University of Virginia Library

A Textual History of Wallace Stevens' Three Travelers Watch a Sunrise
by
Richard Alan Schwartz

In July, 1916, Poetry Magazine published Wallace Stevens' short play, Three Travelers Watch a Sunrise. It was later reprinted in a collection of plays edited by Frank Shay and Pierre Loving entitled Fifty Contemporary One-Act Plays (Cincinnati: Steward and Kidd Co., 1920), and in Opus Posthumous by Wallace Stevens, edited by Samuel F. Morse (New York: Alfred A. Knopf, 1957). These seem to be the only occasions on which Three Travelers has appeared in print. Nonetheless, the play's textual history, especially its prepublication history, reveals useful information about Stevens' method of composition and about his reliance upon the literary judgment of Poetry's editor, Harriet Monroe.

Two typescripts and the author's corrected galley proofs, as well as some of Stevens' correspondence with Monroe, are housed in the University of Chicago's Joseph Regenstein Library, Poetry Magazine 1912-36 Collection (Box 39, Folders 10 and 14).[1] One of the typescripts is free of proof-reading marks and other internal corrections. The other is heavily marked and corrected and was obviously used as the printer's copy. The typescript which is free of corrections has pica type on pages 1-14 and elite type on pages 15-18. Pages 1-14 of the printer's copy are in elite type, and pages 15-19 are in pica. All of the pages typed on an elite typewriter are approximately a quarter of an inch longer than those typed in pica. We may then conclude that there were originally two distinct typescripts—one which was typed on longer paper with an elite typewriter, and one typed on shorter paper with a pica typewriter. The elite page 15 is a continuation of the elite page 14, and the pica page 14 is continued on pica page 15. Furthermore, in the printer's copy, deletions had to be made on the bottom of page 14 (elite) and the top of page 15 (pica) to allow the passages to flow together meaningfully. Thus, it is apparent that pages 1-14 of the elite typescript and pages 15-19 of the pica typescript combined to form the printer's copy.

The logical assumption would be that the pica typescript was the original and that the elite typescript was a revised version, most of which was accepted as printer's copy. However, external evidence shows that this is not the case. In his letter of May 29, 1916 to Monroe, Stevens discusses a new, revised version "which embodies at least some of the suggestions made by

you and Mr. Michelson. I have eliminated the hanging body. On that point, it seems to me that the creaking of the limb of the tree was more likely than the body itself, which for the most part would have been concealed, to create ridicule." The hanging body and creaking limb appear in the initial stage directions in the elite typescript, but not in the pica typescript. In that letter he also mentions, "You will note that the negroes now have no speeches." On page 15 of the elite typescript the Second Negro twice asks, "Is that you, Anna?" This is the only occasion in either typescript on which either of the Negroes has any lines. In the pica typescript this speech is given to the Second Chinese. Furthermore, Stevens goes on to instruct Monroe, concerning the revised version, "If you prefer this form, use it. Or if you desire still further changes in it, let me know. Or, last of all, if you have doubts and think the original form the better of the two, again let me know. . . . personally, I like the new form." In a note at the bottom of his corrected galleys Stevens wrote that the play, "seems all right. My remark that I liked the second version was a fib designed to give you your choice. But I did rather like the green bandanna. The fuss made about the hanging body was proof of something or other. I have made note of Mr. Michelson's protest." [The bandanna appears only in the pica typescript, and where it appears in the printer's copy, Monroe eliminated it.] This note indicates that Monroe selected the original typescript on which to base the printer's copy, and the printer's copy is predominantly the elite typescript. Finally, there are notes, in Monroe's handwriting, in the margins of the elite typescript referring to particular lines which say "omitted in later version" (p. 7), "changed for the better" (pp. 14 and 15), and "This out in later version" (p. 17). With this evidence we must conclude that the elite typescript is the original and the pica version is the revised, but largely unused draft. Thus the elite typescript represents TS1 and the pica typescript TS2.

