Anthropomorphic Bias
David Kyle Johnson
They have hearts, as much as we do. And they have desires, and they want to give as much as we do. And they do, if not more.
Pet psychic customer on Penn & Teller: Bullshit
One displays an anthropomorphic bias when one displays a tendency to ascribe humanlike characteristics, usually mental properties or agency, to things that do not have it.
The most obvious (but harmless) example occurs when your computer malfunctions and you yell at it, as if verbal reprimand might convince it to do what you want. If it malfunctions at an inopportune time, you may even conclude that your computer is out to get you. If so, you are ascribing humanlike agency to your computer that it does not have; you’re displaying an anthropomorphic bias.People are apt to do this with animals as well, especially their pets. Now, of course, I am not suggesting that animals do not share some characteristics with humans - even some mentality. My dog, Alex, for example, feels pain, gets hungry, and likely feels some level of wants and desires. He might even feel affection for me. What our pets do not have is a psychology that is as complicated as a human’s. Their brains simply are not complicated enough. The lady quoted at the beginning of this chapter, who took her dog to a pet psychic, has not only mistakenly concluded that psychics can read minds; she has also mistakenly concluded that her pet has a mind and psychological
life as complicated as a human’s. She has anthropomorphized her dog. I’m not saying that there is anything wrong with treating your pets as if they are your children; I’m guilty of that myself. But concluding that they actually have a mental life equivalent to a child’s would be fallacious.
An example that is a bit more obvious and egregious is when people conclude that ghosts are the cause of unexplained noises or motions.
Although this invokes the mystery therefore magic fallacy (see Chapter 38), it also includes ascribing humanlike agency to something that does not have it: the cause of the noise or motion. One student of mine had a boyfriend who committed this fallacy: “Every time it gets cold in my room I hear a rattling sound. Therefore a ghost must be causing the cold and the sound.” Anthropomorphizing the cause of the noise and cold into a ghost kept him from realizing that it was steam moving through the pipes because the heat was kicking on. The results of one study, by Willard and Norenzayan (2013), even shows that the anthropomorphic bias leads to belief in the paranormal.A similar example is found when we ascribe agency to weather. Even today when lightning strikes, tornados hit, or a hurricane destroys property, we are apt to think that there is some kind of reasoning or purpose behind it. Indeed, it seems likely that this is where belief in gods like Zeus originated. In the absence of an explanation for lightning, the ancients concluded that it was caused by a powerful person. A similar kind of logic could lead you to believe that things like the sun and moon are gods. As I have often thought upon seeing a full moon rise on a clear night: “If I hadn’t been taught by science what that was, I’d probably worship it as a god too.” Indeed Stewart Guthrie (1995) has argued that all our belief in the supernatural - including our belief in God - originates from the anthropomorphic bias.
Why are we apt to anthropomorphize non-human objects? It’s likely hardwired by evolution. The explanation often goes something like this: In our hunter-gatherer days, if we heard a rustle in the nearby bushes, we were better off concluding that it was something with a mind (like a tiger) out to get us. We’d run away, and that made us more likely to survive. Withholding judgment on the matter until more evidence was gathered would have been the more rational conclusion to draw - after all, it’s likely just the wind - but given that it occasionally would be a hungry tiger, drawing the rational conclusion would make one less likely to survive.
Although a tiger is not a human - so this is not exactly an example of full-blown anthropomorphizing even though humanlike agency is ascribed to it - it is likely from this kind of selective pressure that our tendency comes to ascribe agency where it doesn’t belong.But we are also apt to anthropomorphize because the tendency to do so does involve a bit of sound reasoning. When something behaves in a humanlike fashion, it is somewhat rational to conclude that it has some humanlike properties. The fact that my dog yelps when I accidentally step on his toe is a solid indication that he feels pain like I do.
Indeed, one of the most famous arguments in philosophy follows this line of reasoning: the Turing Test. It suggests that if an artificial intelligence displays an understanding of language that mimics that of a human, one should conclude that the artificial intelligence actually does understand language like a human does. Some have claimed the argument as fallacious because of the anthropomorphic bias, but I have our argued elsewhere (Johnson 2011) that the argument is sound.
Indeed, the solution to another philosophical argument suggests that such reasoning is sound - the problem of other minds. The problem of other minds points out that each person is only aware of his own mentality; I am only aware of my mind, you are only aware of yours. Consequently, every person has no direct evidence for the conclusion that anyone has a mind but herself. For all you know, the argument suggests, you are the only one that has a mind. The solution is found, however, in realizing two things: (1) because everyone else behaves essentially like you do, they behave as if they are minded; (2) the hypothesis that they are minded is the best explanation for such behavior. Thus I am justified in believing others have minds (even though I cannot prove it by observing their minds directly).
So this rule - if something behaves pretty much like I do, it is likely minded pretty much like I am - seems reasonable.
This is why, if we one day invent androids that behave essentially like humans, I think we should conclude that they have minds. One is guilty of the anthropomorphic bias, however, when one stretches this kind of reasoning too far - when one sees a single or limited number of things that remind him of humanlike behavior and then jumps to the conclusion that the entity in question possesses a humanlike mind. For more on the anthropomorphic bias see Strongman (2007).References
Guthrie, Stewart. 1995. Faces in the Clouds: A New Theory of Religion. New York, NY: Oxford University Press.
Johnson, David Kyle. 2011. “Watson in Philosophical Jeopardy?” Psychology Today, February 15. https://www.psychologytoday.com/blog/plato-pop/201102/watson- in-philosophical-jeopardy (accessed October 2, 2017).
Strongman, L. 2007. “The Anthropomorphic Bias: How Human Thinking Is Prone to Be Self-referential.” Working Papers No. 4-07. Lower Hutt, NZ: The Open Polytechnic of New Zealand.
Willard, A.K., and A. Norenzayan. 2013. “Cognitive Biases Explain Religious Belief, Paranormal Belief, and Belief in Life’s Purpose.” National Center for Biotechnology Information. US National Library of Medicine. Web. http://www.ncbi.nlm.nih. gov/pubmed/23974049 (accessed October 2, 2017).