I did actually think /b/ was the original. It's written in reconstructions.shimobaatar wrote: ↑13 Oct 2018 23:05Well, I didn't write the article, of course, but in historical linguistics, it's generally assumed that a single change in one direction (in this case, [β] > [b] in OHG) is more likely than a number of identical changes in another direction (in this case, [b] > [β] in Old English, Old Saxon, Old Dutch, etc.).GrandPiano wrote: ↑13 Oct 2018 22:27How do we know that it shifted from [β] to [b]? How do we know Old High German didn't retain a [b] that lenited to [β] in the other West Germanic languages?
Of course, it's hard to really know anything for certain when we're looking at the past, but as spanick said, this conclusion fits better with things like Verner's Law, and we can make an educated guess. That's what makes sense to me, at least.
Could one assume that a different form of Grammatisches Wechsel was levelled for all forms in different languages?
Online Etymological Dictionary of "live" (bolding mine):
"Middle English, from Old English lifian (Anglian), libban (West Saxon) "to be, be alive, have life; continue in life; to experience," also "to supply oneself with food, procure a means of subsistence; pass life in a specified fashion," from Proto-Germanic *libejanan (source also of Old Norse lifa "to be left; to live; to live on," of fire, "to burn;" Old Frisian libba, German leben, Gothic liban "to live"), from PIE root *leip- "to stick, adhere," forming words meaning "to remain, continue.""
"Old English habban "to own, possess; be subject to, experience," from Proto-Germanic *habejanan (source also of Old Norse hafa, Old Saxon hebbjan, Old Frisian habba, German haben, Gothic haban "to have"), from PIE root *kap- "to grasp.""