Text: Burton R. Pollin, “Pym (Introduction),” The Collected Writings of Edgar Allan Poe — Vol. I: Imaginary Voyages (1981/1994), pp. 29-36 (This material is protected by copyright)


[page 29:]



The Narrative of Arthur Gordon Pym was composed and published during one of the darkest periods in Poe’s literary career. At the beginning of this interval of some twenty-two months — from the late fall of 1836 to July 1838 — he was the overworked and undervalued toiler on the Southern Literary Messenger; at its close he was about to move to Philadelphia after a bleak sojourn in New York City. Much of this long span is, unfortunately, a biographical blank. Fired from his Messenger post in January 1837 — the very month when the first portion of Pym appeared in that periodical — Poe moved on to New York in February; on March 30 he attended a dinner, given by New York booksellers, at which he figured prominently enough to propose a toast; on May 27 he wrote to Professor Charles Anthon, asking for information which he wished to use in a review of J, L. Stephens’ Incidents of Travel in . . . the Holy Land; in May Harper & Brothers announced the forthcoming publication of Pym and copyrighted the title on June 10; in June Poe also published his story “Von Jung” in the American Monthly Magazine; his review of the Stephens book appeared in the October New York Review; and in December he printed “Siope” in the Baltimore Book for 1838. For the first half of 1838 there is no record of Poe’s personal movements. In May Harper & Brothers announced Pym as “nearly ready” and published the long-promised book at the very end of July; on July 19 Poe wrote, “with breaking heart,” to James K. Paulding, then secretary of the navy, pleading for any sort of job which might relieve him of a “miserable life of literary drudgery.” On August 1 the Harper firm deposited Pym at the New York copyright office; within that month Poe moved on to Philadelphia.

This is indeed a skimpy set of facts; but it is enough, when combined with other data, to make a case for the progress of composition of Poe’s only novel. That case, in brief, is this: that Poe began the writing of Pym as a relatively short serial for the Messenger, without having the present ending definitely in mind; that he began to extend the story shortly after his move to New York; that composition proceeded in discrete stages; that these stages were largely dictated by the source material employed; and that external events and internal textual evidence can help us to follow his progress. [page 30:]

A cautionary note, however, must at once be sounded. Since no manuscripts have survived, we can never be certain that the text which Harper & Brothers copyrighted in June 1837, was, in all respects, that which it published in July 1838. Nor can we know the answers to several other important questions: Did Poe present a “scenario” to the Harpers in inducing them to publish his book? Did he turn in copy by chapters or segments — a possibility which might account for some of the discrepancies? Did Poe and/or the Harpers plan to tie Pym in with the excitement created by the formation of the United States Exploring Expedition? If so, was some deadline set, which Poe was pushing his pen to meet? And — most intriguing of questions — just what sort of book did the Harpers expect Poe to supply?

Lacking positive clues to the solutions of such major problems, we are forced back primarily to the text itself. Close scrutiny of Poe’s manipulation of his several major sources suggests that composition proceeded in four stages. These are: (1) the Messenger text (through 4.3); (2) the “voyage narratives” chapters (the material from the end of the Messenger text through the end of chapter 13-plus, probably, the “Preface”); (3) the “Morrell chapters” (chapter 14 to the end of the book, with the exceptions cited as the last stage; and (4) the “Stephens-Keith” material (the first of the two chapters numbered 23 in the first edition text, the final “Note,” and scattered passages in the Tsalal episode. What follows is a condensed statement of the evidence for these stages.(1)

Stage I: the Messenger text (through 4.3). Poe had written only one sea story — “MS. Found in a Bottle” — before the composition of Pym. What led him to undertake another cannot be determined, though numerous critics have suggested the influence of Defoe’s Robinson Crusoe, a new edition of which he had reviewed in the January 1836, Messenger. (See “Sources” at n. 20). The beginning of the actual writing, however, undoubtedly occurred after he had received for review both Washington Irving’s Astoria (published in October 1836) and the printed text (dated October 10) of J. N. Reynolds’ address to Congress in April 1836, which called for the immediate undertaking of a governmental exploring expedition to southern waters. Poe’s lengthy notices of these two books appeared in the January 1837 Messenger, along with the first [page 31:] installment of Pym; and there are distinct verbal echoes of both works in the sea tale (e.g., 2.1, 2.5, 4.4).

