University of Virginia Library

The Red Badge of Courage Manuscript: New Evidence for a Critical Edition
by
William L. Howarth

On January 27, 1896, Stephen Crane mailed his original manuscript of The Red Badge of Courage to Willis Hawkins, a longtime friend.[1] Hawkins had the loose sheets sewn and bound, and the manuscript survives today in this condition as part of the Barrett Collection at the University of Virginia's Alderman Library.

Crane used five different kinds of legal-cap paper in this manuscript, which represents the second draft of The Red Badge.[2] The two editors who based their texts on manuscript examination[3] have testified that the novel was written on sheets of identical paper. Winterich gives the following uniform specifications: "176 sheets of blue-ruled paper, 12 x 7 11/16 inches, with a red line at the side of each sheet setting off a one inch margin" (Folio, 21). Stallman, who had access primarily to photostatic copies (Omnibus, vii, xv), repeats that the entire 176 leaves were "written on blue-lined paper having a one-inch red-lined margin at the side of each sheet" (Omnibus, 202). His dimensions differ slightly from Winterich's, but he emphasizes uniformity: "each sheet measures 12 11/16 by 7½ inches." A close examination reveals that there is actually a great deal of variance in the five papers' dimensions, lineation, and margins. Each paper has its distinctive physical characteristics, and there is no feature that is uniformly shared by all five.

They have been labeled alphabetically here because they appear in this general sequence in the manuscript's foliation.

Crane's composition of The Red Badge can be reconstructed by distinguishing the five different papers and noting their appearances in the manuscript. His earliest writing, of course, is in the first draft, where he used only two kinds of paper. He began with paper D, but after twelve pages switched to the cheaper C stock for the pages numbered up to 92. Pages 2-92 of the first draft constitute fragments of chapters I-XIII. Whether or not Crane completed this draft is uncertain, since it would be impossible to guess at what point after page 92 he began his second draft. Stallman mistakenly estimates the length of the first draft "from the fact that one of its pages is numbered 149" (Omnibus, 216). Yet '149', although it appears on the verso of page 163, is merely a false start for 149 of the second draft, not the first: '149' is on paper E, a paper which does not appear in the first draft. There are several instances of false starts in the second draft which Crane inverted for later use; these are not to be confused with first-draft pages.

Crane wrote his second draft of The Red Badge by copying and expanding the shorter first draft. He began writing on paper A and used 51 sheets to transcribe the contents of 44 first-draft pages. Because one sheet, a false start for page 33, was removed and later used for page 139, the paper A "run" consists of only 50 sheets, pp. 1-50. He wrote his first five chapters on these sheets before the supply ran out. Four additional chapters—Six, Seven, and what eventually became Eleven and Twelve—were written on 26 sheets of paper B. But, then in order to conserve paper, Crane canceled the first-draft material on the C and D paper and turned over each sheet to write on its clean back side. Evidently he had not stacked the first-draft sheets carefully as he finished copying, for their old pagination rarely appears in sequence in the second draft. Pages 81-85 of the second draft, for instance, have the following first-draft pages on their versos: 29, 67, 66, 71, 61B. Crane wrote chapters Eight to Ten, Thirteen to Fifteen, and half of

Sixteen on 49 sheets of Paper C; the second half of Sixteen and all of Seventeen were written on 8 sheets of paper D. He obtained a fresh paper supply for the final third of his novel, and wrote chapters Eighteen to Twenty-five on 51 sheets of paper E. Some time after he had completed his second draft, he removed a few sheets from the manuscript and replaced others with substituted new material. These insertions and deletions require some critical attention, since they help to outline the growth of the Red Badge text.

