University of Virginia Library

Further Observations on the Text of Dubliners
by
Robert Scholes [*]

This essay is a continuation of a study begun in Studies in Bibliography, XV (1962), 191-205. In the first essay, on the basis of an examination of the various manuscripts and proofs of one story, "The Dead," I attempted to establish a rationale for the textual criticism of Dubliners. In brief, and without the supporting evidence, the rationale is this: The first complete printing of Dubliners was done by the firm of Falconer in Dublin in 1910 for a projected edition to be published by Maunsel and Company. This printing went through three states: 1) a set of galleys, printed from Joyce's holograph manuscript but including many compositorial "improvements" in punctuation (including about one thousand additional commas); 2) a set of early page proofs, somewhat more correct than the galleys; 3) a set of printed pages (referred to as a "late stage" of the printing in these studies but almost certainly the final stage, ready for the binder) thoroughly corrected by Joyce and containing some new improvements to the text made by him in proof-reading.

When the Dublin publisher and printer finally refused to publish the book, Joyce obtained a set of the early page proofs (the second state described above), which became the printer's copy for the actual First Edition published by Grant Richards in 1914. There are two states of the printing of this edition. The Edinburgh printer, The Riverside Press, eliminated the galley stage and provided a set of page proofs for correction. Having been set from the quite incorrect early page proofs of the abortive Dublin printing, these proofs required considerable correction. They were sent to Joyce, who made his corrections hastily,

expecting to see another set of proofs. When the other proofs were not sent to him, he forwarded a set of corrections which he hoped would be made. It was not until two years later that he discovered that not only had the thirty additional corrections not been made but that two hundred of the original corrections indicated on the page proofs had not been made either. The second state of the Grant Richards edition is the First Edition itself. Modern reprint texts of Dubliners differ from this only by a few proof-reader's corrections and some new compositor's errors introduced into the text.

For this second study of the text of Dubliners I have examined and collated all the available manuscript and proof versions of all the stories not treated in my first study — that is, all but "The Dead." Since the materials available for study vary from story to story, each will be treated separately below, after a discussion of the general problems which will face the future editor of Dubliners.

General Problems

One of the major problems for the future editor of Dubliners will be the punctuation of the text. He will certainly want to return to Joyce's own desired punctuation of direct discourse (dashes instead of quotation marks); but will he want to follow Joyce's habitual MS procedure and place a dash both before and after a paragraph in which direct discourse appears, or will he adopt the procedure of the published versions of Portrait and Ulysses and use the opening dash only? I am inclined to favor the latter, as representing Joyce's final procedure — but the question is certainly arguable. The use of commas will present another problem of considerable complexity, owing to the bizarre pre-publication printing history of the book. The compositor of the Maunsel (Dublin) printing added about one thousand commas to Joyce's text. In "A Mother," for example, where we have a full set of galley proofs to examine, we find that the galleys have 87 more commas than the MS from which they were set. The Grant Richards page proofs for the first edition (which were set from page proofs of the Maunsel printing) show that 70 commas were removed between the galleys and the Richards page proofs, and we find that 12 more were expunged for the first edition itself. But some of the total of 82 expunged commas were present in Joyce's MS, and a number of compositorial commas which were not in the MS have been allowed to stand in the text. Should the editor go back to MS or allow the final version to stand, as having been proof-read by Joyce and therefore carrying his approval? We cannot be sure which readings in the first edition are among the 200 corrections Joyce wanted to make, but the

existence of 200 errors certainly undermines our confidence in the first edition. Still, Joyce was given to continual revising in proof, a habit which prevents us from accepting with absolute confidence a manuscript reading over a reading in the first edition. Each case will have to be decided individually by an editor who is deeply familiar with Joyce's habits of punctuation and is willing to proceed eclectically, relying on his judgment.

