Thread:User talk:CodeCat/Proto-Indo-European/reply (17)

You are correct about *dʰr̥- becoming *dur-, but this is rather a problem as there isn't really any reason that it would become *dru-. On the other hand, even Germanic itself has some words that are obviously related to *druhtinaz, most importantly the verb *dreuganan "to serve". So the noun most likely goes back to *dʰrugʰ-ten-os or something similar, from the zero grade of the root *dʰrewgʰ- that the verb also derives from. The verb must derive from such a root, because it is a strong verb; Germanic strong verbs, bar a few exceptions, always derive from PIE verb roots.

There are similar problems with Old Norse drengr. It presumably derives from Germanic *drangiz or *drangijaz, so it would have to derive from an o-grade form of a root, which could be reconstructed as *dʰrengʰ-. Again, there is no reason to assume that *dor- (o-grade of *der-) would become *dro-. The only possibility for a relationship with *der- that I can think of is that *drangi(ja)z actually derives from the zero grade of the root, extended with a suffix: *dr-ongʰ- or the like. But the ablaut and accent relationships of such a noun don't seem to fit what is known about PIE, in particular the appearance of the o-grade. Then again I am not an Indo-European linguist so I don't know for sure, but it does seem like a rather shaky etymology. Deriving it from a root *dʰrengʰ- is far more straightforward and plausible.

Kluge's law is explained in the link I gave you. It explains the origins of Germanic geminated voiceless plosives, -pp-, -tt-, -kk- as resulting from a single plosive followed by -n- which then contracted. The Germanic ancestor of 'rock' is *rukkōnan, which contains such a geminated plosive, and using Kluge's law it can be related to the other words I listed without problems.

One last note... there are a few cases where -ur- is replaced with -ru-, but those cases are all analogical and were 'reverted' based on related words. The most prominent example is the verb, which derives from a hypothetical root *bʰreg- (that does not have any certain cognates outside Germanic, so it could be a later coinage). The three grades would have been *bʰreg- (present), *bʰrog- (perfect singular), *bʰr̥g- (perfect nonsingular), which in Germanic would have become *brek-, *brak-, *burk-. It's obvious that the zero grade is anomalous here, since the vowel and consonant have been swapped. Presumably Germanic speakers would have realised the same, and may have metathesised the consonant and the vowel so that they matched the order they have in the other two grades. So this kind of change does happen, but it only happens based on analogy with related words. It doesn't happen 'out of the blue'.