Over at the New York Times, Carl Zimmer has an article on the difficulties of pinning down hominid species, which illustrates an interesting perspective in biology, but is unfortunately a little too brief. There are a couple of factors at play, and no easy way to resolve them.

The very first thing to bear in mind is that ‘species’ is an arbitrary distinction in many ways. The word was born to differentiate, say, chipmunks and bandicoots, or penguins and ostriches – okay, those are bad examples because differentiating them doesn’t take a lot of effort, but the premise is, these are distinctly different groups of animals that cannot interbreed. If they could mate and produce another critter, then they were the same species. Simple.

Until we got to hybrids. Horses and donkeys are different species, but they can still produce offspring, which is where we get mules (and you always thought it was a plant in Missouri.) Mules, however, are sterile, which is what distinguishes a hybrid, so the definition of species changed to mean that the offspring must be able to make viable babies too, otherwise the two original were separate species. Things remained okay for a while.

But then there’s the ancestral bit – Darwin and Wallace came along and had to ruin the party for everyone else by pointing out that all species were different in the past. We’d known about dinosaurs and such before then, but now came the theory that, for instance, Neanderthals were actually older direct relatives of us Homo sapiens (this was a possibility considered for a while, don’t get ahead of the narrative.) What it meant was that we needed some way to distinguish present-day versions from older ancestors, even if there was a direct lineage, and for convenience’s sake an ancestor with enough differences from the living version could be considered a separate species.

This is the first bit of fun, which produces countless issues the world over every day. There is no point where we can say that two adult Homo erectus produced the first Homo sapiens – it doesn’t work that way, and this illustrates the fudge in ‘species’ that has been accepted because, seriously, what else could we do? We do not have fossil remains that span the whole history of hominids, or anything else for that matter; what we have are spots in the past where we’ve found skeletons with different structures – enough like humans to be considered more closely related than chimps or whatever, but different enough not to match modern humans. And, a key point, we’re not really sure whether there is a direct lineage, or if they’re some family line that split away and went extinct later on.

In the middle of all this came a new factor: populations of existing species that could apparently interbreed, but never do. DNA isn’t much help here, for two simple reasons. The first is, there’s genetic variation in every individual of any given species, no matter how confident we feel about the demarcation – so it’s almost as if we could say that every offspring is a different species from their parents, if we went by genes. That’s not really a help, and of course makes ‘species’ well nigh meaningless. Second, a complete DNA strand, what’s called a genome, is ridiculously long, often billions of base-pairs (paired amino acid molecules,) so it’s not this routine task to look at samples from any two individuals and classify them as ‘same’ or ‘different’ even if we make another arbitrary distinction – say, everything that has this much unchanged DNA sequence is the same species. There aren’t any spots that never change. While we can define human from chimp in that manner, because the long divergence from the period when the two species had just split has produced dependably diverse stretches, when we’re talking about trying to tell apart two insects that appear identical, it’s both too ridiculously complicated and too vague to be useful. So we just take the bugs’ word for it and assume, if they don’t want to get it on, then they know something we don’t, and ‘species’ was changed again to accept this distinction as well.

When it comes to fossil remains, what we typically have is one sample from one area in one specific point in time, and very often not even a complete example of that, since nature isn’t very accommodating in keeping corpses complete and undamaged. But as Zimmer points out, we have a huge variability in Homo sapiens right now, to say nothing of species like dogs or horses. Finding two skeletons with radically different structures does not mean we’re dealing with two different species – we could just be dealing with a Newfoundland and a Yorkshire terrier. And then of course, there’s the possibility that we’re even seeing a rare mutation, such as dwarfism or gigantism, and if the skeleton is incomplete then there’s the opportunity for even more examples such as microcephaly. Finding a set of bones that does not match any samples we’ve already found does not mean we have a new species.

If, for instance, the fossils come from a time period far enough removed from any others, they’ll get their own name on the simple fact that there’s nothing that stops natural selection, so over enough generations the species is going to be different. But there are periods when several different species of hominid were concurrent; there was more than one proto-human existing at the same time, like Homo sapiens and Homo neaderthalensis. The other side of the coin is the case with the “Hobbit,” Flores man, or Homo floresiensis – pick whichever name you like. The multiple examples of these remains, existing at the same time as Homo sapiens, have traits indicating descent from H. habilis or erectus, along with traits of much older Australopithecines. This presents numerous possibilities: reverse adaptation, very early isolation with convergence (the simultaneous development of similar traits among unrelated species,) a population of genetic anomalies while still H. erectus, examples of island dwarfism… the debate goes on.

So the species names we affix to fossils is arbitrary in many ways, and does indeed change as new remains are discovered, or even new ways to examine existing fossils. One of the developments of all this uncertainty is cladistics, which classifies fossil remains not by individuals so much, but by the development of body-type variations (phenotypes.) 400 million years ago, all animal fossils were solely of fish, but 375 million years ago, lobe-finned fish appeared in the record; somewhere between those two times, the beginnings of the limbs that would allow walking on land appeared. It doesn’t matter which, exactly, species developed it, or indeed how many – it’s just enough to know the bracket when it occurred.

