In the German word "Winter" (winter), why didn't the High German Consonant Shift change 't' to 'z' /ts/, so as to be "Winzer"?

German "Winter", just like English "winter", comes from Proto-Germanic *wintruz, and Proto-Germanic *t, in most cases, changes to 'z' /ts/ in German. That's why English 't' usually corresponds to German 'z': two-zwei, ten-zehn, tooth-Zahn, tongue-Zunge, twig-Zweig, toe-Zeh, too-zu, tame-zahm...

  • 1
    Have a look at the exception mentioned in the Wikipedia article on the High German consonant shift. Apr 16 at 15:37
  • Funny enough, this correspondence also works for Winzer, which is German for vintner. Though those two words came in via Latin vinitor. Apr 16 at 16:03
  • 2
    @Oliver The deciding factor here is that when vīnitor was borrowed (with early loss of the second syllable, as *wīntor ~ *wīntur), there was a vowel between the t and the r, so the t wasn’t protected and was free to become ts regularly. In Winter, the relevant forms at the time the consonant shift happened started with wintr(u)-, with no vowel between t and r. Apr 16 at 16:52
  • 5
    @curiousdannii is the German Stack Exchange really better suited than the linguistics one? The details of the conditioning of the High German Consonant Shift certainly seems like a linguistic question to me
    – Tristan
    Apr 17 at 8:54
  • 5
    Agreed with @Tristan. This question should not have been migrated. This is exactly the type of etymology question we do want on Linguistics, one that asks about the systematicity of sound change. Apr 17 at 10:51

2 Answers 2


The following r protected it.

See also (along with numerous instances of initial *tr being preserved) Old High German holuntar "elder tree" < Proto-West-Germanic *holantr > Dutch heulenteer.

Recall that Proto-(West)-Germanic has no *sr clusters (such clusters having developed into *str), something that continues in many of the modern Germanic languages as well (although across transparent morpheme boundaries it may be allowed).

As such, the expected result of the High German Consonant Shift would have given a phototactically illegal *tsr cluster. The usual resolution of the illegal cluster would be **tstr which is rather unwieldy so it should not be surprising that instead the High German Consonant Shift simply failed to apply in this environment.


The origin of OHG afalter "appel tree" is contestible (compare *-dóru, Teer "tar", but previously argued to reflect *-ter, cf. Holunder [Kluge/Mitska; EWAhd]).

The simplest solution to your problem is Frankish. The recorded vocabulary of Charlemagne is small, but it contained calendric items. Lenz "spring, lent" is a good example, because the *t is argued to be from *tīna (dwds.de/wb/Lenz), which is quite suspicious of Latin diem, dieus, OHG ziu, engl. Tuesday, but the fact is that common germanic reflects *dagaz "day" and this stifference remains poorly explained, albeit rigorously adhered to.

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.