There will also be problems in substantive readings for the future editor of Dubliners. The question of "lost" improvements in the late stage of the Maunsel (Dublin) text has been considered in my earlier essay (and some additional material of this kind will be presented below) but there are questions even more complex than this which the editor will have to face. The final MS of "Ivy Day in the Committee Room," for example, which was the printer's copy for the Maunsel (Dublin) printing, bears a number of marginal corrections in Joyce's hand. Some of these marginal corrections appear in the page proofs and finally in the first edition; others seem never to have been printed. (They might, however, have been incorporated into the final Dublin state, of which, for this story, there is no known copy.) Here again, and in other similar cases, the editor will have to rely on his critical judgment and decide whether Joyce thought the better of his marginal improvement and deliberately omitted to include it in the printed text, or he simply lost track of it, and now the editor must execute Joyce's long-delayed intention. All this suggests a generalization about the problems of an editor in relation to bibliographical theory. It seems that no theory of textual transmission, however carefully and thoroughly worked out, will absolve the editor from using his judgment and his critical faculty. He will have to take risks. In the present instance we have an unusual amount of material available. What would we not give to have two MS and two proof texts for one of Shakespeare's plays (as we have them for "Ivy Day in the Committee Room")? But even such a miraculous acquisition as this would probably leave the text of the play a long way from settled.

Individual Stories

"The Sisters"

Date of composition: This was the first story of Dubliners to be written. It appeared in the Irish Homestead on 13 August 1904. Joyce revised it slightly in June 1905 for submission to Grant Richards, and revised it extensively before submission to Maunsel in April 1909.

Texts available: 1) Published Irish Homestead text; 2) 1905 MS; 3)

revised MS, used for Maunsel printing; 4) late stage of Maunsel page proofs; 5) Grant Richards page proofs; 6) First Edition.

The development of the final text: This aspect of "The Sisters" has been treated thoroughly by Marvin Magalaner in Joyce, The Man, the Work, the Reputation, (1962), pp. 82-87; and Time of Apprenticeship (1959), pp. 74-86.

Lost improvements to the Maunsel text: The two major changes from manuscript in the late stage of the Maunsel text which were not picked up in the Richards text seem of slight importance:

"An Encounter"

Date of composition: The ninth in order of composition, this story was completed by mid-September 1905.

Texts available: 1) MS for Maunsel printing; 2) late stage of Maunsel printing; 3) Richards page proofs; 4) First Edition.

The development of the text: The MS and page proofs show that Joyce continued correcting and improving right up to the end, but most of these changes are small — a word or two at most. The eyes of the peculiar stranger were "sage-green" in the MS and only acquired their more sinister "bottle-green" hue in the Irish proofs. This improvement was not lost in the Richards text, but a few were and are listed below.

Lost improvements to the Maunsel text:

"Araby"

Date of composition: The eleventh in order of composition, this story was completed in September 1905.

Texts available: 1) late stage of the Maunsel printing; 2) Richards page proofs; 3) First Edition.

The development of the text: Though the absence of MS evidence makes speculation somewhat conjectural, this story seems to have satisfied Joyce as completely as any, for there are no major variations among the three texts available. One punctuation change apparently introduced by the compositor for the Richards edition should probably be corrected:

Lost improvements to the Maunsel text: None.

"Eveline"

Date of composition: The second in order of composition, this story was first published in The Irish Homestead on 10 September 1904.

Texts available: 1) Published Irish Homestead text; 2) late stage of the Maunsel printing; 3) Richards page proofs; 4) First Edition.

The development of the text: Joyce revised fairly extensively between the Irish Homestead and the Maunsel versions, and continued tinkering with an odd word or two right up to the First Edition. The changes from the Homestead text are mainly of two kinds. In revising the narration of Eveline's reverie he sometimes replaced words or phrases not quite appropriate for her with more suitable material, and he sometimes added material for the sake of its naturalistic or symbolic point. Consider the examples below. The first consists of an

entire paragraph from the Homestead version and the late Maunsel version.

Irrelevant matter is pruned away. The word "secrete" which is inappropriate to the thought processes of Eveline is removed, and the Blessed Margaret Mary Alacoque is inserted. When we learn that this saint paralyzed herself with self-inflicted tortures but was cured miraculously when she vowed to dedicate herself to a holy life, we can see that Joyce is not merely adding to the naturalistic description of the home of Eveline but presenting the reader with a symbolic parallel to her own life of emotional paralysis. In other revisions the speech of the edgy Miss Gavan is sharpened a little and we are given more detail on Eveline's past relationship with her father. Compare the following two passages:

