Ala'a Melebari: dissertation defense (May 17)

Ala'a Melebari defended her dissertation with the title "The Interaction of Animacy and Morpho-syntax in Arabic" on Wednesday, May 17. Congratulations!
 
This dissertation investigates the ways that (IN)ANIMACY distinctions interact with various sub-systems of the human language faculty, in particular, morpho-syntax. In Modern Standard Arabic (MSA), morpho-syntax and ANIMACY can be pit against each other directly on the same set of target words, allowing a close inspection of the timecourse of the availability of different information in the integration of words into phrasal level structure. That is, although animate and inanimate singular nouns and plural animate nouns require matching GENDER and NUMBER agreement (e.g., on demonstratives, adjectives, finite verbs … etc.), plural inanimate nouns trigger feminine singular agreement. This state of affairs presents the language comprehension mechanism with a conflict in which the ANIMACY properties of specific nouns render grammatical what would otherwise be a morpho-syntactic violation. Findings from two experiments show that whereas singulars and animate plurals demonstrate uniform response accuracy and short latencies (and replicate previous ERP findings (a LAN) from similar paradigms (e.g., Barber & Carreiras, 2004; Gunter et al., 2000)), our inanimate plurals (mismatch cases), show longer latency effects (and a striking polarity reversal of a LAN-type response for the "correct", but morphosyntactically mismatched cases). To that end, I argue that the processes underlying morpho-syntactic feature matching are initially blind to the lexical/semantic and ANIMACY properties of nominal expressions and heavily dependent on syntactic cues, but follow-up ‘reanalysis’ processes in the sense of Freiderici (1995, 2002) and Faussart et al., (1999) result in consequently licensing the mismatches, rendering them grammatical, which in turn is in line with serial “syntax first” accounts (Marslen-Wilson & Tyler, 1980; MacDonald et al., 1994; Freiderici, 1995, 2002). I also argue, based on behavioral and ERP results, that the mismatch cases provide further evidence for arguments supporting distinct internal representations for NUMBER and GENDER (e.g. Ritter, 1988, 1991, 1993; Faussart et al., 1999, Igoa et al., 1999; Barber & Carreiras, 2004, 2005) and that the NUMBER and GENDER agreements in the mismatch cases should be treated as two separate phenomena constituting a case where ‘normal’ agreement and ‘failed but resolved’ agreement are obtained.