1 NEEDED RESEARCH IN AMERICAN DIALECTS: VARIATION IN MORPHOSYNTAX Jim Wood and Raffaella Zanuttini Yale University INTRODUCTION The goal of the volume is to sum up research done in the past two decades and chart the way forward. The goal of this chapter is to provide an overview of the new sources of data, new methodologies, and new lines of research that have been developed in the past two decades focusing on the the study of morphosyntactic variation in American English. This overview will also serve to chart the way forward for research in this area, as it will highlight discoveries that are being made and questions that have emerged and are ripe for further investigation. Since the pioneering work on linguistic variation by Labov (1966, 1972), much of the research on dialect variation in North America has centered on methodological approaches that require what has become known as a sociolinguistic variable. A sociolinguistic variable involves two or more variants that are essentially interchangeable in a given context. For example, the progressive -ing suffix can be pronounced as [Iŋ] or [In], and the choice has no semantic consequences. Researchers determine the envelope of variation—the contexts where all the variants can occur in principle—and use statistical methods to study how different linguistic and non-linguistic (e.g. social) factors condition the choice of one variant over another in a corpus of spoken or written language. This method has been used productively to study numerous syntactic constructions, including expressions of stative possession (have, have got) (Tagliamonte et al. 2010), necessity modals (must, have to, have got to, etc.) (Tagliamonte and D’Arcy 2007), intensifiers (so, really, very, etc.) (Tagliamonte and Roberts 2005; Tagliamonte 2008), quotative constructions (D’Arcy 2004; Tagliamonte and D’Arcy 2004), comparative adjectives (more angry, angrier) (D’Arcy 2014), complementizer deletion (I think that/Ø she. . . ) (Blondeau and Nagy 2008), and particle shift (pack the car up, pack up the car) (Röthlisberger and Tagliamonte 2020).