Skip to content

Objecting to the prevailing tide

My friend and costudent of Gareth Nelson, Malte Ebach, persists in his campaign to bring some clarity to evolutionary and systematics thinking. I may not always agree with him, but he’s often clearer on these matters than his targets, so I refer you to his recent post on paraphyletic thinking. All I will do here is make it a bit more accessible to nonspecialists.

The initial concern is over a recent edition on transitional fossils of the journal Evolution: Education and Outreach, which is published online and open access by Springer. Here, we have the view expressed, apparently, by Donald Prothero (I haven’t read the articles in question, due to other obligations, so this is a general discussion, not a specific one), that we have intermediate forms between the rest of the okapis and the giraffe. Therefore creationists are wrong. Ebach objects to this for reasons that are a bit subtle to nonspecialists. So let me comment on his comment.

He says

We are not, never have been and never intend to be creationists. We do not defend creationist dogma or any other idea disguised as creationism (i.e., Intelligent Design). We do, however, point out where evolutionary biologists make unwitting mistakes, which may fuel further creationist attacks. Creationists are getting cleverer and are starting to see through arguments supporting paraphyly (i.e., ancestors, transitional fossils, transformations etc.). We warn our fellow evolutionist colleagues not to fall into this trap. We also urge them never to intellectualize the arguments of creationists by entertaining them at any level (i.e., defending paraphyly). Doing so makes matters worse…

The “we” here are a group of taxonomists often dismissed as “pattern cladists”, who make the unchallenged claim that a cladogram is not, without further a history. Although early critics of these taxonomists, sometimes another group of taxonomists who were called cladists (“process” cladists) and sometimes a group known as “evolutionary systematists”, attacked the “pattern” cladists as being creationists or sympathisers, in fact the general claim they made was this:

All that a cladogram does is summarise relationships between taxa. These relationships could have been caused by an indefinitely large number of possible histories. Hence, we cannot use cladograms to identify ancestors. For technical reasons, ancestral taxa appear in a cladogram as a neighbouring branch, or as a “sister taxon”. But it is very likely that we simply do not have any evidence of the ancestral taxon, but only of the general group in which the ancestor species must have existed. So identifying ancestors through taxonomy (by drawing “the tree of life”) is a serious conceptual error.

If you doubt this, note that for a group with a mere 10 taxa, the number of cladograms, just using that number of taxa, is 34,459,425. As Ebach and his colleague David Williams, show in their recent book Foundations of Systematics and Biogeography, even with only three taxa in a cladogram, six phylogenetic trees, which is to say possible histories, can be drawn, and the number goes up superexponentially as you add taxa.

So “transitionals” are merely taxa that have some traits that are intermediate between two other groups. They are not ancestors. I rather doubt that Prothero intended them to be read that way, but of course imprecision leads to chinks in the armor. Likewise, the recent fossil over which so much fuss was made is not proof of evolution, or our ancestry, or whatever. That is proven by the fact that we can group taxa in such a way that the best explanation for it is common ancestry. Creationists ask for impossible proofs based on a probably deliberate misunderstanding of evolutionary theory, so as Ebach says, why bother?

But another attack he makes is on paraphyletic groups. This is a little more technical, so bear with me. I mentioned above a group of systematists who travel under the misnomer of “evolutionary”. They agree that some genealogy (that is, phylogeny or tree of life) is needed to do taxonomy, but that one can also use overall similarity of organisms. So even if birds evolved from dinosaurs, they are enough unlike dinosaurs that we can place both in distinct groups, because birds are “more derived”. On the cladistic view (both kinds) this is simply a mistake. Cladism treats a “natural” taxon as one that cuts the phylogeny in a single place. What is left over after you do that is not a natural group, but a “paraphyletic” group. It is, in effect, the trash left over after you take out the groups that interest you.

The reason why paraphyly is not natural is that the criteria used to excise the “interesting” group are subjective. That is, they tell you more about what interests the systematist* than about the organisms themselves. If things evolved, then only monophyletic groups are informative about the organisms (which is to say, they were evolved from a single ancestor, so the similarities they present are homologies, or shared derived characters, and this allows you to make inferences of an inductive nature).

