On the nature of the subject–object asymmetry in wh-question comprehension in aphasia: evidence from eye tracking. Issue 4 (2nd April 2016)

• Record Type: Journal Article
• Authors: Hanne, Sandra; Burchert, Frank; Vasishth, Shravan
• Is Part Of: Aphasiology. Volume 30:Issue 4(2016)
• Page Start: 435
• Page End: 462
• Publication Date: 2016-04-02
• Subjects: wh-questions -- subject–object asymmetry -- sentence comprehension deficits -- online sentence processing -- anticipatory eye movements; Aphasia -- Periodicals; Aphasia; 616.8552
• DOI: 10.1080/02687038.2015.1065469
• Languages: English
• ISSNs: 0268-7038

Abstract

Abstract : Background : Individuals with aphasia (IWA) show deficits in comprehending object-extracted declaratives while comprehension of subject-extracted structures is relatively preserved. It is a matter of debate whether this subject–object asymmetry also arises for comprehension of wh-questions. Successful comprehension of wh-questions critically entails correct resolution of a filler–gap dependency. Most previous studies have used only offline accuracy measures to investigate wh-question comprehension in aphasia. Online studies exploring syntactic processing in real time are needed in order to draw inferences about gap-filling abilities in IWA and to identify the point of breakdown in sentence comprehension. Aims : This study aimed at investigating processing of subject and object who-questions in German-speaking IWA and in a group of controls by combining an offline and online method. We further aimed to explore the impact of case-marking cues on processing of wh-questions. Methods & Procedures : Applying a variant of the visual world eye-tracking paradigm, we measured participants' eye movements while they performed the same offline task, which is frequently used to assess comprehension of declaratives (sentence–picture matching). Outcomes & Results : Concerning online processing of who-questions in controls, we found anticipation of the most likely post-verbal theta-role immediately after processing the case-marked wh-pronoun in both subject and object questions. In addition, we observed an unexpected advantage of object over subject questions in terms of processing time. The offline results for IWA revealed that there were three heterogeneous patterns: (a) symmetrical comprehension with equal impairments for both question types, (b) asymmetrical performance with better comprehension of subject than object who-questions, and (c) a reversed asymmetry with better comprehension of object as compared to subject questions. For online processing of both types of who-questions, IWA showed retained abilities in postulating the gap and in associating the filler with this gap, although they were slower as compared to controls. Moreover, similarly to controls, they anticipated the most likely post-verbal theta-role. Conclusions : For controls, the findings provide evidence for rapid resolution of the filler–gap dependency and incremental processing of case-marking cues, reflected in early prediction of upcoming syntactic structure. We attribute faster processing of object questions to faster alignment of the anticipated element with a semantically more salient character. For IWA, the online data provide evidence for retained predictive abilities in processing of filler–gap dependencies in wh-questions, but prediction was delayed. This is most likely attributed to delayed integration of case-marking cues.