Throughout his revision of the novel, Crane was preoccupied with the problem of developing effective conclusions for each chapter. Sometimes he dropped pages of exposition and did not replace them, preferring instead to close the chapter with an explicit bit of dialogue or action. This procedure accounts for most of the missing second-draft pages: 66, the last page of chapter Seven; 86-89, the end of chapter Ten; 126, the final page of chapter Fourteen; and 165, the last page of chapter Twenty-one. Crane also removed an entire chapter, pp. 98-103, because it detracted from the dramatic pace of his novel. At other times he re-wrote chapter endings on the backs of first-draft sheets (paper C and D) and inserted them in the place of his original endings. Physical evidence indicates that these sheets are later additions to the manuscripts, and thus underscores the significant fact that each appears as the final page of a chapter. Crane re-wrote the endings of chapters Five, Six, Eleven, and Eighteen on paper C, and the ending of chapter Twenty-five on paper A. For most of the chapters these facts are uninformative, since there is no way of knowing to what extent Crane revised the original ending. In chapters Five and Six, however, the revised endings can be compared with their surviving first-draft counterparts. Page 60, the end of chapter Six, corresponds to the text of page 52 in the first draft. The variants between these two versions are critically significant, because they indicate some of Crane's developing sense of theme and method. Besides softening the general's enthusiasm from "eternal damnation" to "blazes", Crane added a significant paragraph in the revised second draft:

This added comparison of the general and the sun links the officer to the awesomely indifferent Deity that broods over the novel's battleground, and the suggestions of his exultant worship introduces a recurring motif that Crane builds around his personification of War as a "blood-red god".

Paper differences in the manuscript testify only partially to Crane's revision of The Red Badge. There is more substantial evidence, of course, in the actual changes he made in the text: the deletions, cancellations, and insertions. An editor must understand the genealogy of these changes, especially if he wishes to print those variants which represent Crane's final intentions. Seven different states of writing appear in the two manuscript drafts: six are in Crane's hand, and the other one is in a later non-authorial hand.[5] In the first draft, there are two authorial states: Crane first composed with a blue ink in medium strokes, and later revised with the same. For purposes of distinction, "1st-I" will stand for the composition, and "1st-II" for the revision. Any changes made within the same state will be indicated by lowercase letters: Ia, Ib, Ic, etc. There are four authorial states in the second draft:

• 2nd-I—original composition (transcription and expansion of 1st-II): a brown-black ink in thin strokes;
• 2nd-II—first revision of 2nd-I: a blue-black ink in broad strokes;
• 2nd-III—last revision of 2nd-I: black pencil in Crane's hand;
• 2nd-IV—pagination changes and large textual deletions: blue pencil or crayon in Crane's hand.

The second draft also contains one non-authorial state:

The growth of Crane's text from its earliest (1st-I) to its latest (2nd-IV) authorial state can be reconstructed by noting the succession of states and the kind of revision that each makes. A similar study of the non-authorial revision can determine how much the manuscript was altered after it left Crane's hands. Each writing state displays its particular concerns in the revision. For instance, 2nd-II shows the earliest changes from realistic to allegorical characters: "Henry Fleming", "Wilson", "Jim Conklin", and "Simpson" become "a youthful private", "a blatant young soldier", "a tall soldier", and "a corporal". II also begins the task of altering much of the young hero's dialect. III helps to refine some of the name changes ("a blatant young soldier"

becomes "a loud soldier"), makes large deletions in some of the novel's "philosophical" chapters, and initiates some inconsistent revision of dialect that survived in the 1895 edition. The dialect changes created a major textual problem that will be discussed below. IV, Crane's blue pencil, appears in only five chapters (Ten through Fourteen). It is used there to cancel extensive passages and renumber both chapter and page numbers. On one occasion, the blue pencil changes a "Fleming" to "the youth", but ordinarily it does not make "literary" revision. V appears infrequently throughout the manuscript. The identity of this hand is extremely doubtful, but it seems to be that of a typist or publisher's reader, because most of the notations are concerned only with typographical matters: capitalization, spelling, or punctuation.

