University of Virginia Library

4. The Texts of the English Review Galley Proofs

As I have argued, Conrad corrected and revised three sets of galley proofs: one for the English Review; one for Harper and Brothers; and one for Methuen. Revision on one set of proofs was then followed by transcription onto the other two sets. Most revisions were transmitted to all three publications, But Conrad's transcription was not always successful because, as the figures quoted above indicate, a significant number of changes failed to be transcribed. Subsequently, the uncertainty surrounding Conrad's intentions in this process of transcription complicates conclusive attribution of Conrad's authority to many variants. Nevertheless, close examination of specific variants further supports the textual scenario I have proposed, revealing the effect that Conrad's method of transmission had on the texts under consideration.

Approximately 580 substantive variants from TS appear in all three texts, representing Conrad's successful transcription between each set of English Review galley proofs. However, a number of cases show that Conrad revised each proof differently, perhaps inspired after the initial revision on one set of proofs. In TS we read this description of Madame de S—: "She disregarded it. Her carmine lips moved with an extraordinary rapidity". While the English Review text remains the same, Methuen's text replaces "moved" with "vaticinated" (220). Harper and Brothers' text retains the sentence, but adds, "She vaticinated" (221) The addition of such an unusual word only in both book publications strongly suggests Conrad's involvement, but the variation between them indicates that Conrad treated differently the various copies of each galley proof from which each text derives. Such a development is seen most strikingly in the last chapter of the novel where the original reading of Sophia Antonovna's "indulgent voice" from TS became "soft" for the North American Review, "prudent" for the English Review, "cautious" for Harper and Brothers and "guarded" for Methuen. From these and other examples found throughout the novel, one can see that Conrad's transcriptions were not always simple copying, but sometimes developments of earlier readings. Each set may have been used as a draft before a final revision was incorporated on one preferred set of galley slips.[25]

This method of transcription often left the Harper and Brothers text without new readings incorporated in the English Review and Methuen. These variants are detected when all texts are compared, revealing agreements between Harper and Brothers and the North American Review that suggest a common relationship.[26] At first glance, this seems to contradict the picture drawn above of the transmission of the text. But, the agreements between Harper and Brothers and the American serial can be explained by identifying the document on which the change was made. If one assumes that the typescripts sent to the North American Review and the English Review were copies of TS, one would expect agreement between the serials except where Conrad made changes on the English Review galley proofs. For example, "vanishes" (TS29) reads "vanished" and "do not ask questions" (TS 120) reads "do you ask questions" in both American texts. The transmission of the latter example probably proceeded like this: The reading "do not ask questions" was incorrectly typed in April 1910 or changed by Robert Garnett during his corrections. The new reading, "do you ask questions" was transmitted to the North American Review and would have appeared in the English Review, had it not been corrected by Conrad on the galley proofs of the English serial. If this correction was not transcribed to the set of proofs prepared for Harper and Brothers, the earlier reading would remain, producing an agreement with the American serial.[27]

The three sets of galley proofs that passed through Conrad's hands were treated differently, and, from the variants that appear through comparison, one must assume that the set for Harper and Brothers did not receive Conrad's full attention. This is most evident in readings that are unique to both English texts. Almost 200 variants fall into this category, further supporting the argument that Conrad did not correct and revise the galley proofs for Harper and Brothers to the same extent as the English sets.[28] For example, the TS reading that has Miss Haldin "walking alone in the main alley" (TS182) reads "walking alone in the main valley" in both English texts. Because this change was not transcribed to the set of proofs prepared for Harper, the reading in the text of the American first edition agrees with the unchanged reading in the American serial and TS. Many small readings such as this fall into the pattern of agreement between the English texts, especially in the second half of the novel when galley proofs were sent to Harper and Brothers with no corrections or revision.