Here we have not only the interesting addition of the palpitations and the father's past brutality but a significant change in the syntax of the last clause. The formal "were it not for her dead mother's sake" gives way to the "only for her dead mother's sake" in which we can catch the living rhythm of the father's speech. Though the account is narrated rather than dramatized and the discourse indirect rather than direct, the narrative takes its color from the idiom of the characters rather than from any narrative personality. Through countless little changes of this kind, Joyce carefully eliminated his own personality from Dubliners, as he developed a system whereby the events and characters presented in the narrative rather than any assumed narrative persona determine the diction and syntax of the narrative prose. This elimination of the narrator as a personality does away with the need for consistent narrative idiom and paves the way for the experiments of Joyce's later fiction.

Lost improvements to the Maunsel text: None.

"After the Race"

Date of composition: The third in order of composition, this story was first published in The Irish Homestead on 17 December 1904.

Texts available: 1) Published Irish Homestead text; 2) late stage of the Maunsel printing; 3) Richards page proofs; 4) First Edition.

The development of the text: Though this story saw fewer revisions than either of the others which appeared first in The Irish Homestead, Joyce did make quite a few changes in it for the Maunsel printing. The most heavily revised passage occurs midway through the story, in the paragraph describing Segouin's dinner. For comparison, here is the passage, first as it was in the magazine, then as it became in the late stage of the Maunsel text.

Joyce's intention here in electrifying the lighting was probably to add to the impression of meretriciousness which is one of this story's chief characteristics. In a similar revision he altered "That night the city wore the air of a capital" to "That night the city wore the mask of a capital." An air does not imply the deliberate desire to deceive which a mask does. And, finally, he pointed up Jimmy Doyle's desire to be deceived in the closing paragraph, by adding the material which follows the comma: "He knew that he would regret in the morning but at present he was glad of the rest, glad of the dark stupor that would cover up his folly." Joyce also made a number of other minor changes, improving a word here and there. Some of these did not find their way into the Richards text and are noted below.

Lost improvements to the Maunsel text:

"Two Gallants"

Date of composition: The thirteenth in order of composition, this story was finished and sent to Grant Richards to be added to the original twelve on 23 February 1906.

Texts available: 1) Fragment of setting for Richards, 1906; 2) Late stage of Maunsel page proofs; 3) Richards page proofs; 4) First Edition.

The development of the text: Being a late story, this one probably was not revised so extensively as some of the early ones. We have two non-consecutive pages from the typesetting made for Grant Richards on 17 April 1906 which differ in no significant way from the later text.

Lost improvements to the Maunsel text:

"The Boarding House"

Date of composition: The fifth in order of composition, this story is dated 1 July 1905 on the only surviving manuscript. Joyce mailed a copy to his brother Stanislaus on 12 July.

Texts available: 1) 1905 MS; 2) Late stage of Maunsel page proofs; 3) Richards page proofs; 4) First Edition.

The development of the text: This story was extensively rewritten between the 1905 MS, which was signed "Stephen Daedalus" and apparently intended for the Irish Homestead, and the Dublin printing of 1910. Several aspects of the rewriting warrant commentary. In one respect the rewriting parallels that of "Eveline" discussed above. In eight significant substantive changes the intent is obviously to make the language more colloquial, more appropriate to the events being narrated than to the more lofty tone of the narrative persona, "Stephen Daedalus." Thus, "obliged to enlist himself" becomes "had to become"; "attacked his wife" becomes "went for his wife"; "started a boarding house" becomes "set up a boarding house"; "an amateur boxer" becomes "handy with the mits"; "she had been specific in her enquiries and Polly had been decided in her answers" becomes "she had been frank in her questions and Polly had been frank in her answers"; plain "Lyons" becomes "Bantam Lyons"; "the loss of his job" becomes "the loss of his sit" (colloquial for situation); "had a bit of money put by" becomes "had a bit of stuff put by." In other revisions, Joyce is busy at the usual phrase-sharpening, and in one case he is at some pains to make his irony less heavy-handed. The last sentence, originally reading "She remembered now what she had been waiting for: this was it" becomes "Then she remembered what she had been waiting for".

