Art History Without Geniuses

I think that the question “How do creative ideas appear?” isn’t appreciated enough by the academics. It should get more attention. The reason is simple: if we better understand what creativity is and how it works, we can become better problem-solvers. Cancer? Poverty? Donald Trump? The biggest problems that the humanity faces may become manageable if we figure out the mechanism of inventive problem-solving, a.k.a. creativity.

Figuring out this mechanism isn’t easy, though — for two reasons. Reason 1 is boring: tackling this problem requires yet unprecedented levels of collaboration between neuro- and cognitive scientists, biologists, humanists, artists, etc. — that is, between those who’d prefer not to collaborate and instead stay in their comfy cocoons. But this is a standard issue in research, so let’s leave it aside.

Reason 2 is less boring. Creativity already has an “explanation,” which, for many, may appear sufficient: a genius.

In this post, I will reinstate the obvious: the notion of a genius is outdated, and we need a newer explanation of creativity. Also, I will say something less obvious: this explanation can be partly inspired by the evolution theory.

“Genius” is alfa and omega in many conversations about creativity. How did Einstein come up with E=mc²? He was a genius! How did Tolstoy manage to write his (four-volume) War and Peace? He was a (diligent) genius! How did van Gogh come up with the Sunflowers? He was a (slightly mad) genius! Genius! Genius! Genius! A tautological “explanation” that doesn’t expain anything. Instead of answering what creativity is, it asks us to stop asking. Geniuses are simply “more clever” and “more gifted”… They create ideas out of thin air. End of discussion.

This is probably the best illustration (that I can think of) of the idea of genius. Einstein’s brain is unique — that’s why we should dissect it and acrefully examine. There must be something inside these convolutions that makes him better than us, mortals…

The idea of a genius reminds creationism: it’s a simplistic answer to a complex problem. Creationism may not look harmful at first, but it offers a distorted understanding of nature, of its laws and forces. Similaly, the notion of a genius distorts our understanding of culture, of its principles and mechanisms. Instead of a proper picture of how innovations emerge, the standard genius narrative is a fairy tale about prodigy kids with unique “gifts.”

Surely, not many scholars of culture believe in geniuses, as not many scholars of nature believe in god: most of them stick to more rational views. (Steven Johnson’s Where Good Ideas Come From is an eloquent advocacy of this rational perspective.) Nevertheless, the creationist idea of a “genius” keeps thriving in two different domains. One, unsurprisingly, is popular culture. Movies like The Social Network or The Imitation Game depict innovators as unique talents, misunderstood by us, simple-minded. Another domain, more surprisingly, is the humanities research. Take a look at this plot:

Literary inequality: the most “prestigeous” authors in world literature. Prestige is mesaured by the number of academic papers on each author in several major U.S. journals. The plot shows only the initial 50 names: I had to cut the long tail of 1,440 authors. (The data was provided by J.D. Porter from the Stanford Literary Lab; his paper on literary prestige vs. literary popularity will appear soon in the Lit Lab pamphlet series.)

Immediately, I have a bunch of questions. Why is Shakespeare four times more prestigeous than Joyce? Or, even more: why is Joyce 25 times more prestigeous than Henry Miller? Two, maybe three times—I get it! But 25 times?

If we address these questions to literary critics, most of them probably won’t confess: they won’t tell that it’s because these more prestigeous authors are “geniuses.” But the truth is somewhere near. What else than a miracle of a genius can explain such an astonishingly unequal distribution? What could possibly be the reason to dedicate 99% of scholarship to a small group of authors who died decades or even centuries ago? This unequal distribution even has a name: the canon. And “genius,” the very idea of it, underlies the canon.

We shouldn’t blame literary critics too much, though. Apparently, this inequiality of attention is present not only in the academe. Here is another plot, made on the basis of readers’ rankings on GoodReads.com.

This plot is based on two user-generated lists of the “best cyberpunk books” on GoodReads (list 1, list 2). As a proxy for populaity, I took the number of people who rated each book. Say, as of today, Gibson’s Neuromancer was rated 217,832 times. Then, I added up the ratings for each author — but only for the books that were on the lists.

Authors, again — but this time those who belong to a tiny fraction of the literary field: the cyberpunk genre. This plot is similar to the previous one, only this time instead of a “monotheistic” distribution we see a triumvirate. William Gibson, Philip K. Dick, and Neal Stephenson completely dominate the genre. For the majority of readers, they are the cyberpunk. Probably, a similar picture may be found in other genres too: a few central figures, known by virtually everyone, and the majority of others, known by virtually no one.

For most readers, cyberpunk equals to William Gibson, while most other authors are forgotten. A skewed distribution… And William Gibson equals to Neuromancer, while most other books are forgotten. Inequality, again…

Now, evolution theory. If we acknowledge that Gibson is a genius, we acknowledge that, for example, he deserves to be 64 times more popular than Rudy Rucker (another cyberpunk author, not a genius). This would be equivalent to the evolutionary idea of natural (or, in this case, cultural) selection. Gibson’s books were selected by the readers, who considered them much more worthwhile than the books of his competitors.

