This paper purposes to review current studies on ellipsis and gapping and propose a new solution for the remaining problems. Syntax applies just to constituents, not to non-constituents. Feature checking motivates syntactic movement. We should bring back these points to make clear recent debate on gapping constructions. There are two trends in studies for gapping. Winkler(2010) suggests that ellipsis including gapping is movement. In contrast, Kim(2015) explains gapping by PF deletion. I adopt Kim(2015)’s proposal for gapping. And also I present a new solution to the remaining problems. For this solution, I turn back to Pesetsky and Torrego(2001). According to their feature checking theory, PP and CP are stable whereas NP is not. Therefore, I can explain why PP is preferred to NP in gapping constructions. I have reviewed Lasnik(2014) and Kim(2016) on PP preference in gapping constructions. Those solutions are just stipulations. Both cannot explain why it happens. Therefore, I propose a new solution through structural symmetry and stability in feature checking. This solution is cost free and goes well together with the whole grammar.