‘Interpretation’ means discovering a hidden meaning, but there can be no question of attempting this while such an attitude is maintained towards the dream-performance. Look up the description of dreams given in the writings of Wundt, Jodl and other recent philosophers: they are content with the bare enumeration of the divergences of the dream-life from waking thought with a view to depreciating the dreams; they emphasize the lack of connection in the associations, the suspended exercise of the critical faculty, the elimination of all knowledge, and other indications of diminished functioning. The single valuable contribution to our knowledge about dreams for which we are indebted to exact science relates to the influence upon the dream-content of physical stimuli operating during sleep. We have the work of a Norwegian author who died recently—J. Mourly Vold—two large volumes on experimental investigation of dreams (translated into German in 1910 and 1912), which are concerned almost entirely with the results obtained by change in the position of the limbs. These investigations have been held up to us as models of exact research in the subject of dreams. Now can you imagine what would be the comment of exact science on learning that we intend to try to find out the _meaning_ of dreams? The comment that has perhaps been made already! However, we will not allow ourselves to be appalled at the thought. If it was possible for errors to have an underlying meaning, it is possible that dreams have one too; and errors have, in very many cases, a meaning which has eluded the researches of exact science. Let us adopt the assumption of the ancients and of simple folk, and follow in the footsteps of the dream-interpreters of old. First of all, we must take our bearings in this enterprise, and make a survey of the field of dreams. What exactly is a dream? It is difficult to define it in a single phrase. Yet we need not seek after a definition, when all we need is to refer to something familiar to everyone. Still we ought to pick out the essential features in dreams. How are we to discover these features? The boundaries of the region we are entering comprise such vast differences, differences whichever way we turn. That which we can show to be common to all dreams is probably what is essential. Well then—the first common characteristic of all dreams would be that we are asleep at the time. Obviously, the dream is the life of the mind during sleep, a life bearing certain resemblances to our waking life and, at the same time, differing from it widely. That, indeed, was Aristotle’s definition. Perhaps dream and sleep stand in yet closer relationship to each other. We can be waked by a dream; we often have a dream when we wake spontaneously or when we are forcibly roused from sleep. Dreams seem thus to be an intermediate condition between sleeping and waking. Hence, our attention is directed to sleep itself: what then is sleep?