Workshop: Phonological externalization of morphosyntactic structure: Universals and variables

• Hisao Tokizaki

The goal of this workshop is to clarify the processes of externalization that map hierarchical morphosyntactic structure to linear phonological objects. We aim to construct a theoretical model where morphosyntactic structure is universal while variations between languages are limited to phonology. Government and Binding theory tried to capture universal properties of languages in terms of principles, while the variation between languages was ascribed to parameters in the computational system. From 1990s, generative linguists started to ascribe language variation to a detectable system that lies outside of narrow syntax (cf. Chomsky’s (2001) Uniformity Principle). Boeckx (2014) argues that parameters should be attributed to morphophonology and the interface at PF. Some studies have proposed replacing syntactic parameters with phonological differences between languages. Richards (2010) proposes the condition on the prosody of wh-questions, which is extended as Generalized Contiguity in Richards (2016). For example, Richards argues that the pro-drop parameter is derivable from the position of metrical boundaries. AUTHOR (2011, 2013) argues that the head-directionality parameter and the compounding parameter can be derived from word-stress location in a language. Eliminating syntactic features can also be done in terms of conditions on linearization. Kayne (1994) proposes Linear Correspondence Axiom (LCA), which assumes that asymmetrical structure is linearized properly at PF. Given this antisymmetry thesis, the relation between structure and movement has been investigated by Moro (2000), Barrie (2011) and Bauke (2014). These studies can also do away with syntactic parameters that are built on the syntactic formal features triggering movement. We would like to discuss these topics with participants of different backgrounds because we need to investigate the mapping from syntax to phonology from various points of view. We aim to show what kind of information is necessary for externalization. When the externalization process is identified, we can reveal the properties of core syntax.