With the hominids, it’s a little more complicated, because they split off from the other primates in Africa and migrated to the rest of the world, but it’s not clear if this happened once, or in waves, and how many hominid variations existed at any time, nor where they went. Fossils are rare things; the conditions to make a skeleton fossilize rather than simply break down in the soil or sunlight, and then retain them for millions of years, are very specific and don’t happen too often.

And then there are the geological conditions to make finds even possible. Sediment builds up over time, or erodes away. If it erodes away through the fossil remains, wave goodbye – we won’t be finding them unless we’re there exactly when it happens. Otherwise, we have to dig through millions of years of accumulated soil, perhaps now metamorphosed into rock, to get to even the right layer. You can’t just dig a hole, go deep enough, and find hominid remains. Pull up any map, close your eyes, and pick a spot. Did you hit a cemetery? If not, then this displays the difficulty with random digging (and we pile bodies all together neatly, unlike our distant ancestors.) So the trick is to find someplace where rock from the right timeframe is being naturally exposed, and search in the broad fields available therein. This is why the Great Rift Valley in Africa, where Olduvai Gorge resides, is the site of so many finds. I’ve taken advantage of this myself by picking through the crumbled shale exposed when a glacier cut a huge scar through central New York.

Yet, we can’t do this for all time periods to present, because the conditions that preserve remains are rare. The fossils that I was finding existed in one layer, presumably when mudslides or river delta deposits buried dead animals in low-oxygen conditions, preventing bacteria from breaking things down. Earlier, and later on, the mudslides weren’t happening, or if they were, the remains still decayed completely. So, if we wanted to track the migration of the hominids from Africa, we couldn’t just go five hundred kilometers away from a known find and see how long ago more hominid remains can be found; the geology might be all wrong, and random digging is highly unlikely to produce anything.

So, we end up with tiny samples in time, pinpricks in both location and history where we have some particular phenotype. If it falls between the time periods for Homo habilis and Homo erectus, and has a body plan that seems halfway between the two, then we can tentatively say it demonstrates an intermediate species. Often, what we have are some traits of this one, and some traits of that one, but few if any traits that split the difference. It also needs to be noted that evolution does not necessarily proceed in a straight line, as if there’s a goal; selection depends on conditions and environment, which may change, especially for a species migrating across multiple continents. So traits may wander, develop at different rates, and so on. Now, imagine if we found a partial skeleton: portions of the hips, one upper leg and knee, a big toe, maybe some shoulder bones. Everything about the legs seems to indicate an upright walking stance. Then, from a time period .2 million years later and further south, we have a species with toe bones that indicate tree-climbers. Are they of different descent? A local environmental adaptation of the same lineage? A genetic fluke?

We don’t know. And this is true of so much of paleontology – we can only surmise based on limited evidence, and try to find more evidence that supports or destroys the hypothesis. For the time being, we give the fossils a name, knowing that this might change later on. That’s how it goes. [The example given above, by the way, is pretty much the case with Australopithecus afarensis and Au. africanus, though I admit the time separation might be different.]

What this seems to imply is that our impression of human development could be entirely wrong, something that creationists gleefully latch onto since the evidence for their own ideas is nonexistent (yes, any flaw in modern science is reason to deny it in its entirety, but flaws in scriptural accounts are ignored wholesale – pathetic, isn’t it?) But uncertainty as to how such finds fit together does not change the facts that any finds of such body structures tells us selection is a very real thing. No hominids of any kind have been found, or ever will be found, in 30 million-year-old strata. No Homo sapiens will be found alongside Au. afarensis. While we may not know which, if any, given fossil species from 2 million years ago is our direct ancestor, we are extremely comfortable in saying that the body plans in evidence at those times fits exactly into expectations of development from the common ancestor with the great apes. And of course, any competing theory also must explain why we have these remains, and why they progress in shape just like natural selection predicts.

What it all comes down to, however, is not considering any given fossil species to be as definitive as modern species – there’s just not enough information. There are species that are 99% likely to have gone extinct, like Paranthropus boisei – the skull structure was radically different from nearly everything else found, and nothing since shows evidence of developing from there towards Homo sapiens, or indeed any later species. But most other species are not so well determined, and lineage remains nothing but speculation right now – we cannot say which evolved from which, or which (if any) is our direct ancestor. This illustration shows the time periods that we have determined so far, but implying any particular path from it is nothing more than guesswork. Perhaps one day, we’ll have enough evidence to be pretty confident, and in the meantime any one of these may disappear, absorbed into another species due to more information, or we may even split off another line based on closer examination of existing fossils, to say nothing of any new finds. It’s an interesting, but undeniably contentious, field of study.

Coming up: Why do we bother with trying to find out?