The Platypus Fallacy.

I see, with rather alarming frequency, a major fallacy creeping in to discussions of human evolutionary history and how one may infer details about it. Specifically, there is a tendency to examine the traits of one or a few non-human species and to draw conclusions about the origin of human traits purely from these observations. Recent examples include observing some chimps carrying nuts and drawing conclusions about human bipedalism, or watching monkeys flap their lips and linking this to the evolution of human speech.

The fallacy here is to assume that the non-human species is “primitive”, such that it can be used as a proxy for a distant human ancestor. This is a fallacy because “primitive” and “derived” refer to individual traits, not entire species, and because the comparison being made is between two modern species, not an ancestor and a descendant. Both lineages have, by definition, been evolving for exactly the same amount of time since diverging from a common ancestor.

I am going to call this the “Platypus Fallacy”, based on an example that I have used in my lectures. Consider the following quote, by one of the authors of the platypus genome sequence paper:

“The platypus is a very ancient offshoot of the mammal tree, so it was 166 million years ago that we last shared a common ancestor with platypuses, and that puts them somewhere between mammals and reptiles, because they still maintain quite a lot of reptilian characteristics that we’ve lost; for instance, they still lay eggs. So we can use them to trace the changes that have occurred as we went from being a reptile, to having fur to making milk to having live-born young.”

Now, if a platypus were asked to summarize the situation, he could just as well say this:

“The lineage of which humans are a part is a very ancient offshoot of our mammalian family tree, so it was 166 million years ago that we last shared a common ancestor with humans, and that puts them somewhere between mammals and reptiles, because they lack a lot of specialized characters that we have gained but the ancestral amniote also lacked; for instance, they have no electroreception, no bills, no webbed feet, and no venom. So we can use them to trace the changes that have occurred as we went from being a reptile, to having fur to making milk to having our specialized features.”

Of course, if a platypus did say such a thing, I hope his colleagues would encourage him to avoid committing the “Human Fallacy”.


28 comments to The Platypus Fallacy.

  • As a microbiologist I run into this sort of thinking (that bacteria/archaea/protists are “primitive”) a lot from people studying plants and animals — as if microbes somehow stopped evolving at the time of our common ancestor.

      (Quote)

  • h pinxteren

    I guess the platypus fallacy goes for almost all ‘precursor’ arguments, e.g. like goo-goo gorilla’s  is supposed to be a precursor of ‘motherese’ ref: American Journal of Primatology, DOI: 10.1002/ajp.22039 

      (Quote)

    • Every single one.

        (Quote)

      • h pinxteren

        You mean, primatology is platypus fallacies ‘all the way down’. And same is  true for evolutionary psychology?

          (Quote)

        • No, I mean any example of looking at a trait in one modern species and drawing conclusions about the evolution of a different modern species is a Platypus Fallacy.

            (Quote)

          • h pinxteren

            any example of looking at a trait in one modern species and drawing conclusions about the evolution of a different modern species is a Platypus Fallacy
            Raymond Tallis wrote a book about this: “Aping Mankind” (2011), or the idea that human beings can be understood essentially in biological terms: we are our brains, essentially ape brains (e.g our inner ape, 2006 F. de Waal).

              (Quote)

          • Amphiox

            It would not be a Platypus fallacy if you could find evidence from another source independently supporting the hypothesis that the trait in question really is primitive in that particular species.

            For example if it is a trait that humans have and a monkeys don’t, and every other primate species more or equally closely related to monkeys and humans also don’t have it, then we can infer with fairly high confidence that the trait is primitive in the monkeys and derived in the humans.

              (Quote)

          • It would not be a Platypus fallacy if you could find evidence from another source independently supporting the hypothesis that the trait in question really is primitive in that particular species. For example if it is a trait that humans have and a monkeys don’t, and every other primate species more or equally closely related to monkeys and humans also don’t have it, then we can infer with fairly high confidence that the trait is primitive in the monkeys and derived in the humans.
             
            Agreed, there’s nothing wrong with properly implemented phylogenetic comparative methods. However, that’s not what the fallacy refers to. 

              (Quote)

  • Good post — I like having a name for this fallacy.

    One note: venom might be a retention of an ancestral trait: http://www.app.pan.pl/archive/published/app51/app51-001.pdf

      (Quote)

  • Brian Axsmith

    I see your point, and I mostly agree. However, it is still true that the platypus lineage diverged before the loss of egg laying. So it is still relevant to understanding the transition from egg laying to more derived modes of reproductions seen in most other mammals. If you specify what characters you are interested and don’t just assume the platypus is frozen in time, it is still a valid approach.

      (Quote)

  • Playtpus still lays eggs, so we can study platypus to understand how placental development evolved.

    Chimpanzees still have fur, so we can study chimpanzees to understand how hairlessness evolved.

    Crocodiles still have four legs, so we can study crocodiles to understand how bird wings evolved.

