¹ Compare, for instance, the sound of the bell on the garden gate at the narrator's home at Combray, described near the beginning of S i, 26, with the recollection of the same sound in carefully modified terms at the very end of T R ii, 259; or the pair of passages discussing the ideal object of the affections of Charlus: S G i, 267 and T R i, 169. Examples of this sort of variation on reappearing themes abound all through Proust's work. Another type of duplication is that in which the author develops an idea already suggested, acknowledging at the same time its earlier mention: cf. T R ii, 70 and 240, or T R ii, 72 and 75. Attention should perhaps be called to a case wherein the narrator takes care to remark upon the close resemblance of M. Verdurin's conduct to that of the duc de Guermantes, pointing out the intended parallel in the words: “… imitant ainsi sans le savoir le Prince de Guermantes” (P ii, 35). Here of course he has in mind not the prince but the duke, the prince's brother. (See G ii, 244; S G ii, 1, 46–47; and S G ii, 1, 133–134.) And in one passage (quoted in full in note 14) the author, uncertain as to whether he is about to repeat material required to elucidate what he intends to narrate, prefixes to his exposition a frank avowal of his doubts.

Note: For references to the different volumes of A la Recherche du Temps Perdu, the abbreviations used in Les Cahiers Marcel Proust (Paris: Libr. Gallimard, 1927), ii, 25, have been employed; the pagination is that of the current edition of Proust's novel.

² Five such parallel passages, difficult to prove unintentional because lacking in verbal agreement, are listed as follows: A. The narrator tells of a visit to the writer Bergotte made in order to grant a favor to his mistress (P i, 257); later Charlus calls on Bergotte with a similar intent (P, ii, 26–27). B. This example concerns a plot manœuvered by Morel to obtain the dismissal of the Verdurins' coachman (S G ii, 3, 99–101); Saint-Loup later utilizes the same idea (A D i, 88–89). C. An indiscriminate expression of opinion regarding Pelléas et Mélisande, evincing ignorance of music, is attributed first to a titled neighbor of the Cambremer family (S G ii, 2, 31), then to a relative of that family (S G ii, 3, 177). D, E. Dr. Robert Proust, the author's brother, has called attention to two cases (see foot notes, A D ii, 50 and 55); that is, two anecdotes related by Mme de Guermantes (P i, 47 and 48) and repeated by her with variants (A D ii, 49–50 and 55). Since we find no such “coincidences” in the earlier portions of Proust's work, it seems improbable that such a scrupulous writer would have permitted all to remain unmodified in a definitive text.

³ As it has not been possible to consult the author's MSS these conclusions are limited in extent and must be subject to revision.

⁴ Italics in the passages quoted are employed to facilitate comparison.

⁵ Evidently a misreading of Proust's handwriting; “décrire” probably intended, as in the other version.

⁶ Perhaps a misprint; should be “parlât-elle.”

⁷ Attention should perhaps be called here to another passage in the same volume bearing certain similarities of detail to the first passage above. (See T R ii, 12.) The sentence beginning “… que la moindre parole …” and including such expressions as “vases,” “rempli … d'une odeur …,” etc., seems to foreshadow vaguely the one quoted above which begins: “Dans la moindre sensation. …”

Compare also this further variation on the sentence, found at the end of Sodome et Gomorrhe ii:

Moi qui ne m‘étais jamais éveillé sans sourire aux choses les plus humbles, au bol de café au lait … je sentis que le jour qui allait se lever dans un instant, et tous les jours qui viendraient ensuite ne m‘apporteraient plus jamais l'espérence d'un bonheur inconnu, mais le prolongement de mon martyre. S G ii, 3, 219.

The words “les plus humbles,” (cf. first column above), “bol” (second version), etc., seem to imply that the source of this fragment is composite. Since the sentence is separated from the remainder of the material accompanying the idea in the longer passages, it would appear to have recurred to the author from memory. Probably, then, he was not aware that he was repeating what had been used elsewhere in his book.

⁸ Not only as to style and content, but also from the standpoint of principle, judging from what he wrote to his friend J.-E. Blanche, Correspondence générale de Marcel Proust (Paris: Plon, 1932), iii, 139: “En principe, je trouve, et par raison de doctrine, absurde, de préférer une première version, une esquisse, etc.”

⁹ This case of repetition and the three which follow are mentioned by Alfons Wegener in an appendix to his dissertation entitled Impressionismus und Klassizismus im Werke Marcel Prousts (Frankfurt am Main: Carolus-Druckerei, 1930). See p. 110–113.

¹⁰ Similarly, in a recent work based on Proust's additions to his original second volume inserted since the Grasset proofs of 1913, (see Comment Marcel Proust a composé son roman [New Haven: Yale University Press, 1934], p. 71, note 5) Professor Feuillerat reveals that Proust had added just such an interpolated phrase in his proofs in order to “explain” an inconsistency. In this case, the author appears to have had the same purpose of avoiding a sacrifice of material. Also see p. 103–104 where Mr. Feuillerat discloses another of these additions to the definitive text found lacking in the document at his disposal.

¹¹ See his letter to M. Paul Souday, Corresp. générale, iii, 93, dated “17 juin 1921,” wherein he speaks of the recent appearance of Guermantes ii, which preceded by one year the publication of Sodome ii: “Mais si je corrige déjà très mal mes épreuves, quand un livre comme celui-ci parait imprimé directement d'après mes indéchiffrables brouillons, mes éditeurs ont beau avoir la gentillesse de surveiller de leur mieux cette impression, elle est terriblement fautive.”—Cf. 47 Letters from Marcel Proust to Walter Berry (Paris: Black Sun Press, 1930). Lettre 38.

¹² See the photostat reproductions of some of these documents in Pierre Abraham's Proust, éd. Rieder (Paris, 1930).

¹³ Writing to J.-E. Blanche, Corresp. générale, iii, 127, at about the year 1918, Proust speaks of “… ma fatigue, mes yeux, la hate fébrile de terminer mon livre avant de mourir. …” Elsewhere, in a letter to Blanche (idem, p. 113) for which the date given is “1917,” he complains pathetically of his failing memory. This condition was doubtless due to the excessive use of veronal. See Les Cahiers Marcel Proust, i, 54, note.

¹⁴ It is known that Proust wrote lying down in bed, and that he kept his papers about him in considerable disorder. Under these circumstances, an invalid to whom the least effort was torture must at times have found it too wearisome an undertaking to search through his piles of almost illegible manuscripts to find and compare an earlier passage with the one at hand. In fact, at one point in the text, interestingly enough, he frankly admits to the reader his uncertainty as to whether he has previously treated a particular social attitude resulting from the Dreyfus affair, and he interrupts the narrative to remark: “Je ne sais si j'ai dit … pourrai-je le demander au lecteur comme à un ami à qui on ne se rappelle plus, après tant d'entretiens, si on a pensé ou trouvé l'occasion de le mettre au courant d'une certaine chose? Que je l'aie fait ou non. …” P ii, 44–45. This clearly demonstrates how troublesome Proust must have found it to check back, since he actually renounced doing so on this occasion.

¹⁵ Dr. Robert Proust, Les Cahiers Marcel Proust, i, 19, writes: “Déjà à ce moment [l'époque des Plaisirs et les Jours, vers 1896] il commençait à prendre des notes pour la rédaction de ses fameux Cahiers qui devaient constituer le grand cycle: À la Recherche du Temps Perdu. …”