SOURCE: “P/P … Tekelili: Pym Decoded,” in English Studies in Canada, Vol. 14, No. 1, March, 1988, pp. 82-93.

[In the following essay, Smith discusses Pym using Roland Barthes's critical method of “decoding” and deems the work “a metafictional classic.”]

Modern readers of The Narrative of Arthur Gordon Pym seem to have evolved into two distinct and contradictory classes. In the first category are what we might call the “hoaxers,” who take Poe at his word about his “silly book” and solve all the cruxes of the text on the basis of a perceived intent to hoax the public with a potboiler adventure fiction. The hoaxers assume that Pym is a deliberate parody of fiction in violation of Poe's well-known critical opinions concerning the tale and its effect; for them it is a longer example of the genre of “A Predicament.” In the second category we have the “Freudian Fryers,” who accept a parodic element in the narrative, but who see it as a superficial overlay, rather like the borrowings from “Mellonta Tauta” in Eureka. They prefer to emphasize the evidence for Poe's self-conscious symbolizing of the text—the shift of dates to achieve a nine-month voyage and the like—thereby emphasizing the archetypal structure of the narrative.¹ As a result, criticism of Pym is deadlocked and unproductive, rather like the squabble between the “squatters” and the “dandies” among Mark Twain critics. One side has all the biographical evidence in place, but is forced to the conclusion that Pym is pretty poor stuff; the other side struggles with the uneven and crude suggestions from the text to prop up the implications that Pym is a masterpiece.

This paper proposes an approach to the work from a more modern perspective, a reading of Pym in something of the manner of Roland Barthes's decoding of Balzac's “Sarrasine,” S/Z. I say “something of the manner” advisedly, because Barthes's technique in S/Z was itself a parodic example of exegetical overkill. The Tel Quel edition is 278 pages long, of which 31 constitute a reproduction of the text of “Sarrasine.”² To keep those proportions with Pym would require, according to my hand calculator, no less than 1,354 pages of the size of The Collected Works (Boston: Twayne, 1981). Barthes's technique in S/Z may be adapted to a longer work like Pym in other ways, however. Where “Sarrasine” was, for Barthes, “passing” as a “straight,” realistic, heterosexual fiction that needed exegetical undressing, Pym is a problematic direct contradiction to the laws of verisimilitude and Poe's own strictures on the art of fiction. Therefore, I propose to use Barthes's division of the techniques of fiction cited in his textual “lexias” as semiotic “tumblers” to the lock of Pym, or, perhaps a better analogy, code-bodies of unequal weight to be juggled simultaneously as we proceed through the narrative. There are two ways to perform this act of explicatio upon the text: one is to write five exegeses in parallel columns according to the five codes; the other is to presume equal reference to all five codes and move randomly through them. This paper will attempt variations on both.

For Barthes, the process of fiction consists of manipulation of five semiotic codes. The proairetic code includes all the action of the narration and is approximately equivalent to the less fancy notion known as plot. The hermeneutic code is the puzzle of the text that continues the reader's interest—approximately equivalent to the notion of suspense. The cultural code consists of the narrative's use of the reader's expectations concerning the relationship between cause and effect as they appear in the narrative—their etiology. The connotative code is the contribution to the narration of the structure of the work—its division into parts, breaks in chronology, place, perspective, point of view. Finally, the symbolic code is the code of the text itself insofar as it is a recursive generation of its own content. Now, although Barthes does not say so, it is obvious that these cryptanalytic categories are not of equal weight. Yet it is not immediately obvious how they differ; it takes greater sophistication as a reader to recognize and decipher connotative or symbolic coding, but it is equally obvious that proairetic and hermeneutic decoding are just as essential to a successful reading. For this analysis I would like to suggest a slight modification of Barthes's decoding. Let us treat each of the five codes as programs in a computing system, each summonable on a Goto-Return or If-Then basis. They will then constitute a “stack” of programs (to descend further in cybernese) that can be called up at will over any part of the text, but will remind us constantly of where we are in the decoding process as well as where we are in the text. Our base reading program is proairetic, since, as Barthes writes, it “is never more than the artifice of reading: whoever reads the text amasses certain data under some generic titles for actions … and … it unfolds as this process of naming takes place.”³ When the “process of naming” includes a suspension of certainty until a question can be answered, the hermeneutic code is called up and may remain resident until determined; when the “artifice of reading” contradicts our normal expectations of cause and effect, we may call up the cultural code with the same privileges; the same technique will work for problems of form with the connotative code and for problems of recursiveness in form and content with the symbolic code.

