Abstract

Register here

Abstract

Morphosyntax is one of the areas in heritage grammars that is often subject to change compared with a given baseline (e.g., Montrul 2016, Polinsky 2018). The dynamic nature of this area makes is a fertile domain for investigating how mental grammars change across the lifespan of an individual speaker and across generations of speakers. In this talk, we will specifically focus on grammatical gender and use this as a case study of how to model complexity in heritage speakers and beyond. Establishing a working definition of the dynamic complexity of linguistic structure and the accompanying operations responsible for generating these structures is a major challenge for formal approaches to language. This challenge is even more daunting when modeling the grammars of multilingual speakers, due to the dynamic and integrated nature of these grammars (Putnam et al., 2018).

Adopting Miestamo’s (2006, 2008) systemic definition of complexity, we provide an overview of how the connection between atomic linguistic elements can be neatly captured in an exoskeletal model of grammar. An exoskeletal model calls for a separation of the mechanisms responsible for generating syntactic structure and the insertion of lexical items (i.e., morphotactic units) into said structures. Notably, this formalism allows us to propose a new typology of possible outcomes in heritage grammars, a typology which distinguishes between features and the functional sequence itself, and whether or not these are retained or lost. To make this argument, we will present a case study of grammatical gender assignment in language mixing environments.