In dealing with the problem of context sensitivity, in general, and indexicality, in particular, a few distinctions need to be made. The traditional Kaplanian framework can be adopted to deal with indexical reference on one side, and in distinguishing between pure indexicals and demonstratives, on the other side. To deal with various deviant cases, like answering machines and post-its, where an indexical is not used to refer to an item pertaining to the context of use, it suffices to appeal to the distinction between deictic (independent) reference and anaphoric (dependent) reference. This reflects the fact that indexicals can be used either in an independent (deictic) way or in a dependent (anaphoric) way. In order to fully appreciate the way indexicals work within natural languages, one has to distinguish between the various ways utterances can rely on context. With the distinction between presemantic, semantic, and postsemantic context exploitation, it is possible to further highlight how indexicals constitute a particular semantic phenomenon and how the traditional, Kaplan-inspired, treatment is still the best game in town when discussing how references get fixed with the use of an indexical expression.