The Narrative of Arthur Gordon Pym: we are instructed by the title to consider the work not as a novel or a story but linearly, and, at that, to concentrate on the act of narration as well as the events being narrated. Thus the title calls for proairetic and symbolic decoding simultaneously. Moreover, we have barely passed the title and begun to read the Preface when we discover that our narrator, Mr. Pym, is at odds with his “editor,” Mr. Poe, and that the opening section had been first presented as fiction but is here restored to its original narrativity. The Preface concludes with several assertions about the inescapability of the conclusion that it is indeed Mr. Pym's narrative, not Mr. Poe's, that the latter's contributions are not pertinent to matters of fact, and that the least tutored of readers will be able to distinguish between the styles of the two putative authors. The entire text is then concluded with a Note by an unnamed third person whose voice is quite distinct from the deceased Pym and the skeptical Poe, and who attempts a striking exegesis on a hieroglyphic basis and then concludes with a neo-biblical injunction: “I have graven it within the hills, and my vengeance upon the dust within the rock” (208).

This tripling of voices has a curious effect upon the narrative. By excluding the editorial function to the character of Mr. Poe and the exegetical function to the author of the Note, “Pym” preserves a proairetic self which keeps his narrative exclusively eventful and (apparently) direct. Such a division, of course, is suitable only for the hoaxable reader, in which class no modern reader would place himself except for the temporary delectation of the event, with full critical faculties available to step out of the hoaxing mode whenever necessary. Hence, the modern reader is, like the tripartite narrators of Pym, in tres partes divisa from, if not the word go, the second word of the title. Moreover, the reader is “stacked,” in the modern cybernetic sense, like a computerized filing system, and must “push” or “pop” himself from one stack to another as the situation of the text requires. The “lowest” level of reader activity is utterly naïve proairetic, and few texts deliver quite so much action in so little space. However, we are not very far into the narrative before our proairetic reader is challenged over the very possibility of event. When the Ariel is dismasted, the sloop is described as “under the jib only … boom[ing] along before the wind” (60). No reader who can define the words sloop and jib can pass that comment by without moving from the level of event to the other decoding levels. A sloop has a single mast; when it is dismasted, the jib as well as the mainsail must fall; the narrative is therefore impossible.⁴ Even a proairetic reader cannot read an impossible event without moving out of that decoding mode into the cultural code.

The first conclusion one wants to draw is that “Homer nodded,” that Poe simply forgot that a sloop is single-masted; the merely proairetic reader is probably happy enough to leave it at that. A more likely response invokes the hermeneutic code: obviously Poe knew that both jib and mainsail are attached to the mast of a sloop; this “error” is to be treated like all the other errors and improbabilities of the text as a puzzle which requires a solution. Perhaps, indeed, the solution is in the connotative coding; Pym in the Preface complains that Poe, by presenting his narrative as fiction, has misrepresented him. Perhaps this incident is exactly what he had in mind. Alas, he also comments that “no fact is misrepresented in the first few pages which were written by Mr. Poe” (56). Perhaps, then, the solution lies in the other two codes. Perhaps the patent impossibility of the dismasting of the Ariel is intended to undercut the straightforwardness of the narrative structure self-referentially, creating a lucus a non lucendo self-destroying dada narrative—in other words, a hoax. Perhaps. Perhaps not. At this stage all the reader can do is keep her decoding stacks ready.

