Evolving Thoughts

Basic concepts: Ancestors

Some ideas one might think are pretty clear. The notion of an ancestor is one of them. But I am astounded how few people understand this simple idea in the context of evolution.


The basis for evolutionary thinking is the notion of an evolutionary tree, or a historical genealogy of species. It looks somewhat like the diagram in the header, which is a rendering of the first evolutionary tree from Darwin’s Notebooks. One species is the ancestor of another if it is lower in the tree diagram.

That seems simple enough, right? Well ancestry has a few wrinkles.

The first wrinkle is that an ancestral species of a modern, or extant, species might be extinct. Moreover, this means that if it is extinct, no similar modern species counts as an ancestor. One common misconception is that because our common ancestor with, say, lemurs looked like a mouse lemur, the modern mouse lemur is our ancestor. It is not. No matter how much a modern (or extinct!) species looks like another species, no matter how much resemblance there is, it is not the same species. Darwin was explicit about this:

We can clearly understand why a species when once lost should never reappear, even if the very same conditions of life, organic and inorganic, should recur. For though the offspring of one species might be adapted (and no doubt this has occurred in innumerable instances) to fill the exact place of another species in the economy of nature, and thus supplant it; yet the two forms-the old and the new-would not be identically the same; for both would almost certainly inherit different characters from their distinct progenitors. [Origin of Species, 1st edition, page 315]

The idea that species are just groups of organisms that resemble each other was never a widely held view among naturalists before or after Darwin. Lamarck, and those who followed him, thought that all evolving lineages went through similar grades, and so that every existing species was “the same” as some extinct one, but that was never the view held by evolutionary biologists since Darwin. Ancestors were themselves the result of a unique historical process, and no more than Rome can reappear as the same empire it was in 64BCE can a species re-evolve.

A second problem with ancestors is identification. Often, both in the press and in some scientific texts, if something “looks like” it was an ancestor, then it was. This occurs most commonly in human evolutionary scenarios. Was Homo erectus our ancestor or was Homo ergaster? We do not know for a simple reason – we may not have yet found our ancestral species that immediately preceded Homo sapiens. In fact, we may never find it, as we haven’t found the ancestors of modern chimps, gorillas or orangs (though some think we have protogorillas). Since the absence of data is not data itself, we can only assess the probabilities – and this is not as easy as the rhetoric implies it is. It is likely that ergaster is an ancestor of Neandertals, but not of our present species. That is the best we can do. Possibly, given how difficult it is to identify a single species from paleontological remains, ergaster is not even a real species. It might be part of a much wider species that included erectus, and we might be part of that species as well (with some modifications that don’t inhibit interbreeding if only there were some individuals around to experiment on). Or it might be that all the individuals that are called erectus or ergaster are in fact members of many smaller species, but the evidence of that is not fossilised and we’ll never know for sure.

“Cryptic” species, which resemble each other so much even experts can’t tell them apart unless they are observed not to interbreed, are common. “Polytypic” species, where the range of morphological variety is so great, as they are in domestic dogs, that individuals would almost vertainly be classed as different species if not for the knowledge that they do interbreed, are also common. So working out what is a species, let alone what is an ancestor, is very difficult and ultimately, I think, impossible to know for sure.

Even in a famous case like Darwin’s finches on the Galápagos Islands, which are thought to have come from a species still extant in South America, we cannot identify the parent species because it might have been one of five or six, and the evidence doesn’t narrow it down. We know it was one of them (or maybe a closely related but now extinct and unpreserved species), but finding the culprit is harder than proving O.J.’s guilt to a jury. All kinds of likelihoods, but no knockdown evidence.

A third confusion is the notion of “the” ancestor. Every species alive has literally thousands, if not millions, of ancestral species going back to the evolution of the major group – Animals or Plants or Fungi, to name the three larger eukaryotic kingdoms. Each ancestor forms a node on a phylogenetic tree, and there are a multitude of these branching points. Basically, an “ancestor species” is a species that gave rise to at least two modern or later species.

Some useful terms – the cenancestor is the last ancestor shared by all life. The concestor, a term introduced in Richard Dawkins’ recent The Ancestor’s Tale: A Pilgrimage to the Dawn of Evolution, is the last species that is in the lineage of any two modern organisms. So, the concestor of humans and chimps lived about 5 or 6 million years ago, while the concestor of us and, say, brewer’s yeast, lived around one and a half billion years ago or so. The concestor of all life, the cenancestor, lived around 3.85 billion years ago.

A fourth confusion is the notion of “more evolved” species. As demonstrated by the recent media meme of chimps being “more evolved” than us, no species is more or less evolved except in some specific sense, like (as in the chimp case) the number of selected genes over a period. All species at a given time have exactly the same evolutionary duration, and on average, probably the same number of ancestral species, as their nearest relatives. And “primitive”, as I have noted before, doesn’t mean more complex or “more perfect”. It simply means that the ancestor has fewer of the changed traits of its descendants, and more of the general traits of the group as a whole.

So, I hope this brings some clarity to the notion of an ancestor. We all have them, but we don’t all know who or what they are, or talk about them over dinner.


