The central concern of this paper is whether the Japanese FNQ is always a distributive operator, as Nakanishi (2004, 2007, 2008), and Kobuchi-Philip (2003, 2007) claim. Based on various interpretive facts, we instead argue that such an analysis is not necessarily correct, and that the interpretive ambiguity pertaining to FNQs (between distributive and non-distributive readings) can be resolved if the semantic ambiguity arises because of the existence of two different types of FNQs (as quantificational determiners and as quantificational adverbs). To validate this assumption, we will address interpretive aspects relevant to FNQ quantification from the viewpoint that natural languages may adopt two kinds of quantification, namely, D(eterminer)-quantification and A(dverbial)-quantification, as put forth in Partee (1995, 2008). This assumption might appear unattractive from the viewpoint of theoretical economy. However, the advantage of this hypothesis is that it is possible to maintain the generalized quantifier analysis and perfectly offer precise and uniform interpretation rules that derive the meaning of sentences involving Japanese FNQs that are attested. (C) 2014 Elsevier Ltd. All rights reserved.