I began a phonology, but it seems my sources were too contradictory, and old, so that they didn't distinguish sufficiently between phonological, morphological and analogical changes, or at least had a different idea about them from modern sources. Anyway my first attempt at anything tends to get messy! ☺
All attested Romance vowel systems presuppose an early Vulgar Latin system (3) where the length distinction of the Latin (0) was replaced by distinctions of quality, or by a distinction of tenseness. The exception is the Sard-African system (2) where the length distinction was simply lost.
All of (2, 4, 5, 6) can be derived from (3), although in the case of (2) it is more reasonable to assume that it derives directly from (0) through loss of the length distinction. In the case of the Corsican system (2) it is however most likely that it derives from a system parallel to (1) where ĭ had merged with ĕ and ŭ had merged with ŏ before the loss of the length distinction.
Old French (and Rhodrese) vowel developments in stressed open syllables
|Old Rh. i-uml||i||i||i||ɛə||yœ||øy||y|
|Old Rh. u-uml||ɔə|
- ^ I owe the ideas of connecting lines to show the interrelationships of the vowel systems and of color-coding the vowel qualities to Henrik Theiling's page about Latin vowel shifts. He in turn derived his data from this page; I guess that's share-alike as it should be!
- In case anyone wonders why my colors differ from his it is because I once read somewhere that most people with vowel → color synesthesia percieve [i] as yellow, [u] as blue, [a ɑ]-like sounds as red and intermediate sounds as interwediate colors. I avoided to use green colors because my forgotten source reported that most vowel → color synesthetics associate green hues with front rounded vowels and brown hues with central and back unrounded vowels, neither of which occur in this chart. Not surprisingly [ə] was reported to be associated with gray!
- ^ "Afrae aures de correptione vocalium vel productione non iudicant". Augustine De doctrina christiana 4.24
- ^ My only reference for such a system in Corsican, which seems to be contradicted by other sources, is a somewhat unclear passage in
- Meyer-Lübke, Wilhelm. Einführung in das Studium der romanischen Sprachwissenschaft. Heidelberg: C. Winter, 1920, § 90 p. 105.
- (In the edition at the Internet Archive [PDF] it is on pp. 133-134.)