The Interaction Between Low Sluicing and Case Licensing

Under the movement-and-deletion approach to sluicing (Merchant 2001), a key question is how strictly the elided constituent must match its antecedent. While some proposals allow syntactic non-isomorphism allowing various cleft sources beyond full clausal antecedents (Barros 2014; Merchant 2001; Van Craenenbroeck 2010), this talk examines an understudied ellipsis pattern in Turkish where wh-remnants surface with case, copula, and tense morphology (wh-CASE-COP-TNS), unlike more familiar wh-CASE remnants. I refer to this construction as low sluicing. On the surface, low sluicing appears cleft-like: its remnants (i) obligatorily host a copula, and (ii) disallow accusative and subject genitive case.

Despite these similarities, I argue that clefts cannot be the sole source for low sluicing. Evidence from island sensitivity, adjuncthood restrictions, multiple remnant constructions, and unexpected case patterns (contrasting with Uzbek: Gribanova 2013; and Japanese: Hiraiwa & Ishihara 2012; Saito 2004) points instead to a full clausal source. Building on Ince (2006) and Van Craenenbroeck & Lipták (2008), I propose that low sluicing involves ellipsis of a deverbal projection licensed by various functional heads, parallel to VPE in English.

Yet this analysis raises a further question: why do accusative and subject genitive cases fail to surface on the remnants? I argue that this follows from the interaction between the movement-and-deletion analysis and a configurational theory of case assignment. This investigation not only contributes to debates on ellipsis licensing and syntactic identity, but also offers novel diagnostics for case theory supporting configurational case assignment in Turkish and revealing typological parallels with VPE.