We begin this final chapter with a summary of the main argument of the book (9.1), namely, our resolution of the grammar/pragmatics paradox, according to which a grammar/pragmatics division of labor is drawn based on a code versus inference distinction. We then address a few questions which I propose pragmatists apply themselves to in future research. One question is, if grammar consists of codes and pragmatics consists of inferences, and the two stand for two distinct cognitive phenomena, why is the code/inference division of labor sometimes difficult to apply? (9.2). Another issue we may wonder about is why any attempt to reduce or explain the grammar/pragmatics division of labor by reference to some grand-design principle (such as the monolithic topic approach or the (non)truth-conditional criterion) is doomed to fail. In other words, why is the code/inference distinction not co-extensive with any of these only reasonable criteria (9.3). A third question is, given the grammar/pragmatics division of labor here espoused, does grammar shrink in scope in favor of a larger pragmatics, or is it the other way around, namely, we end up with a larger grammar at the expense of a reduced pragmatics? (9.4). We end the book with a brief discussion of two topics traditionally considered pragmatic, whose status may seem paradoxical within the code/inference approach. First, are all extragrammatical discourse patterns pragmatic, or are some of them neither grammatical nor pragmatic (9.5)? Second, what kind of grammatical status should we assign to conventional uses involving extralinguistic concepts?