As we know, most German vowels have a 'tense' (or long) pronunciation and a 'lax' (or short) pronunciation.

Most of the time, which pronunciation should be used can be determined by the context that that vowel appears in. Long vowels are doubled, followed by an H, or followed by a single consonant. Short vowels are followed by more than one consonant. E is pronounced short at the end of a word.

Of course, exceptions to this exist. In 'regnen' and 'Mund', the stem vowel is pronounced long even though the rules say it should be short.

But I was just thinking: are there any minimal pairs for the lax/tense distinction? The German wikipedia article mentions spuken vs spucken. But all the other minimal pairs it mentions are for consonants. There is of course the Mann vs man distinction, but I can't think of any others.

If there isn't minimal pairs for a vowel's lax/tense pronounciations, then why would they be considered separate phonemes? Without minimal pairs, they would be more akin to allophones, minus the fact that Germans are actually aware of the differences in pronunciation like they are for the different pronunciations of CH. German even has names for the two sounds of CH: the 'ich-laut' and the 'ach-laut' (Laut means 'sound').

  • 2
    Much of the "context" that you mention is not phonemic, but orthographic: in "standard" German, there is no phonemic consonant length, and "h" after a stressed vowel is usually silent. Long vowels can occur before more than one consonant: this is not so common in monomorphemic words, but occurs frequently in inflected forms (e.g. "rast", an inflected form of "rasen", has long a, unlike the noun "Rast" which it seems has short a). May 29, 2018 at 8:32
  • I was once an upper intermediate in German, I know full well how German is pronounced. Yes, I know h is often silent, and that the way vowel length is marked didn't evolve naturally but was part of a spelling reform to make it easier to figure out how to properly pronounce words. Of course, it isn't full-proof, obviously. I was more asking if there were minimal pairs in existence that would justify calling all the lax/tense distinctions in vowels separate phonemes. I can only find evidence for two.
    – user19661
    May 29, 2018 at 8:36
  • 1
    To me, the fact that the rules for pronouncing "short" and "long" vowels German are based on spelling and the identity of the specific word in addition to phonological context seems like sufficiently convincing evidence that vowel length is not allophonic in German. I don't understand why despite this evidence, you've adopted a skeptical attitude until you see more minimal pairs. I would expect there are many more; e.g. Massen and Maßen are one example that I know of from discussion of how to capitalize ß. May 29, 2018 at 8:47
  • The pronunciation varies by dialect. The standard language does not have an official pronunciation, but in any version of it there are many vowel distinctions like bahn*/*ban, Ahn*/*an... May 29, 2018 at 9:37
  • 3
    @A.M.Bittlingmayer: I think that example was supposed to be Mond May 29, 2018 at 9:44

2 Answers 2


The distinction between long and short vowels is historic, not merely orthographic. It goes back to proto-Germanic and in many cases to proto-Indo-European. As for minimal pairs, they are not rare. You can start with “Wolle” /ˈvɔlə/ versus “(zum) Wohle” (ˈvo:lə).

  • It might help to add the IPA of those, since the orthography is a bit opaque (i.e. one might guess that those are actually something like /wolle/ and /wo:hle/, with different consonants).
    – Draconis
    May 29, 2018 at 17:32
  • @Draconis. Good suggestion.
    – fdb
    May 29, 2018 at 17:59

There are many.

  • denn/den
  • Zinn/ziehn
  • (dem) Sohne/(die) Sonne

As for the word Mond, it's not irregular as you guessed. That syllable is structured just like gehst or klebt:

onset  nucleus  coda  suffix
M      o        n     d       - Mond
g      eh       s     t       - gehst
kl     e        b     t       - klebt

Germanic languages tend to have a rule that if the nucleus is long, then a coda is short, and if a coda is long1 then the nucleus is short. So Grewendorf posits another slot at the end of the German syllable, the suffix. It's mostly occupied by flectional morphemes, but in the case of Mond (and some others, like Vogt) it's occupied by the last segment of the lexical morpheme. You could have a look at Sprachliches Wissen by Grewendorf if you want to know more about this.

1: (say, a consonant cluster, or, in the case of some other languages, a geminate consonant)

  • "Geminate consonant"? In German?
    – fdb
    May 29, 2018 at 13:21
  • @fdb, okay actually I was thinking a little more generally when I wrote this since much of it applies to Norwegian and others too. May 29, 2018 at 13:26
  • @fdb Not sure if geminate is the right term now, but to my ear, coda in denn forexample is longer than in den. (I could be wrong, I'm not native). But so how would you describe it? May 29, 2018 at 13:27
  • The double consonants in NHG are purely orthographic. For example, NHG Kammer continues MHG kamer, from OHG kamara, from Latin camera. None of these words was ever spoken with a geminate /mm/. The spelling with <mm> is an orthographic device to indicate that the preceding vowel is short.
    – fdb
    May 29, 2018 at 13:31
  • 1
    Turns out I must have confused something somewhere to think that NHG has geminate consonants. OHG did though, and my guess is that the current orthographic conventions are inherited from that system. So I'll edit the text. May 29, 2018 at 16:32

Your Answer

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