Friday, August 15, 2008

Pluralism- Optimal versus non-optimal

Once people get passed the tired, unimaginative, recycled, anti-evolution claims that creationists often make (snore), there are still some common obstacles to a deeper understanding of how evolution has produced the wondrous diversity of life that we see around us every day. Primary among these misunderstandings are the monistic view that evolution produces perfectly optimal phenotypes, and the related idea that natural selection is the only process that shapes evolution.

Of course this point is also a bit tired and unimaginative of me to point out, since Gould and Lewontin famously wrote on this subject almost 30 years ago. And even then, they were not even considered original by everyone, as they were accused by some of strawmandering, of erecting a simplistic caricature of peoples' views on evolution to be argued against. But never fear, I think I do have an original contribution to offer - I think I've stumbled upon a really nifty example to help understand a pluralistic view of optimal versus non-optimal evolution. Like Gould and Lewontin's famous spandrels (architectural objects that originate because of constraint instead of optimal functionality) I provide a non-biological example. I hope to try out this example on my students in next year's Macroevolution course.

The Panglossian barriers of Joshua Tree

Somehow these examples of the struggle between optimal and non-optimal come to me while traveling. My previous best personal example illustrates how people by nature tend to think adaptively. We tend to ask what something was made for - even though many things could simply arise as by-products, not as optimal "for" anything. I found myself falling into this Panglossian trap myself while driving in Joshua Tree National Monument. This is a place I've been to often, as it marks the half way point between my home in Santa Barbara, and relatives' home in Tucson. We usually arrive late in Joshua Tree, leaving SB after 6:00pm to avoid the traffic of the Inland Empire. When we arrive, Joshua Tree is dark, and the star gazing is amazing. Driving in, our mini-van headlights are no match for the vast darkness of the desert, and so the road becomes a singular focus. Some times of the year kangaroo rats dart just out of our path, but others the monotony of the dirt road induces a hypnotic trance. It was in one of these trances that I began to wonder the function of the ridges of dirt that edged the road. Along the edges of the dirt roads in Joshua Tree, sand lays in a neat pile, sloping from a maximum of foot or two in height gradually to the surface of the road. Except for the tan color, it was as if I had just followed a plow through a fresh coat of fallen snow.

[I'll paste a picture here when I find one].

Perhaps these ridges are an attempt to keep animals off the road, serving as a barrier between the raw wilderness and these capillaries of civilization that penetrate the desert. "Wait a minute, what am I thinking??", I thought, jarred from my trance (or was I?). Why to the road ridges have to be "for" anything? It makes so much more sense if these are simply by-products of making a road in the desert. My snow analogy is probably right on. Probably to make and up-keep the roads, plows simply carve a path through the desert, resulting in a pile of desert sand on either side as a by-product. These are not adaptive barriers, these are simply by-products of a need for efficient auto travel. Yet it was so natural for me to assume there is some function, and I so easily came up with an explanation, just so. I can see how one could see the natural world in a similar and Panglossian fashion.

What's even more wondrous is that similar by-products can be later elaborated, to give an even more distinct, yet ultimately false, impression of optimality. Growing up in Wisconsin, we had one glorious winter when the long driveway of our farm house was plowed by a neighbor (as opposed to shoveled, or cleared with a snow-blower). Just like the sand in Joshua Tree, the plow of my childhood left a by-product of a flat clear path, in the form of piles of snow flanking our driveway. Through the winter, these banks piled high, and they packed tight. My brother and I soon found an outstanding use for these by-products. We created the most fantastical snow fort Germantown Wisconsin has ever seen. We spent hours tunneling through the banks, building rooms where the snow piled wide. The plow packed the snow tight enough that we could build shelves and windows in the walls and sunroofs in the ceilings. So spectacular was our fortress, it was hard for our friends not to wonder whether the banks were plowed for the express purpose of creating our winter wonderland. But similar to Gould and Lewontin's elegantly painted spandrels in the chapel of San Marco, and just like the sand banks of the Joshua Tree desert, my childhood snow bank was simply a by-product of a need for efficient car travel, no matter how optimal our resulting snow fort turned out.

The Pluralistic spires and arches Utah

It's true that some structures sometimes arise as by-products. But what I find even more appealing is a pluralistic interplay between optimality and non-optimality. Natural selection may be an optimizing force, but it is not all powerful. There are constraints on the system. Driving through Utah this week, I saw a perfect illustration of this in the inspirationally beautiful wild lands of Utah.

Water, when pulled by gravity, will find an optimal path to the lowest point. And water, when it courses over rock, gradually wears that rock away, leaving a trace of its path. If all rock were created equal, canyons would be arrow straight, as water would continually trace a single optimal path to the oceans. But canyons are complex because all rock is not created equal. As a result, water faces constraints to optimality. When softer rock lies adjacent to harder rock, the softer breaks away more quickly, producing a circuitous path to the lowest point. The harder rock constrains and controls the outcome, in combination with the optimizing pull of gravity on water. The spires, arches and canyons are stunning examples of the intricate interplay between optimization and constraint. With constraints imposed by heterogeneous rock alone, the structures do not appear. Similarly with water acting optimally on homogeneous rock, the structures do not appear. It is only through the interaction that this beauty arises.

Bryce canyon photo by James Gordon, posted on flickr.

The geology is amazing in its own right, and I am sure I have not done it justice by my oversimplification (e.g. I've ignored wind and ice erosion). But my point here is to focus one how these pluralistic ways of thinking can inform us of how evolution proceeds. Nothing in nature works without constraint, without trade-off. Only by seeing natural selection as one player in a complex (though undeniably natural) interplay of various processes, do we gain a more nuanced and pluralistic understanding of how evolution has shaped our world.

[Caveat: These are purely non-biological examples, meant as analogies to prime one's pluralistic thinking. However, where the rubber hits the road for evolution is with biological examples, but that is beyond the scope of the current post.]


Kevin Zelnio said...

The more I ponder it, I think pluralism is the best option regarding species concepts. It recognizes the diversity of life history traits among various types of organisms, bacteria to plants to animals, asexual to sexual. This is why I like to see descriptions use morphology and molecular data in tandem.

Perhaps it might be useful to think of a species as a amalgam of concepts. Or maybe a principal component utilizing all the information: biometrics, phylogenetics, morphology characters, etc. Then use discriminant analysis to separate statistically different species. But that might be too much work ;)

Todd Oakley said...

Kevin - I agree that the pluralistic species concept seems to make the most sense. I remember a good lecture at the evolution meetings by de Queiroz on this idea. I think it was the presidential lecture, either for Sys Biol or maybe Evolution.

I guess the biggest difficulty for me with species is the attempt to make something continuous into something discrete. I'm always reminded of the judge in the People v Larry Flint. When asked how to define what is offensive, he replied something like "I don't know, but I know it when I see it".

Species are the same. Obviously a fly and an ostracod are different species. But there are numerous ambiguous cases out there, somewhere between being one species and clearly being two species.

Kevin Zelnio said...

That may be where a more statistical approach would be useful in species delimitation. When you just don't know if 2 populations represent the same thing, through all the measurements and characters into some components and do some discriminant analysis and see if group A is different or the same from group B. The drawback is you need a large sample size. As someone who describes deep sea stuff myself. I'm lucky to get 5 specimens of the same thing! So perhaps not applicable in all cases.