But the major revision to the early version of this story consists of an insertion some ten lines long. The nature of the insertion throws light on an interesting aspect of Joyce's technique. Joyce is often praised or blamed (depending on the critic's predilections) for the ambivalence or ambiguity of his fiction. When the artist refuses to provide any authoritative commentary, the critics tell us, we are free to believe whatever we want, and to seek for authorial intention is to commit one of the graver critical fallacies. But to this observer it seems that Joyce gives us our heads expecting us to use them. The insertion in question illustrates how we are to proceed. In the original version there was room for some quibbling about the extent to which each of

the two principals was seducer or seducee. The added ten lines provide no commentary, but they give us the bit of evidence we need to resolve the problem with considerable certainty.

Lost improvements to the Maunsel text:

A manuscript reading which should probably be restored: "There must be reparation made in such cases" is the MS reading for a line which all the printed texts have as ". . . in such case" (VC 64, JC 70). The error presumably crept in during the Irish printing and was never picked up. If Joyce had wanted the singular he would probably have used the standard "in such a case", but there is no evidence that he wanted any change from the MS.

"A Little Cloud"

Date of composition: The fourteenth in order of composition, this story was written in the first part of 1906 and mailed to Grant Richards in July of that year.

Texts available: 1) late stage of Maunsel page proofs; 2) Richards page proofs; 3) First Edition.

The development of the text: Like "Two Gallants", which was also written after the first dozen stories had been submitted to Richards, this one probably required almost no revision for the Maunsel printing. Joyce expressed great pleasure in it in October 1906 while he was still working on revisions for some others.

Lost improvements to the Maunsel text: The one change is so slight as hardly to deserve the word improvement, but Joyce made it, probably feeling the word expressed Mrs. Chandler's attitude more accurately.

"Counterparts"

Date of composition: The sixth in order of composition, this story was written almost simultaneously with "The Boarding House," and finished by 12 July 1905.

Texts available: 1) the 1905 MS, nearly complete; 2) late stage of the Maunsel page proofs; 3) Richards page proofs; 4) First Edition.

Development of the text: Like "The Boarding House" this story was revised considerably before the Maunsel printing. The most interesting revisions are those for the long pub scene. Here Joyce added considerable color and life just by making more specific the expressions used for drinks. Thus "a drink" becomes "a half one"; "a drink" becomes "tailors of malt, hot"; "drink up" becomes "polish off that"; "said they would have theirs hot" becomes "told Tim to make theirs hot"; "hot specials" becomes "whiskeys hot"; and "one more" becomes "one little smahan more." Here, as in other revisions, the narrative indirect discourse is enriched by the phraseology of the speakers, and the effect is one of dramatization as the narrator adopts more of the speech of his characters. There are many other revisions in individual words and phrases, but the most interesting is a long passage almost entirely rewritten. In the manuscript the episode of the girl in the pub reads as follows:

This was one of the passages Grant Richards objected to as likely to cause trouble, and especially he objected to the girl's changing the position of her legs. Joyce leaped to the defense of his text, invoking the "Areopagitica" and insisting that without details such as this, "Dubliners would seem to me like an egg without salt." But when he came to revise his text after losing his argument with Richards he took it upon himself to expunge the offensive legs, discovering that he could do all he wanted to with arms and eyes. The vastly improved (if censored) passage reads as follows in the late Maunsel text: In the revised version we are not told so bluntly what Farrington is thinking, but we are brought much closer to his point of view. Naturalism has given way to impressionism in the passage. Even the girl's accent seems different when seen from Farrington's perspective rather than the narrator's. Where the narrator and reader heard Cockney, redolent of slums and sordidness, Farrington hears London, as exotic to him as Araby to the little boy of the earlier story. The reader can form his own opinion of that blue and yellow color combination, and he can judge the woman by the way she brushes against Farrington, but he is not told that her glances are "bold." One is tempted to come to the conclusion, however reluctantly, that if Joyce had taken some of Richards' advice instead of battling him on principle at every point, he might have saved himself a lot of grief and not done Dubliners any harm.

