On the Acquisition of ImplicatedPresuppositions: Evidence from French Personal Pronouns

Numerous studies grounded in spontaneous and elicited production alike have uncovered that French personal subject pronouns (e.g., clitic je, tu, il, etc.) as well as strong pronouns (e.g., moi, toi, lui, etc.) are acquired very early, by the age of 2 (Clark, 1998; Pierce, 1992; Hamann et al., 1996; Jakubowicz & Rigaut, 1997; Legendre et al., 2010a; Schmitz & Mueller, 2008; to name a few). These studies have also revealed that 3 person singular pronouns (il, elle) tend to be produced first, before 1 and 2 person pronouns. The fact that these pronouns appear in speech very early does not entail that young children can interpret them correctly. One reason to express doubt is that personal pronouns can be used as deictic/indexical expressions which get their meaning from the extra-linguistic context of their utterance. First and 2 person pronouns (I, you) respectively refer to speaker and hearer while third person (he/she) is traditionally defined as referring to neither (Jespersen, 1924; Lyons, 1977). Very little is known about the course of acquisition of these interpretational properties across all persons, with a few notable exceptions such as Brener (1983) for English, and Girouard et al. (1997) for French. The present study seeks to help fill this gap from a particular theoretical perspective in formal semantics/pragmatics pertaining to the computation of inferred meaning. In particular, we follow Heim (1991) who couches the interpretational difference amongst personal pronouns in presuppositional terms: 1 and 2 person pronouns lexically presuppose the existence of speaker and hearer while 3 person pronouns have only an implicated (or non-lexical) presupposition of anti-participant. In Heim's theory lexical presuppositions are part of the lexical meaning of pronouns while implicated presuppositions are derived in much the same way as implicatures. Heim posits a grammatical principle (Maximize Presupposition or MaxPresup) which forces a speaker to use the expression associated with the strongest presupposition possible that is compatible with his/her knowledge. This entails that during interpretation, a hearer computes presuppositions by comparing members of the person scale. Computing alternatives in the domain of scalar implicatures has been shown to be hard to acquire by young children. For example, 7-9-year-olds are reportedly more likely than adults to accept the pragmatically infelicitous Some giraffes have long legs because they fail to generate the implicature associated with some, namely not all giraffes (Noveck, 2001; see also Chierchia et al., 2001; Papafragou & Musolino, 2003; etc.). By separating two kinds of presuppositions (lexical vs.