I've recently noticed that in English words starting with "ge-" or "gi-", when the "g" is pronounced /ɡ/, they tend to be etymologically Germanic, while the words where the "g" is pronounced /dʒ/ tend to be etymologically French (or possibly Latin). I'm sure there will be the odd exception here and there, such as words that come from neither Germanic or Latin sources, but is this an accurate observation?


  • get - from Old Norse geta
  • giggle - unknown - perhaps from Middle English gigen
  • girl - probably from Old English *gyrle, *gyrele, a diminutive form of Proto-Germanic *gurwijaz

via French / Latin:

  • gentle - from Old French gentil
  • giraffe - from French giraffe, from Arabic
  • gist - from Old French gist

1 Answer 1


The fun answer is:

English should not have any inherited words beginning in ge-/gi-.

The reason is the so-called Anglo-Frisian palatisation that shifted of original Germanic *ge-/gi- to ye-/yi-. Whenever there is a word beginning in ge- or gi- some explanation is needed:

  • It might be a loan word from Old Norse. This applies to the most common cases (get, give)
  • Levelling has unified a split pattern to restore the original g in the anlaut (goose/geese might be an example, to begin (began, begun) might be another one)
  • It might be a loan word from some other language
  • It might be an onomatopoetic word (this could apply for to giggle)
  • The vowel e/i might have developed from another vowel (a, o, or u) at a time when the Anglo-Frisian palatisation was no longer active

The words beginning in /dʒ/ are all loans, this sound wasn't originally in Old English and was imported from Norman French.

  • 2
    The explanation for "geese" that I am familar with is that by the time umlaut of this kind occured, palatalization of velars next to front vowels was no longer an active sound change. I guess it could be called analogy/leveling but there are no exceptions that I know of Commented Jul 11, 2017 at 19:04
  • 1
    E.g. on Google Books I found "A caeg to Old English syllable structure" which mentions this: books.google.com/… Commented Jul 11, 2017 at 19:15
  • And does the same apply to ga-/go-/gu-? I see many examples like gander, gather, game, ghost, good and gallows that did not shift, and a few like gate that are apparently yate in some dialect, and there are simply very few English words like yard that start with y- or ya-/yo-/yu-. A few others like yoke are odd like year in that they were j-, then supposedly g- for a time, then y- (ie *j-) again. Seems like a hypercorrection in medieval writing and/or a battle between different continental dialects. Commented Jul 15, 2017 at 5:44
  • @A.M.Bittlingmayer: The word yoke (High German Joch, Latin iugum, Greek zygon) never had an initial /g/ (maybe except for some obscure dialects or hypercorrections) and I suppose the same holds for year (High German Jahr) . Yard (High German cognate Garten) is different, maybe it had an /e/ vowel at some time in at least some forms (High German plural is Gärten with an open e sound). Commented Jul 15, 2017 at 14:18
  • @A.M.Bittlingmayer BTW, spelling variation is expected at a time when all original ge/gi-words are shifted and there is no difference between ge/gi and ye/yi.The same happens with Spanish (confusion of b and v) today. Commented Jul 15, 2017 at 14:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.