“…When L2 speakers encounter an unanticipated word or morpheme rather than an anticipated one, one can expect an error signal that will subsequently serve to slightly inhibit the anticipated form in favor of the encountered form when the same message in context is required next. However, L2 speakers appear to be less likely than L1 speakers to predict upcoming forms, even when they demonstrate knowledge of the forms during production and in offline tasks (Grüter, Hurtado, Marchman, & Fernald, ; Ito, Martin, & Nieuwland, ; Kaan, ; Kaan, Dallas, & Wijnen, ; Kaan, Kirkham, & Wijnen, ; Lew‐Williams & Fernald, ; Martin et al., ), perhaps due to greater cognitive load from increased self‐monitoring (Levelt, ) or the need to inhibit their L1 (Green, ). To the extent that L2 speakers are less likely to predict upcoming grammatical choices, they will have less opportunity to learn from predictions that are subsequently falsified by the speech that they encounter.…”