Being a participant matters: Event-related potentials show that markedness modulates person agreement in Spanish

Abstract

The present study uses event-related potentials to examine subject–verb person agreement in Spanish, with a focus on how markedness with respect to the speech participant status of the subject modulates processing. Morphological theory proposes a markedness distinction between first and second person, on the one hand, and third person on the other. The claim is that both the first and second persons are <i>participants</i> in the speech act, since they play the <i>speaker</i> and <i>addressee</i> roles, respectively. In contrast, third person refers to whomever is neither the <i>speaker</i> nor the <i>addressee</i> (i.e., it is unmarked for person). We manipulated speech participant by probing person agreement with both first-person singular subjects (e.g., <i>yo</i>…<i>lloro</i> “I…cry-_{1ST PERSON-SG}”) and third-person singular ones (e.g., <i>la viuda</i>…<i>llora</i> “the widow…cry-_{3RD PERSON-SG}”). We also manipulated agreement by crossing first-person singular subjects with third-person singular verbs (e.g., <i>yo</i>…<i>∗llora</i> “I…cry-_{3RD PERSON-SG}”) and vice versa (e.g., <i>la viuda</i>…<i>∗lloro</i> “the widow…cry-_{1ST PERSON-SG}”). Results from 28 native speakers of Spanish revealed robust positivities for both types of person violations, relative to their grammatical counterparts between 500 and 1000 ms, an effect that shows a central-posterior distribution, with a right hemisphere bias. This positivity is consistent with the P600, a component associated with a number of morphosyntactic operations (and reanalysis processes more generally). No negativities emerged before the P600 (between 250 and 450 ms), although both error types yielded an anterior negativity in the P600 time window, an effect that has been argued to reflect the memory costs associated with keeping the errors in working memory to provide a sentence-final judgment. Crucially, person violations with a marked subject (e.g., <i>yo</i>…<i>∗llora</i> “I…cry-_{3RD PERSON-SG}”) yielded a larger P600 than the opposite error type between 700 and 900 ms. This effect is consistent with the possibility that, upon encountering a subject with marked features, feature activation allows the parser to generate a stronger prediction regarding the upcoming verb. The larger P600 for person violations with a marked subject might index the reanalysis process that the parser initiates when there is a conflict between a highly expected verbal form (i.e., more so than in the conditions with an unmarked subject) and the form that is actually encountered.