I am very disappointed. There is this site called How Stuff Works that I’ve run into a few times, that has nice, short, kid-friendly summaries of, obviously enough, how stuff works. I hadn’t used it much, but it seemed like a cool idea…until a reader suggested I take a look at the section on how evolution works.
The author has a very, very poor understanding of basic biology, and it looks like the essay was simply spun off the top of his head, with a few quick glances at some websites. The author, Marshall Brain, is an electrical engineer and computer scientist, and it shows, embarrassingly enough.
The whole general introduction is thin and strange and far from how a biologist would discuss it, but rather than going over everything, I’ll focus on one section as an example, a summary of “Holes in the theory“. While giving far too much emphasis to problems than is appropriate, this section has another serious flaw: his holes ain’t holes. All this section is is an airing of the author’s ignorance.
He lists three “holes”, and the first is this one: How Does Evolution Add Information?. Uh-oh. He thinks this is a problem? What could he mean?
Evolution’s mutation mechanism does not explain how growth of a genome is possible. How can point mutations create new chromosomes or lengthen a strand of DNA?
First sentence: dead wrong. Variations in genome size are trivial, and happen all the time through the mechanisms of duplication and deletion, and the inheritance of translocations. He writes about transposons and polyploidy as possible mechanisms, but misses the big ones. And of course, point mutations aren’t the only kinds of mutations present. I’m sorry to say that Mr Brain doesn’t have even a basic knowledge of genetics, so it’s extremely troubling to see him presenting himself as an authority on this subject.
His second question is even worse, reflecting some innumeracy and lack of logic skills that I would expect from an engineer. He’s bothered by “How Can Evolution Be So Quick?“. I’d expect a good mathematical discussion from a population geneticist about rates of mutation and fixation, etc., that would show that actually, rates of evolution are well within the bounds of reason, but instead, he uses several peculiar examples that just don’t make sense, and worded misleadingly to cast false doubt on the idea. Here’s one example, the growth of the human brain.
Modern human brain size averages about 1,500 CCs or so. In other words, in about 2 million years, evolution roughly doubled the size of the Homo erectus brain to create the human brain that we have today. Our brains contain approximately 100 billion neurons today, so in 2 million years, evolution added 50 billion neurons to the Homo erectus brain (while at the same time redesigning the skull to accommodate all of those neurons and redesigning the female pelvis to let the larger skull through during birth, etc.).
Let’s assume that Homo erectus was able to reproduce every 10 years. That means that, in 2 million years, there were 200,000 generations of Homo erectus possible. There are four possible explanations for where the 50 billion new neurons came from in 200,000 generations:
- Every generation, 250,000 new neurons were added to the Homo erectus brain (250,000 * 200,000 = 50 billion).
- Every 100,000 years, 2.5 billion new neurons were added to the Homo erectus brain (2,500,000,000 * 20 = 50 billion).
- Perhaps 500,000 years ago, there was a spurt of 20 or so closely-spaced generations that added 2.5 billion neurons per generation.
- One day, spontaneously, 50 billion new neurons were added to the Homo erectus brain to create the Homo sapiens brain.
None of these scenarios is particularly comfortable. We see no evidence that evolution is randomly adding 250,000 neurons to each child born today, so that explanation is hard to swallow. The thought of adding a large package of something like 2.5 billion neurons in one step is difficult to imagine, because there is no way to explain how the neurons would wire themselves in. What sort of point mutation would occur in a DNA molecule that would suddenly create billions of new neurons and wire them correctly? The current theory of evolution does not predict how this could happen.
That’s just plain goofy. I don’t know why he claims there are just four possible explanations, unless he thinks round numbers have some kind of deep significance. All he’s really saying is that there is a range of possible rates for brain growth, from slow and steady, generation by generation, to abrupt and sudden. We can actually look at the fossil data to get an idea of the rates, as we can see below.
He’s also guilty of grossly typological thinking—”the Homo erectus brain”, “the Homo sapiens brain”—and making the mistake of assuming that each child gets 250,000 neurons added to it. This is nonsense, of course. What we have is a wide range of variation in a population, and under selection, those with larger brains do statistically better at surviving and reproducing than those with smaller brains. Is there such variation? Of course! And Brain even includes a diagram of skulls that demonstrates this, with ranges of 1200-1600 cc listed for the modern human skulls.
To put that in context using his simplistic measure of neuron numbers, that means some perfectly normal kids are born today that will have a cranial capacity about 400 cc greater than other kids, or about 25 billion more neurons than some of their fellows. 250,000 neurons is trivial; it represents about 0.00375 cc. Yet, somehow, a variant with 0.00375 cc greater cranial capacity is something that makes him uncomfortable, and for which he can see no evidence, while blandly and (apparently) obliviously mentioning the fact that there is 400 cc of variation in the population.
As for his concern that he can’t imagine how point mutations (again with the point mutations; he really needs to brush up on some genetics) could generate more neurons, and he doesn’t understand how they’d also get wired up…that is a legitimate question and one of great interest to many of us. It’s actually not hard to see how mutations in regulatory genes could cause a cascade of effects leading to changes in rates of cell division in the brain; one candidate is ASPM. Wiring up additional neurons is only a problem if you are under the faulty impression that every connection is specified by the genome. This is not the case, but instead are negotiated by cell:cell interactions during development, a phenomenon called plasticity. Our genomes aren’t full of micromanaging details about what single cells should do, but instead encode general extensible rules for sorting out cells and synapses—variation is easily handled.
His third “hole” in evolution, Where Did the First Living Cell Come From?, at least is asking a genuine question, one that scientists actually wonder about, unlike his first two. However, having just read a book on abiogenesis, I can say that the details in this section completely miss all the real issues, and it basically amounts to a statement that cells are really complicated, and Marshal Brain doesn’t understand how they came to be, and also hasn’t bothered to do his homework to see how scientists are addressing the issues.
As I said, I’m very disappointed. I don’t think I could recommend that site to anyone, given the deep flaws in an entry on subject matter with which I am very familiar. I’m also not at all impressed with the author list, which contains no one with any substantial science credentials (the head guy is trained in engineering, not science, strictly speaking). On top of all that, the site contains overly credulous articles on ESP, Tarot Cards, UFOs, and witchcraft. The overall impression is of excessive babbling with no quality control and no competence to evaluate many of the claims they’re writing about.