University of Virginia Library

As the recovery of women's neglected writing continues to receive the attention and energy of feminist scholars, it seems inevitable that the best of this material become subject to the same textual scrutiny and high editorial standards accorded the works of traditional canonical figures. Naturally, since forms of scrutiny and conceptions of standards vary, women's writing will also become fodder for purely textual disagreements. Charlotte Perkins Gilman's short story "The Yellow Wall-Paper" already bears out this speculation. In the 1993 annual American Literary Scholarship, Lawrence I. Berkove lamented that a new casebook edition,

A full defense of his claim that the Radcliffe holograph is "more accurate" and would prove the "best possible" copy-text for a scholarly edition was beyond the scope of Berkove's survey. Truly, with few exceptions,[1] the manuscript has received remarkably little attention by editors and scholars. The dozens of editions that critics, teachers, and students have been using for decades are all based on the first published version (New England Magazine, January 1892) or versions derived from it. The purposes of the present essay are to introduce new manuscript information essential to the production of any critical edition of "The Yellow Wall-Paper," to argue the wholly nonauthoritative character of the magazine printing (hereafter designated as NEM), and to propose that Gilman's undervalued, handwritten document (hereafter designated as MS) must supply the copy-text for a critical edition that holds the textual intentions of the author paramount.

More than a century of "The Yellow Wall-Paper"'s textual history can be interpreted as either a textbook case of corruption and disintegration, or as a case of social construction, depending on one's theoretical perspective. A

critical edition edited by Julie Bates Dock and published by Penn State University Press is due to appear in April 1998, and, judging from Dock's praise of Jerome McGann and her assessment of documentary priorities in the January 1996 issue of PMLA, the edition will accept a collaborative model of authorship and take NEM as its copy-text.[2] I will demonstrate that such an edition, whatever its merits, will not adequately provide a text of the story that feminists would be expected to require: one preserving Gilman's own writing prior to its appropriation by publisher's readers, editors, and compositors, most of whom were probably males.

The story appeared in at least eight editions during Gilman's lifetime, and another dozen or so before publication by the Feminist Press in 1973 cemented its position as a seminal work of the women's movement in America.[3] Working backward, collation of all later appearances during Gilman's

lifetime shows that each is derived either from the 1899 Small, Maynard book edition (hereafter designated as SM, itself derived from the magazine) or NEM. I have found no diary entry or correspondence to support the possibility of Gilman's intervening (by revision or correction) in these later editions. Therefore, no appearance postdating 1899 has been judged to have any authority.[4]

Documentary evidence concerning circumstances of production of SM is scarce. Gilman was already seeing her great brainchild, Women and Economics, through the press with Small, Maynard & Co. during 1898, and frequently mentions reading proof for that book in her diaries.[5] From late 1898 through Spring 1899, Gilman was on a lecture tour that included stops in Illinois, Ohio, Missouri, Tennessee, Alabama, and North Carolina. From May 4, 1899, through September 1899 she was on a lecture tour in England, and SM (with variant title The Yellow Wall Paper) was almost certainly out by the time of her return to America. No mention of the book itself, let alone reading proof for it, appears in the diaries or correspondence of this time period. Proof was read almost exclusively at the Boston offices of Small, Maynard & Co. for her other books, but Gilman was unavailable while SM would have been at press. Finally, while the story's (and Gilman's) modest popularity in the final years of the century might have warranted a gamble on the thin volume (sold at fifty cents per copy, of which Gilman received ten percent), this venture was not nearly as important to Gilman or her publisher as her other current books. According to her diary of September 26, 1899, Small offered her terms for "another book" (presumably Concerning Children): "500.00 down, 15% to 5000 and then 18%." These numbers dwarf anything the little story could have made, and tend to explain why it goes unmentioned. Also, SM contains a claim to be "reprinted from The New England Magazine of January, 1892, by permission of the publisher, to whom the thanks of the Author are due." Collation of the two appearances confirms this claim, although some variants explainable as compositorial error appear in

SM. Therefore, on the basis of internal evidence and lack of external evidence supporting authorial intervention, the 1899 book edition, SM, is treated here as a nonauthoritative derivation.

This leaves only NEM and the fair-copy MS as documents with possible authority. The situation surrounding the first publication is complex, and several circumstances that only indirectly bear on it must be appended to a narrative whose basic facts are already familiar to Gilman scholars.