In terms of Three Travelers' compositional history, then, TS2, except for its final five pages, is largely irrelevant. Thus, except for those instances where elements from TS2 appear in the conflated typescript, this paper will not be concerned with that version. We should note that Monroe was responsible for all the additions and deletions, and for the selective process concerning which materials from TS2 would or would not be incorporated into the conflated version, as well as for the very choice of which typescript would serve as the basis for the printer's copy. Stevens consented to most of these changes; those he disapproved, he changed in the galley proofs. So this Monroe-corrected-composite of TS1 and TS2 can be said to represent Stevens' final intention, except as amended in the galleys. Therefore, I would recommend that composite of pages 1-14 of TS1 and pages 15-19 of TS2, as corrected by Monroe, as the copytext for a critical edition of the play. For the sake of clarity, I shall refer to that composite as the conflated typescript (CT); by "printer's copy" I mean CT as corrected by Monroe.

Besides using pages 15-19 of TS2 as the basis for the printer's copy, Monroe also incorporated particular words and lines, or pieces of information,

which appear only in TS2. Likewise, she incorporated into CT particular materials from the rejected pages of TS1. Most of Monroe's changes seem designed to render the play more dramatic. This procedure may have been inspired by Stevens' May 29 letter stating, "I desire to have the play a play and not merely a poem, if possible." Most of the alterations involve stage directions, encouraging gestures or limited movements. A few are directions specifying the emotions a speaker is to exhibit.

There is one significant change taken from TS2. On page 2 of CT the three Chinese are first described. In TS1 the descriptions are very brief and limited to the Chinese's physical appearances. In CT, though, Monroe has added other details from TS2 that reveal something about their personalities as well. The First Chinese is not only short and fat, he is also "quizzical". The second is additionally described as "a man of sense and sympathy", and the third is "intent" and "detached". This additional characterization is useful because it helps the reader and the actor further grasp the exact nature of each of these individuals. Had Stevens intended the play to be "about" these people it would seem that these statements should be unnecessary, that a good playwright would show us that the characters posses these attributes, and not have to tell us. Stevens accomplishes this to a certain extent; but without these explicit adjectives it is difficult to be too precise about the Chinese, especially the First Chinese.

Three Travelers is not essentially a play about characters though. It is designed to demonstrate a concept, and the characters are agents toward that end. In his May 29 letter Stevens says, "The point of the play, by the way is . . . in the last sentence of the final speech. . . . The play is simply intended to demonstrate that just as objects in nature affect us . . . so, on the other hand, we affect objects in nature, by projecting our moods, emotions, etc." The sooner Stevens can have us understand what the characters' personalities are like, the sooner he can employ them toward his artistic goal. Thus the decision simply to state, rather than show, what these characters' personalities are is not an artistic flaw; instead, it accentuates the artistic intention. Furthermore, in incorporating the additional information into CT, Monroe is more efficient than Stevens. In TS2, after describing the First Chinese as short, fat, and middle aged, Stevens introduces a new sentence, "He has a quizzical but not too broad manner." Monroe converts that into a single adjective, "quizzical", and inserts it into the previous sentence. Thus she makes the text a bit more concise and smoothly flowing, without deleting important information or otherwise detracting from the aesthetic effort. This is typical of her treatment of many of the insertions from TS2 into CT.

The question naturally arises why Monroe selected pages 15-19 of TS2 as the basis of her preferred text, instead of using the entire TS1. Again, she seems to have been acting with Stevens' intentions very much in mind. In his May 29 letter Stevens discusses the revised typescript: "The most important change has been in the climax. This is still lacking in shading away. To some extent, this is intentional, because I do not desire to become involved

in the story or characters of the man and the girl." Apparently, then, Stevens was concerned about the play's conclusion and these specific comments imply that he believed the ending in TS2 was, indeed, superior. Monroe's selection of those later pages of TS2 for CT reflect Stevens' feelings on this point.

