Have you ever been walking in a dark alley and seen something that you thought was a crouching person, but it turned out to be a garbage bag or something similarly innocuous? Me too.
Have you ever seen a person crouching in a dark alley and mistaken it for a garbage bag? Me neither. Why does the error go one way and not the other?
Human beings are intensely social animals. We live in hierarchical social environments in which our comfort, reproduction, and very survival depend on our relationships with other people. As a result, we are very good at thinking about things in social ways. In fact, some scientists have argued that the evolutionary arms race for strategic social thinking—either for competition, for cooperation, or both—was a large part of why we became so intelligent as a species.
Back then, if you saw something that looked like a person, by golly it was a person.
This affinity for social reasoning, however, has resulted in systematic quirks in human reasoning about the non-human. This happens in two ways. First, we tend to see humanlike agency where there isn’t any, a common form of pareidolia. Many people view the sun as happy, for instance, and in religions the world over, diseases are seen as curses cast by witches. This effect has been argued to be one of the main reasons religions exist at all: People imagine that there must be supernatural beings behind the scenes, making the world work the way it does.
Second, we are more prone to believe in explanations when they are couched in terms of the everyday psychology people use to explain and predict people’s behavior. Teachers sometimes take advantage of this, using “anthropomorphic” glosses on natural phenomena to help their students learn (e.g., “the water wants to find its level.”)
Why would we evolve to have a systematic error like this? Like most biases, it takes advantage of patterns in our environment to help us (or, more accurately, paleolithic people) reproduce and survive. In the environment where humans first evolved, mistaking a log for a lion is much safer than mistaking a lion for a log, favoring the survival of those who err on the side of seeing agency in many places. And for a hunter-gatherer at greater risk from wild animals and interpersonal violence than we face today, living things tend to be more dangerous than non-living things. We tend to see agency in everything, and children have it more than adults, suggesting that it has an inborn element.
There are some interesting ramifications of this. In the 1990s, human-computer-interaction researchers Reeves and Nass replicated social psychology experiments, but rather than interacting with other people, participants interacted with computers. For example, the researchers put a blue ribbon around a participant’s arm and a blue piece of paper around a computer’s monitor.
Participants were told that that computer was on their team, and that another computer, adorned with red paper, was on the other team. Participants believed that the spell checker on the “teammate” computer caught more errors. This is because we think about computers (or characters in fiction, or gods) using the same reasoning processes we do when we reason about other people. That experiment is just one of many fascinating (and often hilarious) examples.