Douglas G Kilday acnasvers at
Wed Apr 4 01:48:51 UTC 2001

Stanley Friesen (29 Mar 2001) wrote:

>At 10:26 AM 3/25/01 Douglas G Kilday wrote [responding to Dr. White]:

>>Your proposal was /trosy-/ > /trohy-/ > /trooy-/ with later loss of /y/ when
>>not re-analyzable as suffixal. The first form, being foreign, evaded your
>>postulated metathesis, so no /troys-/.

>>I just don't like the necessity of invoking foreignness in order to evade one
>>sound-change, then insisting on "naturalization" in order to obey subsequent

>Why, as far as I can see this happens all the time.  A borrowed word ceases to
>be perceived as foreign after a few generations, after which time it is
>subject to any *subsequent* (or even concurrent) sound changes in the
>borrowing language.  Consider for instance ME 'chief', borrowed from Medieval
>Norman French and then subjected to the Great Vowel Shift just like a native
>English word.

Yes, phonetic naturalization of loanwords must be normal, otherwise
languages which borrow extensively would accumulate hundreds of phonemes. My
objection was to the claim that */trosy-/ would have evaded the general
metathesis /sy/ > /ys/ (postulated by Dr. White to explain non-lengthening
in Epic Greek genitives) because it was perceived as a foreign word, and
_then_ would start behaving as a native Greek word, even though it contained
the sequence /sy/ which had just been metathesized to /ys/ in all native
Greek words! This is beyond coincidence; it requires a conspiracy by the
Demons of Phonetic Change to confuse moderns. Thank goodness for those
enlightened migrationists who know in their hearts that Trojans are
phonetically altered Tyrrhenians.

If this sort of thing had happened with 'chief', the pronunciation [c^yef]
vel sim. would have evaded the Great Vowel Shift because it was perceived as
French, all ME-speakers being acutely aware that French words are exempt
from English phonetic shifts. Only later would the "normal" pronunciation
[c^Ef] be acquired, which would rhyme with the more recent borrowing [s^Ef],
and there would be no [c^iyf]. If someone has examples of foreign words
which actually did evade the GVS this way, it would largely deflate my
objection to this stage of Dr. White's scenario.

More information about the Indo-european mailing list