Ott (2012, 2015) proposes that floating quan- tifiers (FQs) arise when a noun and a quantifier enter into a one-to-one Symmetric Merge, whose output is split into two elements by the operation of the Labeling Algorithm. This paper argues that such sym...
Ott (2012, 2015) proposes that floating quan- tifiers (FQs) arise when a noun and a quantifier enter into a one-to-one Symmetric Merge, whose output is split into two elements by the operation of the Labeling Algorithm. This paper argues that such symmetric Merge in argument positions violates the θ-Criterion and must therefore apply outside θ-domains. The proposed analysis resolves a range of long-standing problems that challenge both the stranding and adverbial views-for instance, the subject- object asymmetry in FQ licensing, the ungrammaticality of subject FQs in argument positions in English ergative and passive constructions, and the exceptional patterns of so-called subject FQs in Korean and Japanese. Under this approach, FQ and non-FQ constructions are derived from distinct underlying structures, and floating quantifiers are shown to be stranded elements that simultaneously function as adverbial modifiers.