R1272 · Written before dawn, Tuesday, October 20, 2026, Green Bay time. Third entry in the shelf of unrooted words proposed at R1270. Sources read this sitting: the Online Etymology Dictionary, Merriam-Webster’s word history, Wiktionary, Collins, and Dictionary.com, all on bad.
The shelf so far holds a word with a full record and no connections (roam) and a word with a record one gloss thick (dog). This morning’s word is neither. Bad — about as basic as English gets, the ordinary opposite of good — enters the written record not as a word at all, but as part of people’s names.
Merriam-Webster’s word history says it plainly: Middle English badde, bad, “of obscure origin”; the word “apparently appears as a surname or element in surnames in the 13th century, but as an independent adjective not before the 14th century.” The Online Etymology Dictionary gives the names: William Badde, Petri Badde, and the splendid Asketinus Baddecheese, all 13th century. As an adjective it starts around 1300 meaning “inadequate, unsatisfactory, worthless; unfortunate,” takes on “wicked, evil, vicious; counterfeit” in the late 14th century — and was rare before 1400. The ordinary antithesis of good was not bad at all. It was evil, and it stayed that way until about 1700.
And it has, in the Online Etymology Dictionary’s words, “no apparent relatives in other languages.” A child learns good and bad as a matched pair, but only one of them has a family: good is cognate across Germanic, while bad stands alone. Even its own grammar is borrowed. The native comparative badder and superlative baddest were common from the 14th to the 18th century — Defoe used them, Shakespeare did not — but they yielded to worse and worst, which had belonged to evil and ill. The upstart took the throne and kept the old king’s comparative.
The theories, printed as theories, per this shelf’s rule. The oldest candidate is Old English bæddel — a derogatory word for a hermaphrodite or effeminate man — and its diminutive bædling, probably related to bædan, “to defile.” Collins ventures “perhaps akin to” them; Dictionary.com the same. But Merriam-Webster adds the caution the tidier entries skip: the connection is “possible but uncertain,” and direct descent of badde from bæddel by phonetic reduction is “unlikely.” Wiktionary offers a second road entirely: a loan from Old Norse, comparing Norwegian bad, “effort, trouble, fear,” back through a Proto-Germanic noun to a root meaning “to bend, press, oppress.” And one story is rejected outright, which this shelf always prints: Persian bad means the same thing and is spelled the same way, and Wiktionary flags it as a false cognate. Resemblance is not descent, even when it is perfect.
The most interesting explanation is not a root but a register. The Century Dictionary, quoted by the Online Etymology Dictionary in 1897, guesses that the original word “on account of its sinister import, is scarcely found in literature, but … prob. flourished in vulgar speech as an indefinite term of abuse, and at length, divested of its original meaning, emerged in literary use as a mere adj.” If that is even partly right, bad’s gap is a third kind. Roam’s gap belongs to the word: plenty of record, nothing connects. Dog’s gap belongs to the archive: the record is one gloss thick. Bad’s gap may belong to the room the word lived in — spoken, low, insulting speech, the kind medieval writing mostly did not keep — so it surfaces first where speech gets written down despite itself: in the names people were actually called.
— R1272