Why are Proto-Germanic *ga- and Proto-Indo-European *ḱóm cognate?

I know that Proto-Germanic *h and Proto-Indo-European *k are cognate by the Grimm's law.

I know that Proto-Germanic *g and Proto-Indo-European *k may be cognate by the Verner's law (excluding a starting position)

But what about Proto-Indo-European *ḱ?

When is Proto-Indo-European *ḱ cogante to Proto-Germanic *g?
When is Proto-Indo-European *ḱ cogante to Proto-Germanic *h?

Does "stress" matter?


You're asking two different questions.
Germanic does not distinguish the two series of palatals and palato-velars. So this distinction is irrelevant for the next question.
Now, how come *ḱom- became *ga- in Germanic, for example in past participles *ga-Root-ana-? This issue is a bit debated, and it is a kind of limit-case of Verner's Law. This Law is perfectly clear for the root and next-to-root consonants. k > h means the preceding vowel was stressed, k > g means the preceding vowel was not stressed (and was logically on the second or third vowel).
In the case of word-initial *ḱom-, it was probably not stressed, so in that pre-stress situation, it seems that the Law k > g also applied.

  • Isn't it kind if paradox that the unstressed positiin would be more pronnounced, if I may assume that *g is more sonorant than *h? Much rather, I have come to think that "go" had a reasonable influence, if only for the past-participle particle, because "go" is a supplements tense with an auxilary verb in several simple tenses; granted, that's mostly simple future tenses Fr "je vais dormir", G "Ich geh' schlafen", but: it's not easy to explain the p.p. either way; any kind of semantic explanation for a sound shift is better than none, in my humble opinion. – vectory Nov 15 at 1:50
  • although, ge- in German also appears in nouns that are not always easy to explain as deverbals, e.g. Gestein "rock formation", Stein "stone" or archaic Gestirn "star constelation", Stern "star" (but cp adj. starr "stiff", PIE *ster-), indeed unstressed on the first. I don't know if there were any compouns with stressed vowel at the boundary; instead, I only know Kalkstein "chalcolith". But that's what I would look for, perhaps with a-stems. – vectory Nov 15 at 2:12
  • also cp together, gather? zu-ge- is a frequent collocation; cp gedeihen "become, grow well", Gedei und Verderb (idiomatic) is according to Pfeiffer from " ie. *tenk- ‘(sich) zusammenziehen, fest, dicht werden’"; fits perfectly, if the coda is left unexplained. cp gediegen "mallable" for another word about joining things, in the broadest sense. The older any overlap with gather, the better the chance for it to precede the sound shift. Also cp uncertain *tewth- > "dutch"? cp zugezogen "moved new to town" for my initial ansatz. – vectory Nov 15 at 2:36

Within what we consider "Indo-European", one of the oldest and largest splits was between the "Centum languages" (Germanic, Italo-Celtic) and the "Satem languages" (Indo-Iranian, Balto-Slavic). In Centum, the "accented velars" (like *) merged into the "normal velars" (like *k); in Satem, they turned into fricatives instead.

So, by the time we're talking about Germanic, the Centum-Satem split had already happened, and there was no difference between * and *k. The two had merged and become indistinguishable.

  • Why do you capitalize satem and centum? – Alex B. Nov 14 at 3:33
  • 1
    @AlexB. Hm, never actually considered why I do that. Maybe because I'm treating them as sort-of-"families" of languages? Not sure, actually. – Draconis Nov 14 at 5:05
  • trivially, because one would read centum with [c] not [k]. Read with [k] it must be a proper name. Not quite, surely I'm jesting, but it's an assortment of reasons. By the way, "normal velar" is too biased an expression. Unaccented, if that's not misleading. – vectory Nov 14 at 15:34
  • Also, you do not provide an answer. – vectory Nov 14 at 15:38
  • @vectory I've also heard them called "plain velars" vs "palatovelars", or (my preferred theory) "uvulars" vs "true velars". But the least ambiguous description, to me, is the one that doesn't imply anything about the phonetics: the ones we write with acute accents, and the ones we don't (whatever the actual sound may have been). – Draconis Nov 14 at 15:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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