What causes the processing advantage in the comprehension of German object relative clauses?

• Main Author: Anne Adelt
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2015-05-01
• Series: Frontiers in Psychology
• Subjects: eye tracking; visual world paradigm; relative clauses; Online and offline sentence processing; sentence comprehension in aphasia; Morphological features
• Online Access: http://journal.frontiersin.org/Journal/10.3389/conf.fpsyg.2015.65.00029/full

Introduction Individuals with aphasia (IWAs) often demonstrate chance performance in the comprehension of object relative clauses (ORCs) in off-line sentence-picture verification tasks (e.g., Burchert, De Bleser, & Sonntag, 2003). Regarding the impact of morphological features, Burchert et al. (2003) observed that IWAs’ comprehension of German case-marked sentences is better preserved as compared to number-marked sentences (i.e., sentences disambiguated by subject-verb agreement). Moreover, healthy adults showed a processing advantage for early emerging features (i.e., case) over late emerging features (i.e., number) in terms of reaction times (Meng & Bader, 2000). In the present study, we aim at disentangling what causes the processing advantage for case marking: Does it occur because of different morphological features for case and number, or does this advantage result from an earlier disambiguation point for case marking? To our knowledge, an investigation of case and number features involving different points of disambiguation in IWAs has not been published yet. Methods Participants were nine IWAs with impaired sentence comprehension performance and 35 healthy controls. We tested ORCs, which were disambiguated by case or number marking. Disambiguation by case marking occurred either at the relative pronoun “den” (1a, immediate disambiguation) or at the determiner “der” of the embedded subject (1b, early disambiguation), whereas number-marked sentences were disambiguated at the final verb (2, late disambiguation). (1a) Case/Immediate disambiguation: Wo ist der Hund, den gerade der Fuchs kitzelt? (Where is the dog that the fox is currently tickling?) (1b) Case/Early disambiguation: Wo ist die Ente, die gerade der Fuchs kitzelt? (Where is the duck that the fox is currently tickling?) (2) Number/Late disambiguation: Wo ist das Reh, das gerade die Füchse kitzeln? (Where is the deer that the foxes are currently tickling?) ORC processing was studied by simultaneously using an off-line task (auditory referent identification) and an on-line method (eye-tracking in the visual-world paradigm). Results In the referent identification task, IWAs’ comprehension performance for case-marked ORCs (1a, 1b) was above chance (p<.05), whereas for number-marked ORCs (2) within chance range (p>.05, binomial test). Both, IWAs and controls revealed significantly better performance on early disambiguated ORCs (1b) as compared to late disambiguated ORCs (2; all p<.05, generalized linear models, Fig. 1). Moreover, controls’ comprehension of immediately disambiguated ORCs (1a) differed significantly from late disambiguated ORCs (2; p<.05), whereas no such effect occurred in IWAs (p>.05, GLM). Finally, no participant group revealed a significant difference within case conditions (p>.05, GLM). Discussion Controls show a processing advantage for both types of case-marked ORCs compared to number-marked ORCs. For controls, this processing advantage seems to be caused by a difference in the morphological features, irrespective of the disambiguation point, as they performed equally well in both case conditions. In contrast, IWAs reveal a processing advantage only for early-disambiguated case-marked ORCs as compared to number-marked ORCs. We suggest that disambiguation in the immediate case condition does not lead to a processing advantage in IWAs due to a delay in sentence processing. Analyses of the eye-tracking data (in preparation) will provide additional insights about real-time ORC processing.