We know for sure that Gilman, then Charlotte Perkins Stetson, sent a copy of the story from Pasadena, California, to William Dean Howells in Boston on August 28, 1890. This is confirmed by both her manuscript log[6] and her diary. The diary entry for August 24 states that Gilman "finish[ed] copy of Yellow Wallpaper," so she may have fair-copied an existing original for the express purpose of enlisting Howells in placing the story. This supposition is supported by an entry under "June [1890]" in the manuscript log stating that "The Yellow Wall-paper" was sent to Scribner's, but the entry has been lined out. The implied documentation of one manuscript in June and another completed in August suggests the existence of (at least) two manuscripts. Whether the story was originally sent to and rejected by Scribner's, or Gilman reconsidered before sending it, is unknown, but she evidently retained a copy while Howells had the story. No second manuscript apparently survives, but its original existence may be significant, as will be considered in the discussion of MS.

Howells passed the story on to Horace Scudder of the Atlantic Monthly, who attached a handwritten rejection card when he returned the manuscript to Gilman, which reads: "18 October 1890. Dear Madam: W. Howells has handed me this story. I could not forgive myself if I made others as miserable as I have made myself! Sincerely Yours, H. E. Scudder." Gilman must have been affected by this ambiguous rebuff, since she kept the card (which bears the words "(returning mss.)" in her hand) and recorded her umbrage at the incident years later in her autobiography, The Living of Charlotte Perkins Gilman.[7]

With the manuscript back in her hands, she sent it (or another copy) promptly out "to Mr. Austin" on October 26, according to both her diary and manuscript log. Henry Austin, whose name is attached to the name "Traveller Literary Syndicate" in the manuscript log, had written to Gilman weeks before, apparently soliciting manuscripts as a literary agent. Her diary records receipt of a letter from him (which has not been located) on September 23, and on September 27 she had sent "all this week's mss. to Mr. Austin" minus "The Yellow Wall-Paper," which was still going the Howells / Scudder route.

The autobiography records that Austin

Gilman's original letter to The New England Magazine does not survive, nor does their reply. The author is obviously relying on her memory of these events, as evidenced by her quoting her letter "to this effect" and her inaccurate recollection of the publication date, which was actually January, 1892. As for Austin's role in the placement of the story, there would normally be no reason to doubt Gilman's word if it were not that a slightly different account descends to us.

Howells reprinted "The Yellow Wall-Paper" in his 1920 collection The Great Modern American Stories: An Anthology.[8] In "A Reminiscent Introduction," he recalls

Since Gilman's account of Austin's agency appears to conflict with Howells' self-promotion, and since hers is supported by some (though hardly conclusive) documentary evidence and his by none at all, it would be tempting to accept the former, though one would wonder why Howells would misrepresent the facts, even under the guise of "reminiscence." However, given his affinity for dramatic phrases here and throughout the "Introduction," it would be difficult to define "corrupted the editor" without some additional information. Fortunately, indirect information sheds some light on this whole situation, demonstrating that Howells' and Gilman's accounts are not mutually exclusive, and more importantly, that the manuscript Howells "shiver[ed] over" was Gilman's only means of controlling the text of her story.

First we must step back in order to sort out the relationships of several people. On March 1st, 1890, while Gilman was composing "The Yellow Wall-Paper," her manuscript log records that she submitted a poem, "Similar Cases," to The Nationalist, a periodical edited by, as it turns out, one Henry Willard Austin. The poem was published just over a month later in the April Nationalist. Meanwhile, on March 11, Gilman sent her short story "The Giant

Wistaria" to The New England Magazine, edited jointly by Edward Everett Hale and Edwin Doak Mead. The manuscript log records that "Wistaria" was sent "via Walter" (Charles Walter Stetson, the husband from whom Gilman was then amicably separated, who was living in Providence, R.I., at the time).

Later, in early September and while Howells was perusing a holograph of "The Yellow Wall- paper," "Similar Cases" was reprinted in The New England Magazine. This demand for the poem was certainly a windfall for an unknown writer, though further entries in Gilman's records explain it. On September 16, Gilman sent another poem, "An Anti-Nationalist Wail," "to Uncle Edward Hale," and it was promptly published in December's New England Magazine. Scattered entries in both Gilman's and Stetson's diaries confirm that "Edward Everett Hale and his wife, Emily Baldwin Perkins Hale (the sister of Frederick Perkins, Charlotte's father), frequently invited Charlotte to visit them in Boston" where Gilman always seemed to enjoy herself immensely.[9] Gilman apparently meant to make the most of this relationship-through-marriage: on October 26 she sent "Mer-songs, etc. to E. E. Hale" (diary), otherwise known as "Uncle Edward (Traveller Literary Syndicate)" (ms. log). The "Mer-songs" weren't accepted, but her short story, "The Giant Wistaria," was, and appeared in the June 1891 issue of The New England Magazine. By then Hale had left the publication, but retained close ties with Mead, who remained. With the publication of "The Yellow Wall-paper" in January 1892, that would make a total of two poems and two short stories placed there in a sixteen month period, during which time her "Uncle Edward" was either co-editor or a friend of the editor there.