One of the most intriguing groups of variants produced by Conrad's use

of English Review galley proofs is the unique readings found in the published English Review text. In chapters three and four of Part Four, 30 omissions and variations are unique to the published English Review text. For example, a passage that appears on TS"R15" reads,

This passage appears in all published texts except the English Review serialisation, suggesting that Conrad might have deleted it from one set of galley proofs, but failed to transcribe the change to the sets of proofs sent to Harper and Methuen. Another example is found several pages later. The narrator describes Razumov's state of mind: "It was as though he had stabbed himself outside and had come in there to show it—and more than that as though he were turning the knife in the wound and watching the effect" (TS"S20"). The emphatic text that follows the dash in this passage is omitted only in the English serial.

This collection of variants complicates the attribution of authority because of its inconsistency with the dominant patterns of variation found in the rest of the novel. A second, longer, omission involves the narrator's analysis of the attraction between Razumov and Natalia and shows Conrad's different treatment of the three sets of proofs:

Higdon cites this variation as evidence that Conrad was consulting TS when preparing for book publication. Higdon argues that this passage is "no simple addition but instead is a complex reworking of a passage in the typescript cancelled before the English Review was set" (Higdon, 1991b, 177-178). It is highly unlikely that this passage was cancelled before the English Review was set because it was transmitted to the first American edition which, I argue, used English Review galley proofs as setting copy. The entire passage quoted above is found in Harper and Brothers' text. In Methuen's text the passage between "and it was impossible" and "She was no stranger to him when he saw her first;" is omitted. If Conrad was preparing three sets of galley proofs, the set for Harper and Brothers was untouched and the sets for the English Review and Methuen received different treatment.[29] It is most likely that the

collection of unique omissions and additions in the English texts derive from two different sets of galley proofs, representing different moments of intention. However, because the majority of the variants unique to the English Review are found in one section of the October instalment, these variants might be the result of Methuen's "beastly muddle".[30]

Conrad's acceptance of Methuen's text might be seen as a case of passive authorisation, but his continued dissatisfaction with Methuen suggests that he had some misgivings about the text of the first English edition. Unique readings in the Methuen text may indicate Conrad's continued involvement after corrected English Review galley proofs had been sent to Methuen as setting copy. For example, the Methuen text reads "a long time" instead of "many years" (TS2); "gouty invalid" for "gouty subject" (TS15); "other, flinging off his cap" was expanded from the simple "other" (TS21) to match an image of Haldin's hair several paragraphs later; and "gloomily" replaced "in a gloomy murmur" (TS35). Approximately 290 readings unique to the first English edition suggest that Conrad corrected Methuen's proofs; or, alternatively, he corrected and revised the English Review galley proofs for Methuen to a greater extent than the galley proofs he sent to the English Review. Of course, many of these changes might have been editorial or the work of a compositor. But, while the majority of unique readings could be attributed that way, most readings suggest Conrad's hand. For example, in the closing paragraphs of the first chapter of Part Third the following passage was omitted only in Methuen's text:

Higdon attributes the excision of this meditation on journals to Conrad, arguing that it was done "perhaps in a move to make his narrator more limited, less perceptive, more self-deluded about his true attraction to Razumov's story" (Higdon, 1991, 176). Such interpretations can allow attribution of Conrad's authority to larger variations. But, without further evidence, one can only attribute Conrad's authority to smaller variants with trepidation.

The deadline set by Methuen arrived and Under Western Eyes was published, ending Conrad's opportunities to correct and revise, and establishing

the version of the novel familiar to most readers.[31] The four published texts of Under Western Eyes reveal characteristics that give each text a unique status. The text of the North American Review remains closest to TS and probably resembles the text of the typescript "corrected" by Robert Garnett. The texts of the English Review, Harper and Brothers and Methuen are the result of Conrad's inconsistent revision of English Review galley proofs. There is little doubt that Conrad wanted Methuen's edition to represent his final intentions. But the mechanical processes of typing and typesetting that preceded its final state complicated this outcome, and undetectable errors may have been incorporated in Methuen's text. As the case of Joseph Conrad's Under Western Eyes stresses, readers must not disregard the physical and mechanical processes of textual production in their analyses of the history of texts, because these processes can have a significant effect on the way an author transmits a text to readers. Behind the stability of the text of Under Western Eyes familiar to readers is a complex network of transmission that poses many interpretative challenges; challenges that, when fully engaged, will offer a better understanding of Conrad and his texts.