Crane's revision of The Red Badge was methodical, yet not particularly thorough. He transcribed the entire first draft by copying and revising (2nd-I). Then he began changing names of characters in II, but upon reaching chapter Nine, he also decided to normalize the dialect of Fleming and Wilson. From chapters Nine through Twelve, Crane made both the name changes and dialect revision in II. He skipped over chapters Thirteen through Seventeen and continued the revision from Eighteen to Twenty-Five. When he finally began touching things up in III, he corrected those chapters that were neglected in II, but also decided at this time to make some additional major revision. In chapters One through Four, he smoothed out a few name changes and, in addition to Fleming and Wilson, normalized the dialects of other major characters. He canceled long descriptions of Fleming's thoughts at the ends of chapters Seven and Ten and completely removed chapter Twelve. In chapters Thirteen through Seventeen, Crane made name and dialect changes in III because he had not touched these chapters during his earlier revision. Possibly he had not revised them in II because he had decided to re-work them extensively, or was already doing so. By this time he had evidently decided against normalizing the dialects of other characters, for only Fleming's and Wilson's were altered here in III. He canceled another descriptive passage at the end of chapter Fourteen and removed its final page. There are paper changes and editorial markings in this section of the manuscript which indicate that Crane did some re-writing on it before and after revision II. When he later decided to insert the section, he tried to bring it into accord with the rest of the text by making the name and dialect changes in III. His effort was casual rather than scrupulous, for some names and dialect passages were left unchanged, and a few of them reached the 1895 edition in this form.

Crane's soft blue pencil (IV) was later busy in this area marking canceled passages and re-ordering chapter or page numbers. These changes seem to have been made for someone—a printer, perhaps—who was not familiar with the manuscript. IV canceled only those long passages which had already been canceled in III; the blue pencil may have been emphasizing that this material was not to be copied. Wherever he had removed pages, III or IV supplied "bridges" to make the pagination sequence clear: 66/7, 86/90, 98/104, 126/7, and 165/66. Crane dropped his original chapter Twelve, pp. 98-103, and then later moved up chapter numbers to cover the gap. For some reason he stopped after changing Sixteen to Fifteen, and as a result the final chapter of the manuscript is numbered Twenty-Five, whereas only 24 chapters appear in the corrected 1895 edition. V, the final revising hand, appeared in chapters throughout the manuscript to make occasional marginal notations and grammatical or typographical corrections within the text itself. These markings could have been made by a typist, editor, or possibly even a later scholar.

The physical details which have been described in the preceding sections of this paper support the kind of textual evidence that is needed for a critical edition of The Red Badge of Courage. Any editor who assumes the responsibility of examining the novel's manuscript will have to recognize the existence therein of five different papers and seven writing states. These facts are important textual testimony, for without them he could not follow the manuscript's stages of growth. The importance of paper types as evidence of revision has already been discussed. A similar study of the writing states would be of great value to Crane's readers and critics. There are at least three kinds of substantive variants in the manuscript that have definite literary significance: variants between the two drafts, between the three authorial states, and between the authorial and non-authorial states.

The changes Crane made between his first and second draft need to be closely examined. He composed the second draft, as it has been said, by copying and expanding the first draft. Although his changes were usually extensive, he frequently transcribed the early version without revising. Sometimes he transcribed too closely, and his new version became garbled. On page 68 of the second draft, for instance, appears the sentence,

The error here was not due to Crane's ignorance, but to his scribal fallibility. In the first draft he had written the sentence as follows: He had copied the first two words into the second draft when the slightly different version occurred to him. He stopped, canceled "All", changed "his" to "His", wrote the new variant "accumulated", and then continued to copy the sentence without further revision. Crane may have had his grammatical lapses, but he was not always incorrigible.

Other errors appear in the second-draft text that resulted from inaccurate transcription. For instance, on page 55 of the second draft is this puzzling sentence:

The Appleton editors emended the passage as follows: Although this reading is sensible and fairly clear, it does not match Crane's first-draft version, which he evidently miscopied in the second draft: A full-scale collation of the first draft might reveal other readings that were unintentionally dropped from the second draft and need to be restored.

The changes made between early and late authoral states (I, II, III) need to be studied, since the readings Crane canceled are often just as significant as those he inserted. Sometimes the variants demonstrate his stylistic manipulation of even the most ordinary phrases:

At other times, they show his conscious effort to replace explicit statements with suggestion:

Crane's revision did not always do his text justice. On one occasion, the meaning of an image was obscured by his cancellation:

If he had preserved the earlier version, the image of wagons as "sundried boulders" would seem more appropriate than it does now. The alterations Crane made in his climactic scenes are particularly important, because they demonstrate his ability to develop or subdue thematic implications. Critics who have written interpretations of the Jim Conklin death scene will be interested in the following revisions:

Another series of revisions demonstrates Crane's ability to improvise variations on a symbolic refrain. In the unforgettable "chapel" scene, the youth discovers the corpse of a soldier seated upright against a tree. The corpse's eyes repel him, but he is particularly horrified by its ant-covered mouth:

Only a few pages later, the youth stands on a hill above the furious battle-ground and gazes downward:

Crane probably revised the second passage because the youth's similarity to the corpse had become too obvious. Then, too, it was more to his purpose to ally the corpse and the "tattered soldier", since the youth later abandons the tattered soldier to a certain death. Crane's initial description of the tattered soldier, which occurs within a few pages of the chapel scene, deliberately echoes his earlier description of the corpse:

The most significant revision Crane made in his second draft was the change from specific to anonymous characters. In state I there are four characters who are identified by name: Henry Fleming, Jim Conklin, Wilson, and Simpson. These were all changed in states II and III to "the youth", "the tall soldier", "the loud soldier", and "a corporal". The proper names were allowed to remain only when they were spoken by other characters. In state I everyone calls the hero either "Flem" or "Flemin'". Crane changed these forms to "Henry" and "Fleming" only once in the manuscript, yet the change appeared consistently throughout the 1895 edition. There are many unchanged

names in the manuscript that were "silently" corrected when published, but Winterich's assertion that Crane made name changes only on a few pages and left the rest up to a typist (Folio, 23) is inaccurate—and unfair to Crane, who was a reasonably industrious reviser. In the entire manuscript there are actually only a few pages that bear no sign at all of name revision. As the rest of the manuscript attests, Crane's hand was directly responsible for at least 80 per cent of the several hundred name changes that appeared in the first edition.

Some of the changes he made raise interesting critical questions. Why, for instance, was he so careful to make his hero the only "youth" in the book? In state I of the second draft, "Fleming" was "a youthful private" who listened to the debates of Jim Conklin and "young Wilson", and marveled at the profane powers of "the youthful lieutenant" of his company. In state II "Fleming" became "the youth" and Wilson temporarily became "the blatant young soldier" or "the loud young soldier". Crane had decided to isolate his young hero by the time revision in II reached chapter Five, for he changed "the youthful lieutenant" there to "the lieutenant". He skipped over that crucial block of chapters, Thirteen through Eighteen, and continued revising in II. He canceled "youthful"—the lieutenant's familiar epithet—throughout chapter Nineteen, but was less consistent in the following two chapters. In Twenty he canceled "youthful" twice but neglected to change it three other times, and in Twenty-one he did not alter its single appearance. Somehow these four unchanged "youthful lieutenants" reached the 1895 text, despite their obvious inconsistency with Crane's recorded intentions. In the remaining chapters Crane was careful to remove all suggestions of youth from the lieutenant. When he began final revision in III he went back to the early chapters (One through Four) and made similar changes on Wilson, who went from "the blatant young soldier" and "the loud young soldier" to "the blatant soldier" and "the loud soldier". Moving then to the unrevised chapters, Crane changed a few more of the "loud young soldier" epithets. Although he canceled several "youngs" in chapter Thirteen, he left a few in the text, perhaps in preparation for the ironic contrast that he was developing between Fleming and Wilson. A single revision in chapter Fourteen may explain his deliberate effort to make Fleming the book's only "youth":

On the battle-field that is the world of The Red Badge, a man's character is measured by his ability to profit from experience. Wilson moves toward a maturity that Fleming will never grasp, and Crane chose to indicate his hero's pathetic inadequacy by labeling him the only "youth" in a company of men. After Fleming's recognition of Wilson's accomplishment, Crane bestowed upon the latter a distinctive new title: "the youth's friend" or "comrade"—never again was he to be known as "loud" or "young". In chapters Fifteen through Eighteen, Crane made the isolation of Fleming complete by canceling "young" or "youthful" in III wherever they were used to describe the lieutenant.