The linking of "Uncle Edward" to "Traveller Literary Syndicate" provokes interest. The specific business name, about which nothing has been discovered, sounds much like one of the literary agencies whose advertisements offered to "undertake every kind of work required between author and publisher" and some of which are glued to the inside back-cover of Gilman's manuscript log. She has there clipped ads for "The Writer's Literary Bureau" and "The Co-operative Literary Press," along with a clipped letterhead from the "American Press Association." Recalling that, according to her diary, Gilman had received a letter from "Mr. Henry Austin, 'Traveller Literary Syndicated'" on September 23, 1890, one is reminded that Henry Willard Austin published Gilman's first poem and Hale reprinted it in short order. It seems reasonable to conclude that Henry Willard Austin and Henry Austin were the same person, and that an agency (the formality of which is not known) consisting of Austin and Hale (and perhaps others) was formed. The letter Gilman received from Austin does not survive, but he was evidently soliciting manuscripts for publication: four days later she prepared and sent "all this week's mss. to Mr. Austin," and "The Yellow Wall-paper" followed them when Gilman received the manuscript back with Scudder's rejection on or before October 26.

The extent of Gilman's familiarity with Austin is not known. That the agent was indeed the same man who edited The Nationalist is further supported by Gilman's diary entry of February 14, 1891: Henry Willard Austin's slim book of poetry, Vagabond Verses, arrived in Gilman's mail as a gift from author to author. Austin evidently admired "Similar Cases" (as did Howells and Hale) and hoped for mutual appreciation.

This is the same man who, Gilman claimed, may have later stolen her payment for NEM. That she indeed never received payment seems almost certain: first, a page headed "1892 June" in her manuscript log reads "I have out, printed and unpaid" followed by a list of 27 items, some of which have been crossed out (apparently as payment came in). The second entry is for "The Yellow Wallpaper New Eng. Mag. March 1892" and has not been crossed out. Gilman was so impecunious at the time that she recorded the smallest amounts of received money in her diary, and nothing is mentioned in connection with the story.

However, the young author did receive a check for $14.00 from The New England Magazine on August 18, 1891, two months after the publication of "The Giant Wistaria" and five months before NEM appeared. It seems far more likely that this was payment for "Wistaria" (which does not appear in the manuscript log as "unpaid") than an advance for NEM. Besides, Gilman claimed in her autobiography that the latter story's publication was a matter of forty dollars.

And what role did Howells really play? As editor / critic Dock has pointed out, New England Magazine co-editor Edwin Doak Mead was Howells' younger cousin-by-marriage, and had been brought to Boston by Howells as a teenager (58). Howells' "Reminiscent Introduction" is cryptic at best, and could be taken for a claim that he received Scudder's rejection and then exerted pressure on his cousin to print the story. This scenario is not actually incompatible with Gilman's account, and the "handy compromise" critics Thomas L. Erksine and Connie L. Richards outlined (and that Dock seems to deplore as irresponsible scholarship) goes farthest in resolving the facts with the perspectives of all concerned.[10] Scudder likely spoke directly to Howells about the story's inappropriateness for the Atlantic Monthly,[11] and Howells learned somehow (through Mead, Hale, Austin, or Gilman herself) that the story had gone on to Mead's office. If he put in a word for the story, Gilman may not have known about it or felt it important enough to mention in the autobiography, given her publishing history and pre-existing connections at The New England Magazine. Conversely, Howells may not have reckoned properly with Gilman's own connections and given himself more than his share of credit. And, of course, the story has its own estimable merit.

