A number of interesting research problems arise from current attempts to incorporate abstraction mechanisms in programming languages. Some of them are central issues in current research projects and others are direct extensions of current work. Several problems in each of these areas are outlined below. A third section presents a series of speculations about how this work might evolve in the future. Since this is supposed to be a session to spark research ideas, I have written brief descriptions of a number of topics and not delved very deeply into any one. Most of these problems can probably be solved in more than one way. Hence the fact that a problem is being worked on should not deter anyone from looking at it from another viewpoint.