As the proairetic reading continues the number of gross improbabilities and even grosser absurdities increases. They are of various kinds. There are what we might call pedantic irregularities of narrative, like the extended disquisition on stowage, which seem to fulfil some sort of verisimilitudinous purpose but do so to our limit of tolerance. Even a naïve reader must conclude that perhaps his leg is being pulled. Then the events of the narrative themselves do not become any less unlikely after the extreme improbability of the Ariel episode, in spite of the dead-panned delivery. Pym's entombment, the struggle with the mutineers, Peters' appearance and character, the shipwrecked mariners' lashing themselves to the wreck so tightly they are almost cut in two: the cumulative effect of the proairesis is to build suspense of a most unusual kind. One begins to anticipate something like a punchline—“and then he awoke and found it was all a dream,” to take the most common of such conclusions. But instead the reader goes on and on, it seems, southward with the Jane Guy, to the island of Tsalal, to the south pole of J. N. Reynolds and John Cleves Symmes.

Again we must pause in our narrative reading, the accumulation of hermeneutic clues being too great to allow us to bypass the stack. It is true that every writer of fictions in the largest sense has what we might call “freedom of cosmology”—the possibility of placing his fiction in a galaxy long ago and far away. Milton with his Ptolemaic solar system, Lucas in his galaxy, Reagan on his Star Wars: each has the right to make of his universe what he wills. But Poe's choice of the Symmes configuration of an earth with “holes at the poles” is so common in the canon of his writing and so heavily emphasized in Pym by its appearance, unresolved, at the conclusion of the narrative that it becomes a central hermeneutic problem. That position is reinforced by the ancillary clues that accompany the problem. The congruence of names, Poe + Symmes = Pym, the increasing dissociation with the laws of nature as the narrative moves southward, the regression in historical time to biblical suggestions in the names Tsalal-Solomon and the Hebrew “roots” in the hieroglyphic byplay: all tend to emphasize the super-nature that surrounds Poe's use of Symmes's cosmology. They emphasize it. They do not explain it. The effect is to hold back the conclusion of the proairesis and turn it into a hermeneutic problem. Pym becomes a narrative without a conclusion, a fictional Möbius strip that forces the reader out of the proairetic code into the other decoding modes willy-nilly just to make sense out of the narrative.

One other proairetic quality of the text is also germane to this consideration of the non-conclusive conclusion. Much of the action that Pym engages in has existential as well as proairetic purpose. Like Sartre's Oreste and Barth's Todd Andrews and Jacob Horner, Pym is sometimes cursed with freedom of choice, suffers from the post-modern disease of cosmopsis, the inability to choose to act, and seems therefore to choose, perforce, inaction. Choice and non-choice, action and inaction interweave contrapuntally through the episodes of the proairesis until they become themselves a structure of the work. From the dismasting of the Ariel (when Pym can do nothing to save himself) to his fall from the cliff into the arms of Dirk Peters, Pym explores the limits of action upon event existentially. The effect is to merge the proairetic with the hermeneutic. With the appearance of each new event Pym's response begins with the rational and then, with each turn of the screw, becomes more and more adroitly rational. His comfortable hiding place aboard the Grampus turned into an entombment, he deduces from the persistence of the larboard tack that the Grampus has proceeded normally from port; the dream-monster licking his face and hands he concludes must be his dog, Tiger; then, when taxed to his limit to read Augustus's message, he manages a climactic synthesis of ratiocination: he has Tiger retrieve the pieces of the torn paper; he fits them together like a jigsaw puzzle, trusts his sense of touch to determine which side has been written on, and uses the last of his phosphorus efficiently to read the seven concluding words of the vital message.