On the evidence of the two portions of Pym which appeared in the Messenger, it appears that Poe originally planned the tale as an episodic sea adventure-though how far it would have been extended cannot be known. (For the theory of a hoped-for juvenile audience as well, see “Aims and Methods.”) The only clue is that the Messenger did not at this period run lengthy serial fiction — for example, “Sully, A Tale,” ran only about seventeen pages in three installments in the 1837 volume. It therefore appears improbable that Poe had at this stage plotted out a story that would have reached the book version’s length of 201 closely printed pages. Clearly he would have published as much as T. W. White, the editor, would have allowed; miserably paid for his editorial duties, he had arranged with White to receive an extra sum for each page of creative work. But White fired Poe on January 3, 1837, and the tale of Pym came to a sudden and unexplained end in the February issue. Just why White published an abortive tale at all is something of a puzzle. Perhaps he wanted to get his money’s worth for work for which he had already paid the author; more likely, although he disliked the man, he recognized the value of Poe’s name-for in an editorial he credited “Arthur Gordon Pym, a sea story” to Poe and also placed his name by the title in the table of contents. This attribution would force Poe to devise a convoluted “explanation” of the “true” authorship in the preface to the book version.

Stage II: the “voyage narratives” section; from 4.4 to the end of chap. 13. Poe and his family settled in New York by February of 1837. The move to the metropolis was probably based on the hope that the renown he had won on the Messenger would gain him a post on one of the “monthlies of Gotham,” which he was to toast at the March 30 booksellers dinner. (For the invitation from Hawks, see “Sources” at nn. 13 and 14). There is, however, no record of any steady employment, literary or otherwise, during his entire nineteen-month stay. It is reasonable to suppose, therefore, that he at once returned to his sea tale, with the intention of turning it into a book-length narrative (the Harper firm had written Poe in March 1836, that their readers definitely preferred a long, connected narrative to a group of short tales — particularly if those pieces had been printed previously in magazine form). At an early point he reworked some of the Messenger material, making these significant changes: the ages of Pym and Augustus were altered to make them less juvenile; the time taken for the refitting of the Grampus was increased; the date of sailing was also set ahead; and this sentence was [page 32:] added just before the recounting of the Ariel episode: “I will relate one of these adventures by way of introduction to a longer and more momentous narrative.” The “Preface” also seems to have been an early addition to the text, since its chief import is to explain away the fact that White had spoiled Poe’s apparent plan of anonymity by naming him as author; this was a problem which clearly was of more concern to Poe and his publishers in the first part of 1837 than in July 1838, when the book finally appeared. Poe may have made later changes in this Preface — its date, “July, 1838,” obviously was added as the book went to press — but it is curious that he allowed Pym to continue to refer to his returning to the United States “a few months ago,” a statement that would have been “true” in 1837 but not in 1838.

The best evidence for the delimitation of this stage is Poe’s heavy reliance on one type of source — books of voyage narratives, specifically collections of short pieces which nearly always dealt with marine disasters. His first unquestionable use of this material occurs in the very next paragraph after the end of the Messenger installments (see n. 4.4B); regular dependence on these sources begins in chapter 6 (after Pym’s adventures in the hold) and continues to the end of chapter 13, when Poe shifts to a new source. It should be noted that this grouping of chapters contains a number of discrepancies and promises of actions which never occur; moreover, the story line gives no hint of the marvels to come in polar waters. Indeed, as the drifting sequence ends, we are told only that the rescue ship, the Jane Guy, is “bound on a sealing and trading voyage to the South Seas and Pacific.”

Stage III: “the Morrell chapters”; chapter 13 to the end, with the exceptions noted in Stage IV. It cannot be determined when Poe first read Captain Benjamin Morrell’s Narrative of Four Voyages, which had been a Harper best-seller since 1832; but he could have seen its name — with its tantalizing subtitle — in any Harper catalog since the date of its publication. Poe’s borrowing in these chapters is so blatant that it is hard to believe that it went unnoticed by the Harpers themselves, But the publishers appear to have had a rather lax attitude toward free use of other writers’ words. Morrell’s book was itself the product of a ghostwriter who heavily padded out Morrell’s logbooks with material drawn from a variety of sources. And J. L. Stephens, whose Incidents of Travel would soon be a big hit, was urged to take advantage of the Harpers’ own wares in preparing his narrative of his travels.(2) [page 33:]

Poe had frequently incorporated borrowed information without credit in his earlier tales and reviews, and he probably had no qualms when he began to make heavy use of the Morrell material, in chaps. 14 and 15. These pages, loaded with detail, describe an erratic and rather dull cruise, which lasts from August 7 to December 12. But at the beginning of chap. 16 (on page 138 of the 201 pages of the first edition) Poe finally began to make good on the promise so luridly held out on the title page of adventures in deep southern latitudes. In addition to the accounts in Morrell about early exploration in the area, he again turned for data to Reynolds’ Address, a source last specifically used in the Messenger installments. For two chapters the drive toward the South Pole is headlong; but at 18.2 the Jane Guy arrives at Tsalal, and for the remainder of the text the pledge of discoveries “still farther south” is all but forgotten. The Tsalal episode is entirely self-contained, and it is not until the beginning of the very last chapter that we are recalled to the fact that all this action occurs in the “eighty-fourth parallel of southern latitude.”

