KFLC 2012

Mark J. Elson

University of Virginia

                                    The Fate of Proto-Romance *Cw in Italian

            In this paper, I will re-visit the question of the inter-relationship in Italian between the fate of Proto-Romance *Cw and the position of the stress with respect to it. Superficially, the facts are simple. A correlation exists: If *Cw was post-tonic, we find CC, but if pre-tonic, we find C. The reflexes are found most commonly in the preterite of the E conjugation, and for that reason, are often designated rhizotonic (= post-tonic; e.g., 1s ténni < *ténwi) and a-rhizotonic (= pre-tonic; e.g., 2s tenésti < *tenwésti). There are, although relatively infrequently, instances of *Cw outside the verbal system which corroborate the correlation, but there are also instances, verbal as well as non-verbal, which appear to contradict it. I considered the latter in a previous presentation, attempting to demonstrate that all of them have, potentially, explanations, and therefore need not be understood as counter-evidence to the stress-related correlation. In fact, 1p ténnemo > tenémmo, an attested analogical change from rhizotonic with CC to a-rhizotonic with C in the first person plural preterite of tenere ‘have’, leaves virtually no doubt of the psychological reality of the correlation, and many, including such well-known scholars as Rohlfs, assume the validity of the correlation without establishing explanations for the forms they themselves cite as counter-evidence.

            The mere existence of a stress-related correlation, however, does not mean that it was relevant in the origin of CC versus C as reflexes of *Cw. Other factors could have given rise to the differing reflexes thus creating the correlation, i.e., the correlation as such need not have been a cause, but a result. This is Maiden’s hypothesis, which claims that the position of *w in the syllable played a role in its fate. I will argue in this presentation that although a correlation, for which there is evidence in the form of 1p ténnemo > tenémmo exists, that it was in fact a result, not a cause, and that the cause, or origin, of the correlation is to be found in a consideration of the evolution of Latin hiatal strings.

(Hispanic/Romance Linguistics; J. Rini and D. Tuten)