Manuscripts on Microfilm: The Disturbing Case of Proust - Anthony R. Pugh
I have three anecdotes, three for the price of one: together they indicate that there is more than one aspect to the business of manuscript versus microfilm. One of the anecdotes even suggests that we should be happy that some manuscripts have been preserved on film.
For the last ten years or so I have been working on the manuscripts of Marcel Proust, trying to establish the chronological sequence of everything that he wrote in preparation for his great novel A la recherche du temps perdu, from 1909 (when it first took shape, and the pillars were set down) through 1911 (when Proust completed a first draft) to 1913 (when a typescript of half the novel was ready for a publisher; half this typescript was set in proof and half the proof published). My examples come from the first book of the novel, Du côté de chez Swann, the part entitled "Combray"; from the central portion, Le côté de Guermantes; and from the last part, now La fugitive or Albertine disparue.
The "manuscripts" are essentially exercise books, with the addition of a few episodes written on loose sheets and of a seven-hundred-page typescript that covers about half the novel as it was envisaged in 1911. Proust wrote on the recto pages of his exercise books, frequently passing from one book to another, and maybe another, and back again to the first; he used his verso pages for subsequent additions. Occasionally he saved himself the trouble of recopying by giving a whimsical cross-reference (a sketch of a boat, or a butterfly, for instance), but generally, as he refined his prose and reorganized his episodes, he rewrote the whole text.
This material was all in private hands until the early sixties, when the Bibliothèque Nationale in Paris acquired the lion's share (other exercise books have come into the library since then). It was not in good order, with exercise books falling apart, and so the books were all "restored," which means essentially that they were rebound, with the loose pages firmly fixed in place. A practical method of classification was devised, assigning each rebound exercise book to the part of the novel to which it seemed most closely related; thus the books that work on the overture were numbered 1–7, the books on "Combray" 8–14, the books on "Un amour de Swann" 15–19, and so forth.
For twenty years scholars had a field day with this material, gradually analyzing the contents, identifying problems, proposing solutions. But the day came, inevitably, when it was decided that too many fingers in the Proustian pie were having a destructive effect on the precious pages and that henceforth consultation would be done on microfilm, in order to preserve the manuscript. It is not impossible to obtain permission to see the originals, but all routine work has to be done first under glass. In the process, things can get overlooked.
One disadvantage of microfilms, as everyone who has used them knows, is that all sense of the physical reality of the book disappears. A microfilm is like a black-and-white reproduction of an oil painting. The words are there, but everything has become intellectualized, distant. For the scholar in search not only of evidence but also of suggestions, some clues are no longer discernible—the color of the ink, for example. Ten years ago I realized that the first part of the "1911" typescript, including nearly all of "Combray," must have been made toward the end of 1909, and I asked a young Japanese doctoral student, Akio Wada, who was writing a thesis on the evolution of "Combray," if he had considered that possibility. "Yes," he replied, "and, what is more, I can prove it." His proof, which he would not disclose at the time but which became public knowledge once his thesis had been defended, was that a note in Proust's personal carnet, which we can certainly date December 1909 or January 1910, indicates in red ink that there were passages added on certain pages. On the typescript all these pages have marginal additions, and they are all in red ink. Red ink is very rare in Proust. Luckily Wada noticed the red ink in time, before the manuscripts were taken from him.
With microfilms you proceed one dreary page at a time, without the sense of the whole, which you get physically from the exercise books themselves. My second anecdote suggests that the mechanical process of microfilm reading encourages laziness. It was always said that there was a gap in the manuscript version of the central portion of Proust's novel, Le côté de Guermantes. The manuscript version is found in exercise books 39–43 and 49 of the Bibliothèque Nationale collection. The last page of 43 ends in mid-sentence, and the first page of 49 begins in mid-sentence, and they are not the same sentence. The assumption, never questioned, was that we were in the presence of two different versions, the first version missing only its last exercise book, the second version known only through its last exercise book. Or maybe it was the other way around; there was some dispute about which was the first version and which the second. But nobody doubted that there were two distinct versions. We were, however, in the presence of a modern-day "golden tooth." The gap between 43 and 49 is of only a few lines, as we see if we compare the draft with the published text, and it seemed to me unlikely that 43 and 49 did not belong together. That the remaining portions of two different versions should fit almost exactly was too much of a coincidence. Evidently a few lines, perhaps one manuscript page, was missing. Could the missing text be found? More easily than I had imagined; for when I looked at the microfilm of book 43, I realized that when Proust had got to the end of his last page (writing, as usual, on the recto), he continued on the facing verso. But once again he reached the end of the page without finishing his sentence, and he picked up a new book (our 49) and completed it there. The scholars who had studied Guermantes had evidently not thought to wind back the reel to the previous image. It can get quite hot and soporific in the microfilm room, but even so. . . . Can you really afford not to bother to read the verso pages of a crucial manuscript?
My third example comes from later in the novel, the part that never made it to the 1911–12 typescript. Here we find that the microfilm, far from obscuring the evidence, actually leads us to the truth about the original version. This part of the novel, reworked by Proust during the war, is one of the rare instances where he simply cut out pages and included them, without recopying them, in his new manuscript. Another Japanese scholar, Jo Yoshida, had spotted and identified five such pages. When I was trying to reconstitute the original 1911 version, using the microfilm, I noticed a sixth page, of which only the margin (which was empty) remained in the original exercise book; the rest of the page had been pasted into the wartime exercise book. But in 1911 Proust had written something on the back as well, going of course all the way to the right-hand edge of the page, so when he cut along the margin line of the recto, which was the text he wished to reuse, the writing on the verso was divided about one inch from the end of each line. The text could be reconstituted by putting together the two portions of the divided page, but this is quite unpractical on microfilm, even if one could for a few minutes use two adjacent readers. So I preferred to add this question to the list of questions I needed access to the originals to answer. Imagine my surprise when I had the original 1911 exercise book in my hand and could find no trace whatever of the page I wanted to see. I can only assume that the one-inch margin was "tidied up" when the exercise books were rebound and that fortunately this tidying up occurred after the microfilm was made. I have no proof of this, however. It shows why the Japanese scholar counted only five such pages: in the early eighties, he was working straight from the originals. It also presents a horrendous prospect to the conscientious scholar, who would have to compare every page of the originals with every page of the microfilms to be really sure that nothing was missed. This is impossible, practically speaking, but it shows just how careful one has to be.
This final anecdote leads to another conclusion. Scholars are responsible for their evaluation of the evidence put before them. The onus is on libraries to give very complete bibliographical descriptions of their manuscripts and to suppress no evidence, however untidy it may be. Scholars need to contemplate the mess and work it out for themselves. The assumptions of librarians and archivists may not be beyond reproach, and it is dangerous to present assumptions as facts and tidy up the evidence.
University of New Brunswick