Wednesday, March 13, 2013


"This pelican looks like a urinal. 
Go home, evolution, you are drunk." 
Urinalification is so lol.

But it's calling evolution drunk that really got you, isn't it?

Steven Pinker (as of the late 90s) might say that personifying evolution's funny because it juxtaposes or mixes up what are normally separate rules or domains of our intuition. Personification's funny the way slapstick comedy's funny: Without any warning (but with a well-placed banana peel) a guy goes from being a dad to being a bag of molecules.When we give nonhumans human potential, Pinker says that we’re marrying “our intuitive psychology to our intuitive biology.” [1]

Personification is the same sort of thinking that enables us to discover, to know, to understand how the world works. It’s story, analogy, metaphor, unseen imaginative b.s. that we misfire like crazy. And even though we've overcome much of our reliance on it, we tend to preserve the spirit of it anyway. 

The powers of life, death and destruction aren't the only things worthy of personification. We'll give agency to a triangle or square. Just shapes. Being two-dimensional and moving around is all it takes. We are constantly up to this and as Dawkins said, 

“The reason we personify things like cars and computers is that just as monkeys live in an arboreal world, and moles live in an underground world, and water striders live in a surface-tension dominated flatland, we live in a social world. We swim through a sea of people. … We are evolved to second guess the behavior of others by becoming brilliant intuitive psychologists. Treating people as machines may be scientifically and philosophically accurate, but it’s a cumbersome waste of time if you want to guess what this person is going to do next. The economically useful way to model a person is to treat him as a purposeful goal seeking agent with pleasures and pains, desires and intentions, guilt and blameworthiness. Personification and the imputing of intentional purpose is such a brilliantly successful way to model humans, it’s hardly surprising the same modeling software often seizes control when we’re trying to think about entities for which it’s not appropriate, …”[2]

Although many other animals have desires and pleasures and pains and generally have experiences—at least they do as long as we assume we do—we should extend Dawkins’ reflection to our personification of nonhuman animals. There's very good reason we do this as scientists. It's not a terrible null hypothesis that if you observe an animal behaving like a human could, then it might be thinking like a human could. This is the argument by analogy or as I call it doing demonstrates knowing. How else could you reverse engineer cognition? You're a human, you think like one, you know how humans think from a scientific perspective better than you know how any other animal thinks. This makes sense. It's a good default. But that's all it is. And scientists get that. They taught me all this after all. (Reading a lot of Shettleworth, Povinelli, and Tomasello lately.) But the rest of us who don't work on animal behavior and who have very little experience with real animals at all are taking this perspective literally without caution. For example, maybe we should perk our ears but cock our heads skeptically when the website for Dognition (a service rooted in good science that sells cognitive testing for your furry friends) says, “Learning who your dog is as a ‘person’ will help you make the most of the time you spend together.”

Sometimes anthropomorphism smells like money and sometimes it makes us angry for no good reason. For example. Ed Yong wrote a piece "Male frog extracts and fertilises eggs from dead female" about an Amazonian frog that bleeps to death and still procreates. (No survival necessary--take that, Darwin!) Yong didn't personify the frogs but that didn't stop a reader. 

I have a hunch that our habit of anthropomorphizing (combined with some empathic and also narcissistic tendencies) makes it much harder for us to understand reality, to understand scientific explanations for the universe, and to stumble onto new ones.

I wonder a lot about how our language encourages these mistaken habits. I've written before about how I don't use "force" to describe processes and mechanisms of evolutionary change. It's one of my f-words of evolution.

