Fake Banner
    Confusions On Evolution, Creationism, And Falsifiability
    By Adam Retchless | February 20th 2010 09:15 AM | 5 comments | Print | E-mail | Track Comments
    About Adam

    I get paid to study microbial evolution. These writings are for my non-scientist friends and family (hi guys!) and anyone else who sees any value...

    View Adam's Profile
    Assorted creationists claim variously that creation theories are falsifiable and that evolutionary theories are not falsifiable. Here, I want to quickly point out a few flaws that I see in their arguments. I will discuss these issues in a more general manner in another post that I am currently writing.

    Creationism is falsifiable: This essay from the Discovery Institute provides two arguments, neither of which are satisfying.

    First, they look at the idea of irreducible complexity (IC, a component of Intelligent Design theory), and claim that it can be falsified by finding an evolutionary explanation for the structure in question. This does not match my understanding of "falsifiability" because IC does not make any predictions of its own, it just asks to be compared against an evolutionary explanation. As such, they are actually asking "What is more plausible, evolution or design?". This can be a productive question, but is not the gold standard of falsifiability. Even within this framework, the proponents of IC rely on an unjustified bias, assuming that as long as there is not an airtight evolutionary explanation, the IC explanation is valid (i.e. this is a "god of the gaps" argument). However, if we are just working with plausibility, the evolutionary explanation already trumps the IC explanation simply because we've observed countless instances of mutation, but no instances of intelligent design. Mutation is a more plausible explanation in all situations, without any further research. (Update: To clarify, with no known mechanism of ID, the explanation is completely implausible and therefore there is no way that we could examine existing life to see if its structure is more consistent with ID or evolution). Scientists gain nothing by considering the possibility of IC. For what it's worth, an evolutionary explanation has been proposed for their example of the bacterial flagellum.

    The second example they give, the Privileged Planet hypothesis, may be theoretically falsifiable, but seems totally impotent as science. Based on what is written in the Discovery Institute essay (I haven't read the book or watched the movie), there is no reason to expect any of these predictions to be tested within my lifetime. Furthermore, the predictions all seem to boil down to "we expect alien life to be like known life", which is not a risky prediction, and therefore is not informative. At best, the privileged planet hypothesis is a dormant scientific hypothesis waiting for that century when we can test it. At worst, it is a scam to draw money from people who really want to legitimize ID.

    Evolution cannot be falsified: This article at CreationWiki is just a mess, but what can you expect from a small wiki?

    First, they claim that historical theories cannot be tested. This is untrue. Historical theories are regularly tested by a back-and-forth process that involves new observations of the patterns created by historical processes, and predictions based on processes known to occur in the modern world. For instance, a hypothesis about the relatedness of two organisms can be tested by making a prediction about how similar their DNA sequences will be (based on observed mutational processes) and then sequencing their genomes (new information) and seeing whether the prediction is upheld. For what it's worth, evolution is not the only historical science--astronomy and geology both rely on historical models. I also don't know why historical sciences would be treated differently than history in general, as if the testimony of witnesses is trivial to interpret (despite changes in languages, the diverse motives that people have when they write, and even the limits of human memory).

    Second, when saying that only "subtheories" can be tested, they totally misrepresent what subtheories are (i.e. hypotheses). In evolutionary theories, the "subtheories" are not limited to things like evolution in a test tube (actually a fact, not a theory), they include things like asking whether birds are more closely related to mammals or dinosaurs (or bacteria). They could even include questions such as "Are birds related to mammals or dinosaurs at all". To use their "rocketship" analogy, this is like asking whether the ship landed in Asia or North America, and whether there were multiple ships.

    The rest of the article seems to be nit-picking over details of evolutionary theory. The authors treat the evolving nature of evolutionary theory as evidence that it cannot be falsified. This viewpoint ignores the fact that the theory is not just refined to explain some fact and then left alone; any new hypothesis makes a variety of predictions which are then tested. These predictions involve both the historical artifacts (e.g. fossils and genome sequences), and the ongoing processes that allow populations to change (e.g. population growth, how quickly forms can change).

    In the end, creationists complain that when one evolutionary hypothesis is rejected, it is replaced with another. They see this as evidence that "evolution" as a whole cannot be falsified. They are misguided on two points. First, a theory is not tested as a whole--it is tested in parts and the theory fails when so many of its parts have been demolished (or co-opted into other theories) that it can no longer be recognized as a theory. This is what happened with Spontaneous Generation, which required a few centuries between the recognition that it doesn't apply to flies, and its final death when it was conclusively demonstrated not to apply to microbes either. Their second error is their failure to recognize that evolution simply requires that life can change and that it has had sufficient time to change (i.e. the Earth is old).

    As long as those conditions are met, and there is no plausible alternative mechanism for generating the diversity of life, some form of evolutionary theory will have a central place in biology.


    Thanks for a great article.
    If you notice, the only sites that can get away with such nonsense science are the sites that don’t allow public comments. Too many people are on to them and will be quick call them on their dishonesty.

    It’s just incredible how the folks at the Discovery Institute continue their bold face lies and keep a straight face doing it.
    They are already suing the California Science Center on grounds of freedom of speech.

    Keep an eye on them, their next gig will be to try to change the definition of science so that they can gain entry into science classrooms and science museums.

