Abstract
Paul Valéry's lectures at the Collège de France were published by combining preparatory drafts of the lessons with verbatim transcriptions taken by a stenotypist paid by Gaston Gallimard, who had a very serious plan to publish them. Maurice Blanchot said in 1942 that the Poetics lecture had " not yet " been published, and Paul Valéry's notes corroborate the existence of such a project during his lifetime, as they provide an outline of the lecture and speak of " chapters ". But Valéry found it difficult to go back to his work and refine it for publication (" la forme coûte cher ", he claimed), and he didn't have time to go back to his notes before his death to publish his Poétique. Valéry also gave Georges Le Breton permission to publish the summaries of the first year's lecture (1937-1938) in the journal Yggdrasill. The accuracy of these summaries, which cover only part of the first year, is confirmed by existing transcriptions.
Our edition, which aims to present a clean, finished text rather than a diplomatic edition, intended for fluent reading, required a great deal of work on the transcriptions of the lecture by the stenotypist. Real-time note-taking can lead to errors and misunderstandings. The editors have attempted to recover this lost word by interpreting the transcripts, which serve as relays, a very important concept for Valéry.
A relay is an intermediary element in a transmission that removes any direct relationship or proportion between the initial moment and the final moment. There is no common measure between what precedes the relay and what follows it. This concept is very important for Valéry in characterizing the work of art. Language is a relay : there is no common ground between the author's (or producer's) intention and the reader's (or consumer's) understanding. Everything passes through a conventional system, which is language, and which means different things to the former than to the latter. To a certain extent, transcription is also a relay. The text it produces is partly (only partly) independent of the spoken text.
Thus, an expression such as " la triade de Egger ", which we read on one of the sheets typed by Gallimard's stenotypist, has to be corrected to " triade de Hegel " ; similarly, the " marquis de Saxe " is actually the " marquise du Châtelet ", and so on. We also had to eliminate a number of awkward orality effects (tics, anacoluths). When work was carried out on Valéry's preparatory notes, any missing words had to be re-established, marginal (handwritten) annotations had to be introduced into the body of the text (typed or handwritten), words in red or capital letters had to be interpreted, and so on.
