Tangled Bankruptcy: Response to Nathaniel Jeanson

Nathaniel Jeanson, newly of the Institute for Creation Research, gave a lecture in Boston last night entitled “Evolution: Bankrupt Science, Creationism: Science You Can Bank On.” This shouldn’t come as a surprise to anyone familiar with how creationist arguments work, but there was absolutely nothing new about his talk. As a freshly-minted doctoral graduate from Harvard’s molecular biology program, Jeanson is the embodiment of the very bleeding edge of so-called creation science: a dull intellectual wasteland rehashing decades-old arguments long since refuted, unable to be fertilized even by graduate-level training at an Ivy league university.

I attended the lecture with the Boston Skeptics. It was was mostly one stale, nonsensical creationist talking point after another, spanning everything from geology to astronomy to biology (in case you hadn’t guessed, by “evolution” he meant “every field of secular science that challenges Young Earth Creationism”), with bits of pieces of pure absurdity sprinkled in for flavor. I had to chuckle a bit when he brought up Irreducible Complexity; and here I had though Behe had gone out of style. There was, however, one claim he made late in his talk that caught my interest as a biologist: he suggested that the variation between species in a protein called cytochrome c (“cytC”) actually refutes common descent.

I hadn’t heard this particular talking point before, but apparently even this is nothing new. In fact, one of the first hits that come up when you Google “cytochrome c homology” is to a creationist webpage that cites a book called “Evolution: A Theory In Crisis” by Michael Denton, published in 1986. (Denton’s argument also apparently appears in the infamous creationist textbook, “Of Pandas and People”.) Denton compared the amino acid sequence of a bacterial version of cytC to the cytC sequences of several fungi, plants, and animals, and found that across the proteins had roughly the same degree of shared sequence across the board. From this, creationists assert:

If evolution were true, then the further organisms have evolved from bacteria, the greater change there should be in Cytochrome C. . . . From the data in table above it is evident that the amino acid sequence of the Cytochrome c molecules of all of the species in all of the groups of organisms are equally isolated from that of the bacterium Rhodospirillum rubrum. Thus there is no basis in this data to indicate that any group is intermediate between other groups. All are equally isolated from all other groups. This data supports the biblical record of creation of each “kind” separate from all other “kinds.”

Jeanson tried to make essentially the same point, but with so much additional obfuscation as to lead me to believe he wasn’t merely profoundly ignorant, but downright blatantly dishonest. He first presented a chart similar to the following that shows how cytC from several species compare to human cytC1:

Species Human
Human (Homo sapiens) 100%
Chimpanzee (Pan troglodytes) 100%
Dog (Canis lupus familiaris) 89%
Zebra fish (Danio rerio) 84%
Fruit fly (Drosophila melanogaster) 78%
Yeast (Candida albicans) 66%

CytC is a very old, highly conserved protein, appearing in animals, plants, fungi, and even many unicellular organisms. Since it’s so prevalent, it’s an excellent protein to choose when we want to compare protein sequences between two different organisms like this. These are exactly the kinds of numbers we would predict based on what we know about evolution. Humans and chimps, being very closely related, have identical cytC. Dogs, our fellow mammals, have less shared identity than chimps, but more than fish. Our fellow vertebrates all match us more closely than does the invertebrate Drosophila, which in turn, being a fellow animal, matches us more closely than does the fungus Candida.

We can represent these relationships using a branching figure called a “dendrogram”:

simple-dendrogram

Note that, the more closely two species are related, the less distance you have to trace back along the dendrogram before you reach a common horizontal. And Jeanson baited the audience by freely admitting that this was a fair interpretation of the data presented. However, he followed up his single-column chart with an expanded chart similar to this:

Species Human Chimp Dog Fish Fly Yeast
Human 100%
Chimpanzee 100%
Dog 89% 89%
Zebra fish 84% 84% 90%
Fruit fly 78% 78% 84% 85%
Yeast 66% 66% 66% 66% 67%

This chart, Jeanson claimed, refutes the evolutionary model and necessitates the creationist hypothesis of discontinuous species. It’s the same flawed argument as seen above, wherein each row of this table is akin to the Denton table. Each species is approximately “equidistant” from all the species above it, which supposedly discredits evolution (nothing could be further from the truth, as we shall see).

