Ellipsis in the phrasal comparative: evidence from correlate constraints
This paper bears on the debate (cf. Lechner 2015) whether single DP-remnant comparatives are underlyingly clausal (reduced clause analysis, RCA) or base-generated as DPs (direct analysis, DA). Setting aside čem-comparatives, exhibiting clear CP-like behavior (see Pancheva 2006), I focus on genitive comparatives. No evidence for additional structure has been proposed for the Russian genitive comparative, which might invite the simpler DA. However, here I present a novel constraint and show how it falls out from RCA, but not DA. Among the two available versions of RCA, I argue for the small clause (Pancheva 2009) rather than the full clause (Merchant 2009) analysis, since it allows positing less structure (avoids unmotivated TP and CP layers) and does not involve DP-raising out of a finite clause, not independently reported for Russian.