Many of the findings Dresler (this collection) mentions in his commentary are not about dreaming, but rather about sleep, its different stages, and their potential correlates, effects, and functions. While it is encouraging that there is much evidence about the functions of sleep that relates to memory and learning, and that emotionally significant information seems to hold a special place, most of those studies have very little or nothing to do with dreaming as a subjective experience. In most of the sleep studies, whether or not the sleeping participants have been dreaming or not, and what their dream contents have been, is irrelevant for the hypotheses being tested (e.g., whether a certain stage of sleep enhances memory consolidation of particular types of stimuli) and usually remains unknown. In sleep studies purely objective neurophysiological and behavioural phenomena are investigated with objective measures. In contrast, in dream studies purely subjective phenomena are explored by collecting subjective introspective reports describing the contents of phenomenal consciousness. Modern theories of the functions of sleep are undoubtedly quite strong as scientific theories of sleep and its relationship to some neurocognitive mechanisms of memory and learning, but they are not in any direct sense theories of dreaming. Of course, any proposed theory of dreaming should be at the very least consistent with the leading theories of sleep, because the phenomenal level of organization supervenes on the lower, neurophysiological level. However, the opposite is not necessarily true. As Dresler (this collection) points out, lower-level functions can be carried out independently of the higher, phenomenal level of organization. Thus, we would like to strongly emphasize that the merits and the predictions of theories of dreaming primarily have to be tested by using data that reflects subjective dream contents, not the objective features of sleep.