On 2007-05-15 20:56, tgpedersen wrote:
>> This time the change can be dated very precisely using phonological
>> evidence (it's relative chronology with respect to other changes)
>> combined with the direct testimony of 17th-c. grammarians.
>
> They report a change?
They report new contrasts and new pronunciations unknown to their
predecessors. For example, early grammarians compare the vowel of <love>
to French <ou> and German short <u>. According to Butler (1633) <sun>
and <soon> still have "the same" vowel. However, Wallis (1653) already
says that <cut, but, dull, come, couple, love> as well as <burst, burn,
turn> etc. are pronounced with a vowel like French <eu> and almost
identical with French fem. <-e>, presumably a kind of (rounded?) schwa.
In the 18th century Portuguese authors identify the vowel of <love> with
Portuguese "dull" <a>.
Note that the lowering _must_ be dated after the Great Vowel Shift,
because words like <blood> and <flood> (as well as <stud, month, glove,
rudder> and others whose orthography has changed), originally with long
/o:/ (raised to /u:/ by the GVS and _subsequently_ shortened to /U/),
underwent the change.
The lowering is older than the shortening of /u:/ in <good> and <foot>,
though some words like these used to vary between /u:/ and /U/ at the
time of the lowering, hence later hesitation between /u:/ (also with new
shortening producing /U/) and lowered /V/, e.g. <soot> /sUt ~ su:t ~
sVt/ in various accents.
Piotr