Seeing Ducks

Duck tiles

I can vividly recall the bathroom floor linoleum in the trailer where I lived as a kid. It was a cheap design of rectangles and squares, each an earthy shade of brown, yellow, grey, or peach. To me they looked like ducks. Almost any two adjacent quadrilaterals could be seen as a duck profile. One shape was the bill, the other the head. Sometimes additional shapes formed the body and tail. I would stare at those ducks endlessly, usually while sitting on the toilet. But it was only a simple geometric design. They didn’t really look very much like waterfowl, my brain just created pictures out of the tessellations. You’ve probably done this, too. Humans find meaning in random noise all the time, a phenomenon called apophenia.


Perhaps the biggest challenge in biology is distinguishing signal from noise. In evolutionary biology, we want to know which changes were favored by natural selection and which were nonadaptive chance events. In genomics, we want to separate functionally important genetic variation from functionally unimportant genetic variation. These kinds of problems are tougher in biology than in many other scientific disciplines because biological systems are inherently so chaotic. To make things harder, our brains are inclined toward apophenia. It’s easy to see patterns that aren’t actually there. But good scientists know how to step back and let noise be noise.


Neutral genetic variation is a particularly difficult concept for most people. Maybe it’s because the power of DNA gets so overhyped. Suppose I were to randomly choose a hundred-letter DNA sequence from the genome of, say, an iguana. Then I find the equivalent sequence, similar but not identical, from a cobra. The average person would probably assume that the differences underlie some of the obvious dissimilarities between snakes and lizards, like the ability to grow legs or ears. But that would almost certainly not be the case. Most genetic substitutions just occur by chance and don’t really have any noticeable effect. The same is true if I did the experiment with a human and a chimp. Or with two humans. It takes a lot of work to demonstrate that a DNA variant does anything at all. Not every site in your genome is important for making you who you are.


DNA sequences are far removed from our everyday experience, so here’s a more concrete example. Imagine walking through a mature forest and looking at the leaves. Here in the Pacific Northwest, you might see maples with wide jagged hands, incense-cedars with splaying scales, hemlocks with flattened needles, manzanita with thick green teardrops, birch leaves seemingly trimmed with pinking shears, holly-like Oregon grape, and broad ferns composed of many smaller leaflets. Every leaf has a different shape. In fact, leaves are so distinct, they are one of the most useful traits for identifying a plant. But why is every plant unique? They all inhabit the same forest and experience the exact same climate. Some get slightly more light, wind, or moisture than others, but there are a limited number of such microclimate combinations that could occur in the same place. And for the most part, plants don’t use their fronds to find mates, hide from or scare predators, attract pollinators, or huddle into herds. The enemies of foliage, most of them caterpillars, are not deterred by shape. In other words, there is no evolutionary pressure for leaves to transform, stand out, blend in, or differ from the neighbors. They should all converge on the optimal shape for light capture and stay that way. But in over 400 million years, that hasn’t happened. Instead we just get endless forms most beautiful.


Some leaf diversity is a result of natural selection, but most of it is surely neutral. Plenty of other physical traits, fingerprints for example, show similarly inconsequential variation. Differences in form are not necessarily differences in fitness. Some features, like flowers or animal hides, do respond directly to changes in the environment or in the brains of animals that interact with them. Polar bears really are white in order to hide in the snow. But if neutral processes can generate such a plethora of leaves, we can’t conclude that every stripe and spot on a pelt is a consequence of natural selection. Before inferring the evolutionary reason for a particular trait, make sure you’re not just superimposing a duck where there isn’t one.


When we think of evolution, we often picture species climbing a “peak” of fitness. In the classic Darwinian view, species have the bodies they do because that’s what brought them to the top of the peak. But fitness might be more like a long meandering ridge with branching offshoots. A crisscrossing maze not unlike the lines of my old bathroom linoleum. As long as you are up out of the valley of death, there are a lot of equally high places where you can wander around. Oaks and hawthorns might just be growing on distant segments of the same ridge. So might snakes and people.


Not even scientists are immune to apophenia. We really want data to be meaningful. This is especially true if we’ve invested time, money, and emotion into generating the data. I occasionally  see colleagues touting a genetic difference they observe as having a major impact on fitness, even though their evidence is flimsy. That’s why science is set up with checks and balances, like peer review, to keep us from overinterpreting. We can, and should, try to find explanations for the structures of the natural world. But not so speculatively that we can’t see the forest for the stories. Sometimes what you see is really a duck. Sometimes it’s just disconnected lines and corners. Either way it’s beautiful.

One thought on “Seeing Ducks

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s