All these circumstances are significant beyond satisfying the curiosity of Gilman scholars: they are crucial to the textual situation of "The Yellow Wall-Paper." An overall picture emerges. The text of "The Yellow Wall-Paper" as transmitted via manuscript through The New England Magazine seems to have gone beyond Gilman's control: the story was placed through an intermediary agent(s) who, enabled by personal ties, circumvented usual submission procedures. These procedures would normally have included an author's continued textual control after acceptance. However, no proofs are mentioned in any source, and Gilman lived in California, 3000 miles away from the publisher, at the time and was beholden to the agency of her husband and others when dealing with Boston's literary community: MS bears the return address of Gilman's husband, Walter, in Providence, not Charlotte's in Pasadena. As to internal evidence of authorial control, NEM's variants from MS are almost uniformly corruptions and can be attributed to compositor error or editorial intervention. Finally, Gilman was never compensated for the work.

I believe "The Yellow Wall-Paper" embodies an instance of what Fredson Bowers has called a "single authority textual situation."[12] That is, only one document survives over which Gilman can be demonstrated to have had textual control. If, indeed, demonstration could be made that Gilman corrected proofs for NEM or a later edition of the story, then even if the proofs did not survive, that appearance would gain authority and a critical edition would probably need to be edited eclectically. Since preponderant evidence suggests that Gilman did not correct proofs at all, "The Yellow Wall-Paper" fits the situation Bowers describes in which "The ideal copy-text will ordinarily remain any preserved holograph manuscript [MS] close to the print derived from it."[13] It remains possible, however, that the copy-text itself does not fully reflect Gilman's textual intentions. (The deviations may include such simple matters as slips of the pen or transcriptional errors that entered as she copied from an earlier draft.) The challenge for the editor is to assess Gilman's textual intentions and to adopt readings that most accurately reflect them. Exercising critical judgment based on an understanding of Gilman, the editor may in fact find helpful suggestions in the readings of later editions, even though these texts do not in themselves possess authority.

Some subsidiary concerns must be addressed before granting the above premises. First, is MS the actual document used as printer's copy at The New England Magazine? Recall that in June 1890 Gilman's manuscript log indicates that she may have sent the story to Scribner's, and only after copying the story in late August did she send a manuscript to William Dean Howells. Logically, then, at least two manuscripts existed while only one is known to survive. The essential question becomes: If MS and the lost manuscript differed, and MS was not the printer's copy, must it then defer as copy-text to NEM (which may have derived from a lost document more closely reflecting Gilman's textual intentions)?[14]

This takes some sorting out of conflicting evidence and argument. Cursory examination of MS shows that it is certainly a fair-copy in Gilman's hand with few corrections (and those are in her own hand.) It consists of fifty-nine separate leaves, approximately six by nine inches and blue-lined. The verso sides are blank except for occasional ink blots from the facing rectos, a sign of speed in fair-copying. That it was definitely intended for circulation to publishers is confirmed by the heading of the first page:

 

Mrs. C. P. Stetson. (about 6000 words-)

Box 401 Pasadena Cal. to be returned to Mr Charles Walter Stetson at the Fleur-de-Lys Providence R. I.

This is also in Gilman's hand. The portion beginning "to be returned" refers to Gilman's husband, the sometime-agent for Gilman living in proximity to the New England publishers. This portion is rendered in red ink (as opposed to the black of the manuscript proper) and was probably inserted after the document was returned from Boston to California by Scudder. The word count (written in black ink at the time of original transcription) is, of course, to this day a requirement of many manuscripts submitted for publication.

On the other hand, the document bears none of the telltale signs of printing house handling, such as thumbprints in ink or take marks. Aside from the expected foxing (brown oxidation), some leaves have been stained brown by a chemical that does not, however, appear to be printer's ink.

Closer examination reveals that at some points the handwriting has been clarified in a distinctly different shade of black ink from the original rendering. In most cases the lazy or hasty endings of words have been redone, but in one case the change is intriguing. On MS page 17 a sentence reads "I never saw so much expression in an inanimate thing before, and we all know how much expression they have!" "They" has been crossed out and "inanimate things" has been substituted in Gilman's hand, creating a deliberate repetition. However, the magazine printing retains "they." Although this wording suggests that another manuscript was used that did not have the authorial correction, the number and nature of authorial changes in MS argue at the least that it represents an advanced stage in the composition of the story. Collation of MS and NEM reveals 73 substantive variants and 334 in accidentals (including 110 paragraph alterations), but hardly one of these 407 variations defies explanation as compositor's error, regularization, or "correction" by a printer's reader of MS or a manuscript similar to it.[15] Conversely,

explaining many of these variants as Gilman's own changes to proof or to another manuscript would be difficult indeed.