But that wasn’t enough for Jeanson. To further confuse the audience, he replaced the dendrogram with a deliberately obfuscatory “model” in which every species eventually has a line drawn to every other species signifying the percent shared identity. After 3 iterations, his so-called “n-dimensional model” would look like a triangular pyramid (use your imaginations to get the 3D effect):

n-dimensional-crap

He claims, then, that since you have to add a new dimension for each species you add to the “model,” it ultimately becomes unwieldy, and it makes more sense to think of the species as discontinuous rather than inter-related. The problem is, his “model” is nothing of the sort. It tells us absolutely nothing new. It’s simply a confusing way of illustrating the data presented in the chart, with no effort made to actually interpret the data. The only reason it requires multiple dimensions to illustrate is because he made the completely arbitrary decision not to let the connecting lines overlap. He could just have easily used crisscrossing lines, or (hey!) left it in chart format! Same data, no added value, just different means of visualizing it.

Why, then, mangle the presentation of the data? To hide the truth that the data fits our earlier dendrogram beautifully! In fact, this additional data is precisely what makes our dendrogram possible in the first place.

The creationist cytC argument fails because it falls for the Great Chain of Being fallacy. They imagine fruit flies as being more primitive than zebra fish, which are in turn less evolved than dogs, etc, with humans being the most “highly evolved.” But all of the species being studied here are extant species, alive in the here and now. Zebra fish are not an intermediate species between humans and yeast, though the last common ancestor between humans and fish is. The human lineage has changed a lot since that common ancestor, but so has the fish lineage!

The measure of shared identity between cytC tells us roughly how long the proteins have had to change since they diverged. The point at which the human and yeast lineage diverged is the exact same point at which all fungi and all animals diverged, so we shouldn’t be surprised to see that the similarity between yeast and fish is similar to between yeast and dogs. Similarly, the point at which the human lineage diverged from fruit flies is the exact same point at which all vertebrates diverged from all invertebrates. And since the vertebrate-invertebrate divergence shows less change (~81% match) than the animal-fungus divergence (~66%), it fits our model with the animal-fungus divergence occurring further in the past.

This analysis of the data is precisely where the dendrogram comes from. You can’t get a very accurate tree structure just by looking at how all species relate to a single organism. You need to compare as many species as possible. This is where Jeanson’s table completely backfires; whereas Denton only ever looked at one divergence (prokaryotes-eukaryotes), Jeanson was kind enough to present us with data from five (animal-fungus, vertebrate-invertebrate, fish-mammal, canine-primate, human-chimpanzee), making for a much more interesting dendrogram.

And I suspect that Jeanson knows all this, which is why he insisted on being deliberately confusing. I kinda botched my attempt to challenge him during the Q&A session (at that point I was still struggling to discern just how he was getting it so wrong), but I had the opportunity to approach him afterward and press him on the fact that the data supports the evolutionary model. First he tried (as he did during the Q&A) to say that the dendrogram was a 2-D projection of his n-dimensional model, which was pure nonsense. The two diagrams have nothing in common, the lines drawn mean completely different things. And after some further discussion, when I (repeatedly) asked him whether he would admit that the data presented fits the evolutionary model, he refused to give a yes-or-no answer. If that’s the lot he’s chosen, then fine. It’s his life. I’ll be content to marvel in how the science really works.

And it really is a marvel. CytC is a prime example to use when tracing evolutionary heritage. But the really remarkable thing is that, when we take similar approaches to other proteins, they confirm the same evolutionary heritage! Disparate proteins fall along similar dendrograms, indicating that the inter-relatedness is not an accident, but follows an overarching pattern of common descent. Furthermore, we can use the amount of divergence in these proteins’ sequences as a very rough estimate of time since species diverged, and we can compare that with the account given by the fossil record.

Evolution has absolutely no problem interpreting the data that Jeanson presented. In fact, it thrives upon such data. If Jeanson wants to lie and obfuscate about that, I’m sure the ICR will pay him handsomely to do so. But I prefer the truth.

(For more on Jeanson’s talk, see Rebecca’s post at Boston Skeptics)

1: Whereas Jeanson gave values of percentage difference between proteins in his lecture, I list here percent identity. Numbers were obtained by hastily copying cytC sequences from NCBI Entrez Protein and comparing them using NCBI BLAST.