"Untranslatable" Is a Marketing Word

The word "untranslatable" sells. It promises readers access to feelings no English speaker has ever had, to concepts that lie beyond the reach of their native tongue. The claim is almost always false. Translators do not face a blank wall when they encounter toska, cafuné, or Schadenfreude. They face a menu of four ordinary moves—borrow, calque, paraphrase in place, or compensate elsewhere—each with a price attached to tone, rhythm, or structure. The harder problems are not the exotic nouns that populate viral lists. They are the grammatical obligations that English cannot replicate without rebuilding the sentence entirely.

English · 1199 words

The myth of the perfect equivalent

Barbara Cassin's Dictionary of Untranslatables, published in French in 2004 and in English translation in 2014, uses the term "untranslatable" loosely. The Oxford Journal of Linguistics review of the work notes that Cassin applies the label to terms either left untranslated as they move between languages or to terms "typically subject to mistranslation and retranslation." The category is procedural, not ontological. Nothing in the dictionary claims that these words cannot be rendered into another language. It claims that their movement is troubled, that they accumulate baggage.

This is not how the concept travels in popular culture. The BBC Culture article "Eleven untranslatable words," updated in 2022, presents definitions it calls "real" and notes that the collection had grown to 60 words from 30 languages. Scientific American, in a 2016 piece titled "The Magic of 'Untranslatable' Words," reported that one such list had reached 601 entries. Every supposedly untranslatable word on every one of those lists can in fact be translated, at the cost of a phrase instead of a word. The lists persist because they circulate. They do not persist because they describe a genuine linguistic boundary.

The items themselves are familiar to anyone who has seen these articles. Toska, Russian, defined as "great spiritual anguish." Tartle, Scottish, "the moment of panic when you forget someone's name." Ilunga, Tshiluba, "a person who is ready to forgive any abuse the first time, to tolerate it a second time, but never a third." Prozvonit, Czech and Slovak, "to call a mobile phone only to have it ring once, so the other person calls back and the caller avoids paying." Cafuné, Brazilian Portuguese, "the act of running your fingers through someone's hair." Schadenfreude, German, "pleasure at another's misfortune." None of these resist translation. All of them resist translation by a single word. That resistance is not the same thing.

What translators actually do

The translation-techniques literature, summarized in an IJHAMS survey on translation procedures, names the standard moves. Borrowing means not translating at all: the source word enters the target text unchanged. Schadenfreude did this in English, moving from German to common use without alteration. Calque translates each part of the source expression literally. "Superman" is a calque from the German Übermensch. These two strategies preserve form. They sacrifice naturalness, or they wait for naturalness to arrive through repeated use.

When borrowing and calque fail, translators turn to meaning. Paraphrase, also called lexical expansion or descriptive translation, unpacks the source term into a phrase or clause. Toska becomes "a longing with nothing to long for," or "the ache of vague desire," or whatever the context demands. The cost is length and rhythm. The gain is precision. A reference page on designation strategies for non-equivalence lists this alongside adaptation, modulation, and transposition as common solutions when no equivalent exists.

The fourth move is compensation. The IJHAMS survey defines it as the technique used when meaning is lost and "a word or expression cannot be translated directly." The loss is acknowledged. The meaning is recovered elsewhere in the text, or in a footnote, or in an expansion that the original did not require. A translator who cannot fit prozvonit into a single English verb might describe the action in the narrative, then later use the space saved to thicken the characterization of the person who does it. The information survives. Its distribution changes.

Why "equivalent" is the wrong test

The obsession with single-word equivalents misidentifies the problem. Translators do not fail when they replace one word with three. They fail when they cannot match the range, register, or grammatical packaging of the source without distortion.

Cassin's dictionary, again, is instructive. Its entries track how philosophical terms—polis, subject, reason—shift in meaning as they cross languages. The "untranslatable" here is not a word that has no translation. It is a word that has too many, each loaded with a different philosophical tradition. The translator must choose, and in choosing, betray.

An article on translatability and poetic translation poses the operative question: whether an element is obligatory or optional in the target-language text format. This is the distinction that separates manageable problems from hard ones. Optional elements can be dropped, added, or relocated. Obligatory elements must be accounted for, one way or another.

The harder cases are grammatical

The viral lists concentrate on nouns. Nouns are easy to borrow or to paraphrase. The difficult cases are the grammatical categories that English speakers rarely notice they are missing.

The IJHAMS survey notes that translation shifts can involve changing the grammatical category of a word, and that this can be compulsory or optional. When a language requires its speakers to mark evidentiality—whether information was witnessed, inferred, or reported—the English translator faces an obligatory choice. English does not mark evidentiality grammatically. The information must be added lexically ("I saw that," "apparently," "they say that") or lost. The loss is structural. The sentence in the target language cannot replicate the source's economy.

Honorific systems present the same problem. Languages that encode social relationship in every verb form force translators into constant expansion or constant flattening. The Japanese translator who renders a business conversation into English must either add explanatory phrases ("speaking to a superior") or strip the social information entirely. Neither choice preserves the source. Both are translations. Neither is neutral.

These cases do not appear on the viral lists. They are not exotic. They are routine. They are also invisible to readers who measure translation success by whether a cool foreign word has been matched by a cool English word.

Why the lists survive

The lists survive because they serve a purpose unrelated to translation. They offer readers the pleasure of lexical tourism: the sensation of having brushed against another culture's interior life without the work of learning its language. The 601-word list reported by Scientific American in 2016, the growth from 30 to 60 languages noted by BBC Culture in 2022, the endless replication on social media—none of this documents a discovery about language. It documents a market.

Bluntness of that kind is the outlier. Most publications do not correct the record. They expand the list. The formula is too effective: a foreign word, a poetic definition, the implication that English is impoverished by comparison. The reader feels enlarged. The translation problem is hidden beneath the romance.

Translators know what the work actually requires. It is not the hunt for magic one-word matches. It is the choice of which loss to carry, and where in the sentence to place it.