Lost improvements to the Maunsel text:

"Clay"

Date of composition: The fourth in order of composition, this story seems to have cost Joyce more pains than most. In November of 1904 he began a story, "Christmas Eve," which he abandoned half finished, apparently because the idea for another, "Hallow Eve," had superseded it. (The MS of "Christmas Eve" has been found and the fragment published in The James Joyce Miscellany, Third Series, 1962.) "Hallow Eve" was completed and sent to Stanislaus Joyce for possible publication in The Irish Homestead in January 1905. James Joyce may have done some revision on the story at this point. It was in his mind much and mentioned frequently in his correspondence during the next few months, while the fifth and sixth stories were not written until July. By September the title had been changed to "The Clay." In November of 1906 Joyce was working on the story again, adding the name of the laundry where Maria works. At some later date

he must have reconsidered the title again, dropping the article from it.

Texts available: 1) Late stage of Maunsel page proofs; 2) Richards page proofs; 3) First Edition.

Development of the text: Until some manuscripts are found we will have to be content with the knowledge that Joyce worked on the story carefully, though we do not know the nature or direction of his improvements. The last change he made, as noted below, suggests that he added some of the details which make Maria appropriately witch-like, a cloak being the proper garment for a witch.

Lost improvements to the Maunsel text: On two occasions, Maria's garment is described as a "raincloak" in the late Maunsel text rather than as a "waterproof" which is the reading of the modern editions.

"A Painful Case"

Date of composition: The seventh in order of composition, this story was first written in July of 1905, its original title being "A Painful Incident." A draft was sent to Stanislaus for copying in August of that year. A second manuscript exists, dated 15 August 1905. In August of 1906 Joyce wrote that he had three paragraphs, and then five pages, to add to the story, but the August 1905 manuscript was printer's copy for the Maunsel printing, and there are no lengthy revisions which seem to have been inserted in this text, though this text is longer than the earlier manuscript. The most likely possibilities are these. First, the manuscript dated August 1905 may have been actually written in 1906. This is likely, not so much because Joyce might have misdated it accidentally, though this is possible, but because he often placed the date of original composition on a work which he revised much later. "Tilly" in Pomes Penyeach is backdated in this way, and A Portrait carries both the date of the first draft and that of the last. A second possibility is that the revisions to the story were simply never inserted. This seems to me the least likely alternative.

Texts available: 1) Heavily revised first MS; 2) second MS with few revisions, dated 15 August 1905; 3) portion of the late stage of the Maunsel printing; 4) Richards page proofs; 5) First Edition.

The development of the text: Joyce's revisions have been covered thoroughly by Marvin Magalaner in Time of Apprenticeship (1959), pp. 87-96.

Lost improvements to the Maunsel text: None.

"Ivy Day in the Committee Room"

Date of composition: The eighth in order of composition, this story was completed in all its essentials in a draft dated 29 August

1905, but Joyce continued making minor revisions until the First Edition was printed.

Texts available: 1) MS dated 29 August 1905; 2) a later MS used as copy-text for the Maunsel printing; 3) an intermediate stage of the Maunsel printing; 4) Richards page proofs; 5) First Edition. The intermediate stage of the Maunsel printing consists of pages rather than galleys, but retaining the period in "Mr." and other indications that it is not the "late stage" of proof which is available for most of the other stories in Dubliners. This is the only sample yet discovered of the state of the text which was postulated as a hypothetical "C1" in my study of the text of "The Dead." This is the state of the Maunsel printing from which the Richards edition was set. Such errors as "Parke's" for "Parkes" and "revenge" for "renege" made by the Irish compositor in setting from Joyce's holograph MS are repeated in the Richards page proofs and finally corrected in the First Edition. No doubt if a late stage of the Maunsel text were available we should find the corrected reading there too. This copy of the intermediate state of the Maunsel printing was not itself copy-text for the Richards, for it contains numerous corrections in Joyce's hand which were not made in the Richards page proofs. A virtually uncorrected set of proofs must have been sent to Richards.