The most puzzling revision Crane made in his second draft was an added change of dialect in the early chapters. When he first wrote The Red Badge, Crane tried to imitate in his characters' speech the contemporary rural dialects he had heard in New Jersey, New York, and Pennsylvania.[6] In state I of the second draft all of his characters, including Fleming and Wilson, spoke this curiously-spelled English. In the early stage of revision II, Crane maintained the uniformity, but in chapter Nine he decided to normalize Fleming's dialect. He probably did so because he wanted to emphasize, either for purposes of irony or verisimilitude, Fleming's educational background. Interestingly enough, he had decided only four chapters earlier to make Fleming the book's only "youth". Throughout the rest of state II, Crane seems to have changed only the dialect of Fleming and Wilson, possibly because he hoped to suggest contrasts within the characters' external similarity. When he returned to the early chapters with a black pencil (III), he brought the dialect of Fleming and Wilson into conformity with the normalized standard of the other chapters, but for some reason he also decided to revise the dialect of Fleming's mother, Conklin, and the lieutenant. This revision was quite inconsistent with the preceding state (II), but Crane may have envisioned making such changes throughout the entire manuscript. He did not proceed beyond chapter Three, however, for in chapter Nine Conklin's dialect was left untouched—despite the fact that his symbolic role was at its height there. While the lieutenant's dialect was altered by III in chapter Three, it was left unchanged in chapter Seventeen (where Crane revised only in III) and for all of his later appearances. The dialect changes Crane made for the additional characters were far from consistent, as the following passage from the 1895 text indicates: 'You watch out Henry, an' take good care of yerself in this here fighting business. You watch out, an' take good care of yerself. Don't go a-think-in' you can lick the hull rebel army at the start, because yeh can't. Yer

jest one little feller amongst a hull lot of others, and yeh've got to keep quiet an' do what they tell yeh. I know how you are, Henry.' (1895, p. 8)

The later changes made by Crane and another hand (IV and V) in the manuscript also require critical analysis, because they are related to some of the omissions and corrections that appeared in the 1895 text. As a whole, the later blue-pencil cancellations (IV) seem to be consistent with other changes that Crane made in III. In state I of the second draft, he had written several interior monologues for Fleming which he later canceled in III. He probably made these cuts because he realized that the novel's themes could be presented more dramatically in action rather than exposition. He must have realized also that the high bombast and "philosophical" pretension of the monologues were inappropriate for his young hero. The cancellations made Henry seem more probable because his thoughts became compatible with his age and education. Most of these passages appeared at the ends of chapters. If they covered more than one page, Crane would cancel the portion that appeared on the first page and simply remove the second. He did this at the end of chapters Seven, Ten, and Fourteen. Chapter Twelve was so impossibly grandiose that he decided to drop it completely.[7] Of the three chapters that dwell on Fleming's "dark night of the soul" (Ten through Twelve), Crane left only Eleven unrevised to carry the burden of the soldier's lofty introspection. Because the blue pencil canceled only those passages that Crane had already canceled in III, there seems to be no contradiction between the two states. The final manuscript state (V) deals only with a few of the text's accidentals; therefore its relation to critical problems is of lesser importance. The identity of this non-authorial hand is extremely doubtful; it may be that of an Appleton editor, a typist, or some other contemporary of Crane's. The markings consist of penciled marginal question marks, grammatical corrections, and a few insertions that Crane later re-traced in ink. The latter markings may indicate that someone read parts of the manuscript before Crane had completed its revision.

The question of a critical edition of The Red Badge of Courage

hangs upon its editors' choice of a copy-text. Every text since 1895, including the Winterich and Stallman "manuscript editions", has been based on the Appleton first edition as a copy-text.[8] Winterich and Stallman add to their editions a few manuscript passages that are missing from the copy-text, but they do not attempt a full-scale collation of the manuscript. Winterich briefly notes the existence of a first draft, but he prints none of its significant variants. Stallman reproduces some canceled passages from the first and second draft in footnotes, and he corrects a few—very few—of his copy-text's errors by inserting readings from the second draft. In instances of conflict between the second draft and the 1895, and even when both versions are obviously incorrect, the two editors usually prefer to reproduce their copy-text. They follow this practice because they assume that the 1895 represents Crane's final intentions.

