Muke Tever wrote:
> R A Brown <[log in to unmask]> wrote:
> True, though by the second century AD at latest the grammarians were 
> certainly remarking on the qualitative difference between the values of 
> the vowels when short and when lengthened.   

I've little doubt that by the 2nd cent such things were 
becoming more noticeable. I think of the 1st centuries BC/AD 
as the 'high' Classical period. As time went on it is clear 
that changes in the spoken language were filtering through 
to the written norm. Even in the Classical period, I would 
expect to find a gradation of styles from the 'purist' 
language of a Ciceronian speech through to the language of 
peasants in the market place. Indeed, Cicero's letters show 
a rather different level from the language of his speeches 
and 'philosophic' works.

I have no doubt that Caesar did not speak to the common 
soldier in the same style in which he wrote "De Bello 
Gallico" etc.

> (Allen also points to 
> early hesitation between |e| and |i| in inscriptions, like _trebibus_ 
> for _tribibus_; I don't agree with all his examples though, since some 
> of them [like _Philumina_ for Φιλουμένη] have the same outcome as the 
> ordinary Latin vowel *reduction* processes would produce, like _redigo_ 
> from _ago_.)

Agreed - tho _trebibis_ would suggest a conflation of 
stressed [I] and [e:] in unblocked syllables.

> Well, the spurious diphthong ει [e:] had certainly been raised to [i:] 
> by that point in some (Νειλος = Nilus) but not in all environments 
> (Αλεξανρεια = Alexandrea) so I assume a distinction at least in some 
> places may have been possible (though perhaps not likely).

Nor of course _Aeneas_ <-- Αινειας. Indeed, ει before a 
vowel is generally rendered as [e:] in Roman borrowings. It 
may be that ει had shifted only to [I:] by this period, 
which he Romans heard differently before consonants and 
vowels. Or more likely, I think, the raising of ει [e:] to 
[i:] had not happened to same extent before vowels where, 
maybe, it was more like [I:].

>>> Now, according to Allen's _Vox Latina_, there _was_ a distinction 
>>> between two types of long E in ante-classical Latin--and some vulgar 
>>> Latin of the classical period, which classical writers labelled as 
>>> "rustic"--but one of them (which had previously been written as |ei|) 
>>> had merged with /i:/ in Classical Latin.
>> Umm - I should like to see the evidence that Allen gives. It's true 
>> that pre-classical |ei| is just plain |i| in the Classical language, 
>> e.g. deico --> dico /di:ko:/ "I say". But I was under the impression 
>> that |ei| actually meant [ej]   in the pre-classical language, where 
>> in medial syllables earlier -ai, -ei, -oi all became -ei.
> It was a diphthong to begin with; to show it was becoming a monophthong 
> by the 3rd century BC he says in inscriptions it begins to be spelled as 
> |e| (_ploirume_, _dioue_ for _plurimi_, _Iovi_) and original /e:/ was 
> sometimes spelled |ei|, as _decreiuit_ for _decrevit_, while spellings 
> of /i:/ were still distinct.

Looks more to me like confusion and hesitation during a 
period where the diphthong [ej] was shifting towards [i:].

> He also says there is metrical evidence in Plautus and Terence where the 
> genitive singular of -ius words was -i /i:/ but the nominative plural 
> was -ii; the first contracted because the -i of the ending merged with 
> the -i- of the stem, but the second did not because, at the time, the 
> ending was still /e:/.

I'd like to see more specific evidence. The genitive -ius 
remained ambiguous in Classical verse, i.e. it could be 
scanned /ius/ or /i:us/. Spellings of nominative plurals in 
-ii were normal throughout the Classical period (and 
beyond), but no one suggests the final -i was still 
pronounced /e/.

> He also points to those rustic Latin examples, where _vella_ and _speca_ 
> are found for _villa_ and _spica_[1],

Dialect difference?

> and some places where the /e:/ 
> variant survived into Romance (citing, e.g., how French _voisin_ would 
> be from a Latin _vecinus_ rather than _vicinus_.)

I doubt that. The Vulgar Latin of north Gaul does seem to 
have had some peculiarities that were local to itself. I 
suspect _voisin_ has more to do with a local pronunciation 
in which the first _i_ was short, i.e. [wI'ki:nUs], rather 
than any survival in Gaul of sound changes happening in 
pre-classical Latin. It is thought by some that this was 
dissimilation, i.e. avoidance of [i:] in contiguous syllables.

I think the broad pattern of vocalic change in Latin is 
fairly obvious, but the fine details are sometimes open to 

Nid rhy hen neb i ddysgu.
There's none too old to learn.