But each of these triumphs of reason is accomplished to a counterpoint of inaction brought about by a failure of reason. Pondering Augustus's absence in spite of the evidence of the Grampus's progress, Pym falls into a deep sleep “in spite of every exertion to the contrary” (72). Waking from his monstrous dream to a “huge and real monster” pressing against him, he can do nothing to save himself “had a thousand lives hung upon the movement of a limb or the utterance of a syllable.” He feels his “powers of body and mind” desert him, believes himself to be “perishing of sheer fright.” “Making a last strong effort, [he] at length breathed a faint ejaculation to God, and resigned [himself] to die”—at which point Tiger begins licking his face and hands (73). And then, of course, he has to have Tiger “go fetch” the torn note from Augustus for reassembling because he had “childishly torn it in pieces and thrown it away” when he failed in his first reading and spent “many miserable hours of despondency … before the thought suggested itself that [he] had examined only one side of the paper” (78).

These are but a few examples of what seems to be the dominant structural mode of the proairesis. Perhaps it is not possible for even the most literal-minded proairetic reader not to become conditioned to this problem-solving system as part of the connotative coding of the text. The very thematic repetitiveness of the proairesis might well be intended to create a meta-structure to serve as an interpretive window into the text. Certainly the repetition of event a producing puzzle b solved by reason, but only to yield other puzzles n until reason fails and inaction !i! then brings about a final resolution !r! serves as a meta-proairesis of the narrative in a nutshell. One is tempted to reason to a conclusion from this evidence. The process of reason can control events to a limited extent, one might conclude about a text, but when logic yields only more deeply complex puzzles, a fall or retreat into inaction will arrest the process and bring about an end to the sequence with a final (and fortunate) result.

Just such conclusions are borne out by invoking the cultural coding of the narrative as well. Just as the events of the narrative present a meretricious proairesis, so the conclusions that the character Pym draws from events are equally snares and delusions. Once again, the coding is so common that the cultural and connotative codes merge. Augustus's vivacity aboard the Ariel is really a sign that he is falling-down drunk and totally incapable of action. Dirk Peters—he of the bloodthirsty behaviour and horrendous appearance—becomes the most nurturing and supportive of soul-brothers to Pym. The natives of Tsalal, friendly and innocent to all appearances, turn into bloodthirsty assassins. And so on. Just when the evidence becomes over-whelming in favour of the acceptance of appearances, the results are reversed. The proairetic reader who fails to invoke cultural and connotative coding for an explanation becomes like the Charlie Brown who believes Lucy when she tells him that this time she will not pull the football away as he attempts to boot it. Yet what is the poor reader to do? When the “shrouded human figure” with a skin “of the perfect whiteness of the snow” appears to Pym, Dirk, and Nu-Nu at the vernal equinox on the edge of the cataract spilling through the “hole in the [South] Pole” (206) leading to the centre of the Symmesian earth … does the signifier believe have any relevant signified left? Pym is ready to fall one more time—into the arms of the figure, down the cataract, into the earth's centre—and the concluding Note affirms that on this occasion, as on so many past occasions, his fall was fortunate.

The cultural coding of the work seems to present a progressively more complex and intense critical examination into the heart of that “proairetic sequence [which] is never more than the result of an artifice of reading.” If we cannot be certain of the very meaning of such simple signifiers as jib and sloop, if our trust in the good faith of our narrator concerning the likelihood of event makes us suspect that we are reading about a dream, if our very faith in existential action is shaken, what system of belief, what “transcendental signified” can we put our faith in? Even the great imperative, the need for self-preservation, is challenged directly by the cultural coding of the narrative. But, again, although the proairesis suggests the paradoxical need to “fall” into life, the puzzle of what indeed Pym is falling into at the conclusion remains unresolved hermeneutically. Since the puzzle is unresolved, the reader necessarily calls up the next level of decoding and attacks the etiology. But this time he does not limit himself to problems of the significance of simple event, like the definition of sloop and jib, or of puzzle like the meta-structure behind the contrapuntal rational-irrational behaviour of Pym. No, this time the reader must question her own assumptions about life and death.

