2018 Friday Session C 1715

Friday, November 2, 2018 | Session C, Terrace Lounge | 5:15pm

The Acquisition of French Causatives and Parallels to English Passives
J. Borga, W. Snyder

Overview: Since Kayne (1975), the literature on Romance languages has distinguished two types of causative: faire-infinitif (FI) and faire-par (FP). Guasti (2016) draws a connection between these two structures and the English be- and get-passives (BP & GP), respectively. As shown in (1-2), GPs and FPs are more constrained than their BP/FI counterparts: FPs and GPs both require actional verbs (1) and affected objects (2). Guasti conjectures that FPs will be acquired before FIs, parallel to English GPs and BPs. This paper supports Guasti’s conjecture, and moreover shows we can explain this pattern for causatives in terms of the Snyder-Hyams (2015, S&H) Universal Freezing Hypothesis (UFH), introduced to account for late acquisition (age 4+) of BPs.

From Passives to Causatives: S&H explain the different acquisitional timeline of BPs versus GPs in terms of a freezing effect which arises under Collins’ (2005) ‘smuggling’ account of BPs. Under Collins’ strict interpretation of Baker’s (1988) UTAH, an argument must receive its θ-role in the exact same structural environment in active and passive sentences. In BPs, movement of a larger phrase ‘smuggles’ the internal argument (IA) to a position above the external argument (EA); the IA can then move into subject position without violating locality. The UFH says children younger than 4 are ‘frozen’ and can never do syntactic operations (Movement, Agreement) on an XP inside a moved constituent; hence, BPs are delayed. The UFH explains children’s much earlier use of GPs, because GPs often lack an EA and hence do not require smuggling.

Belletti and Rizzi (2012, B&R) argue that FIs with an embedded transitive verb require a version of smuggling, as the dative-causee in the embedded Spec-vP blocks the Agree relationship necessary for case-valuation of the accusative object. In contrast, neither FPs, nor FIs with intransitives, require smuggling: in the latter, no causee intervenes; and in the former, the causee is realized as a non- intervening PP adjunct (Folli and Harley 2007). Thus, the UFH leads us to modify Guasti’s conjecture slightly: in contrast to FIs with a dative causee (‘FDs’, 2c), FIs with an intransitive should not be delayed. Likewise, UFH predicts no delay for the ‘reflexive-causative-passive’ (RCP), which lacks an EA (Labelle 2013).

Investigation: We selected 11 longitudinal, spontaneous-speech corpora for French (CHILDES, MacWhinney 2000); located all faire-causatives; and classified them as FI-datives (with an overt dative causee, and/or a transitive-verb incompatible with FP) or ‘Other’ (causee in an overt par-phrase, or at least a lower verb compatible with FP). The strong prediction of S&H is that FDs will not appear until the same late period when BPs appear (age 4+). Results: (i) None of the 11 children used a single FD prior to age 4 (Binomial p < .0001; see (6)); (ii) as expected, shortly after 4, some children began using FI-datives (3); (iii) other causatives appeared earlier, including RCPs (4) and faire-intransitives (5) before age 3.

Thus, our findings provide new support for the UFH account of English passives, by showing it makes accurate, novel predictions for Romance causatives.