Re: [tied] *g'(h)- > d as aberrant outcome

From: Piotr Gasiorowski
Message: 32525
Date: 2004-05-09

09-05-2004 22:32, alex wrote:

> And this is one of more complicated issue, this Alb. "gj" which asks
> "guess_what_I_was_before", s, g^, j, gl etc:-))))

That's always a problem with mergers. Modern English /i:/ in words
inherited from Old English is the regular development of:

Middle English //e: from OE /e:/ (also from umlauted *o:) or from /e:o/
(< PGmc. *eu);

Middle English /E:/ from OE /æ:/ (< PGmc. *e: or umlauted *ai) or from
/æ:a/ (< PGmc *au < PIE *ou or *au);

Early Middle English /e/ (of various origin) lengthened in an open syllable,

... not to mention several minor or less regular sources.

Fortunately, cognates disambiguate its origin in most cases.

Piotr