A significant difference between the endings of TS1 and TS2 might also have encouraged Monroe to choose the latter's. In CT, upon discovering the dead man but before finding Anna, the Second Chinese says, "Death, the hermit, needs no candle in his hermitage" and then snuffs out the candle while the First Chinese extinguishes the lanterns. This episode does not occur at all in TS1. The addition seems to reveal the Second Chinese's sensitivity, and the First Chinese's acquiesence to this notion shows that even he, the pragmatist, can respond appropriately to, and be affected by, this moment of humanity's invasion. Also, by having the speaker call death a hermit, Stevens highlights the parallel between the earlier conversations and the finale. The audience is virtually compelled by this personification to associate the ending with the earlier portions and to feel that they are bound up with each other.

It seems, then, that Monroe's choice of text and her decisions to incorporate certain passages from the rejected pages of typescript are designed to advance the artistic ends that Stevens had in mind for his play. The cumulative effect of these decisions is to make the play slightly more dynamic, hence more appropriate to actual dramatic presentation; and to organize the parts of the work more efficiently for accomplishing Stevens' stated artistic goals. Monroe, however, not only edited and made decisions of artistic selection from the two typescripts, she also occasionally made changes and additions which were entirely of her own creation—not taken from earlier or later typescripts as Stevens had written them. These become particularly interesting because they represent passages which, although approved by the author, are nonetheless not of his own invention.

Most of Monroe's changes do not affect the play's meaning or emotional impact. Instead, they are designed to make Stevens' stage directions more concise, precise, and explicit. Often the changes involve taking two of Stevens' sentences and fusing them into one, complex sentence. For example, in TS1, 3.23-24, Stevens writes, "He finds a book in the pocket of his costume. He reads." Monroe changes this to, "He finds a book in the pocket of his costume and reads from it." The effect of such changes is to make the prose read more smoothly. As a matter of editorial style, Monroe consistently italicized stage directions, changed Stevens' parentheses into brackets, and spelled out the numbers identifying the speakers. Thus Stevens' "(1st Chinese)" appears as "[First Chinese]" in the printer's copy. Stevens tended to be very fond of commas and he frequently overpunctuated. Monroe deleted much of his extraneous punctuation, and in only five or six cases in the galley proofs did Stevens indicate he wanted the punctuation re-inserted. In a note at the end of the galleys Stevens mentions that he has no copy of the play: obviously, then, he was not correcting the proof against his manuscript. Those cases

where he reinstated his original punctuation must then have seemed especially "right" to him. In lines 5.11, 5.23, and 9.12 of TS1 Stevens had commas after the words: "reality", "moments", and "says" respectively. Monroe deleted them while preparing TS1 as the printer's copy, but Stevens restored the punctuation in the galleys. However, no printed version of the play contains it. So apparently the author's expressed intention was subverted in regard to these accidentals. There are other instances in the galleys where Stevens reverted to his original punctuation and those revisions were honored.

Stevens was sometimes vague in his stage directions and would say "a negro" instead of First or Second Negro. Monroe clarified these situations. On one instance, TS1 14.15-17, Stevens confused the Negro who was the servant to the Chinese with the one who had been searching for the Italian man. Monroe corrected this also. She further deleted some stage directions indicating pauses or extra, superfluous action. The primary effect of these alterations is also to make the text read somewhat smoother.

There are a few instances, however, where Monroe's alterations do substantially affect on the play's overall aesthetic sensibility. Apparently it was at her suggestion and Max Michelson's that Stevens omitted the hanging body in setting the initial stage when he wrote TS2. Instead, the Negro raises his lantern and sees a green handkerchief on the rocks. "The light of his lantern falls on the face of a man. The negro lifts the head slightly and looks at the face. He covers the face again with the handkerchief." Yet Monroe apparently judged this rendition of the scene to be inadequate; so she revised TS1 1.11-13 from "His [the Negro's] lantern light falls on the body of a man hanging to the limb of one of the trees. Only the back of the body is to be seen" to the version in the printer's copy: "Discerning a dark object swaying among the branches, he [the Negro] shrinks back." In both TS1 and TS2 Stevens informs the audience that there is a dead man on stage. This creates dramatic irony—the audience is aware of the body and anticipates its discovery by the characters. In the printer's copy, however, Monroe's change leaves the audience unaware that the man is dead until the characters receive that knowledge—the revelation is thus a shared one; there is no dramatic irony. In a very direct and important way, then, this detail determines the nature of the audience's experience with the play. A viewer's or reader's emotional response is vastly different if he knows beforehand what is to happen. And the discoveries of the body and the girl are the most important dramatic moments of the play, when "the invasion of humanity" reaches the Chinese.

