Investigating change from a perspective of continuity

Word order variation in Dutch two-verb clusters continues to attract the attention of linguists from a variety of perspectives. Most accounts have a contemporary focus and diachronic developments therefore remain poorly understood. Rather than simply marshalling more evidence from more historical sources, I argue for an approach which considers biases in the historical record and how common methodological practices of text selection and data aggregation exacerbate them. Van Coetsem’s (1988; 2000) psycholinguistic model of contact offers a principled way to establish whether the linguistic data represents a perspective of continuity or one of change. When combined with a rigorous understanding of the sociohistorical context, the biases in both the historical and linguistic record can be assessed and this understanding applied to the compilation and analysis of new sources of evidence. I demonstrate the utility of this approach using a 35,000-word research corpus of unpublished archival manuscripts from sixteenth-century Antwerp to examine word order variation in Dutch two-verb clusters and use these findings to suggest directions for future corpus-based research.