Developmental Perspectives on the Acquisition of the Passive

This dissertation investigates the acquisition of the passive. The apparent cross-linguistic delay of the verbal passive compared to other constructions suggests children\u27s knowledge is somehow restricted, leading some to propose the difficulty arises because of syntactic maturation (Wexler 2004, Orfitelli 2012) or because of a heavy reliance on discourse features assigned to arguments (Snyder and Hyams 2008). These hypotheses predict that until relevant mechanisms mature, children will only produce and comprehend a subset of passives (i.e., adjectival), or will only show adult-like performance when felicity conditions are met. Others have suggested children have not overcome the poverty of the stimulus for passives, arguing that frequency of passive in the input determines the timing of acquisition (Kline and Demuth 2010). The prediction of this account is that passive production and comprehension will vary cross-linguistically as a function of the primary linguistic data. ^ I argue against both the maturation and frequency accounts with data from English and Sesotho. In the Sesotho study, I show that 4-6 year old Sesotho-speaking children perform adult-like on certain types of unambiguously verbal passives. Furthermore, I show that Sesotho-speaking children perform no better on any type of passive than their English-speaking counterparts, although the frequency of passives in Sesotho child directed speech is ten times greater than in English. ^ The English study, a targeted grammaticality judgment task where felicity conditions are met, shows that 4-6 year-old English-speaking children accept passives with purpose phrases as grammatical, indicating they have verbal passive syntax. This result further provides evidence against frequency as the main source of passive acquisition delay, since passives are less than 1% of all English child directed speech. The results add to growing evidence that English-speaking children have knowledge of verbal passives when felicity conditions are met, though felicity might not be the sole factor. ^ The overall results indicate children have adult-like knowledge of verbal passives in most cases, though the source of non-adult-like performance requires further explanation. Following Gehrke and Grillo (2008) and Grillo (2008), I suggest that the type shifting required for some predicates to passivize may play a role in the passive delay.