With his customary elegance, erudition, and aplomb, Matt Cartmill1 once again presents us with a counsel of despair. Having long ago decided2, 3 that the morphologies he has labored so meticulously to describe in such exquisite detail have little or nothing to contribute to systematics—and having, in essence, bleakly concluded that systematics itself is an inherently fruitless pursuit—he now joins the devout invertebrate paleontologist Simon Conway Morris4 in promulgating the notion that, given enough time and the appropriate starting point, the phenomenon of parallelism (convergence, if you prefer) virtually dictates particular evolutionary outcomes, including the emergence of “an animal capable of producing pots, pyramids, and the Pythagorean theorem.” At some point, it appears, everything becomes written. But this isn't all. Add to the slow, transformational inevitability of evolution the fact that even the largest and most distinctive of taxa must ultimately trace their origins to “trivial” speciation events, and it becomes almost a matter of logic that those higher taxa themselves should be “either polyphyletic or unimportant.” (It is hard for the reader to resist the obvious extension of this chain of reasoning: If the founders of great evolutionary dynasties may thus be dismissed as epiphenomena at best, how much more unimportant yet must individual remote descendants of those ancestors be, especially terminal ones such as Homo sapiens?)

Well, the “polyphyletic or unimportant” attitude toward major taxa may or may not have some obscure philosophical merit. But I suspect that most biologists will find it not merely devastatingly arid, but hugely counterproductive. Far better, then, to look at higher taxa as the proper subject of biological inquiry rather than of metaphysical posturing. So let's start with a proposition with which everyone should be able to agree, namely that the diversity we see in the natural world is nothing short of phenomenal. Indeed, its sheer luxuriance is probably the biosphere's most striking single aspect. To take just one example, our own Class Mammalia alone harbors forms as hugely dissimilar as whales, bats, and moles. And even when you put this in the context that Mammalia itself is but a drop in the ocean of organic diversity—after all, it is but a single class among sixteen that belong to just one of 38 phyla—it is still hugely difficult to apply such terms as “trivial” and “unimportant” to the differences among them. Indeed, even in the unlikely event that one were to encounter a Neanderthal congener while out for a stroll in the wilds of the Altai Mountains and take note of its differences from one's self, it seems unlikely that either of these terms would immediately spring to mind.

Equally striking is that this luxuriant diversity of nature is clearly structured. Every folk taxonomy in the world screams this.5 What is more, the configuration of the similarities we observe among living things is excellently accommodated by the nested sets at the heart of the preevolutionary Linnaean taxonomy we still use today, a system that has robustly survived a quarter-millennium of close scrutiny. Within the inclusive Linnaean hierarchy, species group into genera, genera into families, families into orders, and so on. This is not simply an arbitrary matter. The names and ranks we give the taxa we recognize may indeed be arbitrary but, if we've done the job right, the taxa themselves are necessarily “real” in the sense that each one is the product of a unique sequence of irreversible historical divergences. And if we have any confidence at all in evolutionary biology, we can reasonably hope to reconstruct those sequences through patterns of morphological and genomic resemblances. The nested arrangement of taxa that such analyses produce is thus neither an artifact of present perspective nor of taxonomic convenience. The fossil record abundantly demonstrates that such structuring has been present at every point in time since living organisms first began to leave evidence of diversification.6 And the solitary testable hypothesis we have to explain the ubiquity of this particular pattern of resemblance among species, fossil and living, is Darwin's “descent with modification.” In other words, the dichotomous structure of the inclusive Linnaean hierarchy is so far explicable in scientific terms only by successive lineage splittings via speciation.

Cartmill is clearly correct in supposing that lineage splits may often involve rather minor anatomical modifications in the diverging branches (though now we know that structurally small genomic changes may have large developmental consequences). In support of his notion, we can offer the observation that living primate species currently grouped into the same putatively monophyletic genera tend not to differ from each other very much, especially in the hard-tissue characteristics we might hope to pick up in fossils. But over time, and with multiple successive speciations, even minor modifications inevitably add up by a sort of ratchet effect, so that the more inclusive the Linnaean grouping, the more sharply its crown members are distinguished from members of equivalent contemporaneous taxa. Nobody has any trouble distinguishing a lion from a lobster, even though it may be a lot tougher to tell a lion skull from that of a tiger, or the antennae of an American lobster from those of its European cousin.

In view of all this, to regard the ancestor of a major crown group (whales, say, or cephalopods) as “unimportant” simply because it did not differ much in morphology from either its parent species, or from its sister species that went on to evolve into something else or to become extinct, is entirely to miss the point. The significance of an ancestor (which is a member of the higher taxon of which it is progenitor, even if only recognizable as such by virtue of a single synapomorphy), is above all historical. Evolutionary biology is a historical and genealogical science, and phylogenies are given structure only by those essentially historical events we see in retrospect as lineage splits. In this context, ancestral forms are intrinsically more significant for their historical roles, which govern their places in the great Tree of Life, than for the particular morphologies by which we recognize them. It is certainly realistic to conclude that in the early stages of the evolution of most higher taxa, associated apomorphies are far more likely to be relatively unostentatious than to represent “important” leaps. But if the historical independence of the lineage concerned has already been established, then the most fateful step of all will already have occurred. And, however modest, those apomorphies bear crucial testimony to this fact.

