In Search of Lost Works (19) - William Marx (2021-2022)

The final lesson in this series on lost works concludes the cataloguing process and sets the stage for exploring methods of recovery. Four categories of lost works have already been examined: those whose traces are entirely gone, those represented only by reflections, those surviving only in fragments, and those absorbed into later creations. A fifth crucial category is introduced: unrealized works. These are not merely unfinished or fragmentary pieces, but rather concepts fully composed in the creator's mind yet never committed to paper, remaining stillborn within the author's spirit.

Jean-Jacques Rousseau provided astonishing insights into this phenomenon within the fourth book of his *Confessions*. He expressed deep regret over not keeping journals during his travels, noting that walking animated and sharpened his ideas, as remaining stationary hindered his thought process. The combination of fresh air, physical appetite, and freedom from dependency provided an audacity of thought, allowing him to combine and appropriate elements of nature masterfully, often resulting in vibrant descriptions he never fixed to paper.

What vigor of brush, what freshness of color, what energy of expression I give them in myself.

Rousseau questioned why he did not write down these mental compositions during his travels, concluding that the charm lay in the immediate enjoyment of the moment, rather than the prospect of readers. He recognized that ideas arrive when they please, not on command, sometimes in such a flood that ten volumes per day would be insufficient to transcribe them. This intense, solitary intellectual gratification was valued highly, standing in contrast to the physical self-satisfaction that carried hygienic anxieties in his era.

This intense creative outpouring, when unconstrained by the necessary friction of writing, risks consuming itself. This concept finds its literary parallel in Honoré de Balzac's short story, *Le Chef-d'œuvre inconnu*. The painter Frenhofer, in his relentless pursuit of perfection, continuously adds retouches until his masterpiece degrades into an amorphous magma, where only a magnificent foot remains discernible. The moral suggests that uninterrupted creation fails to produce a stable work; the process must halt for form to stabilize.

The sixth category addresses works lost because they were arrested by publication. This act stops the creative process, meaning the work could have evolved into something else had the creator continued. Paul Valéry expressed a similar sentiment, stating that returning to anything previously written would result in an entirely different creation if external circumstances had not broken the enchantment of incompletion. This state of ongoing creation, where the work is still being formed, is crucial.

When an author does not return to an edition, the virtual work falls to the reader to realize. This concept has fueled a significant branch of contemporary criticism. Pierre Bayard is noted as a master of these unrealized work virtualities, particularly in his work on improving failed creations. This critical school also includes Marc Escola, who imagined a sixth act for Molière's *Le Misanthrope*, and Sophie Rabau, who explored variations on Mérimée’s *Carmen*.

The seventh category involves neglected works, which are those that exist in physical libraries but remain lost within mental ones. It falls to the researcher, philologist, or scholar to exhume these pieces and bring them to collective awareness. Works become neglected for many reasons, often failing to enter the canon. This includes minor works by canonical authors or, conversely, the major works of non-canonical authors. Literary historiography frequently relegates atypical works, such as those belonging to aesthetic rear-guards, outside the central field of tension.

Historiography, when defining dominant aesthetic movements, relegates a quantity of atypical works outside the field of tension.

Resurrecting a neglected work is not without problems, as demonstrated by the case of Irène Némirovsky, who died in 1942. Her posthumous, unfinished novel *Suite française* achieved massive success decades later, winning the Prix Renaudot. However, this resurrection provoked controversy in the United States regarding the representation of Jewish characters, suggesting that Némirovsky was being inflicted a second death by being judged by contemporary, hypersensitive standards, effectively erasing her work again in the name of Judaism.

The eighth category concerns works that appear not to be lost because they are still read as complete, existing entities, when in reality they are ruins or trafficked objects whose disfiguration is invisible. While the damage to a sculpture like the Parthenon is obvious, literary texts more easily conceal their amputations. When reading an ancient Greek tragedy, the reader may not perceive the missing references, allusions, or context that made the text fully alive for its original audience.

The final category covers works whose loss is programmed by their very consumption, such as food and cuisine. When consumed, the work disappears, much like a fruit ripening and then being eaten. This ephemeral nature integrates deeply into memory, exemplified by Proust’s madeleine or Karen Blixen’s *Babette's Feast*, where a meal serves as a powerful act of commemoration and gratuitous expenditure. In the Bible, the prophet Ezekiel eats a book, incorporating the word to transform it into prophetic speech.

The desire to recover these lost treasures leads to exploring what might be called the 'dark matter of literature'—the invisible works structuring the visible ones. To investigate these completely lost works, at least two methods are possible: the direct and the indirect. The direct method, also termed the genealogical method, attempts to infer the lost works from the structure and composition process of the surviving texts.

This direct approach has been employed in biblical exegesis since the 18th century, notably by Jean Astruc, who distinguished between the Elohist source (where God is called Elohim) and the Yahwist source (where God is named by the Tetragrammaton). This principle of distinguishing textual strata allows for the potential reconstitution of lost originals. Victor Bérard applied a similar, though highly controversial, method to the *Odyssey*, recomposing the text based on his own interpolations, a liberty rarely afforded to the Bible.

Recent genealogical attempts, such as Julien d'Huy's work in *Cosmogonie* which uses genomics techniques to reconstruct primitive myths from existing versions, highlight the excitement surrounding these revolutionary methods, which possess the appearance of scientific objectivity. However, these endeavors must be approached with extreme caution due to inherent methodological problems that plague all genealogical approaches.

Furthermore, the direct method can only work with works that have left descendants; if lost works generated other lost works, this lineage is untraceable. The most instructive lost works—those most different from what is known—are precisely those that left no discernible descendants.

The indirect method offers a partial circumvention of these issues by focusing on the mechanisms of loss itself, calculating what is missing based on what remains. This involves an interrogation of the negative space within both real and mental libraries. A recent, highly commented example involved applying statistical techniques used in biology to evaluate unknown species to medieval Arthurian romances.

Researchers modeled the diffusion of Arthurian novels across manuscripts, treating manuscripts as specimens. By plotting the number of surviving manuscripts for each romance, extrapolation allowed estimation of those represented in zero manuscripts. These calculations suggested that 32% of Arthurian romances have disappeared, with loss rates varying significantly by language, such as 47% for French romances versus 62% for English ones. Equitable copying across manuscripts increased preservation chances.

A second example of the indirect method concerns Greek tragedy. Only 32 complete tragedies survive from scores written by dozens of playwrights. Analysis revealed that the 24 tragedies selected for the ancient school canon (stabilized around the 2nd and 3rd centuries CE) disproportionately favored unhappy endings for Euripides (8 out of 10 canonical plays). In contrast, the eight 'alphabetical' tragedies—preserved randomly rather than pedagogically—showed the inverse pattern, with seven out of eight having happy conclusions.

The indirect method provides astonishing results regarding the nature and quantity of lost works, even when absolutely nothing remains of them. These results stem from analyzing the structure of the remaining corpus and understanding the historical process by which that corpus was constituted. By accounting for these factors, it becomes possible to extrapolate the structure of the missing corpus, partially revealing that dark matter of literature which underpins all visible creations.

This research into lost works has yielded nearly unexpected results, paving the way for new lines of inquiry and fresh readings of existing corpora. While time constraints prevent further detailed discussion of Greek tragedy analysis, the task now falls to the audience to continue completing these curves through contemplation of those invisible works whose contours are finally beginning to be sketched.