Presupposition of new information as a pragmatic garden path: Evidence from Event-Related Brain Potentials

This study investigates the processing of presupposition in discourse through the Event- Related Brain Potential technique. While theoretical linguistics has largely described the phenomenon of presupposition, there is little empirical investigation, mainly from behavioural studies. Here we employed the Event Related Potential (ERP) technique to search for the brain signature of presupposition as opposed to assertion in discourse. Based on theoretical accounts, we hypothesized that presupposing new information should elicit higher efforts due to the mismatch between the information packaging and the actual knowledge, and to the need of accommodating the presupposed content in the mental model of discourse. We also hypothesized that these efforts could reflect in enhanced N400, similarly to other mechanisms operating at the discourse-context level. Twentyseven participants were presented with passages containing new information packaged either as presupposition or as assertion. Two types of presupposition triggers were selected: definite descriptions and temporal subordinate clauses. Results evidenced a difference between the processing of presuppositions and that of assertions, reflected in a more enhanced N400 for the former. Results also showed that the temporal development of the presupposition effect is earlier for subordinate clauses than for definite descriptions. Differently from some behavioural studies on presupposition, but consistently with the theoretical literature and with other ERP studies on discourse processing, our data offer the first neurophysiological evidence that presupposition is more costly than assertion when new information is presented, with differences in the time development of the effect across trigger types. We proposed to account for the N400 effect induced by new presuppositions as stemming from a pragmatic “garden path” effect, in that, being presented with a new presupposition, the receiver is led down a mismatch between information packaging and discourse representation.