Any argument for a new copy-text of The Red Badge needs to be prefaced by a brief history of the documents that reportedly stand between the second draft and first edition: a typescript, a condensed newspaper version, a printer's typesetting copy, and a set of page proofs. The testimony in this matter is scanty and usually contradictory. The major sources are in reminiscences by Hamlin Garland and Irving Bacheller, the Beer and Berryman biographies, Stallman's somewhat disjointed survey in his Omnibus (201-217), and the volume of Crane's collected letters.[9]

In the early months of 1894, Irving Bacheller, later of the Bacheller-Johnson newspaper syndicate, read the second draft and advised Crane to have a typescript made (Bacheller, p. 276). Hamlin Garland reports that Crane visited him with one-half of a typescript. When Garland

asked about the other half, Crane said that it was in pawn with a typist for $15. Garland lent the money, and Crane returned shortly with the typed second half (Garland, p. 501). He then took the first half, which Garland had finished reading, to the editors of McClure's Magazine, who allegedly accepted the book—for serial publication—on the condition that the later chapters met with their approval. Crane's finances must have been abysmally low at this time, for he was borrowing from any available source—without any particular scruples about obtaining money under false pretenses. He wrote to a friend named "Dicon" (John Henry Dick) and asked him to Shortly after April 22, Crane borrowed another $15 from Garland (Letters, p. 36), presumably to repay Dicon; but according to one of Dicon's associates, the trusting friend was never repaid (Letters, p. 30n). Years later, writing from England to his brother William, Crane recalled:

On February 24, 1894, Crane wrote to a friend that he had sold the book, that "the syndicate people" thought "several papers could use it", and that his New York friends thought "that some publisher ought to bring it out when [i.e., after] it has been shown as a serial" (Letters, p. 41). Beer and Stallman both interpret "the syndicate people" to be Bacheller-Johnson, but that syndicate (as Stallman notes) was not formed until late in 1894 (Omnibus, p. 604). Although "it was in the planning stage much earlier", it seems hardly probable that Bacheller would buy a novel for serialization nine months prior to the actual formation of a syndicate. Furthermore, Crane does not indicate in his letter that he had sold to the syndicate, but that the syndicate had thought "several papers could use it". It would seem, then, that the typescript had been "sold" to McClure's, and that Crane had received some advice from Bacheller on possible newspaper publication. Stallman silently shifts to this position in Letters because of evidence that McClure's held the typescript for over six months (Letters, p. 30). Whether or not McClure's ever actually paid for the novel is uncertain. Berryman conjectures, plausibly enough, that McClure's was enthusiastic, but that Crane's talk of a "sale" was merely a flight of optimism. No one had really bought The Red Badge, and

no one would for several months (Berryman, p. 80). McClure's never published the book, and probably even lost that oft-pawned typescript, for it was not returned to Crane.

After nine months of waiting in vain for McClure's to publish his book, a disgusted Crane took the second-draft manuscript to the newly-formed Bacheller-Johnson syndicate. In a letter written to Crane's wife, Bacheller says that he was given "a bulky manuscript" (Letters, p. 298), and in a later recollection he says that it was "a rather worn and dirty manuscript" (Letters, p. 322). Crane was paid $90 for the novel. "They use it in January in a shortened form", he wrote to Garland on November 15, 1894 (Letters, p. 41). It appeared sooner than he had expected, however: first in the Philadelphia Press, December 3-8; then in the New York Press on December 9. According to Stallman, this shortened version (abridged from 55,000 to 18,000 words) of the second draft appeared in "several other newspapers which have not been identified" (Omnibus, p. 604; Letters, p. 41). Berryman estimates that this version appeared "in 200 small city dailies, 250 weekly papers, and as patent insides in some 300-350 more" (Berryman, p. 94n.).

On December 18, Crane mailed the clippings of the newspaper version to Ripley Hitchcock, an editor at Appleton and Company (Letters, p. 46).[10] In January of 1895 Bacheller sent Crane on a Western reporting trip. Before Crane left he gave the second-draft manuscript to Hitchcock. On the 30th, he wrote the editor from St. Louis, asking for an opinion on the novel, and citing some of the favorable reviews that the newspaper version had received (Letters, p. 49). Hitchcock decided to buy the book, and on February 12 Crane wrote from Lincoln, Nebraska, to accept his terms. He was going on to New Orleans shortly, and suggested that the manuscript "could be corrected there by me in short order" (Letters, p. 51). Hitchcock sent the manuscript by express to New Orleans on February 25 (Letters, p. 52). Crane revised quickly but thoroughly; on March 8 he wrote Hitchcock from Galveston, Texas, to say that he had sent the manuscript back before leaving New Orleans (Letters, p. 53).