I wouldn’t fully rely on this explanation, though. I don’t believe that anyone can be 64 times better than anyone else. Especially so — if we are comparing two professional writers. But, if not selection, what else?

Here’s a figure — quite similar to those above — that could hint an answer.

Popularity of baby names: male names are squares, female names are circles. (Note that both scales are logarythmic.) This data, taken from 1990 U.S. Census, clearly shows that few names are super-popular (Ann, John, Peter… — we all know them), while the majority of names are rare. (From Bentley, Hahn & Shennan 2004)

Some names are very popular, while the majority of others… much less so. What is the underlying mechanism, the reason that created a preference for the “selected” few? Alex Bentley and his colleagues would tell us: the reason is no reason. There is no selection mechanism. There is no force that would select Ann and John over Daenerys or Tyrian. Most names are similar to each other in how “good” they are. For eaxmple, the fact that I got the name Oleg is a result of a pure coincidence: my parents happened to like it around the time I was born. And being called this name doesn’t make my life any better — truth be told. No selection, no survival of the fittest. Random drift. And the most interesting — and counterintuitive — part is this: like selection, drift can lead to skewed distributions, such as this one.

Now, a question: could it be that the first two plots — world canon and cyberpunk — are explained by the same logic as the third one, baby names? Could it be that the main force that made the two Williams — Shakespeare and Gibson — so famous isn’t selection? Could their fame be of a “drifty” kind?

At first, this seems unlikely. Novels aren’t baby names. Anyone who has read both Crime and Punishment and Fifty Shades of Grey (where are all these people, by the way? I don’t see them…) could confirm: some books are much better than others. Thus, cultural selection, not drift, may appear the only reasonable option.

But this selectionist view can be objected. I’m sure than Dostoevsky is a zillion times better than Stephanie Meyer, but I’m not sure that William Gibson is 64 times better than Rudy Rucker. However, the selectionist interpretation would tell us exactly that: 64 times more popular equals to 64 times better.

To me, both these explanations — complete randomness and complete selectionism — seem unrealistic. Is there a third way? I think there is one, and I will explain it on monkeys.

A proof that monkeys can look unhappy and grumpy…

Imagine a group of monkeys living in an imaginary Unhappy valley. The valley is called Unhappy because there are too many monkeys and too little bananas. In the search for food, a group of four monkeys (who happen to be shorter than the rest of their group — just by chance) takes a risky trip to the unknown lands, possibly full of predators. However, they get lucky: they find a monkey-free valley with LOTS of bananas. So, these four settle here: in this Happy valley. They give birth to many children, all of which share the genes of this initial small group. As a result, most of the monkeys in the new colony are short — like their ancestors. And — it is important to stress — they are short not because this trait is adaptive (i.e., it was not selected for), but simply due to chance: the founders of the colony just happened to be short.

In evolutionary biology, such a scenario is called the founder effect. It is an interesting combination of drift and selection. Did the initial four monkeys become successful for no reason? No. They were rewarded for courage and for being adventurous. Did their “shortness gene” become successful for a good reason? No. It was a conincidence — drift.

Do we have the founder effect in literary evolution? Well, let’s turn to the last plot.

The growth of cyberpunk, via book publication numbers. For this plot, I used only the data from the two lists on GoodReads, mentioned above. It’s a small sample — far from a comprehensive picture — but even this small sample shows a sensible trend.

This figure shows the growth of the cyberpunk genre. Cyberpunk was dormant until the mid-1980s, then was a quick rise, followed by a relatively stable high publication numbers. Dick’s Do Androids Dream of Electric Sheep? appeared in 1968, Gibson’s Neuromancer — in 1984. That is, they anticipated the rise. Dick and Gibson are generally considered the founders of the genre. They are the monkeys that entered a new territory. They discovered a new “valley,” a new niche with lots of “food”: lots of readers willing to read the stories about cyber-cowboys, evil corporations, and virtual reality.

In other words, their popularity and the artistic quality of their novels might be in a… complicated realtionship. The Wright brothers constructed the first flying airplane, but it wasn’t as fast as Concord. Dick and Gibson were among the inventors of cyberpunk genre, but their novels aren’t necessarily the best cyberpunk novels — despite being the most known novels…

Returning to our geniuses: I have a modest hypothesis. Whenever we hear that somebody is called a “genius,” we must be careful. Quite probably, the reason for this acclaim isn’t that this artist (a writer, a filmmaker, a musician…) is really so much better than others. Quite probably, the reason is different: they happened to be the founder of a new genre or technique.

They happened to be… Which means: their success is both deserved and undeserved. Success as a combination of hard work and luck — this narrative is an old one, but, while confronting anti-scientific myths, we sometimes have to repeat the obvious.