Funes the Memorious and the Useful Lie of Categories

We've evolved to misunderstand our world. Models are making it worse. Here's how to think smarter, by channeling our inner Funes the Memorious.

Before I dive into this edition, I want to say a big thank you to everyone who has supported me in the release of my new book, FLUKE: Chance, Chaos, and Why Everything We Do Matters. It is making a big difference. So much so, that this extraordinary sight can greet you if you check out the Amazon Books billboard near Penn Station in New York City this week. It’s surreal and wonderful—thank you for the part so many of you played by supporting my work.

---

I: Funes the Memorious

In 1942, as World War II was raging, the Argentine author Jorge Luis Borges wrote a short story called “Funes the Memorious.”

The protagonist, Ireneo Funes, is a Uruguayan teenager who hits his head badly after falling off his horse. The brain injury gives Funes a blessing—and a curse. After the accident, the stream of information into his mind is much more vibrant, so overwhelming that the present became “almost intolerable in its richness and sharpness.”

But now, with this newfound and overwhelming flow of vivid sensory experience, Funes is astonished to recognize that his memories have become not just more vivid, but infallible. Nothing is missing, no detail lost. He can replay the tape of his past life just as clearly as he can experience the present.

“He knew by heart the forms of the southern clouds at dawn on the 30th of April, 1882,” Borges explains, “and could compare them in his memory with the mottled streaks on a book in Spanish binding he had only seen once and with the outlines of the foam raised by an oar in the Rio Negro the night before the Quebracho uprising.”

But this attention to detail is a paralyzing curse for Funes. Because he can see that every blade of grass, every leaf on a tree is utterly unique, the notion of generalized concepts baffles him. He becomes physically incapable of seeing the forest, mesmerized by the captivating differences between each individual tree.

“Not only was it difficult for him to comprehend that the generic symbol dog embraces so many unlike individuals of diverse size and form,” Borges writes. “It bothered him that the dog at 3:14pm (seen from the side) should have the same name as the dog at 3:15pm (seen from the front).” As a result, Funes becomes unable to function in the normal human world, burdened by the overwhelming and endless stream of information bombarding him constantly.

Funes the Memorious experiences reality as it actually is—and it’s a crippling curse. We, normal humans, experience the world in simplified snapshots, crammed into generalized, abstracted categories captured by language and represented by symbols.

It helps us survive, navigating an infinitely complex world. But now, we, unlike Funes, have gone too far in the other direction—paying too much attention to the forest and forgetting that every tree matters. In a world overwhelmed with models, our cognitive shortcuts have blinded us not just to the rich details of life, but also to ways to better understand reality.

The cognitive shortcuts we take are sometimes useful. But it’s important to also remember what’s actually true.

II: The puzzle: spot the cognitive trap

When I was writing Fluke, I had a conversation with a friend of mine—an incredibly smart historian. I told him the story that opens the book, about a tourist couple that vacationed in Kyoto, Japan in 1926 and fell in love with the city.

Nineteen years later, the husband—Henry Stimson—intervened to stop the atomic bomb from being dropped on his “pet city.” Hiroshima was bombed instead, all because of a vacation a couple had taken 19 years earlier.

It was, I told my friend, a story that showed the bewildering contingencies of history and life; chaos theory in action.

“Sure,” he responded. “But did it really matter in the grand scheme of things? If Kyoto had been destroyed instead of Hiroshima, the US still would have won the war.”

He’s right. And similar objections are raised whenever we discuss the origin story of World War I. It was triggered by a happenstance assassination in Sarajevo, when Archduke Franz Ferdinand’s car sputtered to a stop right in front of his would-be assassin, Gavrilo Princip. (As I’ve highlighted previously, the Archduke also nearly died ten months earlier in a hunting accident at Welbeck Abbey, in England).

These events, many historians argue, are mere curiosities. The assassination in Sarajevo may have been the trigger, but the forces of history were already set. The Great War was inevitable. Does the exact trigger really matter?

Those making this argument are falling into a cognitive trap—one that’s ubiquitous in the ways that most of us are told to perceive and process the world. And now, as our world becomes increasingly complex, it’s getting us into trouble, producing a dangerous hubris that convinces us that models can solve everything.

Have you worked out what the cognitive trap is yet?

I’ll give you a clue: Funes is right.

III: The evolution of categories

There is no such objective thing as a “weed.” I don’t meant that weeds are figments of our imagination. Rather, the word is a human construct, a way of differentiating plants that are desirable for our purposes from those that are undesirable competitors in the same patch of soil. And it’s useful: if you’re growing crops to survive, separating out the plants we call weeds from those that provide food is essential. If you depend on the soil to sustain human life, mistaking the former for the latter could be a deadly cognitive error.