      (Quote)

  • I understand your point, but I would suggest there are two different issues.  To me, one fallacy is assuming an entire species is primitive as you point out; a second is to assume that the traits being compared are homologous.  False assumption of homology is really what you’re talking about when you make the case above (your example of nut-carrying and lip-smacking–often these examples are not phylogenetic studies that discuss character polarity, but rather behavioral speculation that rely solely on similarity).  Afterall, when one does phylogenetic reconstruction using an extant out-group (for example, using a baboon to polarize a matrix of great-ape/human traits), one makes the assumption that the baboon traits are primitive.  In this sense, there is nothing wrong with treating the out-group as operationally primitive for the traits under study, since it is a necessary assumption of parsimony-based cladistics (Sober’s 1988 book, around page 220 goes over this pretty thoroughly).  Most behavioral primatologists don’t do phylogenetics, they just rely on similarity (i.e., false homology).  For example, I recall some hooplah over a discovery that chimps were sometimes residing in caves in West Africa; this was interpreted as ancient homologous evidence of “sheltering behavior” since humans also use caves.  However, lemurs also shelter in caves (as do many other taxa), so either either cave use is a deep homology, or–more likely–it has evolved numerous times independently. 

      (Quote)

    • Not exactly. “Lip smacking” is not being considered as a homologous trait in two species. It is a trait in one modern species, and is assumed to have been a trait in an ancestor of a different modern species and an intermediate between no lip smacking and speech.

      I am not saying anything about phylogenetic comparative methods — that’s unrelated to the fallacy here. The fallacy is to look at a trait in one species and to use that alone to infer something about the evolutionary origin of a different trait in a different species.

        (Quote)

  • Claudiu Bandea

    I posted the following comment at Sandwalk, and I thought of posting a copy here:
    Clearly, neither modern humans nor modern platypuses are “somewhere between mammals and reptiles“, as they are both on the mammalian family tree and equally distanced phylogenetically from the root of this tree. 
    However, things are more complex, and the Fallacy might not be as bad as it might seem if we asses phenotypic or genotypic characters from a different perspective. In his particular case, although both humans and platypuses have had the same period of time to diverge from their common ancestor, because of differential rate of mutation/selection, generation time, and population size, one of them might have more common phenotypic or genotypic features with the their ancestor (and possibly with the modern reptiles) than the other one. And, therefore, one of them, I’ll say in this particular case is the platypus, might better represent, or serve as model for the common ancestor.
    However, let me bring up an even stronger case, which I call “if evolution could be quantified“, which compares humans to modern bacteria:
    “if evolution could be quantified, based on population size, generation time, and rate of mutation/selection, bacteria are more ‘evolved’ than humans by an enormous factor”
    Whether they are or not, I’ll say again that modern bacteria have more in common with LUCA (the hypothetical common ancestor) than humans and, therefore, they are a better model for LUCA.

      (Quote)

  • Claudiu Bandea

    I maintain that using phenotypic or genotypic features, characters, or traits of one modern species in order to infer something about the evolutionary origin of hologous features in different species, might not be a Fallacy as bad as it might appear. However, using a trait alone to do this might be poor science any way you look at it.
    In your platypus vs human example, I think it is likely that the platypuses have more traits that better represent those of the common ancestor than humans do, and therefore we can learned more about human evolution studying platypuses (i.e. their traits) than vice versa. This brings us to your question:
    Which is more like the LUCA, a mollusc or a fern?
    The mollusks and ferns, which are both multicellular eukaryal organisms, might not be evolutionarily distant enough to reasonably choose which one is more like the LUCA.  Also, their rate of evolution might be too similar to make the point I was trying to make about “if evolution could be quantified”.  However, if we consider either of them, or both of them, versus bacteria or even vs protozoa, I think that bacterial and protozoan organisms might be better models for LUCA than mollusks or ferns.

      (Quote)

    • I’m afraid you are simply committing the fallacy here. Taking the platypus and human example, you are assuming that platypus is more like the ancestral mammal, and I suspect you are doing this because you are focusing on one or two traits (e.g., egg laying).  The ancestral mammal was not aquatic, it did not have a bill, it did not have electroreception, it was not toothless as an adult, it may or may not have had venom, etc. Choose any of those traits instead of egg-laying, and you would have the opposite beginning asumption.

        (Quote)

  • Claudiu Bandea

    I meant “homologous features”

      (Quote)

  • Claudiu Bandea

    I think your Platypus Fallacy was intended to have broad implications/applications, and I just wanted to make the case that that there might be some exceptions.
     
    In the case of platypus vs human example, I put my ‘money’ on Platypus as being a better model for the common ancestor than humans just based on intuition (and fun!), not on specific features, so no fallacies committed here.  However, if you think that humans might have more common traits with the ancestor than do platypuses, that is fine too, as in that case, the human and their traits might be better models, which also makes my point.
    Ryan, you might know, what does the sequence data tell us about the platypus/human story?

