The former seeks to settle the question by help of introspection and memory.

On the one side, it is urged against the theory of unbroken mental activity, that we remember so little of the lowered consciousness of sleep.[73] To this it is replied that our forgetfulness of the contents of dream-consciousness, even if this were unbroken, would be fully accounted for by the great dissimilarity between dreaming and waking mental life.

It is urged, moreover, on this side that a sudden rousing of a man from sleep always discovers him in the act of dreaming, and that this goes to prove the uniform connection of dreaming and sleeping.

This argument, again, may be met by the assertion that our sense of the duration of our dreams is found to be grossly erroneous; that, owing to the rapid succession of the images, the _realization_ of which would involve a long duration, we enormously exaggerate the length of dreams in retrospection.[74] From this it is argued that the dream which is recalled on our being suddenly awakened may have had its whole course during the transition state of waking.
Again, the fact that a man may resolve, on going to sleep, to wake at a certain hour, has often been cited in proof of the persistence of a degree of mental activity even in perfectly sound sleep.

The force of this consideration, however, has been explained away by saying that the anticipation of rising at an unusual hour necessarily produces a slight amount of mental disquietude, which is quite sufficient to prevent sound sleep, and therefore to expose the sleeper to the rousing action of faint external stimuli.
While the purely psychological method is thus wholly inadequate to solve the question, physiological reasoning appears also to be not perfectly conclusive.