Extant taxa cannot be basal

Page 74 from Notebook B

Page 74 from Darwin’s Notebook B: Transmutation.

Shortly after the return of the HMS Beagle to Cornwall, Charles Darwin jotted down the following note:

It is absurd to talk of one animal being higher than another.

In the subsequent twenty years, Darwin either forgot or reconsidered his note to himself.* He shouldn’t have.

The Origin of Species is full of references to higher and lower animals (and sometimes plants), for example (p. 440):

The embryo in the course of development generally rises in organisation: I use this expression, though I am aware that it is hardly possible to define clearly what is meant by the organisation being higher or lower. But no one probably will dispute that the butterfly is higher than the caterpillar.

I dispute it. There is no scala naturae, no universal standard by which organisms are ranked. Biologists and philosophers of biology have generally rejected the idea these last 160 years, but just like Darwin, we keep coming back to it.

Although exceptions certainly exist, the scientific literature has mostly pushed out judgements of which organisms are ‘higher’ or ‘lower’, ‘advanced’ or ‘primitive’, or more or less evolved. But though we change the terminology, we never quite seem to internalize the deeper message that these comparisons are meaningless. We seem to have a deep need to make such judgements, and so we replace the old terms with new ones without much changing their meaning.

The current versions are ‘basal’ and ‘derived’, or occasionally ‘early’ and (I suppose) ‘late branching’. These terms have useful meanings in describing evolutionary relationships, but they are never useful in describing living, or extant, taxa.

Extant taxa can not be basal.

Yet they are often described that way, even in the peer-reviewed, professional literature. The following examples were not chosen because they are particularly egregious, nor because I mean to pick on anyone in particular. They are just the ones I’ve happened across recently (and I hope the authors will remember that if they’re ever on a hiring committee). But they are wrong, all of them.

From Current Biology:

Genomic sequences reveal that the basal lineages Porifera, Placozoa, Ctenophora, and Cnidaria share a large fraction of the genetic toolkits involved in cell fate, patterning, differentiation, and cell-cell communication in bilaterians.

From Nature Protocols:

Among the basal animals, the starlet sea anemone Nematostella vectensis (phylum Cnidaria) has emerged as a leading laboratory model organism…

From Current Opinion in Genetics & Development:

The homeobox and kinase gene families have been further analysed in basal animals

Not unless they’re sequencing Precambrian DNA. From Campbell Biology, undoubtedly the most common textbook for freshman biology classes:

Sponges are basal animals that lack true tissues.

It even shows up in the Wikipedia page for Volvox:

The cells of colonies in the more basal Euvolvox clade are interconnected by thin strands of cytoplasm, called protoplasmates.

So there’s part one of my argument: even the very current scientific literature contains numerous instances of extant taxa being called basal. This can’t be dismissed as a non-issue on the grounds that ‘nobody does that anyway’. These are just a few of many examples, as a quick Google Scholar search will confirm.

The part I haven’t covered yet is why it’s wrong. Basal, in the context of phylogenetics (the study of evolutionary relationships) means closer to the base of the tree. Closer to the base of the tree means further back in time. Living taxa are, by definition, maximally distant from the base of the tree; they can’t be basal to anything (except future taxa). Several published papers have explained this in greater detail than I plan to do, notably Krell & Cranston, Crisp & Cook, and the excellent “Tree thinking” paper by Omland, Cook, & Crisp (there’s also a nice blog post on the topic at For the Love of Trees).

One of the figures from that last paper is a nice visual representation of a problem with the way we often read phylogenetic trees. On a comb-like phylogeny, we tend to call the first species to diverge from the others basal. For example, in Figure 4A, we might call the marsupial Virginia Opossum basal to the placental mammals:

Figure 4A from Omland et al. 2008. Phylogeny of five mammal species focused on primates.

This is very often how people, even biologists, talk about mammals: extant monotremes and/or marsupials are basal to placental mammals. In some cases, this is accompanied by judgements that they are more primitive or even (worst of all) ancestral to the placentals. It’s true that the basal divergence in Figure 4A is between the opossum and everything else, but that is purely an artifact of taxonomic sampling, as Figure 4B dramatically shows:

Figure 4B from Omland et al. 2008. Another phylogeny of mammals including two of the same species, but here the focus of the tree is on opossums.

Both 4A and 4B accurately reflect evolutionary relationships. But if we consider the basal-most species that which first diverges from the others, 4A shows the opossum as basal to the placentals, while 4B shows humans as basal to the marsupials!

Neither is true. Extant taxa cannot be basal.

Humans (in 4B) diverge basally from the other species, but humans are no closer to the base of the tree than, say, mouse opossums. Humans are not more primitive than mouse opossums, and I’m nearly certain that humans are not ancestral to mouse opossums. Now let’s apply that same thinking to 4A:

Humans Virginia opossums (in 4B 4A) diverge basally from the other species, but humans Virginia opossums are no closer to the base of the tree than, say, mouse opossums ring-tailed lemurs. Humans Virginia opossums are not more primitive than mouse opossums ring-tailed lemurs, and I’m nearly certain that humans Virginia opossums are not ancestral to mouse opossums ring-tailed lemurs.

It’s understandable that biologists keep coming back to calling extant species basal, even though we’ve been told time and again that this is wrong. After all, we need some way to verbally describe the evolutionary relationships that a phylogenetic tree represents. The right way to do this is using cladistic terminology, that is, describing the tree topology in terms of sister group relationships. Sister groups (sister species, sister clades, etc.) are each other’s closest relatives, at least among the taxa represented in the tree. Each node, or branching point, in such a tree leads to two sister groups. For example, in 4B, the most basal divergence is between humans and marsupials. Humans and marsupials are sister groups (again, among the sampled species). Virginia opossums are sister to southern opossums, mouse opossums are sister to southern opossums plus Virginia opossums, and red kangaroos are sister to all the opossums.

