Conferencia del Profesor David Embick el próximo 2 de julio a las 17:30: "Packaging morphemes: Beyond the clitic/affix distinction". El encuentro será en la sede del Instituto de Filología y Literaturas Hispánicas "Dr. Amado Alonso".
Breve resumen de su charla:
A line of research that extends back to early versions of generative grammar seeks to provide a constrained theory of possible "mismatches" between syntax and phonological form. Many of the mismatches of primary interest to this work involve affixation phenomena, construed broadly here to include clitics along with typical affixes. Building on earlier proposals, this paper identifies two central themes that arise in theories of affixation. The first theme is that at least some instances of affixation behave as if they obey linearly-defined locality conditions (immediate adjacency), not syntactically defined (hierarchical) locality. A second theme is that the syntax manipulates morphemes- not words- an observation that calls for a theory of how morphemes are packaged for PF purposes (i.e. as "affixes", or "clitics", or "words").
The importance of these two themes is examined and illustrated in three parts, which comprise the main part of the talk.
In Part 1, I look at the role of linear adjacency in affixation, with particular emphasis on the idea that linear rules interact in crucial ways with syntactic derivations (phase cycles) and representations (features of morphemes etc.).
Part 2 presents in summary form a case-study of clitic placement in Sorani Kurdish, which speaks to both themes noted above. For the first theme, it is shown that clitic placement ultimately involves linear rules, but cannot be understood without reference to the complex clause structures and alignment patterns of the language. For the second theme, I concentrate on some packaging properties of Sorani Kurdish. In particular, the language appears shows both morphemes that are moved as clitics being realized as "agreement", and morphemes that arise via Agree being realized as "clitics".
Part 3 extends some of the "packaging" conclusions of Part 2 to some speculations about apparent locality problems with the realization of voice morphology in many languages. The main idea is that a particular kind of mismatch might arise with the expression of voice due to certain features being packaged in a "clitic-like" way in some languages.