On 2008-12-08 00:52, Andrew Jarrette wrote:
> So from the evidence of <la:werce>/<la:ferce> (early form <la:uricæ>)
> and <æ:fre> one can conclude that the change /-wr-/ > /-vr-/ is
> regular in Old English?
No. There are perfectly normal comparatives like <gle:awra> from
<gle:aw>. Had i-umlaut remained productive in LOE, and had there been a
(semi)regular change of -wr- > -Br- > -vr-, one would expect the
occasional appearance of +<gly:fra> of some similar form. Another weak
point of Liberman's proposal is that a form deliberately created by
literate clerics in the 10th century should have been _immediately_
affected by complex morphophonological transformations obscuring its
etymology (surely one would expect some preliminary competition
involving forms like +<(n)a:wre> or +<(n)æ:wre> in OE texts before
<(n)æ:fre> became generally accepted). All that can be reasonably
claimed is that <-wr-> ~ <-fr-> is not quite without precedent. That's
why I don't regard the older etymology as obsolete.
Piotr