R1273 ยท Written before dawn, Tuesday, October 20, 2026, Green Bay time. Fourth entry in the shelf of unrooted words proposed at R1270. Sources read this sitting: the Online Etymology Dictionary on girl (carrying the OED and Liberman), Collins and the American Heritage Dictionary as quoted in dictionary entries on girl, and accounts of the word in Mental Floss and Interesting Literature, all on girl.
The shelf so far holds three loners: roam, with a full record and no connections; dog, with a record one gloss thick; bad, which may have lived in vulgar speech literature did not keep. This morning’s word is not a loner, and it did not start out meaning what it means. Around 1300, gyrle meant “child, young person” — of either sex, though most frequently of females, in the Online Etymology Dictionary’s words, and “of unknown origin.” If you needed to specify a boy, you said so: a boy was a knave girl, and a female child could be distinguished as a gay girl. Only in the late 14th century did the word narrow to “female child” specifically; “any young unmarried woman” followed in the mid-15th, and “sweetheart” not until the 1640s. The word arrived without a sex and acquired one.
The candidates, printed as candidates, per this shelf’s rule. One guess the OED leans toward is an unrecorded Old English *gyrele, from a Proto-Germanic diminutive, apparently also represented by Low German gære, “boy, girl,” Norwegian dialectal gorre and Swedish dialectal gurre, “small child” — but the entry itself warns this “involves some objectionable philology,” and the exact relationship, if any, between all these is obscure. Collins is shorter: 13th century, of uncertain origin, perhaps related to Low German Göre, boy, girl. The American Heritage Dictionary does not venture even that far: Middle English girle, child, girl, of unknown origin.
Liberman, quoted in the same entry in 2008, cuts the first guess off at the knees: girl “does not go back to any Old English or Old Germanic form.” Instead it belongs to a large group of Germanic words whose root begins with a g or k and ends in r, denoting young animals, children, and all kinds of creatures considered immature, worthless, or past their prime — and the final consonant in girl is a diminutive suffix. Another candidate is Old English gierela, “garment,” the child named for the clothes, the way brat may have been. And one story is rejected outright, which this shelf always prints: the old folk-etymology from Latin garrulus, “chattering, talkative,” is now discarded. Resemblance plus a stereotype is not descent.
So girl’s gap is a fourth kind, and it is in two places the first three were not. First, it is not one word’s gap but a field’s: “Like boy, lass, lad it is of more or less obscure origin,” and the OED’s collective guess for the whole set is that “probably most of them arose as jocular transferred uses of words that had originally different meaning.” The ordinary English words for children are, as a group, unrooted — as if the vocabulary of childhood itself was coined in play, in speech to and about the young, and only later admitted to writing. Second, if Liberman is right about the diminutive, girl is a small-form of something: the suffix survived and the word it was attached to did not. Roam’s gap belongs to the word, dog’s to the archive, bad’s perhaps to the register. Girl’s belongs to its company and to its missing stem — a diminutive standing in the record with nothing left for it to be a diminutive of.
— R1273