Structural Priming and the Developing L2 Construct-i-con: A Critical Review of Representation, Association, and Change

Abstract

Construction-based accounts characterize second-language (L2) grammar as a construct-i-con, a network of learned form–function pairings. Structural priming, the influence of a recently processed structure on later production or comprehension, offers a way to examine how this network is organized and how it changes. This critical review brings together within-construction, L2-internal cross-structural, and cross-language priming and attempts to discuss about what each design can show about representation, what factors condition the effect, and when priming supports a developmental claim. There is disagreement over the processes underlying structural priming: residual activation may cause recently accessed information to be temporarily more accessible, but implicit-learning theories view persistent or cumulative priming as adaptation to experience and therefore view cumulative priming as an implicit learning. Since these processes can coexist, a priming effect shows a context-dependent influence, not a permanent network link. Three patterns emerge: first, increased experience tends to lead to generalization across lexical items, although item-specific knowledge remains active; second, transfer between L2 constructions and languages is selective: development may strengthen licensed relations while reducing carry-over based only on a repeated string or an incompatible L1 pattern; third, immediate production, delayed production, and online prediction often diverge. Large effects therefore need not indicate mature representations; low baselines, rare structures, and unexpected verb–structure pairings can all increase priming. The state of an L2 construct-i-con is better determined by what an effect carries, where it stops, and how long it lasts than by its size alone.