Chapter 1

Bridging theoretical and experimental linguistic research

At present, the alignment of linguistic theories and processing models is in its infancy. While scholars in both research groups try to unravel the nature of human language, it appears to be problematic to bring the two enterprises together. Relevant questions include whether the groups – each employing different levels of analysis – ultimately describe different cognitive systems and to what extent the ways in which data are collected (i.e. offline versus online) address the same hypotheses. In this paper I will argue in favour of a single cognitive system that maps linear strings (sounds or symbols) to complex conceptual representations and vice versa, comparable to a “One-System Architecture” as recently proposed by Lewis & Phillips (2015). In the context of ellipsis research, I will show that we should understand the human language system in terms of three levels of description. Further, I will argue for computational linguistics to play a mediating role in bridging theoretical and psycholinguistic methods of enquiry.