Cartmill helpfully offers us an alternative here. If higher taxa are not always unimportant, they may nonetheless be polyphyletic. Or, of course, they may be both. But while it is always good to have alternatives available, here we find the conceptual conflated with the operational. It is uncontestable that there are enormous practical difficulties involved in recognizing taxa at all levels, species as well as higher ranks. Newly evolved species may well be genetically leaky vessels with fuzzy boundaries, even though future developments may show them to have become historically individuated. What's more, even well-established species may be very tough to recognize with confidence on the basis solely of hard-tissue evidence. And while different in kind, the obstacles are just as great in recognizing higher taxa and their relationships. Most such difficulties arise from the widespread occurrence of homoplasy (that convergence again), a phenomenon that has turned out to be even more rampant in evolutionary histories than classical evolutionists like George Simpson had believed.4 As a result taxonomy, frequently derided as a dull clerical exercise, actually turns out in many cases to be a hugely challenging proposition. Still, despite the undeniable operational difficulties, the bottom line here is that if, as all seem to agree, evolutionary histories are the products of lineage splits, then the higher taxa produced by nature are indeed historically and genealogically “real” and, at some point very close to their origin, new clades are already by definition monophyletic. The problem for working systematists is to get all those clades right. But systematists are well aware that, if a particular higher taxon can be shown to be polyphyletic, something is not as it should be. And they also know that in such cases the focus has to be on amending the genealogical tree to make all higher taxa monophyletic rather than on throwing out the baby with the bathwater.

The higher taxon of greatest interest to the egocentric species Homo sapiens is, of course, the family Hominidae (or subfamily Homininae, or tribe Hominini, or whatever—it's still “real”). By now, Hominidae has a very substantial fossil record. But despite the wealth of evidence available, Cartmill appears to regard the search for our family's origins as futile, and to view the possibility of finding any genealogical and historical structure within it as remote indeed. This seems to be the result of taking a valid observation—that it isn't going to be easy to figure out exactly what was going on—to an illogical extreme. Certainly, it does appear highly likely that several late Miocene hominoid lineages independently experimented with upright posture, and that this will have clouded the picture of hominid emergence. But one hopes that many paleoanthropologists will nonetheless doubt the proposition that such complexity precludes seeking the origins of our family in a lineage-split at around this time. Similarly, the fact that rapid brain enlargement evidently occurred separately in multiple lineages of Pleistocene Homo should not compel us simply to throw up our hands in horror in the face of massive homoplasy. In fact, since this unusual and crucial phenomenon didn't happen in any other group, these parallel and independent trends are clearly telling us something very important about the intrinsic nature of our genus Homo. And there is no question that we will have to know what that something was before we are able fully to understand exactly what underwrote the emergence of our species Homo sapiens, an issue that one hopes most of us consider significant, or at least nontrivial.

When Linnaeus wrote nosce te ipsum in place of his customary morphological diagnosis while naming the species Homo sapiens back in the mid-eighteenth century, it was intuitively obvious what “human” meant. There was nothing closely comparable and, given the murky nature of a species that had already for long been the sole hominid on earth, Linnaeus was probably wise to prevaricate on the issue of diagnosis. But after a century and a half of paleoanthropological investigation and the discovery of a multitude of extinct relatives much closer to us than the apes of which Linnaeus knew so little, questions such as “Who are we?” and “How long ago did the first humans live?” have become infinitely more complex, as well as more compelling. And those who ask them deserve better than to be answered with another question.

One final reflection. Homo sapiens is a storytelling species, and from our parochial perspective the human evolutionary drama is literally the greatest story ever told. But you'll never understand the story if you don't know exactly who the protagonists were: all you will ever have is a host of disembodied observations and events to be endlessly shuffled around. Just as without the actors there is no play, without systematics there is no evolutionary story. As a result, rejecting systematics as trivial or unworkable means admitting, or even insisting, that we will never properly understand who we are and how we got here. Rather sad, really.


  1. Top of page
  2. Acknowledgements

I thank John Fleagle for the opportunity to offer this comment, and Matt Cartmill for many years of friendly discussion.


  1. Top of page
  2. Acknowledgements
  • 1
    Cartmill M. 2012. Primate origins, human origins, and the end of higher taxa. Evol Anthropol 00: 0000.
  • 2
    Cartmill M. 1981. Hypothesis testing and phylogenetic reconstruction. J Zool Syst Evol Res 19: 7396.
  • 3
    Cartmill M. 1994. A critique of homology as a morphological concept. Am J Phys Anthropol 94: 115123.
  • 4
    Conway Morris S. 2004. Life's solution: inevitable humans in a lonely universe. Cambridge: Cambridge University Press.
  • 5
    Yoon CS. 2009. Naming nature: the clash between instinct and science. New York: W. W. Norton.
  • 6
    Tattersall I. 2011. Paleontology: a brief history of life. Consohocken: Templeton Press.