/k/,/g/>/h/ in certain contexts in Proto-Germanic?

Discussion in 'Etymology, History of languages, and Linguistics (EHL)' started by berndf, Jun 14, 2013.

  1. berndf Moderator

    German (Germany)
    We have the PGm verb *wakanaN, to be awake, to wake. Attested in Gothic wakan. Reflexes: English to (a)wake and, from causative derivation, to watch; High German (er/auf)wachen, Low German (op)waken, Dutch (ont)waken, Danish vågne and others. On the other hand we have the noun *wahtwo, attested in Gothic wahtwo, Old Saxon/Old High German wahta, modern Dutch/Low German/High German wacht/Wacht. Cognates in Nordic languages are loans from Low German, missing in English (all development stages).

    Both words are obviously derived from the PIE root *weǵ- (alive, awake). PGm. seems to show the suffix -wo with an infix letter -t- (??). Because of the /h/-/k/>/x/ merger as a result of the 2nd Germanic sound shift, High German wachen and Wacht seem to be immediately containing the same root wach- but this is obviously not so as the forms in other West Germanic languages and in Gothic show.

    Now, why does PIE ǵ end up as /h/ in *wahtwo and not as /k/ as in *wakanaN. Is there a /k/,/g/>/h/ shift, maybe in front of /t/ which would also explain mysterious things like bring/brought.
  2. fdb Senior Member

    Cambridge, UK
    French (France)
    I found this in my old copy of Wright’s Middle High German primer, par. 28:

    The following sound-changes took place in primitive Germanic: Every labial + {t} became {ft}; every guttural + {t}
    became {ht}; every dental + {t} became {ss}, which was simplified to {s} after long vowels. This explains the
    frequent interchange in MHG. between {pf, b} and {f}; between {k, g} and {h}; and between {ȥȥ, ȥ} and {ss, s}
    in forms which are etymologically related. (…) Preterites like {wiste}, {muoste} were formed after the analogy
    of preterites like {worhte}, {dāhte}, where the {t} was regular.
  3. berndf Moderator

    German (Germany)
    Thank you very much. So, this basically confirms my idea.

Share This Page