The principle of ​​parsimony — which suggests that simpler explanations are more likely to be accurate than complex ones — can itself be oversimplified. Scientists do prefer simpler explanations over more complex ones, but only if the two explanations account for the available evidence equally well. When parsimony and evidence support different explanations, evidence wins. A recent example from research on animal vision illustrates this evidential trump card …

We humans perceive the world as a continuous rainbow of colors — reds, shading into oranges, shading into yellows, and so on. Yet this subtle gradation is rooted in a discrete biology. The color receptors in our eyes (i.e., cones) come in only three varieties — one sensitive to red light, one sensitive to blue light, and one sensitive to green light. Each of these three types of photoreceptors is most sensitive to light of a particular wavelength. So, for example, the blue color receptors in your eyes might be most sensitive to light of the wavelength 430 nm and would be stimulated to a lesser degree by light at 400 nm or 450 nm. So how do we get from cones that detect three distinct colors to perceiving all the different shades of the spectrum? In the late 1800s, scientists came up with the idea that our eyes and brains process this information by comparing how much each type of cone is stimulated, and encode the difference in levels of stimulation among different cone types as a shade of color. This idea is now known as the opponent process ​​theory. (Note that the idea is called a “theory,” not because it is still being tested — though it is — but because it helps us understand and generate ​​expectations about a range of different neurological phenomena.) Subsequent research has backed up this theory of how color vision works, not just in humans, but in other animals as well … and then along came the mantis shrimp.

Mantis shrimp, also known as stomatopods, are ocean-dwelling crustaceans that look something like colorful lobsters with long, raptorial claws. Stomatopods have complex eyes and can detect circularly polarized light, a feat that no other species can accomplish unaided (we humans, of course, can used special equipment to detect circularly polarized light). When biologists began studying this crustacean’s color vision in the 1980s and 90s, they discovered not the three varieties of color receptors that humans have, nor the four that zebrafish have, nor even the eight that some butterflies seem to have — but 12 different receptor types, each sensitive to a different wavelength of light!

According to the opponent process theory, more photoreceptor types should lead to a better ability to distinguish shades from one another across the spectrum. Two shades of scarlet might look identical to us, but theoretically a mantis shrimp would easily be able to tell the two hues apart. Their world must just be a more colorful place, biologists reasoned. But why? What could the mantis shrimp be doing with all those color receptors? Evolutionary biologists immediately began generating evolutionary ​​hypotheses to explain this ​​observation. Perhaps mantis shrimps’ extreme color vision allows them to identify predators and prey in their colorful coral reef habitats. Or maybe it lets them communicate with other stomatopods of their own species. Or maybe it helps them distinguish their own species from others so that they don’t accidentally mate with the wrong individual.

Amidst this flurry of hypotheses, a team of biologists from China and Australia decided to test mantis shrimp color vision directly. The opponent process theory suggested that the animals would ace the test, but the researchers wanted to check that the theory still held true for this case. After all, good science involves skepticism, and occasionally reexamining and retesting ideas that have proven useful in the past. To investigate the crustaceans’ color vision, the researchers trained stomatopods to associate a particular hue with a food reward and then offered them a choice between the shade they were trained on and a similar shade. If the mantis shrimp consistently chose the shade they were trained to prefer, that indicated that they could tell the difference between the two shades, and if the mantis shrimp chose both colors with similar frequency, that indicated that the animals saw the two shades as the same color.

So are mantis shrimp super-seers when it comes to color? No. In fact, they practically flunked the color vision test and could barely tell the difference between yellow and orange! The opponent process theory generates the expectation that mantis shrimp should perform at least as well as humans on this test, and yet when we actually tested this ability, they did much worse. The researchers concluded that, for mantis shrimp, the opponent process theory must not apply. Instead, stomatopod behavior is much more consistent with the idea that each of their twelve types of receptor works independently with no comparisons made between the different types — i.e., their visual system classifies all shades into one of twelve possible colors and they don’t see the shades that would be made by combining two of those colors. Stomatopods’ color vision system really does sense just a few discrete colors — unlike those of all of the other animals we’ve studied so far. The researchers hypothesize that this simpler system may allow mantis shrimp to make quicker judgments about color. Further tests will be needed to evaluate this idea.

In this case study, biologists had to weigh two potential explanations for color vision in the animal world: (1) that color vision operates as the opponent process theory describes, and (2) that some color vision systems operate according to the opponent process theory and some operate in other ways, such as by sensing discrete colors. The first explanation is simpler and much more parsimonious. The second explanation is more elaborate and requires that we test each visual system separately to understand how it works. And yet, the second explanation is the obvious winner for biologists because the evidence in favor of it is so clear. Parsimony is an important consideration — but not as important as evidence. When the evidence suggests that our previously held, simple idea is incorrect, it must be modified or replaced to account for new observations.