One reason that Stevens originally may have chosen to make the audience aware of the dead man from the beginning is that such dramatic irony would cause the audience immediately, instead of retrospectively, to associate the Chinese's early discussions of the beauty of the court and humanity's invasion of it with their own situation. Because this parallel provides the play's most important metaphor, we can understand why Stevens would wish to

accentuate it. On the other hand, in depriving the audience of the foreknowledge, Stevens and Monroe enhance the play's dramatic power. In the revised version, the audience is in suspense over the outcome of the First Chinese's tale about the sweaty Italian tragedian who ran away with his neighbor's daughter. The periodic creakings of the limb and the troubled expression of the Negroes are, in this case, ominous, but not obvious. Indeed, in the earlier version, they could very possibly have provoked ridicule, as Stevens feared they might. Furthermore, in the printer's copy the audience experiences something of its own invasion of humanity, and this serves to bring home Stevens' point more effectively than dramatic irony allows. The audience members not only intellectually accept the thesis; they feel it too. For these reasons the decision to eliminate the foreknowledge of the hanged man seems to have been a good one. Certainly, whether good or bad, it is crucial in determining the nature of our emotional experience with the play. Stevens, however, did not appear to be aware of the importance of this decision. His remark on the galleys that the "fuss made about the hanging body was proof of something or other" seems to indicate that he did not think the issue was especially serious, and his only comment on it in his May 29 letter deals with the more superficial issue of the creaking limb.

In his May 29 letter Stevens suggests that more shading is necessary for the play's conclusion and he states that, "Possibly further thought might lead to something in the speech commencing, One candle replaces | Another, etc." In TS2 he had already added three more lines, causing that speech to read:

The "you" here is Anna, and the speaker refers to the sunrise replacing the candle. Monroe rejected this addition, crossing out those last three lines in the printer's copy. It is difficult to speculate exactly why she did so, but the effect of her decision is to leave the observation universal, instead of specific. The addition, "That is for you" focuses meaning onto Anna. Its omission allows the passage to serve as a general observation on the way of things in nature. The two lines "Re-coloring, | As they move" are appropriate to the thesis of the play, and it is somewhat surprising that they were deleted.

There is another significant element from TS2 that Monroe did not incorporate into the printer's copy. At the bottom of the last galley sheet Stevens said that his earlier remark that he preferred the second version was "a fib designed to give you your choice. But I did rather like the green bandanna." This is the green handkerchief with which the First Negro covers the dead man's face in TS2. In that typescript the Second Negro removes it when they discover the body, and during the Third Chinese's concluding soliloquy the Second Negro spreads it over the dead man's face again. Stevens

may have liked the bandanna in TS2 because it frames the action: in that version the play begins and ends with a Negro spreading it over the dead man's face. Thus, the play is bounded by the audience's awareness of the presence of death—the ultimate form of humanity's invasion. Monroe probably deleted the bandanna because in CT, which rejects the beginning of TS2, the man is never cut down from the tree; he is still hanging from it when the curtain falls. This situation makes it either impossible or ludicrous for anyone to cover the corpse's face.

