This project sets out to better understand how morphological variation arises from phonological variation and change, and how morphological variation may be preserved or leveled. In addressing these two questions, the project will focus on morphological variation within the stem, in contrast to variation within affixes. The phenomena to be investigated are metaphony as an instance of morphophonological variation within the stem, patterns of stem allomorphy and patterns of stress shifts within the paradigm. The languages we will examine are French, Occitan, Italian and the dialect of Piedimonte Matese (Campania, Italy).Our goal is to better understand the relation between phonological and morphological variation and change by addressing the following questions:o In what ways does morphological variation arise from phonological variation and change?o Under what conditions is morphological variation preserved, how does it spread, and how may it be leveled?In addressing these two questions, we will focus on morphological variation within the stem, in contrast to variation within affixes. The phenomena to be investigated are metaphony as an instance of morphophonological variation within the stem, patterns of stem allomorphy and patterns of stress shifts within the paradigm.The relation between morphological and phonological variation will be investigated on the basis of a number of Romance languages. Although these languages belong to the same language family, they present a remarkable scale of complexity of morphological and phonological variation. Within Western Romance, both French and Occitan have different patterns of alternation, with French being far less transparent than Occitan. Within Eastern Romance, Italian and the dialect of Piedimonte Matese (Campania, Italy) differ likewise in the degree of the abstractness of underlying forms, where Piedimonte Matese may be said to present more abstract patterns of variation. We believe that by investigating a given phenomenon of morphological variation and its historical evolution in more than one Romance language, and by comparing the different synchronic grammars that incorporate the variation, we can show that general principles of phonology and morphology constrain the direction of linguistic variation and of linguistic change.
