Vowel Lengthening in Northern Italy
The goal of this dissertation is to argue that a wide variety of historical and synchronic processes of vowel lengthening in many languages can be motivated as a way of optimizing segmental and prosodic structures. The northern Italian dialects are chosen as the main pool of data for the study of vowel lengthening and its conditionings, since in a relatively small area we find a great variety of vowel length developments. Moreover, some of the vowel developments are compelling for recent prosodic theory. A set of phonological optimization tendencies are proposed to be active in these dialects, and it is shown that the interaction between the prosodic and segmental optimization constraints are resolved by language-particular parametrization. Historical vowel lengthening produced in the central part of the Romania has been interpreted by Romanists such as Lausberg, Leonard, and Weinreich as a result of the so-called 'bimoraic conspiracy' or 'open-syllable lengthening'. That is, while open syllables lengthened and sometimes diphthongized, vowels in closed syllables remained short. Yet, the present dissertation shows that there are many other types of vowel length conditionings in the Northern Italian area which can be explained by optimization tendencies at different levels: (1) Syllabic Optimization (e.g. 'Stressed syllables tend to be heavy', 'syllables optimize their "contact" with other syllables', 'unstressed syllables tend to be light'); (2) Foot Optimization (e.g. 'Feet tend to be analyzed as binary', 'vowels in word-final positions tend to be short', 'sonorous segments tend to license moraic or weight units'); and (3) Segmental Optimization (e.g. 'Sonorous segments tend to be long'). In sum, it is contended that vowel length, a prosodic phenomenon, is constrained by feet structure on the one hand and by segmental conditions based on sonority on the other. These results are in complete accordance with the basic tenets of the recent Optimality Theory, proposed by Prince (1991), Prince & Smolenski (1992), McCarthy & Prince (1993), and Mester (to appear), where phonological conditions and rules operate in order to increase the well-formedness of representations.