When Grammar and Parsing Agree

Theoretical and experimental research on language has nowadays reached an extraordinary level of complexity. Generative linguistics has produced detailed maps of syntactic structures and computations that have significantly contributed to our knowledge of language architecture as an abstract system (Rizzi, 2012; Everaert et al., 2015). Psycho-/neurolinguistics has developed complex experimental designs and techniques that have made it possible to study linguistic processing with millisecond and voxel precision (Friederici, 2002, 2011; Hagoort, 2005; Bornkessel-Schlesewsky and Schlesewsky, 2009, 2013). Unfortunately, the dialogue between the two disciplines has not always been constant, with theoretical linguistics often proceeding without drawing on experimental results, and psycho-/neurolinguistics limitedly relying on linguistic theory (see discussion in Ferreira, 2005; Poeppel and Embick, 2005; Jackendoff, 2007; Embick and Poeppel, 2015). This has resulted in a sharp separation between the formal/computational level of linguistic analysis and the functional/neuro-anatomical investigation of language, limiting the depth with which we can investigate what we know and what we do with language. There have been proposals for the development of research programs in which linguistics and psycho-/neurolinguistics engage in a tighter dialog, with the aim of highlighting the relation between theoretical formalizations, cognitive mechanisms and their neurobiological implementation (Marr, 1982). The current opinion article shows how such a research program can be successfully implemented.