Two important themes form the grounding for the discussion in this chapter. First, there is great value in conducting semantic analysis, as far as possible, in such a way as to reflect the cognitive reality of ordinary speakers. This makes it easier to model the intuitions of native speakers and to simulate their inferencing processes, and it facilitates human-computer interactions via querying processes, and the like. Second, there is concern over to what extent it will be possible to.achieve comparability, and, more ambitiously, interoperability, between different systems of semantic description. For both reasons, it is highly desirable if semantic analyses can be conducted in terms of intuitive representations, be it in simple ordinary language or by way of other intuitively accessible representations.