Friday, August 9, 2013

Tokogenetic relationships

During a discussion about species, a visiting colleague has argued that I am misusing the word "tokogenetic" (see here, here and basically everywhere I discuss phylogenetic systematics). Apparently the way Hennig originally intended the concept of a tokogenetic relationship to be used was not as a synonym for "net-like" but instead to describe the ancestor-descendant relationship, as from mother to daughter across generations.

If that is so, I will from now on try to avoid making that mistake. That being said, my previous misuse, both on this blog and in research papers, does not appear all too problematic because nearly everybody else seems to use the term in the same way, as a Google search easily demonstrates.

The important point, however, is that this insight does not change anything that matters: Regardless of whether they can or cannot be called tokogenetic in the original sense of the word, networks, or net-like structures, definitely do not contain monophyly or paraphyly. Only tree-like, phylogenetic structures can contain monophyly or paraphyly.

This is why anybody arguing that reticulation makes it necessary to accept paraphyletic taxa only demonstrates their own conceptual confusion. As do those arguing that the members of the same sexually reproducing species form a monophyletic group or clade, of course.


  1. I don't understand your last sentence. Probably I have missed a previous explaination of your view.

  2. As I understand the situation, there are two issues, one definitional and one epistemological:

    1. The terms monophyletic or clade only apply to phylogenetic (tree-like) structures, and sexually reproducing individuals do not form a tree-like structure but instead a network (we have more than one ancestor in every generation before us, to say the least).

    2. The justification to consider something to be a clade is that all members of it share certain characters (synapomorphies), or at a minimum there has to be some character evidence, today usually molecular, that can be used to infer their relatedness. That works in a phylogeny, but the problem is that if we are in a net-work of sexually reproducing individuals then there is no such justification. Because of recombination, it is possible that I have not inherited any genetic material whatsoever from my maternal grandfather, so there would be no evidence for my descent from him. Unlikely in this specific case, yes, but over many generations situations like this become guaranteed.