Answers are grounded in this paper and cite their sources.
Traditional linguistic reference works are static artefacts that diverge from the representation of language as a cognitive phenomenon in several ways: Historically, these artefacts are rooted in the intuitions of highly educated speakers, they impose an artificial division between lexicon and grammar, and they lack psychological plausibility. Constructionist approaches break with this tradition and attempt to model language in cognitively plausible networks, known as constructicons , that are intended to mirror the linguistic knowledge as represented in speakers’ minds. However, there are numerous concepts of constructicons, including the concepts of reference and mental constructicons. This dichotomy reintroduces an old problem in a new guise. These two concepts fundamentally differ from each other, as the former describes language as an individual, cognitive phenomenon, and the latter describes language as a social phenomenon, extrapolating over more than one mental constructicon. This paper explores these fundamental differences using the example of German noun phrase agreement. First, it models NP agreement within a reference constructicon framework, followed by a comparison with elicited production data, which reveals substantial differences between reference and mental constructicons.
No takes yet. Share an insight, caveat, or question.
Kligge et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: