In the development of e-Learning systems, there is a recognizable trend
towards more and more adaptive systems aiming to become assistants for the
individual learner. This has led to systems making a lot of decisions and
suggestions without asking the learner, e.g., presenting "best-fitting"
content, by using methods and techniques from the field of Artificial
Intelligence (AI). The increasing complexity of e-Learning systems raises the
question whether such systems should explain their decisions. In this paper,
we try to examine to what extent explanations are possible and useful in a
first pass.