Priscila López-Beltrán Forcada
Ph.D. in Hispanic Linguistics & Language Science
M.A. in Education and Applied Linguistics
NSF Dissertation Improvement Award Recipient
Foreign Fulbright Scholar
The effect of information structure in the processing of word order variation in the L2
Traditionally, psycholinguistic research has supported the view that non-canonical word orders are inherently more difficult to process than their canonical counterparts. However, few of these studies have taken into account the major influence of information structure constraints on word order variation. The present study explores the effects of these constraints on the online processing of active and passive word orders by native speakers of English. Corpus studies have shown that, actives are less discourse-dependent, passives require a given-new pattern. To investigate the effects of information structure constraints in these constructions, 30 participants completed an eye-tracking experiment in which they read sentences containing target nouns which were preceded by referential contexts that manipulated the givenness of the target nouns. Sentences could be active or passive and followed a given-new pattern if the first noun had been mentioned in the preceding context, or a new-given pattern if it had not.
Our results show that total reading durations (in milliseconds) for new-given conditions were longer on the target noun for actives and for passives. Two sample t-tests revealed that these differences were significant for active and passives too. This entails that in both cases, the target nouns in new-given conditions were read slower. All in all, results support the view that non-canonical orders are not inherently costlier, since their comprehension can be facilitated if their discourse constraints are satisfied. Likewise, these findings go in line with theories of multiple-constraint parsing according to which bottom-up structural information and top-down discourse information are integrated simultaneously during language processing.