Deep archives: Irreducible beauty

Peacocks and audience at the Toronto Zoo

Were peacocks designed with this kind of audience in mind?

A while back I was searching for images of peacock feathers on Google, and I stumbled upon this article. It’s a piece by Stuart Burgess, an engineer who is head of the Department of Mechanical Engineering at Bristol University, and apparently also quite an opinionated creationist.

Burgess’ idea is that the peacock’s train feathers “contain an extremely high level of optimum design”, so much so that they provide evidence against Darwinian evolution. He thinks that the aesthetic features of the peacock are so complex, so contingent upon each other, that no step-by-step process of evolutionary change could have produced them. He’s right that these ornaments are highly complex, and that selection for this kind of extreme aesthetic feature presents a bit of a puzzle for evolutionary biology. To claim that the extraordinary complexity must be “irreducible”, however, is a big assumption.

The article provides a lot of amusing examples of twisted logic along the way. For example, one of the features that Burgess finds irreducibly beautiful is the fact that the peacock’s train forms a fan-like shape. This is because “the axis of every feather can be projected back to an approximately common geometrical center” – indeed, the body of the bird that grew them!

He also notes the wondrous spacing of the eyespot feathers: “Even though the display contains around 170 eye feathers, they are all visible and all spaced apart with a remarkable degree of uniformity. All the eyes are visible because the feathers are layered with the short feathers at the front and the longer feathers at the back. The eyes have an even spacing because each feather has the right length.” I suppose it’s equally fortunate that each of my arms has exactly the right length, or that my fingers are nicely ordered such that my thumbs are on the outside rather than in the middle. These kinds of patterns of development are ubiquitous in the animal kingdom, and not something special about peacocks. The problem is that Burgess assumes patterns require something extra special, like a bit of design work by an intelligent being, to produce.

For example, he goes on to talk about the fact that each eyespot is a digital pattern, since it is “formed by the combined effect of many small barbules”, which are the tiny branching structures that grow out of each feather barb (take a closer look here). He says,

Some patterns in nature are formed by natural growth mechanisms, as with the spiral shape of the nautilus shell. However, the eye pattern in the peacock tail requires the precise coordination of independent barbs and this cannot be achieved by a simple growth mechanism. Barbules on adjacent barbs coordinate perfectly with each other to produce the eye pattern.

Burgess claims that “the spacing of colours on each barb must be specified by instructions in the genetic code”. Because we can use a mathematical function to describe the shape of the eyespot, there must be something particularly complex in the peacock’s genome that produces that shape (math is really complex, right?). Indeed, “every detail in the peacock tail must be defined by genes in the genetic code of the peafowl”. And since there are so many fine details, we’re going to need a lot of genetic information.

Though he admits that “it is difficult to determine how many genes would be required to specify the aesthetic features of a peacock tail feather because it is not known how the tail feather grows”, he’s willing to hazard a guess, and it comes to… 20 genes. I think this meaningless number is supposed to shock us as being an impossibly large amount of information but it’s not even particularly high given that animals generally have well over 10,000 genes. It’s true that the eyespot is amazingly complex, and it would not be trivial for biologists to work out how it is produced at the cellular and molecular levels. Regardless, Burgess doesn’t seem to be aware of the fact that biological development can form patterns without having to specify every bit of information in the result somewhere in the genetic code.

Later on, Burgess lets natural selection in the back door. He tries to argue that even if science shows that peahens have hard-wired preferences for more “beautiful” males, we still can’t rule out design by a creator:

If there is a preference gene for aesthetic features, this does not prove that the sexual selection theory is true. The reason for this is that the Creator may have installed a preference gene as a means of ‘maintaining’ beautiful features. Beauty generally gives a disadvantage in terms of escaping from predators. If a peacock lost its colours due to a gene mutation, it would suddenly find itself more protected from predators. This is an example of where a loss of information could be a great advantage in terms of survival. Therefore, it is conceivable that the Creator would deliberately create preference genes for prominent aesthetic features such as colour.

Basically, because natural selection might act to reduce the peacock’s ornamentation, a good designer would put genes for aesthetic preferences in peahens as well – just in case.

I could go on – his explanation of how the colour is produced by the peacock’s feather barbules is just plain wrong (Figure 4) – but reality is actually more complex than he supposes (see figures here)! He also says, “the beauty of the peacock tail can be termed ‘added beauty’ because it appears to be surplus to that necessary to survive” – leading one to wonder exactly what level of beauty is necessary for survival.

A couple of months ago I came across another article by Burgess on peafowl. This one appeared to be a proper scientific paper in the journal Optics and Laser Technology. Published 5 years after the creationist article, this paper is entitled, “An analysis of optimal structural features in the peacock tail feather“.

In it, Burgess rehashes many of the same arguments about the extraordinary complexity of the eyespots. He even reuses the figures from the creationist magazine article. Gone are the references to the intelligent creator and irreducible beauty, however. Instead, he infers that because peacock feathers are simultaneously optimized “in the three areas of structures, optics and aesthetics” they provide “a key lesson for engineers and physicists”. That is, “different disciplines should work together to explore common features that can allow simultaneous optimal design in widely different areas.”

A thinly-veiled attempt to publish an intelligent design rant as science. It brings up a few questions, though. Was the journal, which is apparently peer-reviewed and publishes only original work, aware of Burgess’ other article? And is it possible to plagiarize yourself? (that is, by re-using your own figures without citing where they were originally published)