      (Quote)

    • I don’t think humans are more like the common ancestor of mammals. The very attempt to choose one modern species or another as representing the common ancestor is the fallacy.

        (Quote)

      • Bobo

        I don’t think humans are more like the common ancestor of mammals. The very attempt to choose one modern species or another as representing the common ancestor is the fallacy.
        Wrong.  Go learn some anatomy.  It’s a sure bet that monotremes resemble the common ancestor of mammals more closely than humans morphologically and reproductively.  (Not to mention metabolically, etc.)
        Just because it is traits and not species that are either “primitive” or “advanced” does not mean that one taxon rather than another may not retain more primitive or advanced features. Even though each has been evolving for the same amount of time does not mean that each has diverged equally from the common ancestor.
        More obfuscation from TRG for the mere sake of obfuscation (and attempting to make himself look intelligent).
        Platypuses are more similar to ancestral mammals than are humans.  Enjoy teaching your students lies.
         

          (Quote)

        • It’s a sure bet that monotremes resemble the common ancestor of mammals more closely than humans morphologically and reproductively.  (Not to mention metabolically, etc.)”

          You must have knowledge of the common ancestor of mammals that no one else has to be so sure.  More importantly, the issue is not which one is *more like* the common ancestor, it’s whether *either one* is similar enough to serve as a proxy for the ancestor of both species. (Read what I said again: “I don’t think humans are more like the common ancestor of mammals. The very attempt to choose one modern species or another as representing the common ancestor is the fallacy.”). 

          Just because it is traits and not species that are either “primitive” or “advanced” does not mean that one taxon rather than another may not retain more primitive or advanced features.”
          That’s not the point. Because traits and not species are primitive or derived (“advanced” is not the correct term), you can’t look at one modern species as a whole and infer stages in the evolution of another modern species (that is what the fallacy involves). You still must consider this on a trait-by-trait basis using phylogenetic comparative methods. 
          Even though each has been evolving for the same amount of time does not mean that each has diverged equally from the common ancestor.”
           
          You may be using “diverged” in a non-technical sense. The amount of change in particular traits may differ (be they morphological, genetic, physiological), but in terms of evolutionary relatedness they are exactly equal.  

          Let’s say, instead of platypus vs. human, we compare platypus vs. mouse. Or shrew vs. opossum. Or human vs. whale. Which is most like the common ancestor of mammals?  Or, more relevant to this issue, can you use one of the paired modern species to represent the common ancestor in order to infer evolutionary changes in the other? (Hint: no).

           

            (Quote)

        • I also would suggest that you cut down on the juvenile vitriol as this greatly weakens your credibility.

            (Quote)

  • Claudiu Bandea

    The Platypus Fallacy is a welcomed concept as it points to common errors/abuses made, particularly when assessing the evolution of humans and their traits.  However, we need to remain open to the likely possibility that certain traits of the modern species are more likely to reflect those of their ancestors than the homologous traits of other related modern species with which they share the ancestor.

      (Quote)

  • J. Chris Pires

    The fallacy is that “species are primitive” when all extant species are “equal” with respect to a common ancestor.
    In contrast, character states can be “ancestral” or “derived” – in a phylogenetic context ; but taxa are not primitive.
     

      (Quote)

  • jose

    Looks like the guy confused evolution and cladistics. Basal doesn’t mean primitive.

      (Quote)

  • Lenoxus

    Basal is similar though not identical to primitive, but the fallacy isn’t in the confusion of those concepts, because the platypus isn’t a “basal species” any more than it is a primitive one. The fallacy is in declaring a species to be “basal” (or “primitive”), by virtue of focusing on some traits and ignoring others.
    If the earliest fossils of a common human-platypus ancestor looked identical to the modern platypus, the fallacy would be considerably less fallacious. But there would be no reason to expect such a thing, and in fact we know that’s not the case. Platypi are just as derived/advanced as humans are. (And not just in terms of “extra” traits like venom, but substituted ones, like bills instead of teeth. Adult humans retain teeth, and teeth are something which sentient platypi would probably associate with childhood and by extension with primitiveness.) And while it’s true that traits like egg-laying are more basal, this in itself tells you nothing about the other traits! It makes no sense to think “The ancestral mammal was like X because monotremes are like X.” You need more species and/or traits to make such a case.
    I notice a similar fallacy, at a subtle level, in folk linguistics. Americans often assume that today’s British English is closer to the English spoken in the time of, say, Shakespeare, than today’s American English. But there’s actually a good argument that of all modern accents, Shakespeare’s would be closest to that of (fittingly enough) New England, albeit with an “English” tinge of some sort. Of course, we have no recordings of that time, but we do have writings, and while some original spellings changed in American English, other spellings were actually retained in American English while changing on the other side of the pond (such as “vise” for the tool which is elsewhere spelled “vice”).

      (Quote)

Leave a Reply

 

 

 

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

*