|
Kevin Redmann
Flemming’s (2007) work on the nature of schwa in English
delineated the qualities of two varieties of schwa: (i) the well-recognized
mid central vowel which is indicated by the IPA symbol for schwa and occurs
as the final vowel in such words as sofa and umbrella;
and (ii) a highly variable vowel which assimilates to adjacent segments and
is found in medial positions in such words as suggest and
probable. This paper seeks to update the standard view of Latin
phonology by showing that, in addition to the five vowels typically posited
for classical Latin /i e a o u/, variable schwa, at least, was a part of
Latin’s native phonetic inventory.
Despite the fact that such renowned grammarians as Gildersleeve
(1895), Greenough (1902), and Hale and Buck (1966) did not recognize a Latin
schwa, evidence of vowel reduction and deletion in attested forms strongly
supports an argument for its existence, e.g.:
regō >
sur-rigō > surgō where /e/ > /i/ > Ø
In order to
advance such an argument, however, certain preconceptions about Latin
phonology must be laid aside in favor of a more objective reading of the
data. In particular, all residue of the once prominent belief that Latin
had a system of tonic accentuation like that of ancient Greek must be
completely expunged, and a model based on stress accentuation must be
adopted in its stead. This, in and of itself, is not particularly
controversial, but the way in which Latin vowels are described still bears
the mark of this former way of thinking. For example, those accounts which
frame vowel reduction in terms of quality without also considering a
diminishment of quantity are ignoring the effects of stress accent on
unaccented syllables.
This paper not only opens the door to a whole new area to be
explored in Latin phonology, it also provides significant evidence from the
historical record which corroborates the findings of recent explorations
into the nature of schwa, particularly with regard to its variability and
assimilation to context. |