General programme, activity sheet

Saturday 20 September, 2014 10:00 to 10:30 Oral presentation
Which 'which'-questions are easy? A study from French acquisition
Authors: Anamaria Bentea and Stephanie Durrleman


Object which-questions, contrary to object who-questions, pose challenges to children, but the cause of the difficulties is debated. The grammatical intervention account attributes children’s difficulties to the presence of a [+NP] subject in the interpretive chain formed by the moved object with the gap. This account has been challenged by an approach in terms of processing limitations. The latter claims that which-questions are more problematic because they are associated with an increased processing load determined by properties that are not structural, such as set-restriction. Our study taps into this debate by investigating (a) the effect of set-restriction on the comprehension wh-questions and (b) how this links to working memory (WM) capacities. We tested 119 French-speaking children aged 5 to 11 on object wh-questions varied across three conditions: +NP+Set-restricted (which dog), +NP–Set-restricted (which animal) and –NP–Set-restricted (who). Participants had to choose the correct character identified by a wh-question (1-3). Each question was associated with two pictures illustrating the same action with reversed Agent-Patient roles. WM was assessed through digit-span tasks. The crucial finding (Graph 1) reveals that children’s performance across all ages significantly improves when the wh-object is [–Set-restricted] regardless of whether it is marked +/– NP (all p < .001). Inclusion of the NP feature per se does not entail difficulties for children, contrary to what an explanation purely in terms of syntactic features predicts. Moreover digit-span scores significantly impact accuracy for +Set-restricted (p < .001) but not for –Set-restricted structures (p > .05). We argue in favour of a processing account drawing on the cost associated with set-restriction to explain increased soliciting of WM resources for this operation and improved performance in its absence: –Set-restricted objects are easier to retrieve from memory as they are semantically distinct from the +Set-restricted intervening subject.
Further information:

Submited paper (208 KB)

Paper session F
Place: Main Room

Other activities in Paper session F
09:00 h. to 09:30 h.Oral presentation

Pronominal variation in Uwa Spanish. A possible case of cross-linguistic influence
09:30 h. to 10:00 h.Oral presentation

A syntactic account of differences in the processing and production of code-switched concord and agreement structures
10:30 h. to 11:00 h.Oral presentation

Acquiring Italian accusative clitic placement in simultaneous bilingualism


Organizers, sponsors, collaborators