The development of the text: The future editor of Dubliners will find himself confronted with an embarrassing complexity of riches in the form of authorial revisions to various states of the text of this story. The second MS, printer's copy for the Maunsel type-setting, contains marginal revisions, some of which were included in the Maunsel proofs and some of which were not, possibly because they were made in the MS after it was used by the compositor. The Maunsel proofs contain holograph corrections, some of which appear in the Richards proofs, most of which do not. Some of the overlooked corrections from both of these sources are inserted in the First Edition and some are not. In one case Joyce, who fought so bitterly for every word of his text against Grant Richards, provides a marginal alternate reading for a disputed passage, apparently leaving the decision to an editor or compositor, whoever cared to make it. After the intermediate stage of the Maunsel edition had been set, Joyce went through it, eliminating all the attempts he had made to reproduce the sounds of Irish speech (e.g., "Owl'" gives way to "old"). He apparently decided that it was inartistic to rely on such nineteenth-century devices, and that he would depend on grammar and syntax alone to make his speech Irish. Of course, the Richards compositor did not have the benefit of these

corrections, so the same culling out had to be performed between Richards page proofs and the First Edition. Joyce seems to have started early to indulge in the practice of making heavy revisions in proof much to the distress of all his publishers. Here are some of the major changes and problems related to the text of "Ivy Day."

a) On his entrance, Father Keon's voice is described in the modern editions as "a discrete, indulgent, velvety voice" (VC 126, JC 141). In the first MS the sentence continued, "which is not often found except with the confessor or the sodomite." In the second MS the clause is crossed out, in the interests of either decency or subtlety.

b) In the second MS Joyce inserted the word "lousy" before the expression "hillsiders and fenians". This should probably be added in the modern texts (VC 125 JC 139). He also substituted "Bantam" for "Lyons" (VC 130 JC 146) in the phrase "Of course Lyons spots the drink first thing" — a change which parallels one in "The Boarding House."

c) For a passage relating to Edward VII Joyce provided an alternate reading in the second MS:

The alternate, it seems, was not needed. This, of course, was one of the passages which Joyce and his publishers disputed about so violently.

d) One sentence was added in the margin of the second MS but never printed. It should probably be added to the modern texts. It is a delightful thing in itself and it adds to the characterization of the ubiquitous Bantam Lyons, which seems to have been one of Joyce's general intentions in his revision of these stories. After ". . . a fit man to lead us" insert "Do you think he was a man I'd like the lady who is now Mrs Lyons to know?" The man in question is Joyce's hero, Parnell (VC 132 JC 148).

Lost improvements to the Maunsel text: Since we do not have the late stage of this text, there are none, but the MS changes discussed above should be noted. Perhaps we should also note here one error which has crept into the American text although the First Edition was correct. The phrase "schoolboy of hell" (VC 123) should be "shoeboy of hell". This is not too important, but one critic has already based a clever interpretive paragraph on the misreading. Caveat lector.

"A Mother"

Date of composition: The tenth in order of composition, this story was finished by late September 1905.

Texts available: 1) MS used as printer's copy for the Maunsel edition; 2) galley proofs, the first stage of the Maunsel edition; 3) Richards page proofs; 4) First Edition.

Development of the text: Joyce made a number of revisions of the mot juste variety between the galleys and the Richards text. This kind of revision is typical of all his corrections to Dubliners and is probably the least significant.

Lost improvements to the Maunsel text: Without the late stage of the Maunsel text no lost revisions can be found.

"Grace"

Date of composition: The twelfth in order of composition, this story was called "the last" by Joyce, as it completed his original plan for twelve stories in four groups of three. He began it in October 1905 and finished it sometime before the twelve stories were sent to Grant Richards in December of that year. But during his stay in Rome the following year he did additional research in the Biblioteca Vittorio Emmanuale for the theological parts of the story.

Texts available: MS used for Maunsel printing, very clean copy, probably not 1905 version; 2) Richards page proofs; 3) First Edition.

The development of the text: Between the MS and the Richards page proofs, and, to a lesser extent, between the proofs and the First Edition Joyce made a number of minor revisions. In two of the larger of these we can see him eliminating the superior prose of the narrator in favor of plainer locutions, more suited to the subject matter: "Nor was she an utter materialist for she also believed (to a certain extent) in the banshee and in the Holy Ghost" becomes "Her faith was bounded by her kitchen but, if she was put to it, she could believe also in the banshee and in the Holy Ghost"; and "a confirmed inebriate" becomes "an incurable drunkard." Other revisions are of the mot juste variety.

Lost improvements to the Maunsel text: None. No late Maunsel available.

"The Dead"

See earlier essay, SB XV (1962), 191-205.

Notes