The paper proposes to re-examine some of the aspects of floating quantification, on the basis of some new data in French. The research questions that the paper intends to address are essentially questions about the motivation for Quantifier float. I will propose that FQs move overtly to a dedicated quantifier position, and that QR is motivated by an interpretive requirement and subject to feature-checking mechanisms. I argue that the relevant interpretive property is distributivity. Both the intervention effects and the apparent focusing of FQs can be accounted for under this approach.