5. What the dictionaries say

A word that English has taken into its dictionaries is, by the most institutional measure available, no longer a gap in English. So for twenty-four of the twenty-five words (saudade, the first word studied, was never checked against any dictionary) the project asked whether Merriam-Webster, the Oxford English Dictionary (OED), the crowd-written Wiktionary, and several other general dictionaries had an entry — checking directly where it could, recording a “reported” status where it could not, and leaving some checks, the OED for several words among them, unattempted. The results sort into three tiers, and the tiers do not line up with how well-founded the popular story is.

Adopted. Hygge is the flagship: Merriam-Webster dates its first known English use to 1960, decades before the wave that made it famous, and the OED added it as a headword in June 2017, defining it (per a news report; the OED’s entry pages were never readable) as “a quality of cosiness and comfortable conviviality that engenders a feeling of contentment or well-being (regarded as a defining characteristic of Danish culture)” (hygge). Wabi-sabi has a Merriam-Webster first-use date of 1963 and reported OED coverage from 1962. Schadenfreude is in Merriam-Webster and Wiktionary, and reportedly the OED. Jugaad has three converging senses in the Oxford Advanced Learner’s Dictionary, read directly and labeled “Indian English,” and reportedly entered the OED in September 2017 among dozens of other Indian-origin words (jugaad). Sisu gave the project its only OED entry, for one of the twenty-five words, confirmed from the dictionary’s own metadata — one noun sense, earliest use 1926 — which sits fourteen years earlier than Wiktionary’s earliest citation, a 1940 Time report from the Winter War; the project recorded the disagreement rather than resolving it (sisu). “Lose face,” the idiom calqued from mianzi, and its English-coined companion “save face” have Merriam-Webster entries of their own, though the Chinese word itself has no English Wiktionary entry (Merriam-Webster was not checked for it).

The most striking adoption happened while the genre was still calling the word untranslatable. In March 2025 the OED added Tagalog gigil as a headword, in a batch of 42 words from Ireland, Southeast Asia, and South Africa that the dictionary itself framed as “untranslatable or hav[ing] no direct English equivalents,” and defined it, in wording the project took from a Philippine news report and found corroborated in a European one, as “an intense feeling caused by anger, eagerness, or the pleasure of seeing someone or something cute or adorable, typically physically manifested by the tight clenching of hands, gritting of the teeth, trembling of the body, or the pinching or squeezing of the person or thing causing this emotion” (gigil). A word listed as an English gap in 2023 newsletters and listicles was, less than two years later, English by the most authoritative available standard — the same arc hygge had traced from 1960 to 2017.

Partly adopted. Several words appear in Wiktionary and nowhere else, and the form they take there is telling. Swedish lagom, an adverb and adjective at home, is an English noun in Wiktionary — “the philosophy or ethos of trying to achieve balance in one’s life” — and all three of the entry’s supporting quotations are from the 2017 lifestyle books: English borrowed not the word but the marketed ethos (lagom). Dutch gezellig is an English adjective there, “chiefly in Dutch contexts, rare,” with three citations ending in the Obama remark. Ubuntu has Wiktionary and Wikipedia entries and reportedly several general dictionaries, though the project could not confirm those directly. And two of the shakiest words on the file are in: mamihlapinatapai, which the last fluent speaker, when asked, did not recognize, has a genuine English Wiktionary entry, and iktsuarpok, whose one cited primary source the project could never read, has one with four dated literary citations. Wiktionary attests circulation in English, and circulation is exactly what those two words have (iktsuarpok).

Not adopted. Twelve words show no confirmed entry in any English dictionary the project checked: komorebi, meraki, litost, hüzün, toska, ya’aburnee, teranga, tizita, gemas, sankofa, abhiman, and oodal. Some of the absences are extreme. Meraki has no Merriam-Webster entry, no English Wiktionary page at all, and a Wikipedia title that resolves only to a disambiguation page for a networking company and a television show; toska on Dictionary.com redirects to a Japanese dog breed; teranga on Wikipedia is a genus of cellar spiders; ya’aburnee on Wikipedia is a Halsey album. For komorebi the project specifically ran down a rumor: an OED editor had been quoted, in 2025 news coverage of the update, mentioning it as an example of a word filling a lexical gap, and one reading of the sentence implied it had just been added. Convergent evidence — the batch’s geographic scope, every enumeration of its words, an eleven-word roundup of Japanese additions that year, and the OED’s own “recently added” list read through a workaround — showed it had not (komorebi). Collins, meanwhile, carries user-submitted “new word suggestion” pages for komorebi, litost, and ya’aburnee: the crowd knocking, the door not yet open.

Two lessons. First, dictionary adoption and the soundness of the popular story are independent variables. A word solidly attested in its own language — meraki, with three senses in a standard Greek dictionary and cognates across the Balkans — has no English standing; a word whose meaning may be a lexicographer’s anecdote has a Wiktionary entry. What the dictionaries track is English-side use, and English-side use is what listicles generate. Second, the OED itself was the project’s most persistent blind spot. Its website sits behind a login wall the project’s tools could not pass; a reader-proxy sometimes got through and sometimes silently returned the wrong page; for the last word, oodal, five distinct methods were tried and exhausted without a result (source record). Nearly every OED fact in this essay is therefore a fact about what someone else reported the OED as saying, and the knowledge base marks each one accordingly.