On the adjectival interpretation of passives in acquisition

- Borer and Wexler (1987) argued that children interpret passives as adjectival.
- This assumption underlies much work on the acquisition of passives.
- Catalan has the property of having non-homophonous verbal and adjectival passives.
- In our results children are shown to choose an adjectival reading of the passive.

Borer and Wexler (1987) hypothesised that children interpret verbal passives as adjectival passives in a language like English in which the two passives are syntactic homophones. This has been assumed in the literature, although no direct evidence for it has ever been provided. Here we present two experiments, the first testing comprehension of short and long passives by Catalan-speaking children, the second testing whether such children indeed interpret short passives as adjectival. For this second experiment we capitalise on a particular grammatical property of Catalan: unlike in English, verbal and adjectival passives are not syntactic homophones, since verbal passives select auxiliary ser while 'true' adjectival passives (in Meltzer-Asscher's 2011 terms) select auxiliary estar - although the participle is identical in both passives. Our results show that Catalan-speaking children miscomprehend verbal passives until the age of 6 and assign them an active structure, in consonance with results from many other languages. Second, when made to choose between an eventive and a stative depiction upon hearing a short passive, they choose the stative depiction, consistent with the adjectival interpretation and incompatible with the verbal one (in spite of auxiliary mismatch). Thus we add to the findings in the literature for the first time direct evidence for the adjectival interpretation of verbal passives during the acquisition period.