Jackson Matthews, “A Note on Valéry,” Monsieur Teste by Paul Valéry, 1947, VI-VII
Mallarmé’s work had given Valéry a peculiar formative shock. These marvelous little “crystal systems,” as he called Mallarmé’s poems, struck the terror of perfection into him. Reading them, he could feel nothing but despair (“beauty is that which makes us despair”). He was himself already writing some very good poems indeed, but now his mind was driven past poems themselves to wonder how these “crystal systems” were constructed; the one thing superior to a perfect poem, he thought, would be a full knowledge of how it was made. He was soon to give up writing poems himself and turn his intense powers to the study of “the preparation of these beauties,” the generation of poems in the poet’s mind. Valéry was already coming into possession of his own and proper subject: the mind behind the work.
Paul Valéry, Preface, An Evening With M. Teste, 1925
Who knows but that most of the prodigious ideas over which so many great men, and a multitude of lesser ones, have for centuries turned gray, may be psychological deformities — Monster Ideas — spawned by the naïve exercise of our questioning faculties, which we carelessly apply here and there — without realizing that we should reasonably question only what can in fact give us an answer?
But the monsters of flesh rapidly perish. Yet not without having existed for a while. Nothing is more instructive than to meditate on their destiny.
Why is M. Teste impossible? That question is the soul of him. It turns you into M. Teste. For he is no other than the very demon of possibility. Regard for the sum total of what he can do rules him. He watches himself, he maneuvers, he is unwilling to be maneuvered. He knows only two values, two categories, those of consciousness reduced to its acts: the possible and the impossible. In this strange head, where philosophy has little credit, where language is always on trial, there is scarcely a thought that is not accompanied by the feeling that it is tentative; there exists hardly more than the anticipation and execution of definite operations.