Joseph Emonds and Philip Spaelti: Fully Distributing Morphology: The Phonology and Syntax of Latin Case Inflections

by Joseph Emonds and Philip Spaelti

Certain executions of minimalist syntax use “uninterpretable formal features.” This term raises the question, do there really exist features of morpho-syntax that are {em never interpretable}, that play a role in neither Logical Form nor Phonological Form? Case features are in our view best analyzed as categorical head features that are realized on adjacent DPs. Case features are therefore uninterpretable only when they are not in their base positions; in their base position, they are simply categories such as V and P, and are interpretable. However, lexical features such as declension classes cannot be analyzed as “alternative realizations&rdquo of this sort, and so might be examples of purely “uninterpretable formal features.&rdquo We argue that Latin noun and adjective declension class feature bundles (e.g., [3^rd declension, ablative, singular] ) are all better reanalyzed, on independent grounds, as spell outs of case and number suffixes whose forms depend only on the phonological features of the final segment of a preceding stem. Moreover, in almost all situations, these dependencies are phonetically natural. The “6 declension classes&rdquo of Latin are simply contextual variants fully determined by 6 possible values of preceding underlying final segments: consonants and 5 distinct vowels. That is, we argue that spell outs of features complexes such as [OBLIQUE, ±PLURAL] or [GENITIVE, ±PLURAL] do not depend on arbitrary uninterpretable morpheme class features. We claim rather that such constructs, at least in the well known Latin inflectional system, are entirely superfluous.