  1. #1 Karl
    April 19, 2007

    I am going to be picky here to see if I understand what you mean. You say: “All species at a given time …, and on average, probably the same number of ancestral species, as their nearest relatives.”
    Isn’t this tautological? The nearest relative is the one on the other side of the last branch. So there is just one step back to the concestor. And from there back to the cenancestor they have, trivially, identical paths. So of course they have the same number of ancestral species.
    What am I misunderstanding?

  2. #2 Matt
    April 19, 2007

    John, I like your ideas here- but tell me what you think about the difference between the concept of relatedness and ancestry. Both seem to measure “identity by descent”, but operating on different time scales. You can talk about relatedness between individuals of the same species (or population), and ancestry between 2 different species. Also, relatedness can be defined within an age cohort, but ancestry implies one came before another.

    OK, what I am getting at is is it incorrect to think about ancestry as a measure of genes identical by descent? After all, this is roughly speaking what we are talking about when addressing synapomorphies..

  3. #3 Matt
    April 19, 2007

    Karl, I think you are missing an important verb..

    “…and on average, probably [share, have, ???] the same number of ancestral species, as their nearest relatives.”

  4. #4 John Wilkins
    April 19, 2007

    See? I was right! Nobody understands it…

    Oh, well, maybe nobody understands me. Yeah, that’d be it…

    “Nearest relatives” doesn’t just mean your immediate sister species. It can mean a very large clade indeed; say, vertebrates, in the context of animals or eukaryotes.

    On the number of genes, I think this is a useful diagnostic tool, but not a definition of the underlying facts. Sure, overall gene change is roughly correlated to evolutionary events or distances, but there’s no reason to think that each speciation event has a standard number of gene changes, since speciation in sexual organisms also relies a lot on adaptive specialisation or not, ecological conditions, repair mechanisms, pathogens, and so on.

    Genes are just another form of morphology used to diagnose species (albeit a relatively discrete form, and one with masses of commensurate characters), and so the basic problems of identifying species with the amount of morphological change that held in the pre-DNA sequencing days still hold. Character identity by descent is indeed the basis for constructing phylogenies, but it doesn’t give us the basic nature of the nodal taxa – i.e., species.

  5. #5 Karl
    April 19, 2007

    I think you’re right. but maybe your last comment helped. Let me try again. Vertebrates includes humans, monkeys, bears, dogs, frogs, etc. So that’s where “average” comes in and makes it non-tautological. All of these species “have…on average, probably the same number of ancestral species”.
    Is that it?

  6. #6 John Wilkins
    April 19, 2007

    That’s my belief, yes. All members of a single clade will have, on average, the same number of ancestors. There will be massive variation, of course, and those that have three litters per year are more likely to have more species in their ancestry than a lineage that has one every forty years, but by and large it doesn’t follow that there is much difference over the entire clade.

  7. #7 El PaleoFreak
    April 19, 2007

    “All species at a given time have exactly the same evolutionary duration (…)”

    Yes, but evolution is change, not duration :o)

    “and on average, probably the same number of ancestral species, as their nearest relatives.”

    On average, maybe. But great differences in number of ancestral species are possible, as well as differences on morphological and genetic evolutionary change.

  8. #8 Tom Scott
    April 19, 2007

    Might not have engaged my brain properly this morning – but I’m not getting the significance of your hypothesis that all species within a clade have, on average, the same same number of ancestral species.

    I’m understanding your point as: If you average the number of ancestral species within a clade then the species within the clade will have the same average number of ancestral species (even if there is a wide deviation around that mean); which can’t be what you mean!

    Individual species do have different numbers of ancestral species – and different lines do have different rates of both speciation and number of genes changed e.g. chimps and humans. (Whether or not any given set of changes is deemed to result in a new species).

    In other words different species have ‘got to where they are’ via quicker or shorter routes i.e. via ‘more’ or ‘less’ evolutionary steps… no? I’m not of course implying that more or less implies that the end point is better or worse, more advanced or primitive, just that there were more steps, more changes to get there.

  9. #9 Karl
    April 19, 2007

    John, come back.
    You seem to have some of us confused still or, maybe, even more.
    Are you saying that although the number of ancestors for different species within a clade may vary, the differences from the average for that clade is significantly less than the difference between the averages for different clades?
    Can you give some, very approximate, numbers?
    And then, I have the same question that Tom Scott asked: what is the significance of this observation?

  10. #10 Spaulding
    April 19, 2007

    These sorts of discussions are difficult to reconcile with the common perception of species as discrete, segmented units. Species aren’t such a black-and-white thing, as your article hints. When you think of ancestry as something like a temporal equivalent of ring species, it makes more sense, and shows that the semantics of species divisions are sometimes very artificial.

  11. #11 John Wilkins
    April 19, 2007

    Karl, I will do a separate post on this, soon.

    Spaulding: Whether or not species are deliminable in a given time horizon (and nearly all the time they are) has very little to do with the number of speciation events in a given species’ history.

  12. #12 Scholar
    April 20, 2007

    For some reason, I don’t think those (monkeys) are the kind of ancestors my grandmother was talking about. Did they even have Monkeys way back when our ancestors sailed the Mayflower across the ocean?

  13. #13 John Wilkins
    April 21, 2007

    Sometimes you just have to accept that your ancestors weren’t chosen but just happened, no matter how embarrassing it might be.

New comments have been disabled.