It may seem a bit unwieldy at first, but it’s not hard to get used to. The last two sentences of the previous paragraph fully describe the tree in 4B; it contains no other information. Thinking of living species as ‘basal’, ‘primitive’, or even ‘early branching’ is not just wrong, it reinforces habits of thinking that biologists should have left behind before starting graduate school. Modern sponges are not basal animals, sea anemones are not primitive animals, and duck-billed platypuses are not ancestral to placental mammals. Cladistic terminology not only avoids these pitfalls, it also has the advantage of being right, where ‘right’ means accurately reflecting evolutionary relationships. That, after all, is what we care about.


*Or maybe he couldn’t read his own handwriting. I mean, damn, it’s worse than mine!


Stable links:

Brooke NM, Holland PWH. 2003. The evolution of multicellularity and early animal genomes. Curr. Opin. Genet. Dev. 13:599–603. doi: 10.1016/j.gde.2003.09.002

Darwin CR. 1837. Notebook B: [Transmutation of species (1837–1838)]. CUL–DAR121. Darwin Online, http://darwin-online.org.uk/.

Campbell NA, Reece JB, Urry LA, Cain ML, Wasserman SA, Minorsky PV, & Jackson RB. 2009. Biology, Eighth Edition. San Francisco, Benjamin Cummings.

Crisp MD, Cook LG. 2005. Do early branching lineages signify ancestral traits? Trends Ecol. Evol. 20:122–8. doi: 10.1016/j.tree.2004.11.010

Krell F-T, Cranston PS. 2004. Which side of the tree is more basal ? Syst. Entomol. 29:279–81. doi: 10.1111/j.0307-6970.2004.00262.x

Omland KE, Cook LG, Crisp MD. 2008. Tree thinking for all biology: the problem with reading phylogenies as ladders of progress. BioEssays 30:854–67. doi: 10.1002/bies.20794

Smith CL, Varoqueaux F, Kittelmann M, Azzam RN, Cooper B, Winters CA, Eitel M, Fasshauer D, Reese TS. 2014. Novel cell types, neurosecretory cells, and body plan of the early-diverging metazoan Trichoplax adhaerens. Curr. Biol. 24:1565–72. doi: 10.1016/j.cub.2014.05.046

Wolenski FS, Layden MJ, Martindale MQ, Gilmore TD, Finnerty JR. 2013. Characterizing the spatiotemporal expression of RNAs and proteins in the starlet sea anemone, Nematostella vectensis. Nat. Protoc. 8:900–15. doi: 10.1038/nprot.2013.014


  1. EnlightenmentLiberal says

    The part I haven’t covered yet is why it’s wrong. Basal, in the context of phylogenetics (the study of evolutionary relationships) means closer to the base of the tree. Closer to the base of the tree means further back in time.

    Question from a lay-person:

    Could they be using “basal” in the sense that “the modern population is practically identical to the ancestor basal population in terms of genetics” ? Or perhaps more crudely “the modern population is substantially similar to the ancestor basal population” ? I wonder if this is their actual motivation. I believe that this use could be defensible. It does not run automatically run afoul of the fault that you wish to point out, but it seems that such usage of language is sloppy and can lead to the faulty thinking that you point out.

    Further, I do not know if in this cases, or at large, whether we have enough evidence to even make such a determination. I suppose we could do so based on morphological comparisons to fossils.

    Do I have the jist of it? Do I have any misunderstandings?

    • Matthew Herron says

      I suppose that might be defensible if they’ve done ancestral character state reconstructions and found that the descendent is genetically similar to the inferred ancestor, but population genetics tells us that that’s unlikely to ever be true. I think you’re right that we would rarely have the evidence to make such a comparison.

  2. paxoll says

    Yes basal is a poor and wrong way to talk about extant taxa scientifically. Yes, biologists know this and still use the word even in apparently scientific journals (unless those examples are opinion pieces, oh and cladistically speaking basal is an appropriate descriptive term as in the wiki example). So the question is does this sloppy or wrong way to talk about it lead to actual inappropriate thinking, leading to inappropriate hypothesis and conclusions when analyzing data. I will say without actual evidence but personal experience that it does hurt the unscientific understanding of evolution but in that case I don’t see that using a complicated cladistic terminology would help either as it would simply make the subject more confusing to the lay person. In scientific literature, accuracy in language needs to balance with understanding. I think that colloquial terms used to facilitate understanding is perfectly reasonable.

    It reminds me of a recent conversation I had about climate change and ice ages. Colloquially as used by the average person (especially disney) an Ice age is a period of expanding glaciation, and temperature variations that mark the difference between expanding and contracting glaciation is typically fast over a geological time period. Glaciers have been retreating for 10k+ years, so colloquially speaking the last ice age ended at the time period. Scientifically speaking though an ice age is any time there is a glacier on the earth, since we still have glaciers we are scientifically speaking still in an ice age and have been for 2.58 million years. What this did was lead to a complete misunderstanding and a complete waste of time talking past each other, it also had on my “scientifically wrong” side absolutely zero effect on my conclusions based on the scientific data. So similarly, I don’t think this is a problem unless you can demonstrate that it leads to wrong hypothesis or conclusions in the science.

    • Matthew Herron says

      I think we’re mostly in agreement. I don’t really think the cladistic terminology is complicated or difficult, though, and I think it’s worth the effort to avoid giving the impression that we are descended from echidnas.


Leave a Reply