It is commonly believed that Negative Polarity Items (NPIs) in Japanese must be c-commanded by a sentential negation marker in the same local domain. The book under review challenges this common belief, taking up three kinds of NPIs (rokuna-, sika- and wh-mo-NPIs), and claims that the latter two must occur outside the scope of negation. The present paper reviews this claim and points out a couple of problems from the viewpoint of where negation is expressed in a sentence with an NPI. It will be shown that rokuna- and sika-NPIs, unlike wh-rao-NPIs, constitute a semantic locus of negation, hence a different grouping emerges. A solution will be suggested to reconcile this apparent contradiction.
ASJC Scopus subject areas