When we are children, we get taught how to name things. As Roger Brown writes in his seminal 1959 paper “How shall a thing be called?”:

“The most deliberate part of first-language teaching is the business of telling a child what each thing is called…The dime in my pocket is not only a dime. It is also money, a metal object, a thing, and, moving to subordinates, it is a 1952 dime, in fact a particular 1952 dime with a unique pattern of scratches, discolorations, and smooth places.

When such an object is named for a very young child how is it called? It may be named money or dime but probably not metal object, thing, 1952 dime, or particular 1952 dime. The dog out on the lawn is not only a dog but is also a boxer, a quadruped, an animate being; it is the landlord's dog, named Prince.”

If we take the worldview of Funes the Memorious, in which no two objects are ever quite the same, that dime becomes incomparable, unique, not part of another category. And that makes it impossible to create abstract generalizations, or, in short, to learn.

For the most part, then, the way we decide how to call things is based on their utility to us, an evolutionary shortcut that helped us survive. Our minds have been shaped to have an architecture that lends itself to rapid categorization, learned through experience and imprinted through language.

As the Canadian cognitive scientist Stevan Harnad put it: “All of our categories consist in ways we behave differently toward different kinds of things, whether it be the things we do or don’t eat, mate with, or flee from, or things that we describe, through our language…and isn’t that all that cognition is for?”

For example, every fork is, strictly speaking, unique. But we don’t need to pay attention to the slight bent of a given tine in order to use it. Categories help us figure out how to behave.

If any of our distant ancestors had a cognitive superpower (or supercurse) like Funes the Memorious, we’re not likely to be descended from them. And that’s because being fascinated by the uniqueness of every detail on the fang of saber-toothed tiger makes you lose sight of the fact that the only relevant information you need to pay attention to is that it’s a predator that’s about to eat you. Evolution purged cognitive behavior that fixated on specifics at the cost of being able to process the world through useful categories.

Or, as the psychology professor Eleanor Rosch once observed: “the task of category systems is to provide maximum information with the least cognitive effort.” But it’s not just maximum information, because that would overwhelm us, like Funes. Instead, it’s the task of our brains to find, filter, and process the maximum useful information.

We are category creatures.

IV: Classification systems and chunking up time

Borges, in another short story, wrote of a classification system used by a fictional Chinese encyclopaedia known as the “Celestial Emporium of Benevolent Knowledge.” It classified animals according to this system:

“On those remote pages it is written that animals are divided into (a) those that belong to the Emperor, (b) embalmed ones, (c) those that are trained, (d) suckling pigs, (e) mermaids, (f) fabulous ones, (g) stray dogs, (h) those that are included in this classification, (i) those that tremble as if they were mad, (j) innumerable ones, (k) those drawn with a very fine camel's hair brush, (l) others, (m) those that have just broken a flower vase, (n) those that resemble flies from a distance.”

All classification systems are arbitrary. But the point Borges is making is that this one strikes us as absurdly comical not because it is more arbitrary, but because it is largely useless for human purposes. Some of the categories are over-specific (those that belong to the Emperor; stray dogs), while others are far too general (fabulous ones; innumerable ones). And whether an animal has “just broken a flower vase” is less useful information for us than whether an animal is a predator that has teeth, or whether it produces milk.

To make sense of the world, then, we chunk things up into bits that we can pragmatically process. We do it automatically. If I ask you to describe the start of your day, you might begin with brushing your teeth. But you wouldn’t describe walking into the bathroom, picking up the tube of toothpaste, applying pressure with your right hand to the bottom of the tube, and so on. You’d just say “I brushed my teeth,” a useful shorthand for a more complex reality. (Funes, by contrast, would provide a mechanical description—one that’s certainly more accurate, but far less useful).

This is how we process our social world, too. (In case you missed it, I explain how we make sense of politics through cognitive shortcuts known as schemas). For example, every politician is different—all humans are unique—but it’s still helpful shorthand to know whether a self-professed Republican is “pro-Trump” or “Never Trump.”

Human language is therefore designed to split the jagged edges of reality into smooth categories of individual things that can be separated and neatly labelled. But scrutinize those categories more closely and they suddenly aren’t so smooth.

We think we know what a forest is, but we can’t pinpoint where it ends. Is the bee that provides the crucial services of pollinating flowers still part of the forest when it flies outside of the trees? Which specific tree or blade of grass marks the end of the forest and the beginning of something else?

When I included those few lines in an early draft of Fluke, my editor pushed back with his typical take-no-prisoners New Yorker style:

“I can picture a certain type of reader saying, ‘Hey pal, I can tell you where the forest is not. Ever see an asphalt parking lot? Ever see the Antarctic ice shelf?’ In truth, it does seem that while the exact edges of things are hard to ascertain (their ends tend to trail off in an asymptotic curve), they do give way to other things. There is a point where the table yields to the chair.”

