Komorebi: a word is an act of attention someone decided to keep

Japanese has a word for sunlight filtering through the leaves of trees: komorebi, written 木漏れ日, literally “the sunlight that escapes between the trees.” English needs a whole clause. I have always had a weakness for the words that get collected in weekend features, though not for the reason usually given. Their supposed untranslatability is a marketing line. Almost everything translates, more or less well, more or less quickly, and komorebi is no exception, since I just did it in a single line. What fails to cross from one language to another is not the meaning. It is the habit of noticing.

The Japanese are not the only people who see light come through a canopy. The Portuguese are not the only ones who feel saudade. The Greeks were not the only ones who sensed a difference between time passing and the moment to act. Everyone has access to the same phenomena. But certain languages, at some point in their history, found a distinction important enough to name, and then kept the name alive from one generation to the next. So the right question to ask of an “untranslatable” word is not what it means. It is this: what did someone have to notice before wanting to name it?

A light and an institution

I could line up the usual suspects here. The Greeks separated chronos, the time you count, from kairos, the opportune moment, the one that comes around only once. English has no single word for the second; it has approximations, the opening, the window, the right moment, and combines them as needed. Portuguese has saudade for an absence that remains inhabited, neither nostalgia nor simple longing. Japanese has wabi-sabi, less a word than an entire aesthetic framework in which the marks of time count as qualities rather than flaws. All of that is true and all of it has been written a hundred times. Two examples are enough, and they are not of the same nature.

Komorebi names a phenomenon. It asks for nothing more than an eye and a tree, and it survived because, for centuries, enough people wanted to talk about that particular light to keep the word from falling out of use. This is attention fixed in language at its simplest: someone looked, someone found it beautiful, someone wanted to say so, and the language kept the trace.

Terroir names an institution. The word has no satisfactory English equivalent, which is why English speakers use it as is, in French, whenever they discuss wine or cheese seriously. In two syllables it gathers a soil, a climate, an exposure, a body of inherited know-how, and the idea that a place can own a flavor. No eye alone is enough to see a terroir. It took centuries of farmers, winemakers and cheesemakers before a language judged that combination stable enough to deserve a name, then a body of law, then a system of appellations. Between komorebi and terroir lies the full range of what a vocabulary can archive, from a glance lasting a second to an entire civilization of taste.

I once watched this mechanism run in reverse. I was looking for a word for things that appear simple while concealing a complexity that has been mastered. I thought I had coined it, and then discovered that simplexity already existed, courtesy of a professor at the Collège de France. The experience is instructive. When a language lacks a distinction, you feel an itch, and the itch almost always ends up producing a word. If the word survives, it is because other people had the same itch.

Words change what you notice

Here is what struck me most. I had seen light pass between leaves thousands of times before I learned komorebi. Since learning the word, I notice it more. On a forest path, in a parking lot at the end of a summer afternoon, under the plane trees of a village square. The phenomenon has obviously not changed. My attention has.

A child does something similar when learning the names of trees. At first he sees a tree. Then he learns to recognize an olive, an oak, a cypress, a walnut. As his vocabulary grows more precise, so does his eye. He does not see more trees, he sees more things within the trees. The landscape going by the car window is no longer a green mass; it is a sequence of species he can name, therefore distinguish, therefore see.

There is something counterintuitive here about our relationship to language. We instinctively think of words as a tool for describing what we see. They also teach us what to look at. Linguistics has argued the point at length under the name of the Sapir-Whorf hypothesis; the strong version, in which language determines thought, has been abandoned, while the weak version holds up. In 2007, Jonathan Winawer and his colleagues showed in PNAS that Russian speakers, whose language obligatorily separates goluboy from siniy, light blue from dark blue, discriminate between those two shades faster than English speakers, for whom all of it is blue. Language does not manufacture color. It adjusts the fineness with which we perceive it.

Vocabulary thus becomes an archive of the real. It is not only reality that produces words; it is also words that allow reality to appear. What has no name exists, but slips past. What has a name catches.

Blue arrived late

Colors are the best laboratory for this story, because they have always been there while their names have a date of birth. Greek and Latin had no stable word for blue. The nearest terms, glaukos and kyaneos in Greek, caeruleus in Latin, cover gray, green, dark and gleaming just as readily as blue. Homer describes the sea as wine-dark, an oddity that intrigued Gladstone, British prime minister and Hellenist in his spare hours, enough for him to devote an entire chapter to it in 1858. The sky over Troy was no less blue. No one had found it useful to single out.

French repaired the gap by borrowing. Michel Pastoureau, who devoted a whole book to the color in Bleu. Histoire d’une couleur (Seuil, 2000), points out that our two words for it come from neither Greek nor Latin: bleu comes from the Frankish blao, a Germanic word whose earliest sense hovered between “pale” and “livid,” and azur comes from the Arabic lazaward, lapis lazuli, by way of Persian. A Romance language had to go to the Franks and the Arabs for a way to name the color of the sky. The word then took off in the twelfth century, once dyers learned to fix a blue that would hold, once the Virgin traded mourning black for a blue mantle, and once the king of France put azure in his arms. Technology, faith and power together did what centuries of eyes turned skyward had not.

This lateness is not a French peculiarity. In 1969, the anthropologists Brent Berlin and Paul Kay compared the basic color vocabulary of some twenty languages, later of about a hundred, and found a consistent order of appearance: light and dark, then red, then green and yellow, then and only then blue, with the rest following. The scheme has been contested since, but blue’s place, always behind red, has hardly been: red is the color of blood and fire, blue is the color of nothing you can hold, and a language is in no hurry to name what it does not handle.

