oa Understanding grammar at the community level requires a diachronic perspective

Structuralism and formal grammar have, in the course of the 20^th century, rightfully taken issue with more vague and unfalsifiable just-so stories of some of their predecessors. For all its merits, though, the structuralist-formal strand of linguistics has its drawbacks as well. The classical Saussurean distinction between synchrony and diachrony can be harmful: a purely synchronic description is often inferior to the insight gained from diachrony, not only because grammar is laden with heirlooms and débris of prior structures and because languages draw on a wide variety of pathways to generate new grammar, but also because variation can often only be understood fully in the light of its history. Many cases of synchronic variation are the result of competition between an innovative mutant encroaching on an obsolescent construction. In such cases, the synchronic skew in the proportion of one variant to the other is not arbitrary, but is a reflection of how far the change has progressed. To the extent that one wants to incorporate variation in grammatical description – and there are sound theoretical reasons to do so – the historical perspective is indispensable. In this article four case studies from different corners of Dutch grammar are discussed (on cardinal numerals, on the Big Mess Construction, on bare infinitive complements of auxiliaries, and on the hortative). The case studies together form a plea for the historisation of the science of linguistics, just like biology has been historised, and indeed, as is shown in this article, there are numerous parallels between linguistics and biology.