| dc.description.abstract | Previous studies have shown that there exists an asymmetry in comprehension and processing between subject-extracted relative clauses (SRCs) such as in (1) and object-extracted relative clauses (ORCs) such as in (2). The ORC structure is found to be more difficult to process and comprehension studies have shown that (young) children are in general more accurate on SRCs than on ORCs.
The dog that is biting the cat.
The dog that the cat is biting.
This study looked at the offline and online comprehension of Dutch relative clauses with two lexical DPs by 42 5- and 7-year old children. Dutch relative clauses with two DPs that share the same number feature are ambiguous between a subject and an object interpretation – without contextual or semantic cues. When the DPs are different in number, number marking on the relative clause verb disambiguates between a SRC and an ORC reading. Previous studies in different languages have shown that number as disambiguating cue is relatively weak in ORC comprehension and that children often fail to employ this information and interpret ORCs as SRCs.
The present study had three main goals: (1) to investigate children's error patterns on relative clauses by an act-out task and compare the types of errors made between the SRC structure and the ORC structure; (2) to look at children's sensitivity to number as disambiguating cue during relative clause processing in a self-paced listening task; and (3) to investigate the role of memory capacity as measured by a forward and backward word span task in children's offline and online behaviour on relative clauses.
The first finding of this study was that there were no differences in error types between the SRC structure and the ORC structure, except that children almost always reversed the thematic roles of the DPs in the ORC structure whereas they rarely did so for the SRC structure. Hence, as found in previous studies, children interpreted ORCs as SRCs. Second, at least the 7-year olds did show sensitivity to number marking as disambiguating cue as reflected by higher reaction times at the relative clause auxiliary in the ORC condition compared to the SRC condition. Finally, the second finding was influenced by memory capacity: it seemed that only children with a lower memory span showed sensitivity to number marking in the ORC condition.
Future research with more participants is necessary to further investigate this memory effect. Furthermore, other factors than memory, such as sensitivity to number disagreement in general, should be looked at as predictors of children's online behaviour on relative clauses. | |
