In Germanic languages, the p sound in Proto-Indo-European became f. I have wondered if the p sound means that the word does not come from a Germanic source. This is because words that have p in them usually come from Latin. Does the presence of that letter mean that the word is not Germanic?

  • 3
    If just the presence of just the letter p by itself would make the word non-Germanic, there are a lot of non-Germanic words in German.
    – Mast
    Commented Jun 23, 2020 at 8:52
  • 5
    @Mast There are a lot of non-Germanic words in German. But also German underwent a second shift (the High German consonant shift or) which devoiced the voices plosives. That shift only applied generally to /b/ (turning it into /p/) in High German in the south, but it applied to geminate /bː/ throughout, hence pairs like Eng rib, Ger Rippe. But most of the p’s that appear in Modern German are indeed in loan words from non-Germanic languages (primarily French/Latin), and many of the actual, Germanic p’s are now (p)f’s in German (auf, apfel, etc.). Commented Jun 23, 2020 at 9:53
  • 1
    And than, there are three odd words out there, I quote the High German forms first: Pfad "path", Pflug "plough" (with the verb Pflügen "to plough"), and pflegen "to care" (formerly a strong verb in German) where no outer-germanic borrowing relation is really confirmed. Commented Jun 23, 2020 at 10:54

3 Answers 3


Not always.

Grimm's Law predicts that Proto-Indo-European *b would turn into Proto-Germanic *p. However, Proto-Indo-European *b is vanishingly rare, and some scholars argue it didn't actually exist in the oldest reconstructable forms of the language (only appearing later). Regardless, though, an ancestral *b is probably the source of a few native Germanic words like English "apple", cognate with Russian jabloko and Gaulish abalom.

Grimm's Law also had certain exceptions where it didn't apply. The most common of these is after *s, which gives us native Germanic words like English "spew", cognate with Latin sp, or "sprout", (probably) cognate with Ancient Greek speírō.

  • 2
    There are also some cases of PG *b (from PIE *bʰ or *p with Verner’s Law) followed by a nasal yielding *pp, though the only example I can think of this early in the morning is PIE *upó- > PG *ub- + -nó- > *upp(a)- > ‘up’. Commented Jun 23, 2020 at 6:03
  • 1
    there's also plenty of non-native vocabulary that's solidly reconstrucable at the proto-Germanic stage. This is still (arguably) Germanic, as it is present at such an early point, even if it was borrowed from some other language (whether that other language was IE or pre-IE varies by word ofc)
    – Tristan
    Commented Jun 23, 2020 at 9:52
  • 1
    Well, apple is probably a bad example word since the names of fruits are often wanderwörter. Looking up when and where the cultivation of apples started it is not clear whether Protoindogermanic already could have a word for apple or not. Commented Jun 23, 2020 at 12:13
  • 4
    @jk-ReinstateMonica True, but it does seem to show a fairly consistent /p/ in Germanic and /b/ elsewhere, which implies it was borrowed before Grimm's Law happened.
    – Draconis
    Commented Jun 23, 2020 at 17:22
  • 2
    @vectory When we see a root which has clear cognates with /b/ in Celtic and Slavic, but /p/ in Germanic, that seems very clearly to be the result of Grimm's Law, not its cause. Languages don't completely rearrange their phonemic inventories because "hey it might be cool if we started pronouncing 'apple' with a /p/".
    – Draconis
    Commented Jul 3, 2020 at 15:54

No, because PIE *p does not always become f. It does not in the cluster sp, for example "spin" < *spen, "sprawl" < *sper. Germanic p regularly derives from b, e.g. "deep" < *\dheub. Germanic *swompuz "swamp; fungus" is attested in all branches of Germanic as well as Greek σομφός: the reconstruction *su̯omb(h)o-s is a bit of a problem because of the variability in aspiration. Nevertheless, it's clear that those instances of "p" are in Germanic words. Other examples are "apple" and possible "peg" < *bak

  • Sumpf is probably a variant of Tümpel, rel. dew. The Greek can hardly be directly from *su-, cp. hexa "6", initial s was regularly not maintained
    – vectory
    Commented Jun 25, 2020 at 5:39
  • 1
    As @vectory says, the equation σομφός : swamp is quite doubtful -- one phonological mismatch (p:φ) in an etymology is bad enough, but two (s-:σ-) are probably fatal.
    – TKR
    Commented Jul 3, 2020 at 18:56

Even in the case that the /p-/ is word initial, there are some words of Germanic origin containing it due to some irregularities mainly. As an example I quote the word Patzer (mainly known as a term for a bad chess player via Yiddish, but a more general word in High German) that is related to the regular High German word Batzen "heap, pile, lump", the initial P is a typical Upper German (Bavarian and Alemannic) dialectal feature. Packen "paket" is another example, in this case related to backen "to bake".

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.