I've got similar concerns about our language that describes animal behavior. Take for example this fascinating paragraph written by the primatologist Toshisada Nishida about the chimpanzees he studied at Mahale, Tanzania:
“In an effort to maximize their reproductive output, males and females may cooperate when it is to their mutual advantage. However they may also control, manipulate, betray and even attack each other when they have conflicting interests. Chimpanzees maintain a promiscuous mating system, for which females develop a huge swelling of sexual skin and males large scrota. These morphological features make sperm competition the characteristic features of chimpanzee sexual activity. Males are expected to mate with as many females as possible, while female are expected to be more choosey because reproductive output over their lifetime is more limited. A powerful constraint on female chimpanzees is the need to consider males’ tendencies toward infanticide.”[3]
I added those italics to emphasize the language that could be interpreted to give chimpanzees more cognitive ability and more agency than the author intended or has evidence for. That paragraph, re-written to avoid any overstatement of chimpanzee cognition and any hint of calculated intention by chimps or by Mother Nature, could look like this:
In a system where higher reproductive output means more of whatever traits lead to it in future generations, males and females may behave cooperatively because it has been perpetuated by higher reproductive output in the past and will continue to do so until something changes. However, depending on the circumstance, they may also control, manipulate, betray and even attack each other if cooperation is supplanted by another behavior that is and has been perpetuated due to its increase of reproductive output, instead, in that sort of circumstance in the chimpanzee’s life. Chimpanzees maintain a promiscuous mating system and, because higher reproductive output of females with huge sex swellings and males with large scrota, females have huge sex swellings and males have large scrota. These morphological features exist, as hypothesized, because sperm competition has been occurring over chimpanzee evolutionary history. Males that mate with as many females as possible have, and have had, higher reproductive output, while females mate less frequently and have much more limited procreative potential over their lifetimes, so they appear to be choosier than males about their copulatory partners because doing so (assuming they are actually choosier than males) has and does increase reproductive output. A powerful influence on female chimpanzee mating behavior is males’ tendencies toward infanticide because females who have babies that are killed have lower reproductive outputs.
There. I managed to suck all the fun out of chimp sex. Not only that, but if you made it far enough you must be wondering how I could be so circumspect about female mate choice. That’s something for another day, but let’s just say that we should be circumspect about male choosiness too since they can’t decide which females (or males, or other circumstances) will give them erections.

Oh, hello! Now you’re awake! Why did my translation of Nishida’s paragraph fail to engage you? Here’s a try: Because it wasn’t a narrative and there were no interesting actors with calculating minds raging with intentions. There was nobody to figure out, to empathize with, to love, to fear, to root for. You just don’t tell stories without those kinds of characters.

But listen up close: that’s the true story of most of Earth’s evolutionary history. Agency evolved (maybe), but evolution itself did nothing purposefully, or even whimsically, to cause it.

Left to our naïve devices, how we talk and write about animal behavior affects how we think about their truths, their essences. The animals we put on the big screen are now just CGI manipulated reality--no cartoon drawings necessary for a pig like Babe to politely request sheep to scram, or for the pups in Space Buddies to bark interstellar commands. If only they had our mouths and throats these creatures could tell us about all the complex thoughts and feelings they experience! My mom has named this the Disney Syndrome. Some might blame the fertile “lolcats” and other such memes. We don’t just anthropomorphize animals, though, we personify oak trees, grandfather clocks, and toenail fungus.

Let's not pile all the blame our limited language, our love of agent-driven narrative, our teachers, scientists and Disney. Because how removed we are from “nature” probably affects our perceptions of animals too. Urban and suburban Americans might be the worst about this. In a piece about Joe Henrich's work, Watters writes,
“While studying children from the U.S., researchers have suggested a developmental timeline for what is called “folkbiological reasoning.” These studies posit that it is not until children are around 7 years old that they stop projecting human qualities onto animals and begin to understand that humans are one animal among many. Compared to Yucatec Maya communities in Mexico, however, Western urban children appear to be developmentally delayed in this regard. Children who grow up constantly interacting with the natural world are much less likely to anthropomorphize other living things into late childhood. Given that people living in WEIRD societies don’t routinely encounter or interact with animals other than humans or pets, it’s not surprising that they end up with a rather cartoonish understanding of the natural world.”
Nobody's saying to give it up completely. It's adorable.It's funny. It's lovely, actually. But if we coddle it like crazy in our children and if we don't check it once in a while in ourselves, this cultural phenomenon is going to continue to contribute handily to the scientific challenges that we face with increasing urgency.

Before submitting a scientific manuscript about animal behavior, evolution, climate, or any process that tempts us to infuse agency, make sure it's not drunk.

[1] From Pinker's lecture "How the mind works" published in the Annals of the New York Academy of Sciences

[2] Richard Dawkins's TED talk at about 21:00 -

[3] Nishida, T. (2012) Chimpanzees of the Lakeshore (Cambridge University Press), p. 201


Ken Weiss said...

Another great and thoughtful post, Holly! Your discussion makes me think of Darwin's expressed concern about the phrase 'natural selection'. I wonder if you would think it's appropriate to liken his concern to your topic?

He was reluctant to use 'natural selection' because he thought it implied a 'selector' (God) and he didn't want to personalize evolution. I can't remember if he actually said that's why he often used 'survival of the fittest' instead, as a depersonalized phrase that was first coined by Herbert Spencer, but I vaguely recall that's why he liked that latter phrase.

