Deadlock

The English First Edition was published in February 1921. The American Edition was listed in The Publishers' Weekly for 26 November 1921 (Smyers, 65). In a letter dated "Mar: 9th, 10th or thereabouts 1921" Dorothy Richardson wrote to Owen Wadsworth: "I have strained my eyes trying to prepare, for America, a proof of Deadlock relieved of some proportion of the hundred odd printers errors so confoundingly included in the copy I send you herewith" (Windows on Modernism, 47-48). A year later in March 1922 she again wrote to Wadsworth: "Your letter . . . finds me packing up for you a copy of the American Edition of 'Deadlock.' I want you to destroy Duckworth's travesty & substitute for it this almost perfectly corrected version" (Windows, 57). If Richardson had completed revisions to the proofs for America by around 10 March 1921, it is quite possible in light of the time constraints that she was correcting for Knopf proof sheets from the English Edition and that she never saw proofs of the actual American text, but concluded after looking over the book itself that her corrections had been carried out.

A comparison of this text with the English text does indeed show that the many clusters of points used to indicate breaks in thought or subject or continuity, clusters which seem almost random in the English First Edition, are brought under strict control: three points for a break within a sentence; a period and three points for a break at the end of a sentence. As well, words like regularise and realise are changed to z. But apart from that, it is difficult to make a case for the excellence of the American text.

Of 70 instances entailing a change of word or word form between the English and American editions (but not including spelling errors or word divisions), the Collected Edition follows E in 61 cases and A in 9 only. And of these 9, 6 may be regarded as errors or misprints in E, which are corrected in A and later in CE. For example E40.30 it, A38.15 and CE38.38 is; E161.29 effects, A153.21 and CE124.5 affects. Two involve the phrase different to / different from, a change of idiom Richardson continued to insist upon after 1921. And 1 is an optional usage: E290.32 go and have; A275.32 and CE216.36 go have. What is equally revealing, CE follows E in 4 of its errors. CE38.36 There's isn't is like E40.27 even though A38.13 had been corrected to There isn't; CE164.12 Marie Duclaux is like E217.26, though A206.12 had been corrected to Lucie Duclaux; CE202.18-19 lost the is like E270.32, though A257.20 had been corrected to lost in the; and CE208.28 as wide at the world is like E279.24, though A265.33 had been corrected to as wide as the world. These 4 instances show that the American editor was alert to defects in his copy text. And they confirm what was suggested by the 53 instances in which CE followed E rather than the variant word forms in A: E was used as copy text for the Collected Edition text of Deadlock.

Clearly then, CE follows E and ignores the verbal changes to A. But when it comes to quotation marks, the situation is precisely the opposite. Whereas E erroniously omits a good many such marks, A inserts them. In three cases, however, there is an excuse for the omission of quotation marks in E: a speech is being recorded within another speech; here the convention may be regarded as optional, since the reported speech may be allowed simply to flow into the containing speech. ("Then when I say that what is really wrong is that I have been cheated of my student period and ought to be living on somebody as a student, he says, pairhaps, but you are in life, that is the most important." [E133.28-.32; A126.19-23 and CE104.4-7 with added quotation marks around the inset speech.] ) The two remaining examples are found at E134.5-6 (A126.29-30 and CE104.12-13), and E193.25-27 (A183.22-24 and CE147.2-4, with misprint). But all other instances of missing quotation marks are just carelessness in E. Of 17 such instances, CE follows A in 13, and E only in 4. And all 4 as they appear in A may be regarded as mistakes. At A 182.17, A183.11-13, and A234.5-9 the American editor moves the closing quotation mark to include words which are represented as follow-up thoughts by Miriam at E192.18, CE146.5-6; E193.13-15, CE146.29-32; and E246.25-30, CE185.11-14; whereas at A182.25-26 he does the reverse, moving the quotation mark forward to exclude part of Miriam's speech at E192.26-27, CE146.12-13. One can make out quite a good case for A182.25-26, and further argue that Richardson may have inserted the change on the corrected proofs for Knopf. On the other hand, there can be no doubt that the shift of quotation mark at A234.5-9 is a mistake by the American editor. Miriam could not speak the words in question to Mr. Hancock. Given this fact and the evidence of CE's otherwise reliable emending of quotation marks, I think we have no alternative but to accept the authority of CE and conclude that these shifts in quotation marks in A are unauthorized. This judgment is reinforced by the American editor's making 7 other undoubted errors in reproducing Richardson's quotation marks. None of these errors, of course, is duplicated in CE.

