The image can only be studied through the image, by dreaming images as they gather in reverie. It is a non-sense to claim to study imagination objectively since one really receives the image only if he admires it. Already in comparing one image to another, one runs the risk of losing participation in its individuality.

If there is any realm where distinction is especially difficult, it is the realm of childhood memories, the realm of beloved images harbored in memory since childhood. These memories which live by the image and in virtue of the image become, at certain times of our lives and particularly during the quiet age, the origin and matter of a complex reverie: the memory dreams, and reverie remembers.

To verify images kills them, and it is always more enriching to imagine than to experience.

Any work of science, no matter what its point of departure, cannot become fully convincing until it crosses the boundary between the theoretical and the experimental: Experimentation must give way to argument, and argument must have recourse to experimentation.

One must live to build one's house, and not build one's house to live in.

An excess of childhood is the germ of a poem.

One doesn't read poetry while thinking of other things.