The default nature of teleological reasoning was revealed in a 2009 paper. Deborah Kelemen and Evelyn Rosset asked Boston University students to judge whether certain inaccurate teleological sentences were true. For example, “Mosses form around rocks to stop soil erosion,” and, “The sun makes light so that plants can photosynthesize.” Subjects mistakenly said they were true in 52 percent of cases. And when given a time limit of 3.2 seconds per sentence, the error rate was even higher, at 61 percent, suggesting that when we don’t have time to correct ourselves, our natural reaction is to assume things were created for a particular function.

But that 52 percent error rate among unhurried responders is already pretty high, which means undergrads may not have fully learned that teleological reasoning is unacceptable in science. So Kelemen tested professional physical scientists. Her results, obtained in collaboration with Joshua Rottman and Rebecca Seston, also from Boston University, are currently in press at the Journal of Experimental Psychology: General.

In the first experiment, 80 scholars in chemistry, geoscience, and physics from universities including Harvard, Brown, and MIT were given a test similar to the one in the 2009 paper. They judged the accuracy of sentences such as “Trees produce oxygen so that animals can breathe,” with or without time pressure. Without the pressure they fell into the teleological trap in 15 percent of cases. When rushed, they erred 29 percent of the time. The scientists performed better than the 2009 subjects, but again, a bias was revealed. (Speed increased their error rate by only a tiny margin on control sentences, so there was a specifically teleological bias.) I’m sure any Ivy professor would be comforted to know she’s a creationist at heart.

College students and older members of the community were also tested, and both groups made many more mistakes than the scientists, so it’s not maturation but being a scientist that reduces the bias.

In a second experiment, humanities scholars from the same institutions took part. Their results were similar to the scientists’: they were wrong 18 percent of the time when unrushed, 30 percent of the time when rushed. “There appears to be a limit” to how far education can revise one’s teleological intuitions, the authors write. Extended academic training and/or having the inclination to go into academia gets you pretty far, but focusing on physical sciences over the humanities doesn’t take you much farther.

Kelemen and colleagues also measured academics’ agreement that “Nature is a powerful being.” This mystical “Mother Nature” belief correlated with teleological responses, but it’s not clear whether the Gaia-like worldview increased the intuitive bias or vice versa.

Our “tenacious teleological tendencies” (as the paper calls them) influence more than our explanations of evolution and moss and photosynthesis. It guides how we find meaning in life, as I argue in Chapter 7 of my book. The sense that certain things were “meant to be” or even that “everything happens for a reason” is encouraged by the intuitive assumption that a mind is guiding all the events in our lives. As far as I know, there’s no such thing as divine intervention or karma or destiny (beyond determinism), but even the most skeptical and well-educated of us can feel we were put here on Earth for a reason, or that stepping in dog poo on the way to an interview is just so typical—of course that had to happen to you. We see the world through a lens of purpose: Why is this happening? Why are we here?Who let the dogs out? It’s comforting—except when it leads to paranoia or fatalism. But one cold fact is unavoidable, particularly in science: Sometimes shit just happens. So watch your step.