Stage IV: the “Stephens-Keith” material; the first of the two chapters numbered 23 in the first edition text; the final “Note”; and scattered passages in the Tsalal episode.

Tracing all of Poe’s activities during the preparation of the final segments of Pym is obviously an impossibility. But several facts may be construed as indicating a late change of plan for the conclusion of the narrative: 1) there are discrepancies between the title, as recorded in the June 10 copyright entry and actual events in the tale; 2) two chapters are numbered 23, an implication that an insertion was made between the original chapters 22 and 23; 3) only the first chapter 23 and the concluding “Note” discuss the inscriptions discovered by Peters, and much of this material is suggested by passages in the works by Stephens and Alexander Keith which Poe is known to have been using in May 1837. The key to the timing of the composition of this late material seems to be just when Poe became aware of Stephens’ work and when he contracted to review it. The Harper firm clearly felt it had a best-seller on its hands as soon as Stephens began to turn in copy, for it began to “puff” it by sending manuscript excerpts to prominent journals — probably before the book was even being set into type (see “Sources,” n. 16). Further long quotations appeared in the May Knickerbocker; by this date the two volumes of the work had been set, though the magazine [page 34:] observed that they “are not yet published.” The reason for this delay is not known, but it is not highly significant. Poe could have been working over the Stephens material as early as its periodical publication; he may have later seen proof sheets; and he may have continued to comb it after the book edition was issued. Only one fact sheds much light: on May 27 Poe wrote to Professor Charles Anthon of Columbia College (who was associated with the Harpers in their publication of classical texts) and asked for literal translations of two biblical passages concerning the fate of Edom, a land which Stephens reported he had traversed. On June 1, Anthon replied, giving the desired information, which Poe would insert verbatim into the text of his review of Stephens.(3)

How much Poe himself believed in the literal fulfillment of prophecy — a topic insisted upon in the review — is hard to judge. But we may at least draw up a plausible model for the effect of his pondering over prophetic passages on the concluding sections of Pym. We may begin with the promise given at the close of chapter 17 that Pym will ultimately reveal his role in “opening to the eye of science one of the most intensely exciting secrets which has ever engrossed its attention.” Given Poe’s strong reliance on Reynolds and Morrell, this could scarcely be more than confirmation of their insistence that the southern sea became progressively warmer and that a ship could sail all the way to the pole itself. But this is a “secret” which has already been disclosed in chapter 16, and it would seem a lame conclusion to readers awaiting “apparent miracles” (18.9). Although, as the final “Note” mockingly informs us, “in no visible manner is this chain of connexion complete,” we may construct a possible linking of narrative elements: (1) Poe’s imagination was first stimulated by Morrell’s description of black South Pacific islanders and by their taboo on all things white. A conscious — or unconscious — joining of the themes “black/white” and “taboo/South” was thus formed, to which the language of the Tsalalians apparently has reference. (2) Development of this idea was stimulated by Poe’s use, during the writing of the Stephens review, of Alexander Keith’s Evidence of the Truth of the Christian Religion; this work discusses many biblical passages of prophecy concerning the destruction of civilizations and the dispersal of races to the earth’s farthest reaches. At this point Poe may have composed the first chapter 23, with the figures copied from Gesenius’ Hebrew and English Lexicon of the Old Testament which are not explained until the final [page 35:]

“Note.” It is in this chapter that Poe also first alludes to the remains on Tsalal of a prehistoric civilization, whose inhabitants, it is suggested, have moved on to “the region of the south.” The scientific quest for the pole and the concept of the pole as symbol of some ultimate revelation of man’s history and his fate were now intertwined. (3) Since no convincing narrative resolution of these themes was feasible, Poe attempted in the final paragraphs of the last chapter to create one through the presentation of the idea of a final revelation in the confrontation with the white figure — and, of course, by the maddeningly obfuscatory discussion in the final “Note.”

