In his critical essay James Joyce: Trivial and Quadrivial, Marshall McLuhan makes reference (on page 38 of Interior Landscape) to a passage in critic Paul Valéry’s book Variety V (1945).

The quoted passage is on the nature of language as used by writers. Valéry calls for a reinvention of literary studies which makes use of ancient wisdom and pedagogy—the classical trivium; the figures of rhetoric in particular—to go beyond merely chronicling authors and their lives, and get scientific about language. Those who study today’s large language models will find Valéry’s observations premonitory of their own.

The following is a translation of the full passage (287-293 of Variété V) into English—undertaken, naturally, by a large language model.

The history of literature has greatly expanded in our time and now enjoys numerous academic posts. It is striking, by contrast, how little the intellectual activity that gives rise to literary works themselves has been studied—or how it has been approached only incidentally and with insufficient rigor. Equally remarkable is that the precision brought to bear on the criticism of texts and their philological interpretation is so rarely applied to the analysis of the concrete phenomena involved in the production and reception of works of the mind.

Should any clarity be attained in this domain, its first effect would be to liberate the history of literature from a multitude of secondary facts, from diversions and details that bear only arbitrary and inconsequential relation to the essential problems of art. The temptation is strong to substitute for the study of these subtle problems a preoccupation with circumstances and events which, however interesting in themselves, rarely deepen our appreciation of a work or offer us a truer or more fruitful understanding of its structure. We know little of Homer: the sea-born beauty of the Odyssey is undiminished. As for Shakespeare, we cannot even be certain that the name rightly belongs on the title page of King Lear.

A thorough history of literature ought therefore to be conceived not so much as a chronicle of authors and the accidents of their lives or the fates of their books, but rather as a history of the mind insofar as it produces and consumes “literature.” Indeed, such a history could very well be written without mentioning a single writer’s name. One may study the poetic form of the Book of Job or the Song of Songs without the slightest need to appeal to the biography of their authors, who remain entirely unknown.

But a history of this kind presupposes, as its preamble or foundation, a study whose purpose would be to form the most exact idea possible of the conditions under which literature exists and develops; an analysis of the modes by which this art operates, of its instruments, and of the diversity of its forms. We would not think of writing the history of painting or mathematics, for example, without first acquiring a reasonably thorough understanding of their respective disciplines and techniques. Yet literature, owing to its apparent ease of production—since it uses as its substance and instrument the common language of all, and appears to combine only ideas of no special elaboration—seems able to dispense with any such preparation in order to be practiced or enjoyed.

One does not deny that this preparation may appear negligible: it is the common view that a pen and a notebook, along with some natural gift, suffice to make a writer.

Such was not the opinion of the ancients, nor of our most illustrious authors. Even those who believed their works owed everything to their desire and immediate inspiration had, unknowingly, formed for themselves a whole system of habits and ideas—the fruits of their experience—which imposed themselves upon their creative acts. Though they may have had no awareness of all the definitions, conventions, logics, and combinatory rules implied in composition, and believed themselves beholden only to the moment, their work necessarily employed all these procedures and inescapable modes of mental function. The revisions of a manuscript, the regrets, the erasures, and finally the marked progress evident in successive works—all these make clear that arbitrariness, spontaneity, emotion, and even present intention dominate only in appearance.

When we write, our hand does not ordinarily make us aware of the astonishing complexity of its mechanism and the distinct forces it unites in action. But surely what it writes is no less composed; and each phrase we form—like any complex and singular act—must suit a circumstance that never recurs, involving a coordination of present perceptions, impulses, and images with an entire store of reflexes, memories, and habits. All of this becomes evident with the slightest observation of language “in action.”

Moreover, a reflection just as simple leads us to understand that literature is—and can only be—a kind of extension and application of certain properties of language.

It exploits, for instance, the phonetic qualities and rhythmic possibilities of speech, which ordinary discourse largely ignores. Literature even classifies and organizes these effects, at times employing them with strict and systematic precision. It also draws upon the effects of juxtaposing terms, their contrasts, and creates contractions or uses substitutions that stimulate the mind to produce images more vivid than those needed for everyday comprehension. This is the domain of “figures,” once the concern of classical rhetoric and now nearly abandoned in modern instruction. That abandonment is to be regretted. The formation of figures is inseparable from the formation of language itself; all “abstract” words derive from some abuse or transference of meaning, followed by the forgetting of their original sense. The poet who multiplies figures merely rediscovers in himself language in its nascent state.

Indeed, if one considers things from a sufficient height, can we not see language itself as the supreme masterpiece among all literary works—since every creation in literature is nothing more than a recombination of the powers contained in a given vocabulary, according to forms once and for all instituted?

In sum, the study of which we speak would aim to define and expand the investigation into the specifically literary effects of language—the examination of expressive inventions and the means by which discourse becomes charged with a power beyond its utilitarian function, imbued instead with the capacity to move, enchant, or compel reflection.

Such a study would not only explore the particular techniques and procedures whereby this transformation occurs, but also seek to classify the types of effects thus obtained, and to understand the interaction of these effects with the sensibility of the reader or listener. It would strive to identify the internal logic of literary composition, the mental operations and reflexes employed, the modes of association and anticipation mobilized in the act of writing or reading.

In this way, it might finally become possible to treat literary activity with the same intellectual seriousness accorded to other complex human practices—those of painting, music, or mathematics—which are not studied solely in terms of the lives of their practitioners, but rather as autonomous systems of operation, governed by principles, techniques, and evolutions of their own.

And perhaps then the History of Literature would find its true dignity—not as a chronicle of authors and their adventures, but as the history of a particular form of mental labor, of the configurations and metamorphoses of spirit that result from the sustained use of language as a creative force.