Antal van den Bosch
Computational linguistics offers the option to escape the two basic scientific paradigms of theory and experimentation. In the past decades it embraced the new paradigms of simulation and, more recently, of big data. The endless computational sandbox offered by these two paradigms, plus the availability of immense amounts of digitised linguistic materials (including neurolinguistic data), has enabled us to create entirely new models of language comprehension and production. Computational linguistics is increasingly playing the role of a broker in the language sciences. I review some recent work in the intersection of theoretical linguistics, psycholinguistics, sociolinguistics, and neurolinguistics, where computational models play a mediating role. One characteristic of these models is that they are typically personalised and idiolectal, and I will argue for a renewed focus on this level of linguistic modelling.