The issue of phonological processes in Romance could be tackled in principle in at least two different ways, the synchronic–universalistic and the historical–inductive. One might consider, deductively, a general typology of phonological processes and exemplify them with Romance materials drawn from diachronic change as well as synchronic rules. At the present stage of the debate in theoretical linguistics, this kind of deductive approach would have the disadvantage that, with the blossoming of no-rule approaches to phonology since the 1990s, there is now little agreement upon the necessity of process-based descriptions for synchronic phonology. As Lass (1984:169f.) puts it, in his early discussion of the topic: ‘the only case when process terms can be used in a relatively theory-neutral sense is […] in describing historical change’. Historical change and its effects are the main focus of this chapter.

I shall first address the main changes that affected vowels (§1), starting from the most general context-free changes that reshaped the vowel system as a whole (§1.1). In section 1.2 I review two rather general sets of contextual processes, viz. diphthongization and metaphony. Section 1.3 considers some of the phonological processes that altered vowel qualities as successors to proto-Romance open syllable lengthening, and in section 1.4 I deal with a set of vowel-fronting processes which applied in several rounds over large Romance-speaking areas, especially a-fronting (> [æ] > [ε]) and the rise of the ‘mixed’ vowels [y] (< Ū) and [ø] (< ŏ). Section 1.5 addresses vowel nasalization processes.