The Complete Works of Edgar Allan Poe - Edgar Allan Poe [1579]
'A poem docs not say something-it is something': that doctrine has been held in more recent times.
The interest for Mallarmd is rather in the technique of verse, though Poe's is, as Mallarme recognizes, a kind of versification which does not lend itself to use in the French language. But when we come to Valéry, it is neither the man nor the poetry, but the theory of poetry, that engages his attention. In a very early letter to Mallarme, written when he was a very young man, introducing himself to the elder poet, he says: 'I prize the theories of Poe, so profound and so insidiously learned; I believe in the omnipotence of rhythm, and especially in the suggestive phrase.' But I base my opinion, not primarily upon this credo of a very young man, but upon Valéry's subsequent theory and practice. In the same way that Valéry's poetry, and his essays on the art of poetry, arc two aspects of the same interest of his mind and complement each other, so for Valéry the poetry of Poe is inseparable from Poe's poetic theories.
This brings me to the point of considering the meaning of the term 'la poeic pure': the French phrase has a connotation of discussion and argument which is not altogether rendered by the term 'pure poetry'.
All poetry may be said to start from the emotions experienced by human beings in their relations to themselves, to each other, to divine beings, and to the world about them; it is therefore concerned also with thought and action, which emotion brings about, and out of which emotion arises. But, at however primitive a stage of expression and appreciation, the function of poetry can never be simply to arouse these same emotions in the audience of the poet. You remember the account of Alexander's feast in the famous ode of Dryden. If the conqueror of Asia was actually transported with the violent emotions which the bard Timothy’s, by skilfully varying his music, is said to have aroused in him, then the great Alexander was at the moment suffering from automatism induced by alcohol poisoning, and was in that state completely incapable of appreciating musical or poetic art. In the earliest poetry, or in the most rudimentary enjoyment of poetry, the attention of the listener is directed upon the subject matter; the effect of the poetic art is felt, without the listener being wholly conscious of this art. With the development of the consciousness of language, there is another stage, at which the auditor, who may by that time have become the reader, is aware of a double interest in a story for its own sake, and in the way in which it is told: that is to say, he becomes aware of style. Then we may take a delight in discrimination between the ways in which different poets will handle the same subject; an appreciation not merely of better or worse, but of differences between styles which arc equally admired. At a third stage of development, the subject may recede to the background: instead of being the purpose of the poem, it becomes simply a necessary means for the realization of the poem. At this stage the reader or listener may become as nearly indifferent to the subject matter as the primitive listener was to the style. A complete unconsciousness or indifference to the style at the beginning, or to the subject matter at the end, would however take us outside of poetry altogether: for a complete unconsciousness of anything but subject matter would mean that for that listener poetry had not yet appeared; a complete unconsciousness of anything but style would mean that poetry had vanished.
This process of increasing self-consciousness—or, we may say, of increasing consciousness of language-has as its theoretical goal what we may call la pofsie pure. I believe it to be a goal that can never be reached, because I think that poetry is only poetry so long as it preserves some 'impurity' in this sense: that is to say, so long as the subject matter is valued for its own sake. The Abbe Bremond,