Vertical and horizontal links in constructional networks

Abstract Recent models of constructional networks combine vertical links between schemas and their subtypes with horizontal links between constructions at the same level of abstraction. It remains unclear, however, whether vertical and horizontal analyses express distinct information about the network, or whether one can be reformulated in terms of the other. In this paper, it is argued that vertical and horizontal links do not encode distinct cognitive mechanisms but that they are notational variants for representing a common notion of constructional similarity. The practical advantages of each notation are outlined, and some potential objections to the present account are addressed.