I thought that according to cladistics, only monophyletic groups were valid as species or any other taxa. But extinct species, like say H Erectus must be paraphyletic, as the common ancestor of H Erectus is also the common ancestor of modern man. So if species are frequently paraphyletic, how can cladistics be tenable?

Edit. Also, I'm new to this forum. The people responding here, I'm assuming are actually biologists with a specific knowledge of cladistics, right? I'm a post-grad (philosophy of biology rather than biology per se) so I'm not really looking for links which explain what cladistics is, and certainly not wikipedia. I can access most journals if there is a paper on this subject. I can't find anyone in our dept who can give me a steer on this. I realise thhis is a slightly unorthodox research method, but I'm at a bit of a dead end here!

  • $\begingroup$ Right, thanks. But I don't see that the species concept is the issue here. I'm really questioning whether cladistics is coherent, given that it prioritises clades (which are by definition monophyletic) as being the only valid taxa. Now, as far as I understand, it doesn't really matter which species concept we use. Cladistics doesn't seem to rely on any once species concept (Hennig's Tokogenetic concept aside)but does absolutely insist on monophyly. The vast majority of species, though, over evolutionary history, are paraphyletic, right? Just want to check I'm not making some stupid error here. $\endgroup$ Dec 16, 2016 at 14:33
  • $\begingroup$ For your example; I don't think it is obvious that H. erectus is paraphyletic. The question is if H. erectus is the direct ancestor of modern man (if so H. erectus would be paraphyletic), or if H. erectus and H. sapiens share a common ancestor (which would make them related, but none of them would have to be paraphyletic). Extinct species don't have to be paraphyletic, as you seem to claim. I'm not an expert in this specific case though (human systematics), and this is merely a conceptual point. $\endgroup$ Dec 19, 2016 at 10:19

4 Answers 4


You say that "only monophyletic groups are valid as species"... That is according to a certain interpretation of evolution that is not necessarily mainstream, (you wouldn't be taught that definition of species in Oxford or Yale) and which idealizes cladistics into simple trees, which don't reflect reality of hybridizing species like orchids and hominids, where simple tree like branching has been proven to not exist. Inflexible definitions of phylogeny should be taken with a pinch of salt, because pollen and other reproductive systems cause many species to be flexible. It's unwise to state that definition of species without being well aware of the mainstream one: Ernst Mayr definition.

Strict interpretations of species lead to problems, because Darwin wrote: No one definition of species has satisfied all naturalists; yet every naturalist knows vaguely what he means when he speaks of a species. Generally the term includes the unknown element of a distinct act of creation.

If you check the Wiki page on species, they state 11 contrasted definitions of species which are not necessarily cross compatible. https://en.wikipedia.org/wiki/Species#Standard_definition:_Mayr.27s_Biological_Species_Concept

The definition of species that i was taught had nothing to do with the parental mechanisms that shaped the species differentiation from other species, and that it is a practical concensus to effectively group animals methodically, which doesnt actually reflect natures unmethodical reality.

Seeing as humans interbred with neanderthal and denisovan species, and orchids interbreed very commonly, the following image refers to an idealized phylogenetic model which is not applicable to all reproductive systems, and it isn't applicable to humans.

  • $\begingroup$ I think it's pretty mainstream to say that species have to be monophyletic, but not all monophyletic groups have to be species. In fact, I would even argue that Mayr's species concept is pretty close to that notion as 'reproductively isolated groups' and a monophyletic species (taxon) are not that different (of course without hybridisation across cladistic barriers). Additionally, you should avoid using terms like 'creation', especially when assuming that most natural scientists include this as an 'unknown element' to their idea about species. -1 $\endgroup$ Dec 16, 2016 at 13:14
  • $\begingroup$ Thanks. But it seems, then, like the vast majority of species are actually paraphyletic? Extamt species are monophyletic, and extinct species which are 'dead ends' as it were. But all other species are parphyletic. The species concept isn't the issue here.My qusetio $\endgroup$ Dec 16, 2016 at 14:27
  • $\begingroup$ Sorry. Can't delete the previous comment which I posted by mistake....Yes. That's what I thought. It's just orthodoxy that species must be clades. And that higher taxa must also be clades. I'm pretty sure they would teach me that at Oxford. $\endgroup$ Dec 16, 2016 at 14:38
  • $\begingroup$ Indeed, a research professor who has been very busily involved in naming new species (i.e. nematodes) and phylogeny would rave on and on about all the amazing funny facts about species diversity, like one researcher getting revenge on another researcher by calling a species "johnsonus sillius" and the russian classification system which has 8 times more species than us, and they go through all the reproductive intricacies that mean that clades, i.e. bear species, sometimes intertwine, whereas some species have very specialized reproductive organs with barbs and where clades are better defined. $\endgroup$ Dec 19, 2016 at 14:58
  • $\begingroup$ In short i suggest that the rigorous phylogeny terminology that you quote is useful for genetic measurements and in depth phylogeny study. It's recommended to first learn at least 3-4 alternate definitions of species, and first do case studies of dozens of different speciation routes, physiologies, read about actual animals and surprises and conflicts in research based on certain evolution theories, to make it more of a fun topic that logic of branching terminology. By actually reading topical zoology and botany case studies, i think that someone gains a lot of useful background knowledge. $\endgroup$ Dec 19, 2016 at 15:04

You have to understand that a fossil "species" is very different than the common usage. A species is not really a real thing. Paleontology (and other fields) accepts the vague nature of the classification of species. All life that has ever existed on earth is really one huge messy ring species on the long scale once time is included. phylogeny is a way of imposing a classification system that helps us understand life, any such classification will have arbitrary cut off points becasue WE use them to help understand relationships on the large scale. Species is one such arbitrary distinction, it is convenient and works on the large scale with living organisms just fine, but breaks down on the finer scale or once deep time is included.

For fossils you group by noticeable/measurable differences, while trying to make allowances for differences that should occur within a population (like sexual dimorphism). Paleontologists accept that a "species" is just an approximation, which is why the finer the scale of the cladistic analysis, you rely on individual specimens and not grouping them into a species. Ideally all cladistics would be done this way but it is completely impractical, in extant life becasue of huge population numbers in fossils becasue of incomplete specimens causing missing variables. On an analysis using any group as taxa, including species, it's fine if they are paraphyletic becasue they are only approximations.

You can find a more in depth exploration of cladistics here enter link description here

And a general review of cladistics here enter link description here

  • $\begingroup$ Right. Which is why I'm asking, really. Hennig (and many others since, such as Wiley and Lieberman, Baum and Smith, Ridley, Tudge, Ghiselen etc etc) explicitly claim that Claude based species are the natural (as opposed to merely nominal) kinds of biology. So, as a philosopher, it is of interest to me that this isn't really coherent, and as a philosopher of science it is of interest that this has occurred. It seems to me that it could work if we defined monophyletism synchronically, and allow groups which were at one point monophyletic, but as far as I know, no one does this. $\endgroup$ Dec 17, 2016 at 0:54
  • $\begingroup$ Ah I see a bit more of what your asking. You might want to change your question to reflect this. Part of the difference is how a clade is defined, in Paleontology the most common usage is: X plus Y and there most recent common ancestor and all of its descendants , becasue the assumption is you will never find a direct ancestor, or at least if you do you will never be able to demonstrate it. species and clade are not interchangeable. and by their own definition all defined clade are monophyletic. $\endgroup$
    – John
    Dec 17, 2016 at 2:23
  • $\begingroup$ Species which are defined as a clade have to be monophyletic but that way of defining a species are not commonly used. Mostly because while very accurate is is also largely useless in practice. $\endgroup$
    – John
    Dec 17, 2016 at 2:31

The boring but probably most correct answer is: it depends on the species concept that you accept.

Basic evolutionary theory predicts that paraphyletic species (as defined by a variety of concepts) should be generated under a wide range of processes.

The vast majority of species is described from gross morphological states ("morphospecies"). They can be wholly paraphyletic, as can species under nearly all concepts, except those that insist on monophyly (e.g. phylogenetic species concept of Cracraft 1989, deQueiroz and Donoghue 1990).

A problem to consider is one of the source of data. For example, a species can be monophyletic at a particular gene, but not at great many others. Broadly speaking, this is known as the "gene tree vs. species tree" problem. A particular set of relationships can be recovered from analyses of a single genetic region, drawn from a wide statistical distribution of (for example) genealogies. Ideally we would have the genetic sequence within region(s) that cause(s) populations and species to be differentiated.

Depending on population size, after some amount of time (potentially very large), all genes are expected to be monophyletic in each descendant species, a condition that is then called "reciprocal monophyly."

Separately, you can take a look at https://en.wikipedia.org/wiki/Paraspecies for a straightforward treatment. As for your specific example, here is a great treatment of the subject: http://evolution.berkeley.edu/evolibrary/news/100501_xwoman


Can species be paraphyletic?

Yes, as you have explained in your post.

So if species are frequently paraphyletic, how can cladistics be tenable?

The grouping of lineages into species is necessarily somehow arbitrary. However, the relationship between lineage is free from any subjectivity. The reconciliation you need to go through is to accept that the concept of species, despite being so much used, is of no use to phylogenetic (or only to facilitate naming).

Cladistic and phylogentics is tenable because it does not depend on the concept of species which itself is arbitrary. You can learn more about the semantic of the concept of species in How could humans have interbred with Neanderthals if we're a different species?

  • $\begingroup$ Right. So if species are frequently paraphyletic, why is it regarded as unacceptable for higher taxa (which are usually seen as more arbitrary) to be paraphyletic? As in the idea that reptilia are not a legitimate grouping. How can the fact that even species are not by nature monophyletic be reconciled with cladistics claim that only monophyletic groups are legitimate? Many thanks, by the way. $\endgroup$ Dec 17, 2016 at 1:01
  • $\begingroup$ Probably because as you go to the detail, it becomes harder (impossible?) to really name monophyletic groups only. This is even more problematic as we would ought to rename groups at every small discovery about the ancestry of a lineage. $\endgroup$
    – Remi.b
    Dec 17, 2016 at 5:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .