Pierre Menard, autor del Quijote

What Menard Wrote

On the impossibility of the identical text

"No quería componer otro Quijote — lo cual es fácil — sino el Quijote. Inútil agregar que no encaró nunca una transcripción mecánica del original; no se proponía copiarlo. Su admirable ambición era producir unas páginas que coincidieran — palabra por palabra y línea por línea — con las de Miguel de Cervantes." — Jorge Luis Borges, Pierre Menard, autor del Quijote (1939)

Pierre Menard published almost nothing in his lifetime — a few translations, a sonnet sequence, a study of Leibniz, an invective against Paul Valéry. The work the fictional critic-narrator wishes to discuss is invisible: chapters nine and thirty-eight of the first part of Don Quixote, and a fragment of chapter twenty-two, composed in the early twentieth century, word for word identical to Cervantes' original. Menard rejected the easy path of becoming Cervantes — learning Spanish, recovering the Catholic faith, fighting Moors, forgetting the history of Europe between 1602 and 1918. That, the narrator says, would have been less arduous and less interesting. The harder, more interesting task was to remain Pierre Menard and arrive at the Quixote anyway, through the experiences of Pierre Menard.

The story is a thirty-page argument for a single, vertiginous claim: that the same words, written by a different author at a different historical moment, are not the same text. Cervantes' "la verdad, cuya madre es la historia" — truth, whose mother is history — is, in 1605, a routine seventeenth-century rhetorical compliment to the past. Menard's identical sentence, written in 1934 by a near-contemporary of Bertrand Russell and William James, is, the narrator says, "an astonishing idea": a pragmatist proposition that history fabricates truth. The same words have grown deeper because their author is anachronistically wrong.

This piece tries to literalize the contention. Two semi-transparent layers of the opening of the Quixote drift across the canvas: the parchment text is Cervantes, 1605; the indigo text is Menard, 1934. They share every letter. Where the cursor pulls them into perfect register, the bilingual identity blooms into vermillion — the same words, rubricated by their double authorship. As the cursor drifts outward, the layers separate, and the work shows what Borges insists you should see: that even at zero typographical difference, an unbridgeable thirty-year war has fought itself between them. Click any word, and the overlay shows you the war.

The story is haunting because its joke is now technologically real. When a generative model produces a paragraph of Hamlet, character-for-character identical to Shakespeare's, is it Hamlet? Borges, in 1939, says no — a text is not a sequence of glyphs but a sequence of glyphs read against a context, and the context of an LLM in the 21st century is not the context of an Elizabethan playwright. Menard is the patron saint of a question we will be arguing about for the rest of our cultural lives: when a machine writes the words a human wrote, what, exactly, has been written?

‖ The Literary Heart

"Componer el Quijote a principios del siglo diecisiete era una empresa razonable, necesaria, acaso fatal; a principios del veinte, es casi imposible."

"To compose the Quijote at the beginning of the seventeenth century was a reasonable, necessary, perhaps inevitable undertaking; at the beginning of the twentieth, it is almost impossible."

identical text × Δt(329 years) → meaning differential

The same character string, authored by Cervantes in 1605 and by Menard in 1934, is two different works. Borges' equation: the meaning of a text is the integral of the words against the historical position of the author. Change the author, and the integral — even over the same words — is a different number.