When are Passives Difficult: Cross-linguistic and Methodological Investigations

BA/Leverhulme Small Research Grants 2017-18 Round

Linguistic complexity provides the foundation for models of parsing, language
acquisition and impairment. All language processing fields converge on the general
claim that passives (MaxPATIENT was hit by LeaAGENT) are more difficult to interpret
than actives (LeaAGENT hit MaxPATIENT) due to greater comprehension errors.
Theories variously explain passivization complexity through its noncanonical order of
arguments (PATIENT-verb-AGENT). Two concerns for this perspective underlie our
proposed investigation with healthy adults. First, a puzzle worth consideration:
passives generate more errors in comprehension, but are read faster. Faster reading
times conflict with complexity accounts. Pilot studies by the applicants replicate the
discrepant offline and online data under more controlled conditions. Second, verb type
interacts with passivization in processing relevant ways, not yet investigated in healthy
adults. In unifying language processing fields and methodological approaches, we ask:
'when are passives complex and why?'. In response, parsing theories will require a
more sophisticated perspective on complexity.