It’s actually very hard to maintain all this in mind clearly, which is why scientists so often retreat back to pre-phylogenetic (which to all intents is “pre-Darwinian”, although Darwin himself didn’t quite achieve cladistic purity) thinking. “Transitionals” is a term that can only be used if we actually observe the transition. There must have been some, but whether we can say with any likelihood that out of a very large number of possible evolutionary pathways, this one is true, is moot at best. At worst, history is lost, and we can only test evolutionary historical speculations (i.e., phylogenies) with the data and arrangements of taxa of the cladograms. Moreover, un-evolutionary “groups” formed by paraphyly are to be avoided as best we can. They are a return to the folk taxonomy of prescientific thinking.

* “Systematics” and “Taxonomy” are sometimes treated as being different tasks. But historically and today they are effectively interchangeable, and those who argue for a distinction are usually trying to peddle some line about the nature of “true” science and “true” classification.


  1. If things evolved, then only monophyletic groups are informative about the organisms (which is to say, they were evolved from a single ancestor, so the similarities they present are homologies, or shared derived characters, and this allows you to make inferences of an inductive nature).

    Don’t you mean only holophyletic groups are informative? Paraphletic groups are also monophyletic but have been excluded from that group perhaps because of very distinctive derived traits. (e.g humans from great apes)

    • Snowflake Snowflake

      No, because “holophyly” is monophyly, and paraphyletic groups are not monophyletic. It is exactly that claim that cladistic method rejects. There are as many ways to slice up paraphyletic groups as there are branches in the cladogram, so all the information one is getting from them is whatever it is that interested the taxonomist that was used to construct the group in the first place. Monophyly is real; paraphyly is a construct. Holophyly was Ashlock’s attempt to retroactively seize the terminology back from phylogenetics.

    • With all due respect to Suvrat, usage of the term “holophyly” is a good litmus test that someone is still stuck in the 1970s as far as phylogenetic theory is concerned.

  2. Their post isn’t going to cause me to give up the phrase “transitional fossil” anytime soon, but it is good food for thought. If we are going to use the phrase we should define what we mean by it and not just assume people understand, especially if people not well-versed in evolutionary biology translate “transitional fossil” into “ancestral fossil.”

  3. Jeb Jeb

    Its rather concerning to note the way that such a non -theory, intellegent design effects the internal working of biology.

    It would be intresting to see a group of articles by biologists giving a personal account of how this looming spectre of unreason affects the way you work and think day to day.

    A personal assesment of the impact of these views on working life. A great amount of time which should be given to research must be being wasted in having to deal with these issues.

    The only positive I can see is that clearly you are having to reflect hard on what youre subject is and the definitions you use. Thats never a bad thing. But hardly the ideal set of conditions in which to undertake such work.

  4. Monophyletic: Having traits in common

    Paraphyletic: Not having traits in common

    The fewer the number of traits being considered, the larger the monophyletic group. The greater the number of traits being considered, the smaller the monophyletic group. Therefor all organisms with a nucleated cell are monophyletic for eukaryotic life, but only organisms with nucleated cells and a backbone are monophyletic for vertebrates.

    • Paraphyletic: Having primitive traits in common, that is, symplesiomorphies.

      The emphasis in the characters in common is important in order to distinguish between paraphyly and polyphyly. The latter has homoplastic traits in common.

  5. Lucas Lucas

    Regardin the amount of possible “trees” in a cladistic analysis, you should mention the principle of Occam’s razor also known as the Law of parsimony which cladistics use when they have 34,459,425 possibilities (of course, in this case, asumming that the 34,459,425 possibilities are methodologically and logically possible). I’m not defending cladism not even its results, but I can imagine someone with no theoretical background reading your post and saying – What have they been telling us all this years!!??-.
    By the way excellent blog, I always hang around here but never comment.

    PS: this could be a good start point for a discussion of Phyletic gradualism and Punctuated equilibrium


Comments are closed.