    A thank you from me as well. I've been interested in this subject since the Discovery folks fired their first shot over the bow of real science back in the eighties. A note to Mark: they have indeed made an attempt to re-define science, in the course of a federal court case from Dover Pennsylvania back in 2006. Fortunately it failed spectacularly. For a good description of what happened I would suggest reading Edward Humes' Monkey Girl. I just finished reading it recently and thoroughly enjoyed both the content and the writing style.

    Thanks for the encouragement Mark and TMG. I intend to have another ID essay up this weekend.
    Adam: I am open to correction , but my understanding is that evolution is driven by long-term incremental environmental changes.   Logic demands that it can be driven in reverse - an evolved 'better heat-dissipation system' can evolve into a 'better heat-trapping system' if the local environment cools, for example.

    Here's my take on I.D. v evolution:

    All known living things live in environments which are subject to change.

    To allow for adaptation to such changes there is a mechanism by which, over the course of time, a population changes to match the environmental shift.  Alternately, a population can evolve so as to drift into a differently located environment.

    The only ultimate barriers to such shifts are lethal limits: environmental changes or evolutionary changes of such magnitude or rapidity as to eradicate the entire population.

    If we run the model in reverse we find exactly the same mechanism for evolution - environmental shift causes a population to vary its properties to offset the shift.

    Evolution allows any non-lethal shift.  In principle, if only given enough time, any living thing could evolve or devolve into any other living thing.  I am referring to gradual change, not cross-breeding.

    I.D. requires putting up barriers to change other than lethal barriers.  Proponents of I.D. will not accept the idea of common ancestry for all life, or even all mammals.  They do however tend to accept that at least some changes may occur.  I.D. has barriers to modification where evolution has none.

    The burden of proof is on I.D. proponents to suggest the form and mechanism of such barriers.  If the human race is foolish enough to wipe itself out, what barrier do I.D. proponents suggest stands in the way of any other creature taking our place by evolving human-level intelligence - or even a better intelligence?

    Parsimony - Occam's razor if you will - suggests that the simpler idea is the most likely to be correct.  The barrier to evolutionary reversal - unless proven to exist -  is surplus to scientific requirements.


    The I.D. theory was set up to counter the notion that humans could have evolved from 'lower' creatures.  It requires an evolutionary barrier mechanism which it has never addressed.

    I.D. suffers from mission failure.

    The theory of evolution has no mission other than to offer a scientific explanation of the diversity of life.  In the absence of any better theory, it succeeds.
    Hi Patrick,

    That's a good point regarding the limits of change. If such limits existed, it would be a good case against common ancestry. Some creationists are actually thinking very hard about how to demonstrate these limits.

    Kenneth Cumming writes:

    As a final example, I would now like to review a creationist concept of
    species. ReMine [11]
    has recently put forward an alternate systematic methodology to the
    prevalent phylogenetic systems. He calls it "discontinuity systematics."
    In the scheme, he explains that "species" was merely the Latin word for
    "kind." For various reasons, he coins terms related to the synthetic
    Hebrew word "baramin" to apply to working definitions of this innovative
    system. Frank Marsh [12]
    originally combined the Hebrew root words bare ("create") and min
    ("kind") into the term "baramin." Although there is no direct statement
    to this effect, the term "holobaramin," or its subset "monobaramin,"
    might include or in some instances be synonymous with species. Yet the
    emphasis is upon experimentally circumscribing the continuities and
    locating the real gaps or discontinuities in nature. Empirical research
    on delineating holobaramins is just starting.

    (Emphasis mine)

    I doubt that this line of research will ever advance, because that idea is inconsistent with known genetics (my knowledge is primarily microbiology, but this apples to all life). For instance, this idea of limits to change would be plausible if mutations tended to revert previous mutations. However, in reality, mutations tend to pile up on top of each other. Even a deleterious mutation is rarely reverted; instead, when geneticists mess up a protein (making the organism sick) then select for mutants that are no longer sick, they frequently recover "pseudo-revertants", which fix the damage of the original mutation by changing the genome at a different location (This is a good way to figure out how proteins interact.)

    These experiments also hint at why reversing evolution is not so simple as reversing the selection pressures. If we knew every mutation that had occurred, could engineer every one of them into the genome, and could keep the intermediate organisms alive, then we could probably convert any organism into any other.  However, if we are just relying on spontaneous mutations to generate variation upon which we select, then there is no reason to expect that we could reverse evolution. Mutations can interact with each other in complicated ways, and the first few mutations in an evolutionary process can determine which paths are open or closed for further evolution.

    Another problem with reversing evolution is that much of the selection pressure comes from other organisms, including the mutant's peers and parents.

    The study of the reversibility of evolution at the molecular level is starting to take off, and I've seen a few really interesting papers in the past few years. At the organismal level, the best example that I can think of is the Blount and Lenski paper in PNAS a few years ago (you may have heard of it: it started a big debate among creationists, most of which fixated on unsurprising results and ignored the real point of the paper)

      In the absence of any better theory, it succeeds.
    I'm writing an broader essay that examines this issue. The only think I have to say now is that evolutionary theory could fail (or, could have failed) if if it were found to be incompatible with genetics. But researches put a lot of effort into testing whether genetics was compatible with Darwin's theory (i.e., random mutation followed by selection), and overwhelmingly found that it was.