As for the other 13 misprints and errors in which A corrects E and CE follows suit, it is impossible to be certain who is responsible for the American revisions. My guess is, Richardson included them in the proofs she sent to Knopf. In weighing this issue, one would like to refer to "the hundred odd printers errors" the author found in Duckworth's edition. Yet once one moves beyond the 9 verbal errors found in E and the 13 revisions of quotation marks that might be attributed to Richardson and surveys the entire 516 variants between the two editions, one is confronted by numerous changes in routine punctuation, including well over 300 variants in the use of periods and suspension points. In addition, 190 commas are inserted in the A text; 165 of them are repeated in CE, 25 are not. We may suppose Richardson responsible for some of these added commas in A, but clearly the American editor must have pitched in as well. Moreover, he deletes 23 commas found in E; CE restores 19 of them. So the figure of 100 is too large to include only the changes to words and quotation marks, and too small to cover all the variations in punctuation.

On the more important question of the verbal changes in A, my conclusion is that nearly all, except for the emending of the 9 errors in E, were made by the American editor without consulting Richardson. I count 49 of these, all ignored in the Collected Edition. They usually entail alterations such as: around for round, in the face of for in face of, and on 5 occasions English for British; but also, surprisingly, sighed (A117.8) for gasped (E124.9, CE97.13) . One at least is a fussy correction that is incorrect: The small cluster of tiny soft-toned pearls encircling and curving up to a small point of diamond were set in a circlet of enamel . . . (E22.26, CE26.8); the American editor, seeing what he thought was a singular subject with a plural verb, revised the text to read diamonds were (A21.12); whereas the actual subject being cluster rather than pearls, it is Richardson's verb that is at fault. In addition to the 49 proposed improvements, there are several errors introduced by the American editor: A44.31 sunlight, E47.22, CE43.30 sunlit; A104.6 Eva, E110.15, CE87.18 Eve; A142.6 abruptly blanching moustache, E150.1, CE115.26 abruptly branching moustache; and A289.12-13 upon a fixed ideas, E304.18, CE227.1 upon fixed ideas.

Since the American Edition of Deadlock was thought by Richardson to be "almost perfectly corrected," Dent might have been expected to use it as copy text, especially since it had retained English spelling, had changed the ise words to ize, and had regularized the rather creative punctuation of E, all things that CE would also do. But the evidence is very strong that the English Edition was used as copy text in setting Deadlock for the Collected Edition. Otherwise at least one or two of the numerous American verbal revisions would have slipped through. None did. On the other hand, the successful correction of all the quotation marks in Deadlock may be put down to the attentiveness of Richardson herself or of her editor at Dent. It is also possible that she had retained a copy of the English First Edition marked up with the corrections she had inserted in the 1921 proofs for the American Edition, including revisions of the missing quotation marks, and that she supplied this corrected version to Dent as copy text. Unfortunately there is now no record of any of Richardson's own copies of the Pilgrimage volumes. And there is very little information to show how proofs were prepared for the Collected Edition.

NOTE The American Edition, like the Collected Edition, changes words like apologise and realise to apologize and realize. Since these have already been recorded in comparing the English First Edition with the Collected Edition, they are not noted here.

NOTE There are many irregular points used for ellipses in the English Edition. These were corrected in the Collected Edition. Normally they are corrected also in the American Edition, in which case the irregularities of the English Edition will not be noted here. Where the American Edition does not follow the standard practice of using three spaced dots, thus . . . at the beginning or in the middle of a sentence and a period and three spaced dots at the end of a sentence, this irregularity will be noted. Exception: when the irregularity occurs in both first editions; in that case the irregularity will have been noted in comparing the English First Edition with the Collected Edition.

However, this account of variants is incomplete because the revisions to A to regularize punctuation do not always correspond precisely with the revisions to CE. For example, the following item is overlooked: E83.27 knowledge. . . . . . ; A78.11 knowledge. . . . ; CE68.20 knowledge. Because the A text was regularized with a period and three dots, it was assumed to conform to the regularizing of the CE text. But in this and a significant number of other instances that was not so. These instances were missed because in comparing the English and American texts, CE was not consulted.