Latin verbal system: how perfect and aorist joined in the newperfect?

[Steve Gustafson]

> Didn't diphthongation in Spanish only affect "open /O/", not "closed /o/"
> and wasn't /o/ from /au/ a "closed /o/"?

My understanding is:

Moving from Latin to Proto-Romance, differences in length were levelled, and
in most cases they were replaced with differences in quality.  Proto-Romance
applied a simple rule to vowel length: each stressed vowel is long, each
unstressed vowel is short.

Latin /o/ and /u:/ > /aa/ in Proto-Romance [/aa/ here represents the sound
of  'a' with a circle above it, or the backwards 'c' of the IPA]

[ Moderator's note:
  Represented by [O] in (most versions of) ASCII-to-IPA transcription.
  --rma ]

Latin /o:/ > /o/

Similar developments confounded /ai/ {ae} with /e/;  /e:/ and /i/ fell
together as /e/;  /a/ and /u/ were unaffected.

As to the fate of /au/; it seems to have survived to a late date in French,
if their orthography does not mislead us, and it might.  I think it survives
today in Rumanian, and possibly Sardinian.

Now, the point of the 'fuego' example was to show that CL -short- /o/
in -focum- gets diphthongized when it is stressed; so by the time this
change was happening, it must have been treated as long.  French and Spanish
share a further development in the vowel that results from short
/e/: -pied(e)-.   Long -o-, of course, gets the same treatment when
stressed, but the process seems to be unreliable even in the same
words: -fortem- > -fuerte- in Spain, -fort(e)- in France and Italy.  It
seems to me, therefore, that this change was one that -could- at least
happen sometimes to both /aa/ and /o/.

The difficulty seems to me to revolve around the quality of the vowel
resulting from /au/, and when the change took place.  Since /au/ was always
long in CL, if it became /o/ early on it seems likelier that it would have
shared the fate of /o:/.  On the other hand, the fate of /ai/ {ae} and the
fact that it shares its fate with -short- /e/, though it too was always long
in CL, might suggest the possibility that /au/ would also turn into a short
vowel.  At any rate, it seems likelier that the loss  of /au/ occurred after
the new diphthongs in Spain, Italy, and France came into being.

Exceptions to the diphthongization might also come about because of
reborrowing from the learned language; if the Latin-using monks were more
interested in hours than Giovanni in the grapevines, their word for 'hour'
might prevail over expected phonetic developments.

--
Ante principium erat quaedam testudo; et sola fuit;
et circumspicit, et vidit vicinam eius, quae mater sua
erat.  Et deposuit se super vicinam eius, et ecce: paruit
ei in lacrimis quercum, quae omne die crevit, et tunc
decidit, et fecit pontem.