Abstract
This paper addresses two stages in the search for optimal data to investigate the non-standard subject use of the object pronoun hun ‘them’, which is proliferating in spite of prescriptive stigma. In view of the absence of subject-hun in newspaper corpora, and its scarcity in the Spoken Dutch Corpus, we first compiled an ad hoc television dataset featuring 2449 attestations of standard and non-standard pronoun variants. This exercise did not solve the sparsity issue, but it allowed us to confirm the Contrast Hypothesis that hun is preferred in contexts of involved negative contrast. We then carried out a priming task in three IKEA canteens, in which respondents were asked to reproduce two stories, one of which contained tokens of subject-hun. While the latter did not entice respondents to bypass their norm judgements, the experiment revealed the relevance of the demonstrative die “those” as a possible competitor of zij and hun. It is concluded that the most fruitful resource for the study of non-standard diffusions is data from the copiously available micro-blogging platform Twitter.
