Tuesday, September 10, 2013

Problems with adaptationism

While the basics of evolution are relatively easy to understand, the finer details are surprisingly convoluted. A big debate in the second half of the 20th century concerned adaptationism, or the belief that every trait we observe - be it morphological or behavioral - was specifically selected for and has an adaptive explanation for its existence. Adaptationism argues that natural selection is the only significant means of evolution (the others being genetic drift, gene flow, and mutation), which means that if a trait exists, it is the optimal solution for some question behind it. The debate technically isn't over, with big-name researchers like Richard Dawkins and Steven Pinker still arguing for adaptationism. While many traits in an organism clearly are the result of natural selection, this blog post will highlight problems with trying to extend this thinking to everything you see in an organism.

The beginnings of adaptationism
The groundwork for the debate began with the publishing of two tremendous books, E.O. Wilson's Sociobiology: The New Synthesis (1975) and Richard Dawkins' The Selfish Gene (1976). Sociobiology attempts to explain social behaviors through an evolutionary lens (e.g. how does an animal's aggression lead to it having more offspring in the next generation?). The Selfish Gene is an argument for genes being the unit of evolutionary selection; in a survival of the fittest billions of years ago, a gene that could produce more copies of itself inevitably outcompeted all other variants. (It's quite a bit more complicated now.) 

Big problems arose when researchers began interpreting human social behavior using Dawkins' reasoning for the evolution of genes. These people, who started being called adaptationists, began claiming that any human behavior you see today - cutting someone off while in traffic, some women preferring taller men, smiling when we're happy - had a story behind it. Any human behavior... including suicide, necrophilia, child molestation, torture... had to be there because our ancestors who didn't perform those behaviors were outcompeted by those who did. 

Unraveling the logical missteps
Stephen J. Gould and Richard Lewontin took the first big stab against adaptationism in 1979 when they published a paper critiquing evolutionary biologists for the following four problems:

     1. Relying on story-telling explanations that are impossible to scientifically test

     2. Overlooking other methods of evolution besides natural selection

     3. Overlooking evolutionary and physiological trade-offs and constraints

     4. Claiming that the current way a trait is used is the reason it originated

Let's explain these in simpler terms. A lot of adaptationism utilizes just-so stories. A just-so story, also called an ad hoc fallacy, is an unverifiable explanation for how a behavior or morphology came to exist. It can't be proved or disproved; it's basically the opposite of hypothesis testing because you're looking at the results of something and then forming a question that leads to your answer.

Here's an example of this flawed reasoning. Say you're doing a census of fox populations in the wild and in the city. Interestingly, you notice that foxes in the wild tend to be orange while city foxes are generally a mix of gray and orange. You conclude that foxes in the city have evolved orange-gray fur as a means of camouflage in their predominantly gray landscapes.

Slow down a minute. Yes, this is one potential explanation for what you've seen, but you can't rule out dozens of other equally-valid explanations that have nothing to do with adaptation to the city. What if the genes for fur color are closely associated with genes for neophobia (fear of novelty), and as less neophobic foxes outcompete their more novelty-fearing conspecifics, their fur color gets dragged along as well? What if foxes in cities actually represent a small founding group isolated from groups in the forest, and their gray fur just happened to arise due to genetic drift? And depending on whether you actually handled any of the foxes in your census... what if the gray foxes were just dirty? (Yes, sometimes an explanation this silly has to be considered if your methodology didn't control for it.)

The problem with adaptationism is that you start viewing organisms as a bundle of optimal traits independent from one another. In reality, there are a lot of evolutionary limitations and trade-offs within an organism. For example, zebras don't have machine guns because machine guns were not available as an ancestral trait to be selected for. Similarly, organisms are subject to constraints within themselves. Even though our muscles, bones, and digestive organs don't really do anything to make babies (the only thing that matters from an evolutionary perspective), humans are not huge ovaries and testes running around because we need the other organs to acquire resources, not get eaten, and then get to the point of reproducing. (It's probably for the better, though life would be a lot more entertaining.)

In the example above, you can't test a hypothesis on why dogs have floppy ears. Even if you were to conduct an experiment with lots of carnivore species with floppy ears and somehow found that yes, floppy ears really do help divert water, you still wouldn't be able to say whether that is why floppy ears came to be. One alternate equally-untestable explanation could be that floppy ears were developed to look cute, but then dogs started using them to divert water to the ground. The possibility of a trait developing for one reason and then being used for another is well established; a great example is that insect wings probably originated as a means of thermoregulation.

A case study: human rape
The human implications of Dawkins and Wilson's work spawned enormous controversies, particularly with the eugenics of the Holocaust still fresh in everyone's minds. If every behavior had an evolutionary reason behind it, could we really be blamed for anything we did (i.e.  genetic determinism)? What was stopping the government from rounding up people who behaved a certain way and sterilizing them? At what point would the interpretations of the science start being bent by powerful groups of people trying to control other peoples?

By adaptationist logic, men brutally raping women is an evolutionary adaptation to ensure our genes get passed on. Even today, some people really believe that the tendency to rape a woman is a trait that evolved in men and was passed on from generation to generation because rapists would have outcompeted the gentlemen by siring more children. If it's really an adaptation, no man can really be held accountable for sexually assaulting someone... it's just "in our nature." 

There's a lot wrong about this, but I'll focus on two of the largest logical pitfalls. (For an excellent (and scathing) review by the blog Skepchick that goes in more detail, click here.)

The fact that something exists does not automatically mean it was selected for. Accidentally shooting yourself in the foot with a gun does not have an adaptive story behind it, even though people do it all the time, everywhere. For a behavior to evolve by natural selection, it needs to be heritable, variable, and convey fitness benefits.

And finally, the rape hypothesis is not falsifiable. Simply put, it's non-scientific. How can you prove or disprove a claim about behaviors tens of thousands of years ago? You can claim that spitting helped clean our ancestors' mouths out, but how on Earth would you test that? The ability to disprove a hypothesis is one of the most important aspects of science. If you cannot disprove an idea, it is no more valid than any other story you want to make up.

For simpler misconceptions regarding evolution, click here and here for blog posts I wrote in the past. 

Photo credits:
Selfish Gene - Wikipedia
Fox - The Guardian