Previous studies independently show passives have lower accuracy, but faster reading times, than actives. The online data is inconsistent with predictions from two theoretical frameworks that claim passives are difficult: (1) Good Enough and (2) Syntactic Complexity. To directly compare offline and online processing, we tested English speakers with both offline (comprehension questions) and online (eye-tracking) measures. In addition to manipulating syntax (active, passive), we manipulated the predicate semantics (eventive, stative). Stative passives are temporarily ambiguous between adjectival and verbal interpretations and their availability is more constrained than eventive passives within and across languages. Offline, we found a significantly lower accuracy with passive than active sentences. In line with this, immediately after the verb (prenominal adjective), passives were found to be read for longer durations than actives. Interestingly, at the verb and post-verbal noun an interaction was observed: passives of states were fixated for significantly longer durations than actives, whereas for eventives, the effect was reversed, but not significant. Unlike previous studies, an ecological method (eye-tracking) shows offline and online results converging on passives being more difficult. Moreover, the results enrich our understanding of the processing of passive sentences beyond simple word order by establishing its interaction with predicate semantics.
|Publication status||Published - Jan 2018|
|Event||Experimental Psychology Society - University College London, London, United Kingdom|
Duration: 3 Jan 2018 → 5 Jul 2018
|Conference||Experimental Psychology Society|
|Period||3/01/18 → 5/07/18|