Of course, personification in the form of God is a bit different from anthropomorphizing in the way you discuss so well. But I think it's a reflection of the same sort of thinking by scientists and public alike. We want conscious agency when we see pattern we can recognize in some way, and that is what Darwin was so keen to wean us from. In a sense that may be why we want us humans to be 'in God's image and likeness' so people can imagine that God has a mind like ours.

Holly Dunsworth said...

Yes yes! I don't philosophically separate supernatural explanations from anthropomorphism or personification. Making dog or God in our image, what's the difference?

Ken Weiss said...

Geez, there's a huge difference! If God barked and arfed how would we make any sense of the prayer books, the Bible? Hymns might sound like wolves baying!

At least some religions don't personalize their equivalent of divine essence.

On the other hand, those who do personify God have to say that the crappy nature of so much of the world has to have been intentionally done, by the God to whom we somehow still think we can supplicate for mercy!

Holly Dunsworth said...

Well, jeepers, I'll answer myself. A big one: Dogs are real and have lots of tangible real stuff in common with us. Gods are not and have anything in common with us that we like!

Holly Dunsworth said...


Holly Dunsworth said...

to whom does my dog supplicate when he play bows? I am dog's god.

Anonymous said...

Great post. But I don't believe it is a bad or avoidable habit.

You said: >>>>>>I have a hunch that our habit of anthropomorphizing (combined with some empathic and also narcissistic tendencies) makes it much harder for us to understand reality, to understand scientific explanations for the universe, and to stumble onto new ones.<<<<<<<<

I would rephrase that to: We have a hard time understanding abstract concepts because of our tendency (need even) to anthropomorphize things. Our brain is optimized to process data in such a manner so it comes as no surprise that we tend to reframe in this way.

I do not consider this a bad thing, it's something we have to live with, in order to reason and learn things with increased facility.

No one, not even scientists can be freed of this. Chemists will label certain chemicals as temperamental. Mathematicians named numbers irrational and positive. Einstein's genius was his facility with analogical thinking - breaking down problems and then casting them to common place settings. It is a side effect of how we learn [1], how we naturally use language, and how we evolved. Exploit this, don't lament it.

Only caution I would suggest is to be aware that this is not how things are (as Kahneman often warns), learn the details but still reason with your own private analogies (which is what experts do). Anthropomizing helps novices and children get going. Till they get to the expert's compromise of a collection of faithful analogies - with dipping down to the abstract when necessary. No point arguing how much stronger we would be if gravity were only just a little lighter. heh.

Also I think you might enjoy, if you have not already encountered it.


Steve C said...

Interesting post and comments - thanks. Michael Shermer's book - The Believing Brain- delves deeply into pattern and agency and our human tendencies to find each of those everywhere. Oddly enough, anthropomorphication (is that a word?) doesn't enter into his story. Good read, though...

Anonymous said...

It's common in IT circles to say, for example, "Program A needs some memory to eat the images and spit out spectra, but it can't, because program B is hogging all the memory for walking in circles". I feel like, contrary to this article, this isn't ascribing human-like motivations and actions to programs. Rather, it is the recognition that since humans and programs alike are systems of varying complexity, the same terminology applies to both. It's not about looking at properties of things, deciding they look human, and then using human terminology; it's saying that humans are fundamentally no different than machines and programs and animals and weather systems, and there just happens to be a rich and evocative language that other percieve as limited to describing humans.

matt said...

this was excellent. much appreciated.

Ken Weiss said...

Much of the time this is of course pretty reasonable. We all use colloquialisms. Much of the time we know they are just that. My program 'crashed' or has a 'bug' in it are examples, as is 'junk DNA'. I think the danger is when we can subtly slip in to confusion.

A good example, I think, is the rhetoric of adaptation or gene functions that can be misleading. This is a gene 'for' breast cancer, etc.

I personally think symbols and metaphors are risky in science, no matter how cute or convenient (or relevant to marketing our research). But the main point is to keep our concepts themselves straight. And when it comes to animal behavior, I think that we very much tend to anthropomorphise and are less than sufficiently critical. It is still open whether animals 'think' in ways we recognize as relevant, for example, to consciousness. One can make arguments of many different kinds (and they've been made for ages) in this area, so being careful about it is important.

Personally, I think dogs are conscious, think in ways we would recognize, and so on. I wonder about ants and butterflies...