The latecomers give themselves away even more plainly, since they bear the name of the thing that finally made them noticeable. In French, orange names a fruit as early as the thirteenth century and a color only in the sixteenth, once orange trees had spread through the south and the shade, until then called saffron or filed among the yellows and reds, had an object in plain view to anchor it. Marron comes from the chestnut, violet from the flower, rose from the flower as well. Pastoureau calls these four half-colors: they exist in the language only when propped against a referent. There were violets before violet, oranges before orange, chestnuts before chestnut brown. The word came after the thing, and it came because the thing had grown familiar enough to serve as a standard.

This is exactly the komorebi mechanism, at the scale of an entire language. A distinction exists, everyone sees it, no one names it. Then an object, a dye, a fruit, a Virgin’s mantle makes it suddenly important enough to deserve a word. And from that moment on, you see it everywhere.

Vocabulary as a trace of civilization

When we want to understand a civilization, we look at its monuments, its religions, its institutions, its works and its wars. Far less often at its vocabulary. Yet words are traces too, and traces of a particular kind: they record not what a people did, but what it found worth telling apart.

I am wary of the romantic notion that a handful of words can reveal the soul of a people. Languages are far too mixed, borrowed, plundered and transformed to tell a story that tidy. French is packed with Latin, Frankish, Arabic, Italian and English. Japanese imported thousands of Chinese words and then thousands of English ones. Nobody holds a pure language, and those who claim to are usually holding something else.

What languages do preserve are distinctions. They keep a record of things that generations went on naming because those things were useful, familiar, important, or simply beautiful to them. These are not revelations about a soul. They are readings of attention.

The reverse holds as well. You can read a society in what it stops naming. Every time a precise word is replaced by a vague one, a distinction drops out of circulation, and it becomes a little harder to think what that distinction described. The people who choose the words know this: nobody fires anyone anymore, they “right-size the workforce,” and the missing nuance is precisely the one that would have stung.

What schools erase when they simplify

This is where the subject stops being a traveler’s curiosity. There is more and more talk about students’ shrinking vocabulary, and the PISA results published on September 8 confirm that France has lost further ground in reading comprehension: 456 points, below the OECD average, eighteen points down from 2022. PISA measures a great deal more than lexicon, I know. But all of this gets discussed as a problem of proficiency in French, which is bad enough. It is more serious than that.

Sad, disappointed, humiliated, melancholy, anxious, ashamed, disillusioned: these are not seven variants of a single emotion. They are seven distinct experiences, and it is words that let us tell them apart. The American psychologist Lisa Feldman Barrett named this capacity emotional granularity: the ability to perceive one’s own states finely rather than in bulk. Her work shows that this granularity is learned, and that it is learned largely by acquiring words and their meanings. Vocabulary is not the only factor, but it is the primary instrument for identifying what we feel, differentiating it, and communicating it.

A child whose lexicon runs to “good,” “bad,” “I like it” and “I don’t like it” does not feel fewer emotions than anyone else. He simply has fewer words for understanding why he is hurting, what is worrying him, what is humiliating him, and for explaining it to the people who might help. He is in the position of the child who sees only a tree, except that the tree is him.

This is precisely what literature is for. It does not merely tell stories. It teaches that regret is not remorse, that jealousy is not envy, that solitude is not abandonment, that shame is not guilt. It gives a name to nuances we experience without being able to formulate them. I owe a good part of my own inner lexicon to a few teachers and to George Sand and Victor Hugo, and I doubt that is a rare experience. You come out of a novel with extra words, therefore extra distinctions, therefore a better-lit inner world.

Take Les Misérables, since it is still assigned in middle school. Two short versions exist, and they have nothing in common. The first is the abridgment that the French publisher École des loisirs has issued since the 1980s, launched by the poet Bernard Noël: chapters are cut, Hugo’s own words are kept, a glossary is added at the back. The student reads fewer pages, but he reads Hugo, and when he trips over a word, the word is right there, explained, at his disposal. The second is the “easy French” adaptation, designed for foreigners learning the language at the A2 level and rewritten within a vocabulary of a few hundred words. That second object is perfectly legitimate for what it is meant to do. But it, rather than the abridgment, describes the direction educational simplification tends to take: no longer cutting, but substituting. The rare word becomes a common one, the common word becomes a vague one, and the sentence reads effortlessly because it no longer contains anything to learn.

That is why simplifying the texts we give students is a strange way of trying to help them. When you replace a word judged difficult with a common one instead of teaching it, you are not lightening a sentence. You are erasing a distinction the child might have learned to make, and may now never make. You are taking away a komorebi. The approach starts from good intentions and produces the exact opposite of what it promises: it takes a child who sees only trees and makes sure he never sees more. What is withheld at ten does not come back.

Seeing through other people’s eyes

Every language carves up slightly differently a reality we all share. A distinction that is obvious in one requires a full sentence in another, and that sentence is not one you utter every day, so you end up no longer seeing what it describes.

This is probably why learning a language feels like learning more than words. You discover nuances you never paid attention to, and some of them become impossible to stop seeing. Learning a language may not be only a matter of learning how other human beings talk about the world. It is borrowing, for a moment, their way of looking at it. And passing a vocabulary on to a child is not teaching him to speak. It is lending him, for life, an extra pair of eyes.