Pablos, L. . 1 , Shirley, E. 2 , Erdozia, K. 3 , Laka, I. . 3 , Williams, N. 2 & Saddy, J. D. 2
1 Leiden University Centre for Linguistics. Leiden University. Leiden, The Netherlands.
2 School of Psychology and Clinical Language Sciences. University of Reading. Reading, UK.
3 Dept. of Linguistics and Basque Studies. University of the Basque Country, UPV/EHU. Vitoria-Gasteiz, Spain.
Event-Related Potential (ERP) studies of Negative Polarity Items (NPIs) have focused primarily on the licensing of NPIs in two contexts where the NPI could not be licensed because of (i) the lack of a licensor in the preceding context and (ii) the (inaccessible) structural position of negation in the sentence. Unlicensed NPIs in previous ERP studies have elicited different ERP component patterns depending on the language tested and the specific semantic manipulation employed: N400 followed by a P600 in German, a P600 in English and Italian and a P600 followed by a L-LAN in English for a different semantic manipulation.
The present study tested the on-line licensing of object NPIs such as ‘ezer’ (anything) in Basque in three semantically illicit contexts. The context in (i) which contained a preceding positive particle (‘ba-’). The context in (ii) which had an inaccessible licensor and which differed with respect to previous studies in that its ungrammaticality is detected one word before the NPI. And a third context (iii), which contained a scope violation.
ERP results (n=25, L1Basque) showed that unlicensed object NPIs in context (ii) evoked an early (right) negativity in the time window 100-200 ms. followed by a N400 at 200-500 ms. in comparison to licensed ‘ezer’ in the control. However, no significant differences at ‘ezer’ were found for context (i) or (iii) relative to the control. We suggest the early negativity in (ii) is connected to early effects of lexical and semantic integration (Pulvemüller 1995) and of word class category (Neville et al 1992), whereas the subsequent N400 reflects the semantic anomaly of having an unlicensed NPI. Finally, we propose the lack of a P600 is connected to the ungrammaticality in (ii) being detected one word before the NPI, which generated an early positivity at 100-300 ms at this word.