Date of Award

Spring 2017

Document Type


Terms of Use

© 2017 Elias Blinkoff. All rights reserved. This work is freely available courtesy of the author. It may only be used for non-commercial, educational, and research purposes. For all other uses, including reproduction and distribution, please contact the copyright holder.

Degree Name

Bachelor of Arts


Educational Studies Department, Psychology Department

First Advisor

Daniel J. Grodner

Second Advisor

K. Ann Renninger


One of the most robust findings in psycho linguistics is that object-extracted relative clauses (ORCs) are more difficult to process than subject-extracted relative clauses (SRCs) (Traxler, Morris, & Seely, 2002; King & Just, 1991; Ford, 1983). However, the reasons for this are still controversial. One view attributes this difficulty to experience: ORCs are harder to process because they are more rarely encountered than SRCs (Wells, Christiansen, Race, Acheson, & MacDonald, 2008; Gennari & MacDonald, 2008; MacDonald & Christiansen, 2002). Another view is that the word order of ORCs places a larger burden on working memory than SRCs (Miller & Chomsky, 1963; Grodner & Gibson, 2005; Lewis & Vasishth, 2005). These models predict different loci of processing difficulty along an ORC sentence. Experience-based models predict difficulty at the beginning of the ORC, where the structure can be identified as rare. Memory-based models predict difficulty at the embedded verb, where a long distance dependency must be resolved. There is evidence indicating difficulty in both regions (Staub, 2010). Wells et aI., (2009) conducted a relative clause training study, which found that intensive experience with ORCs reduced (but did not eliminate) processing differences. However, the study did not pinpoint the locus of these training effects. The present study aimed to replicate Wells et aI., (2009), focusing on the locus of any training effects. On a memory-based account, training should have selectively facilitated processing exclusively at the subject region of the ORC. If instead ORC complexity is primarily a function of experience, then training should have facilitated processing over the entire ORC. Training only facilitated processing over the subject of the ORC. In contrast, difficulty at the embedded verb actually increased with training. Though methodological issues discourage definitive conclusions, the results suggest that the role of experience is limited and intrinsic memory constraints underlie ORC complexity.


Sentence processing, relative clauses, statistical learning, working memory, syntax