The Untranslatable Word Is a Design Brief

An editor wanted a footnote for the word "hchouma," and I told her no footnote was going to do the job, because the job the word does was never glossing. It was accusing.

"Hchouma" gets translated as shame, sometimes as disgrace, occasionally as a whole sentence about propriety, and none of those are wrong, they are only incomplete on their own. The word carries at once a judgment on the act, a judgment on the person doing it, and a claim that some witness has now been implicated in having seen it. English has no single word that does all three jobs together, which is where writers usually stop and reach for a footnote, an italic word left in the original with a translator's note bolted underneath, an apology dressed as scholarship.

Treat it as a brief, not a wall

I stopped reaching for the footnote when I started asking a different question: not "what English word means this" but "what is this word specified to do in this sentence, and what would satisfy that specification using ordinary tools." That is the design-brief method. A so-called untranslatable word is not a locked door. It is a list of requirements, and requirements can be met by more than one part working together.

Take a line from a novel I translated, a mother catching her son coming home at four in the morning. In the original, one word carries the whole scene's judgment. My first English draft used "shame" and it sat there, accurate and inert, doing perhaps a third of the work. The second draft spread the job across the sentence instead of loading it onto a noun: "What she said was not that he had done something wrong. It was that he had done it where the Filalis could hear." The accusation, the audience, and the social cost are now distributed across a sentence rather than piled onto a single word, and the paragraph needs no note at the bottom of the page to explain anything.

Why the brief works where the hunt fails

The instinct to hunt for one heroic replacement word comes from treating vocabulary as the unit that carries meaning. But meaning in a sentence rarely sits in one word. It is spread across word choice, rhythm, what gets stated and what gets left to be inferred, who is named and who is only gestured at. When you specify what a word is actually doing, register, audience, threat, judgment, you can usually satisfy each part separately with plain words, and the sentence ends up doing more work than the original single word did, because it is no longer hoping one term will land the whole effect alone.

The case against this

The case against this method is that it flattens something specific into a generic scene, that a reader loses the fact that this particular judgment belongs to a particular social architecture, and that spreading a word's job across a sentence quietly erases where the word came from. That concern is fair, and I do not think it disappears. Something is lost when a reader no longer sees the original word sitting on the page at all.

But I would rather lose the word than lose the sentence's function, because a reader who hits an unglossed foreign term and a footnote has been handed a fact, not an experience. A reader who reads that the son had done it where the Filalis could hear understands the social stakes as they happen, in real time, the way the son himself understood them as his mother spoke. The word "hchouma" is not untranslatable. It is under-specified until you decide what job it has to do in this exact sentence, and once you decide that, English has never been short of tools for the job. It was only ever short of translators willing to spend a paragraph on it instead of a single word.

Where the method runs out

It does have a limit. Some words are load-bearing across an entire book, not a single sentence, a term that names a whole social institution a novel keeps returning to. Spreading that across every sentence where it appears would cost more words than the book can afford, page after page of the same three-part explanation. For those, I do let a term stand untranslated once, early, and let the plot itself supply the brief over the following chapters instead of the sentence supplying it once. The method still applies. It just gets to work at the scale of a book rather than a line.