As usual, he was right. But why is it so hard to pinpoint where exact boundaries lie? (That’s a problem that the great philosopher of language, Ludwig Wittgenstein, grappled with in his writings). These thorny conundrums are why the Ship of Theseus is such a bewildering paradox; if you take a wooden ship and replace every single plank, one piece at a time, like for like, is it still the same ship as the original? If not, when does it stop being the same ship?

We have no good answers.

That’s not because we’re stupid. It’s because we are trying to shove a square peg (categories) into a round hole (an infinitely complex reality). When we try to demarcate any exact boundaries, that’s the moment we are forced to confront the lie: the way we perceive and process the world and the way the world actually is…don’t perfectly align.

Funes the Memorious has no such problem, because Funes can’t understand categories.

V: Breaking free from the cognitive trap

We return, now, to the puzzle I posed earlier about the onset of World War I and World War II. My historian friend was right: the United States likely would have won the war whether the first atomic bomb was dropped on Kyoto or Hiroshima.

But thinking that way is a mistake that blinds us to the real dynamics of cause and effect.

Nonetheless, it’s a way of thinking we are told to follow by all the “smart thinkers” and self-proclaimed data gurus like Nate Silver, who tell us that we should think about “signal,” but ignore the “noise.” The signal, they tell us, comes from the abstractions, the patterns, the generalizations. The noise comes from the “pointless” detail. Which Japanese city got bombed is “noise.” The inevitable Allied victory is the “signal.”

That way of thinking only makes sense if you wrongly imagine the world as being comprised of a series of binary outcomes (win the war or lose the war; war starts or war doesn’t start) that can be crammed into subjective, arbitrary human-made categories.

“World War II,” like a forest, is an imperfect category. When, exactly did it begin? We have conventional answers, but each explanation has tentacles that reach further and further back into history. More to the point, how the war starts matters. Which cities get destroyed matters. It all matters.

No matter how much we pretend otherwise, history—and our lives—aren’t defined by a series of discrete, separable snapshots with a few key variables and obvious cutoff points in terms of timing. Reality is a continuous variable, an unbroken string of causes and effects, a fluid stream of chaotic dynamics—every one of them important—that each contributes to eventual outcomes. After all, the interlocking cause and effect pattern that produced Hiroshima’s destruction on August 6, 1945 pivoted, in quite a large way, on a 1926 tourist vacation.

The noise matters. We ignore it at our peril.

This brings us back to the cognitive trap. The British statistician George Box once quipped that “all models are wrong, but some are useful.” The same is true of categories. As Funes the Memorious might correctly insist, “all categories are wrong, but some are useful.”

We have fixated so much on how we can usefully manipulate variables and process data in models that we’ve forgotten what’s true. We see the world through the funhouse mirror of modeling, in which the details don’t matter much, so long as we understand which categorical box an event belongs in for ease of processing.

But the details—the noise—well, that’s where the seeds of momentous change are most often planted. They are not “weeds” to be removed, but are rather rich offshoots of complexity that can, and do, change our world. Black Swans usually emerge from “noise,” not “signal.”

Consider this: does it matter that September 11th happened on September 11th, 2001 and not September 10th or 12th? The answer, if you ponder it carefully, is obviously yes. Why? Because on September 10th, 2001, there were storms across the East Coast. Some of the hijacked planes might have been delayed, perhaps limiting the damage of the attack.

Or, if the planes had been hijacked on September 12th, there would have been different passengers. What if some of those passengers were less brave, and the plane intended to hit the White House or US Capitol actually destroyed the target? Would January 6th, 2021 have happened if the US Capitol had been reduced to a smouldering ruin two decades earlier? It’s impossible to say, but it’s obvious that small tweaks, even with day to day timing, could have radically changed American history.

The cognitive shortcuts we use to survive are mismatched with the complex reality we now navigate. And a lot of the smartest people around, those who are tasked with making decisions, navigating risk, and avoiding catastrophe have drifted too far from that reality, mistaking simplified models and subjective categories that we call “variables” for the unimaginable underlying complexity of the real world. This isn’t to say we shouldn’t use the models; of course we should. As Borges writes, “to think is to forget differences, generalize, make abstractions.”

But cognitive shortcuts, tantalizing as they are, can also make us lost.

It’s worth remembering that the categories we use to make sense of the world are not true. They can act as a useful lie. But if we don’t remember that they are a lie, we’ll make catastrophic mistakes, misunderstanding the world that we are so diligently trying to usefully navigate.

That is why we must remember the paralyzed wisdom of Funes the Memorious, a character who, in experiencing the “teeming world” as it actually was, became “the solitary and lucid spectator of a multiform, instantaneous, and almost intolerably precise world.”

Thank you for reading The Garden of Forking Paths. If you found this essay interesting or thought-provoking, please consider upgrading to a paid subscription. In doing so, you support my work, make my writing sustainable, and unlock dozens more articles that are for paid subscribers only. Alternatively, you can support my work by buying my new book, FLUKE, or by telling many of your favorite friends, acquaintances, enemies, etc., about this newsletter.