Constructional contamination in language (invited talk)

Constructions are often defined as form-function pairings. The underlying assumption is that the form of a construction is fixed, and uncontaminated -- otherwise the predictable pairing with its form would be hampered. This is demonstrably not always the case, however, neither diachronically nor synchronically. Diachronically, a construction often derives from multiple lineages contaminating one another (see Van de Velde et al. 2013), and synchronically, a construction often displays contamination effects at its fringes. In this talk I will present an overview of several cases of constructional contamination, mainly from Dutch (e.g. the case study in Pijpops Van de Velde 2014). If you are not particularly interested in Dutch, this talk may still be of interest, as it will also deal with methodological issues, especially the use of quantitative approaches to language study. No prior knowledge on either Dutch or statistics is assumed. Pijpops, Dirkirk Freek Van de Velde. 2014. 'A multivariate analysis of the partitive genitive in Dutch. Bringing quantitative data into a theoretical discussion'. Corpus Linguistics and Linguistic Theory (DOI: 10.1515/cllt-2013-0027). Van de Velde, Freekreek, Hendrik De Smet Lobke Ghesquière. 2013. 'On multiple source constructions in language change'. Studies in Language 37(3): 473-489. (DOI: 10.1075/sl.37.3.01int)