Stallman once speculated that the manuscript Crane sent to Hitchcock was "probably the typescript copy or page proof, rather than the actual hand-written manuscript" (Omnibus, p. 645n.). Yet the fact that Crane took the manuscript to Bacheller in November, and submitted newspaper clippings to Hitchcock in December, is good evidence for assuming that the typescript no longer existed and had probably been lost at McClure's. When Hitchcock expressed some interest in the novel after reading the newspaper version, Crane got the

manuscript back from Bacheller and gave it to the Appleton editor just before he left New York in January. As for whether page proof or the manuscript was sent to New Orleans, Hitchcock himself says that publication of the book was held off until Crane had returned to New York to correct the proofs (Hitchcock, p. vi). It seems hardly probable that Hitchcock would have had proofs made before major authorial revision, for this procedure would have been expensive and impractical. In all probability, he sent the second-draft manuscript to New Orleans, not proofs. In a later account, Stallman abandons his conjectural typescript and adopts this position (Letters, pp. 45, 51, 53).

Hitchcock says that the 1895 edition was printed after corrected proof sheets, but he does not explain what document was used by the printer as a copy for the typesetting. Winterich and Stallman both assume that the typescript was used (Folio, p. 23; Omnibus, p. 216), but the typescript seems to have disappeared at the McClure's office, since Crane never submitted it to any other publishers. Winterich rejects any proposal that the manuscript was used, on the grounds that:

Yet this objection has little factual basis, since these markings do appear—infrequently—on the manuscript's pages. A printer would scarcely have taken time to stop and record all of his changes on the copy. Furthermore, the 1895 edition reproduced a number of literal errors from the manuscript.

Crane returned from the West in late May. After a few weeks rest at Port Jervis, he went to New York and began correcting the proofs that had been set from his revised manuscript. Hitchcock's testimony is the only evidence that Crane saw the proofs, but it is obvious that he was in New York through June and July from the return address on his letters, "The Lantern Club", 165 West 23rd Street. If Crane did read proof, he was amazingly docile—or perhaps just hasty—about the obligation. All of the non-authorial revision that was added to the manuscript (presumably after state IV and before the proofs) and in the typesetting (generally in the form of accidental variants) survived in the 1895 text, despite its inconsistency with Crane's intentions and habits.[11] By late July the job had been completed, and in August Crane

left New York for a favorite camping spot in Pike County, Pennsylvania. Hitchcock sent the title-page proof to him there, and Crane sent back his approval on August 26 (Letters, p. 62). A copyright application was filed on September 27, 1895, and a deposit made in the Library of Congress on the following day. Publication of The Red Badge of Courage was announced in The Publisher's Weekly on October 5, 1895.[12]

Without the proof sheets, which carried Crane's last revisions, Winterich believes that ". . . it would be as presumptuous as futile to attempt anything like a variorum edition of the text" (Folio, p. 23). Yet a critical text is a definite possibility, since a thorough collation of both drafts and the 1895 edition would provide a complete history of the novel's revision. This evidence might, in turn, enable an editor to distinguish between authorial and non-authorial variants in the 1895. The first edition contains variants that are, on the basis of the second draft as copy-text, obviously non-authorial. There are perhaps hundreds of incorrect 1895 substantives that can be emended by merely substituting the manuscript reading. A few of the more interesting variants are given below:

These 1895 variants may have come either from printer's errors that Crane did not catch in proof, or from changes made in proof after Crane's corrections. There are also substantive omissions in the 1895 text that seem to be authorial, since they indicate consistent "literary" revision. These silent omissions from the manuscript probably represent Crane's final revision on the lost proof sheets, for they eliminate nothing from the text that needs to be preserved. Winterich and Stallman have both felt that this omitted material should be bracketed and placed in the text; as a result they include dialogue which is clearly irrelevant or redundant, and expository passages which are either similar to or refer directly to the interior monologues that Crane cut out of the manuscript.[13] If the manuscript in its final authorial state (2nd-III) were taken as a copy-text, the editor could emend all of the 1895 corruptions without denying the authority of the 1895 omissions.