From a practical viewpoint, however, all this hardly matters in the face of the present documentary situation. Even if a lost manuscript copy was used as printer's copy by The New England Magazine and subsequently discarded, MS remains the closest surviving document to it and least corrupted incarnation of Gilman's textual intentions. There seems little chance that, given the extent of variation, MS and NEM derive from a common (lost) ancestor. But even if they do (a case of radiating texts, in Bowers' term), Gilman's obvious involvement in creating the transcription (MS) and the absence of any identifiable link between her and the details of the NEM text (whose variations can be explained without her) mean that MS ought to be selected as copy-text.

The issue of tacit consent also deserves some attention. Even if we grant MS copy-text status, some would point out that no evidence exists that Gilman objected to the changes made by the magazine to her story, even expected those changes as "regulariz[ation]" (Dock 55). She then had the further opportunity (in theory if not in fact) to revise for the book publication seven years later. Howells brought the story out again, as did others, while Gilman was still alive. Should this lack of objection on the author's part count as evidence that the story was and always has been a product of social construction, of collaboration between artist and publishers? Certainly in our own time the story is a social artifact "produced" not only by publishers but by critics whose arguments have depended on texts that derive from the magazine edition or the Small, Maynard book edition (i. e. the widely used Feminist Press edition of 1973).

Such editions as have been produced by this line of argument (whether recognized or not) have their uses and will continue to be available. However, an edition that seeks to recover, as closely as possible, the uncontaminated textual intentions of an author will recur to G. Thomas Tanselle's distinction between what authors (especially young, poor, unrecognized ones) will accept from publishers with regard to treatment of their texts and what they would prefer.[16] Tanselle's discussion, a refutation of contrary positions by James Thorpe and Philip Gaskell, deals with accidentals, but logic extends it, in the present context, to substantives. The specifics of Gilman's situation support Tanselle's position.

Gilman appears not to have been invited by The New England Magazine

to correct proofs. Indeed, she may not have known the story had been accepted for publication until after it came out. Her concern over variations from her text was certainly subordinated to anger that she was not paid at all for it,[17] and she would have known the futility of the former "artistic" complaint after the fact of publication. By the time of the book edition, the story was several years old, in danger of being forgotten, and was being brought out seemingly as a low-profit contribution to an author/publisher relationship that was moving in more lucrative directions. And, as has already been discussed, if Gilman had any desire to revise the story, she had little opportunity to do so while the book was initially at press. By the time Howells asked to bring the story out in late 1919 (he didn't ask for alterations either) the story was nearly thirty years old and still obscure. Until the end of her life, Gilman never conceived that the little story would herald her reputation in the future: her 1935 autobiography of 335 pages devotes only 3 to the piece. Her "consent," then, to leave the text of the story as publishers (mis)handled it seems a hybrid matter of early powerlessness and later indifference stemming from that powerlessness. With respect to authorial control, Gilman's experiences with the story certainly do not mirror those of better-known nineteenth-century authors who viewed successive printings as opportunities for revision and correction.[18] However, she may (in this case) be a more accurate representative of vast numbers of unknown writers whose relationships with publishers were more tenuous.

Finally, the most essential questions: Is MS a truly different text from NEM, one that warrants an edition based on it? Because the Feminist Press issued a new edition of the story in 1997, one that corrects the departures from NEM made unintentionally in the 1973 edition, and because Dock's critical edition using NEM as copy-text will also appear in 1998, won't more than enough "good texts" be available to scholars?

The answers lie in the variants:

For the sake of her "health," the narrator has been forbidden to write but keeps a secret journal in an attempt to express herself and retain her sanity. She adopts the subterfuge precisely because spoken conversations (with her husband John, the only available interlocutor) prove repeatedly futile. The narrator frequently repeats herself, perhaps suggesting something about her psychological makeup. The central image of the story, the wallpaper, is the "inanimate thing" referred to in the former example. Before his fainting spell (or death?) at the very end of the story, the narrator associates John with strength, both physical (i.e. his ability to carry her or break down strong doors) and mental (his scientific knowledge and domination of her). The bedroom at the top of the mansion is sometimes occupied by both the narrator and her husband, other times by her alone when John is away all night, as she believes, attending to seriously ill patients. MS uses solitary pronouns that highlight the sense of isolation. Perhaps the magazine's agents wished to efface any suggestions of John's infidelity. No typographical reason is evident for the missing sentence in NEM. It may have been a compositor's eyeskip—or a wish to avoid "suggesting" anything too unpalatable. Clearly two different feelings are being contemplated here. Of course, the narrator is "caught creeping" defiantly in the final scene, and may have been planning the confrontation with John for some time. If anything, she develops an air of pride in her creeping. Though pronoun choice is often considered insignificant, occasionally it may be exploited to good effect or at least interpreted as if it had been. Catherine Golden, apparently without knowledge of the alterations from manuscript, devotes considerable attention to the story's use of "I," "one," "he," etc.: "In introducing 'myself' and 'John,' the narrator intensifies her awkward positioning in her sentence and society; she is not even on par with 'ordinary people like John.'"[19]