The employment of the material from Stephens and Keith apparently ended Poe’s work on his text; there is no evidence that he (or anyone else) ever gave it a careful editorial reading. But just when was the text “completed“? In the May 1837 Knickerbocker the Harpers announced that the “Narrative of ARTHUR GORDON PYM of Nantucket” was “nearly ready for publication,” and on June 10 entered the title for copyright.(4) But the Panic of 1837, beginning in May, upset all publishing schedules. According to the historian of Harpers, all books on their 1837 list were “those which were well advanced by early May or which received special preference because of their authors or sales potential.”(5) Stephens was one of the lucky authors, and months later he would publicly thank his publishers for issuing his book at such a difficult period.(6) [page 36:] We know that Incidents of Travel was “well advanced” by early May — but what of Pym? The only reasonable conclusion is that it had not been sent to the printers before June 10 — and that the Harpers found no compelling reason to press ahead with it. Their copyrighting it (along with four other titles on the same date) may have been no more than a desire legally to protect material for which they contracted.

In any event, they did nothing about Pym until they once again announced it in the Knickerbocker in May 1838, and actually began setting it in that month. The advertisement specified that it contained “Engravings” — the first indication in print of the cuts of the Tsalalian chasms and inscriptions. By the last days of July booksellers were offering it for sale, and within a few days the first reviews would appear.

Poe’s later interest in his story seems to have been largely financial. On February 19, 1839, he wrote to the Harpers, apparently inquiring about royalties. The firm replied promptly: “‘Pym’ has not succeeded or been received as well in this country as it has in England. . . .”(7) In fact, it did not succeed in either country. But this was scarcely a new experience for Poe.



[The following footnote appears at the bottom of page 30:]

1.  The first attempt to work out this sequence appeared in J. V. Ridgely and Iola S. Haverstick, “Chartless Voyage: The Many Narratives of Arthur Gordon Pym,” TSLL, Spring 1966, 6:63-80. This article neglected to assess the impact on Harper and Brothers of the Panic of 1837 and ascribed the long delay in the publication of Pym to Poe’s failure to supply copy rather than to business conditions.

[The following footnote appears at the bottom of page 32, running to the bottom of page 33:]

2.  See “Aims and Methods” at n. 11 for details. Whatever the facts may be concerning the publisher’s cooperation, it is noteworthy that Poe’s chief sources for the [page 33:] later stages of the tale — Morrell, Stephens, Reynolds, Keith — were all published by the Harper firm.

[The following footnote appears at the bottom of page 34:]

3.  Apparently Anthon’s letter went at least temporarily astray. The cover (preserved in the Griswold MSS. in the Boston Public Library) is marked: “not found” and “applied at 76 Charlton.”

[The following footnotes appear at the bottom of page 35:]

4.  For a facsimile of this document see fig. 4. Alexander Hammond in “The Composition of The Narrative of Arthur Gordon Pym: Notes Toward a Re-examination,” ATQ, Winter 1978, 37.9-20, attacks the notion presented by Ridgely and Haverstick, that the long subtitle was added at the time that the book was deposited in August 1838. My own study of the entry in September 1976 leads me to agree with Hammond that the physical appearance of the page suggests that it was made at one sitting — i.e., on June 10, 1837. But some nagging problems remain: Why was the first paragraph so thoroughly inked out, when there is only one visible error (the date at the end of the last line)? Note that later errors are simply lined through. Why was the name entered as “P. Y. M.“? Why does the index to the copyright volume list the title as “Gordon’s narrative“? Who composed the wording of the long subtitle? (For a guess that it was Wesley Harper, see Eugene Exman, p. 113.) Why do the last lines of the subtitle, so prominently displayed on the copyright entry and on the printed title page (“Together . . . rise.”), describe material that occupies only thirty-four of the book’s 201 pages?

5.  Eugene Exman, p, 95.

[The following footnotes appear at the bottom of page 35, running to the bottom of page 36:]

6.  In the “Preface to the Eighth Edition,” dated September 1838, Stephens writes: “When his manuscripts were placed in the hands of his publishers, the whole country was suffering under a crisis of intense severity; commercial business of all kinds was in a state of almost total prostration, and the publishing business was so cast down that the issue of a new book was a mere hazard, in which the chances were very decidedly against the hopes of advantage or even of immunity from loss.” Even so, [page 36:] the Harpers appear to have been unusually cautious. In the August 26 New-Yorker Rufus W. Griswold offered a poetic “Remonstrance” against the Harpers’ present policy of stopping their presses.

[The following footnotes appear at the bottom of page 36:]

7.  Quoted in Exman, p. 114.






[S:0 - BRPIMV, 1981/1994] - Edgar Allan Poe Society of Baltimore - Editions - The Collected Writings of Edgar Allan Poe (B. R. Pollin) (Pym)