From: Andrew Jarrette
Message: 63682
Date: 2009-03-28
> > Not to be intentionally stupid (unintentionally is aI think you, Aydan, were seeking the phonetic explanation of how /au/ could become /e:a/. Well, it was a gradual process over the course of the history of Anglo-Saxon English from its origins on the continent to its eventual position in England. /au/ was fronted to /æu/, something that is common in modern English pronunciations of <ou, ow>, e.g. in many American and in Australian pronunciations (and often further fronted and raised to /Eu/ -- but more later), then the second element was lowered to /o/, i.e. /æo/. This pronunciation /æo/ is represented in one or two very early Anglo-Saxon documents, as in names beginning with <Aeod->, representing later <Ead-> (as in <Eadweard>, "Edward"). Later the second element of /æo/ was further lowered to /A/ (or /a/), depending on which convention you follow), a low back unrounded vowel, so it arrived at the stage /æa/, a pronunciation which is documented in fairly numerous early spellings with <æa> or <aea> for this vowel. But this graphic combination was, I take it, felt to be too cumbersome, so scribes simplified it to <ea>. (Later the pronunciation seems also to have shifted to something more like /ea/ or /Ea/ anyway). The first element also was long when derived from an original Gmc diphthong (*au), but was short when derived from an original Gmc short vowel (*a, before *r plus consonant, and in West Saxon, before *l plus consonant, plus other sources in varieties of Anglian).
> > different matter) but I was not aware that PGmc *aura- had
> > been established as the root of e:ar. But upon tracking
> > down Koebler's Old Norse etymological database, I see now
> > that the etymology is given as *aura-, *auraz (hence the
> > same root Danish o(with umlaut)r, 'sandbank' and as modern
> > English ore). Is there any chance you could explain to me
> > the development of *aura into e:ar?
>
> PGmc. *au regularly yields OE /e:a/, and the final *-az of
> a-stem masculines is regularly lost, so OE <e:ar> is exactly
> what you'd expect from PGmc. *auraz.
>
>