These few substantive variants readily indicate the disparity of the two texts, which contain dozens more. However, they are far less germane to the question of copy-text than the accidentals of MS and NEM, and I present them merely to indicate the essential need for a critical edition based on the manuscript. A copy-text is chosen for the texture of its accidentals; editors disagreeing with or ignoring this crucial assumption will certainly produce different kinds of critical texts. Indeed, all or some of the above substantives from MS could be adopted into an edition, like Dock's, based on a later text, but such a procedure would efface hundreds of authoritative accidentals. Preservation of an author's unique accidental usages, in essence, was a main point of W. W. Greg's famous essay.[20]

Like Dock, some have claimed that Gilman's accidental usages were (or were expected to be) uninformed, uneven, and only improved by intervention of her publishers and their agents.[21] But other scholars who have had the opportunity of examining MS have insisted that the author knew what she

was doing at least as well as the staff of The New England Magazine.[22] As with any author's work, evidence for both positions exists and the question becomes one of preponderance. Again, the proof is in the pudding.

The most significant accidental variants (actually semi-substantives) are the 110 alterations to Gilman's paragraphing, 87 of which were breaks where none exist in MS and 23 of which deleted breaks present in MS.[23] Where MS presents a coherent-looking, well-paragraphed narrative that becomes more and more fragmented as the narrator grows agitated, NEM presents an entirely fragmented, rambling account in which, from the first, the narrator appears unable to hold her thoughts together. No wonder that most critics have buttressed their interpretive arguments by altering her husband John's diagnosis of the narrator's "slight hysterical tendency"(NEM p.648) to one of outright insanity.[24]

Most of the remaining accidentals are comma, dash, or italic additions and deletions, only some of which clarify the text and most of which alter Gilman's emphases. Others are expansions of Gilman's contractions, as "would not" for "wouldn't," probably done purely for form's sake and inconsistently done at that. There are few, if any, spelling variants. More significant punctuation variants, as changes from periods to exclamation points (occasionally vice-versa), are less frequent but, with the fragmented paragraphing of NEM, help support John's diagnosis of the narrator as a

hysterical female. Other punctuation variants positively confuse the story, as when the narrator confronts John and makes her mysterious declaration at the end: Since the name Jane has not appeared previously to this climactic moment, its referent (the narrator's own name, a nickname for John's sister "Jennie," all women, the narrator's "domestic" self, an alter ego created by delusion, etc.) has been a crux of speculation among critics of the story. NEM's introduction of a question mark converts Gilman's ambiguity into downright nonsense, and even noncritical editors have invariably emended it to a period or exclamation mark.

In short, MS contains hundreds of authorial usages, both substantive and accidental, for which no editor in the Greg-Bowers tradition could reasonably justify emendation to NEM's nonauthorial variants. Taken together, MS's usages do present a different enough text to warrant an edition, or "version" as social constructionists have it. If NEM were used as copy-text, a considerable number of Gilman's preferred readings, especially in the matter of accidentals, would never find their way into the reading text and would (at best) be relegated to the apparatus.

To return and sum up, then, as to how the textual situation of "The Yellow Wall-Paper" has been affected by addition of this new information to old. After Scudder returned her manuscript, Gilman's last act of control over her text was to send it to an agent. Its placement in The New England Magazine was effected in some way that precluded any further intervention on the author's part. She did not even authorize its unpaid publication, which subtly altered hundreds of her usages and may have fundamentally changed the work. Although she did authorize later appearances, she never made an effort to regain control but, indeed, had little incentive to do so. Therefore, a critical text should be based on the only surviving authoritative document, the Radcliffe holograph, and should admit only emendations that reflect Gilman's textual intentions more accurately than the obvious errors in MS.