After Monroe completed her revisions to produce the printer's copy, she sent the galley proofs to Stevens for corrections. In an amusing note on the first page she said that she had been correcting the galleys between speeches at the Republican convention, "A delicate 'incongruity'!" To which Stevens replied on the last page, "I think the Republicans are wasting their time in holding a convention." She requested that he be as sparing of changes as possible, within reason. But she did make a few suggestions, most of which Stevens accepted. She recommended that in the initial stage setting, in which there is a dark object "swaying among the branches" (1.10), that "swaying" be eliminated as too definite. Stevens agreed, and it was deleted. One of the revisions Monroe had made in CT was a sentence in which a Chinese enters carrying a Chinese lantern (1.18). Stevens wrote a note on the galleys making it clear that the lantern was not a Chinese lantern, and this too was deleted. In another note he objected strongly to the use of the term "exits" (1.12) and changed the reading from "exits through the wood at the left" to "enters the wood to the left." Interestingly though, in all the printed forms this passage reads, "goes out through the woods to the left." Stevens also stated, "For the sake of consistency say 'to the right' 'to the left' all throughout, instead of 'to the right' in one place and 'at the right' in another." In CT the First Chinese tells Anna's story to his fellows stating "Amusingly" that it was not an elopement (10.26). In the galleys Monroe suggested changing this to "Amusedly" and Stevens agreed. A printer's error changed a line describing Anna, her father, and the Italian as "Three beggars . . . begging from one another" (15.4,5) to "Three beggars . . . begging for one another." Stevens remarked on the galleys that "For instead of from is a most fortunate change," and he retained the inadvertent correction [italics mine]. Also, in CT the Second Chinese first sees the girl and says, "Anna." (16.3, 16.7). In the galleys Monroe suggested changing this to "Is that you Anna?" and Stevens concurred. This is actually the same wording that appears in TS1, when the Second Negro discovers her and asks the question. Finally, in a note on the galley sheet Monroe questioned how one can strike an instrument in an insinuating manner (11.21). Stephen replied, "Please do not change. This may be a heathen mystery but I believe it." The insinuating manner was retained.

Variations between the galleys and the way the play appears in its first American publication in Poetry (A1) are not significant. All of these have already been discussed, except for one accidental. There are, however, some

interesting differences between the Poetry version and the way the play appears in Fifty Contemporary One-Act Plays (A2). Prefacing the play in A2 is a note which states that it has been reprinted from Poetry Magazine. However, there are variations in house styles, as well as accidental and substantive changes. There is no reason to believe that Stevens approved of these changes. In CT, as well as in A1, the opening stage directions are in italics but are not bracketed. In A2 they are both italicized and bracketed. In A1 the character speaking is indicated by italics; whereas in A2 this is done with bold print. In CT and in A1 all stage directions appear within brackets, the first letter of the first word is always capitalized, and there is a period which always falls inside of the brackets. A2 sometimes follows this format, but it often, and for no apparent reason, begins the first words with lower case letters and places the periods outside the brackets. These examples account for most of the accidental differences, but there are two substantives. In CT (10.21) and A1 the Second Chinese speaks to the First Chinese about the elopement, "Confidingly." This appears as "Confidently" in A2. More importantly, in the printer's copy (16.18) and A1 the First Chinese tells the Second Negro to bring fresh water. In A2 he tells the Second Chinese to bring it. This is confusing to the reader, who must wonder why one Chinese has commanded another, and why the second one has apparently simply ignored that order.

The third American edition of the play (A3) appears in Opus Posthumous. This is closer to the A1 version, differing only in the matter of house style and one accidental. The A3 consistently uses bold print instead of italics to designate the speaker, and it consistently uses lower case and no punctuation for stage directions that are not complete sentences. The accidental is the omission of a comma after the word "hill" corresponding to 11.1 in CT.

Thus changes at every stage characterize the development of Three Travelers Watch a Sunrise from its original typescript form to its second and third American printings. Some changes were made by the author, some by Monroe with Stevens' approval, and a few by her without his approval. Some resulted from errors in the reprintings of A2 and A3 from A1. The most significant aspect of these changes is the authority Stevens freely gave to Monroe to select which typescript would serve as a basis for a printer's copy, to select which elements from the rejected typescripts she would incorporate into the printer's copy, and to delete passages from the printer's copy. Monroe seems to have employed this authority judiciously and with Stevens' stated intentions specifically in mind. The end result is an example of a profitable, cooperative relationship between an author and his editor.

Notes