molars, but they are overall much larger in size than either human or ape teeth. They probably stood about 3.5 - 4.5 feet tall.

Early Homo erectus had brains of about 900 cc; this is right about the absolute lower limit for modern humans without getting into severe clinical pathology (the "normal range" would run from about 1200 - 1600 cc). They were totally bipedal, with no obvious climbing adaptations (remember, modern humans are still pretty good climbers, so this doesn't mean they couldn't have spent plenty of time in trees). They were about our height (5-6'), and had teeth not too different from ours.

Individuals that have been labeled as "Homo habilis" are all intermediate in brain size, with a smaller variety at around 500-575cc and a larger one averaging in the 650-750cc range. We have a good estimate of body size only for the smaller variety, and it was small--not much different from a gracile Australopithecus (this means that its relative brain size--brain size for a given body size--was a good bit larger than that of Australopithecus; in fact they were comparable--for body size--to H. erectus). Their teeth and faces are also variable.

Why is Homo habilis important?

Clearly, these specimens appear to illustrate traits we would expect of a species transitional between the distinctly non-humanlike (but also non-apelike!) australopithecines and the distinctly human-like (but not like you or me!) Homo erectus. If only these specimens represented one species, it would be the clear candidate for linking "us" with "them". But more and more scientists agree that there are at least two very similar species mixed together here, rendering habilis something of a conceptual mess as a possible ancestral species.

What's wrong with using this as evidence against human evolution?

Consider this analogy. My last name is Moore, and my family (traced through the male line, by surname) is from Ireland. Suppose I wanted to learn about my male ancestors; what could I find?1

I've never been much into family history, and my father's father (P. J. Moore) died long before I was born. Clearly, though, I probably could find records to trace P.J.'s origins, and discover where in Ireland he/his family was from. With a little luck, there would be parish birth records that would allow me to trace my lineal ancestors back for generations, with record of which village they lived in, who was in the family, who they married, etc. If I wanted to ask who my ancestor alive in 1750 was, I might be able to answer both "taxonomically" by giving his name, and "ecologically" by describing the actual house he lived in, the size of his village, perhaps his occupation--a great deal of specific detail.

But what about 1450? Sooner or later, clearly the written record is going to fade; a fire burned the church with its record book, or something along those lines. If I ask about my ancestor alive at that time, the answer would still be fairly precise: an Irishman, likely living in (say) southwest Ireland. I would know something about his house: he didn't have electricity or running water, likely burned peat for heat, might well have built of sod. I would know that while he might have kept cattle or sheep, he probably would not have had a horse (though he might have) and almost certainly would never have seen a camel.

But what about 450? Somewhere along the line, I'm not going to have much confidence about his living in Ireland; perhaps he was Celtic, but this label gets pretty geographically vague if you go back far enough, and eventually I'm just talking about "northwestern Europe". I do know the odds on him ever having seen a camel are negligible, and I could probably contrast his lifestyle with that of a Greek, or Mongol, of the same period with great confidence.

But what about ... well, you get the picture. Sooner or later I pretty much will lose track entirely, and won't be able to say much beyond "he was a human ancestor of mine".

BUT WAIT: how can I say that? There is no sense in which at that point I can point to some specific person and say "yep, it's him"; given a lineup of Indo-European men I wouldn't have any idea which, if any, of them was my ancestor. My hypothetical ancestor has become a "wastebasket ancestor", an amalgam of a bunch of fairly similar likely candidates who cannot easily be sorted out or lined up in precise ancestor-descendent relationship to me (though they could be contrasted meaningfully with say a Tasmanian, who might usefully illustrate aspects of what it was like to be a human at that time, but would not have had any real chance of being an actual ancestor of mine -- think neandertal for purposes of the analogy).

How do I know I had an ancestor back then, if I can't point to him? Well, in one sense I don't. Maybe sometime around 1014 the first of my line of Moores was spontaneously created near Inisfallen; he wandered into a village, passed himself off as a survivor of the Battle of Clontarf and the rest is history. Hey, it could be true; how could I or anyone else know otherwise? And yet most people would agree it's not very likely. Why not? It would violate all sorts of principles of biology, physics, you name it, not to mention uniformitarianism.

Does my inability to identify a specific ancestor mean I can say nothing about him? Hardly. Assuming he existed, then whatever I can learn about his contemporaries will place limits on what he was like (he couldn't fly, he spoke to his family in some language, he ate cooked food at least some of the time...). I could learn quite a bit about him, relative to people on other continents, nonhuman primates, mammals...

Why is it any different for Homo habilis?

As the fossil record approaches 2,000,000 years old, the trail is spotty and there are several (known) candidates. They resemble each other closely enough that one can easily imagine either being ancestral (sort of like comparing a Viking and a Roman at A. D. 450; either could be ancestral to a given Irishman). They are different enough from others alive at the time (Australopithecus boisei for example) that we can feel pretty good about considering boisei a distant relative, not a direct ancestor (but still interesting, and informative about what hominids of the period were like--just as the Tasmanian I mentioned above would still be part of the broad vision of "human history" though not being ancestral to nearly all those living today [Europeans essentially exterminated the Tasmanians within the last 150 years]).

If we want to understand our ancestors of the time, those specimens that have been called "H. habilis" are the best available candidates; if it turns out that our name for them ("habilis") doesn't correspond to a single biological species, it no more invalidates them as possible ancestors than if I were to find that the Irish of 872 drew rigid social distinctions between Christians and Druids--I might not know which my ancestor was, and so be unable to answer major questions about his life and behavior, but I could still be confident that he was around someplace and most likely one or the other.

The general problem of transitional species

Evolution does not proceed along a linear ladder from one species to the next; if it did, the "missing link" would have a brain size of about 900cc (halfway between chimpanzee and modern) and be "halfway bipedal" (as it happens, bipedalism came first, then we saw major brain enlargment). This means we cannot predict exactly what a given ancestral species will look like in all its features; we can only specify which sets of features must have changed and make educated guesses about which ones changed together because they were functionally linked (say, jaw size and tooth size). Then we look for fossils to test our guesses, refine our theories, and go on.

Another thing about the absence of a linear ladder: it means that at any given point in time, there is not only going to be one possible "rung" on the way to any modern species. If you go to Home Depot and find a ladder with branches coming off it, you'd think someone in purchasing slipped up; but that's what the fossil record looks like. And so at any given "height", there can be several "rungs" that look a lot alike, are indeed close to each other, but only one is on the route to the top.

Taken together, these features of evolution make for a fossil record in which identifying "transitional forms" is always going to be difficult. If you reject anything ambiguous and place clear antecedents and descendants into separate categories, then of course you won't be able to find transitional forms--you've defined them out of existence!

1. While we're on the topic, it seems relevant to note that my father was born in Zamboanga (Philippines); apparently there was a popular song early this century about "Zamboanga, where the monkeys have no tails" (probably a taxonomically incorrect reference to orangutans). This gave my father his lifelong nickname: [the missing] Link. See what I mean, relevant? Back

This page is part of the Fossil Hominids FAQ at the talk.origins Archive.