1090
We have no reason to disguise the fact that in the hypothesis which we have
set up in order to explain the dream-work a part is played by what might be
described as a ‘ daemonic’ element. We have gathered an impression that the
formation of obscure dreams occurs as though one person who was dependent upon a second person had to make a remark which
was bound to be disagreeable in the ears of this second one; and it is on the
basis of this simile that we have arrived at the concepts of dream-distortion
and censorship, and have endeavoured to translate our impression into a
psychological theory which is no doubt crude but is at least lucid. Whatever it may be
with which a further investigation of the subject may enable us to identify our
first and second agencies, we may safely expect to find a confirmation of some
correlate of our hypothesis that the second agency controls access to
consciousness and can bar the first agency from such access.
When the state of sleep is over, the censorship quickly recovers its full
strength; and it can now wipe out all that was won from it during the period of
its weakness. This must be one part at least of the explanation of the
forgetting of dreams, as is shown by an observation which has been confirmed on
countless occasions. It not infrequently happens that during the narration of a dream
or during its analysis a fragment of the dream-content which had seemed to be
forgotten re-emerges. This fragment which has been rescued from oblivion
invariably affords us the best and most direct access to the meaning of the dream. And
that, in all probability, must have been the only reason for its having been
forgotten, that is, for its having been once more suppressed.