Is the need for self-preservation the “categorical imperative?” The use of that term itself suggests that it is not. Yet Poe was unlikely to go to Kant for his cultural coding, considering the unusual spelling he gives the name of that philosopher in “Mellonta Tauta.”⁵ For Poe, values beyond the mechanical and the rational were not likely to be generated by an epistemological “Copernican revolution.” Poe was far more likely to look to science for inspiration in such matters, as Coleridge, we are told, attended the lectures of Sir Humphry Davy when he “ran short of metaphor.” One of the “scientists” Poe knew of was John Cleves Symmes and his theory of a hollow earth; another was Richard Watson, whose essay, “On the Subjects of Chemistry, and their general Division” Poe read and even cited (ambiguously) in “The Fall of the House of Usher.”⁶

Watson's influence on Poe is pertinent to any consideration of The Narrative of Arthur Gordon Pym because Watson presented an alternative to the empirical and mechanical view of nature. The essay on the “Subjects of Chemistry” is a denial of the meaningfulness of distinguishing among the so-called “kingdoms” of nature: the animal, vegetable, and mineral. Arguing from the examples of coral, first thought to be mineral, then vegetable, then recognized as animal; from the observed sensibility of plants; from reports that “in the Cretan labyrinth it hath been observed, that the names of travellers, which have been cut in the rock in former ages, are now in alto relievo, and that the older the dates are, the greater in protuberance, resembling the callus formed by incisions in trees,” Watson arrives at this remarkable conclusion:

supposing, however, that we pay no attention to any of these circumstances, yet cannot we form any judgment concerning the internal state of the earth. The greatest depths to which Miners have penetrated even in mountainous countries, which may be considered as excrescencies from the true surface of the earth, or the level of the sea, have scarcely ever equalled one sixteenth thousandth part of its diameter; a distance altogether insufficient for the forming any probable conjecture about the inward constitution of the globe. The strata of stones, and veins of minerals, which are met with upon the surface, can give us as little information concerning the internal structure of the earth from which these are probably derived, as the contemplation of the scales of a fish, the feathers of a bird, or the Epidermis of a man, would concerning the bones and muscles, the veins and arteries, the circulation of the blood, and the several secretions of an animal body. … All the strata of limestones, chalks, marbles, all gypsums, spars, alabasters, &c. are confessedly of animal origin. The strata of pit-coal, and of all bituminous fossils, of some species of slates, whatever may be thought of argillaceous strata in general, the mould everywhere covering the surface of the earth, and other substances, are supposed, probably enough, to have arisen from the destruction of vegetables; so that I know not whether it would be a very extravagant conjecture which should suppose that all matter is, or hath been organized, enlivened, animated.⁷

That the unresolved proairetic and hermeneutic problems of the text of Pym are at least partially resolved by the suggestion that his cultural coding derives more from Symmes and Watson than from Aristotle and Newton is also borne out by a consideration of some of the connotative coding of the narrative. The cycles we have already considered—reason to inanition to more intense reason to despair and total loss of control to a “fortunate fall”—are paralleled by many other cycles—of burial-rebirth, belief-disbelief-new belief—repeated many times. These cycles are parts of larger cycles observable in the connotative coding and remarked by many critics. Pym's voyage takes nine months and ends on the vernal equinox; the longer voyage of Pym aboard the Grampus and the Jane Guy is a lengthened cycle of the voyage of the first chapter aboard the Ariel. Cycle generates cycle, wheel turns within wheel, inviting, almost demanding that the reader abandon simple proairesis and hermeneutic for the meta-text. “The key to the treasure is the treasure,” Barth's Scheherazade, inspired by her author, remarks. Well, perhaps. Let us at least consider something of that sort with regard to the cycles of Pym's “falls.”

In the organic universe of Watson the avoidance of death is not nearly as important as the continuation of creation. Those coral atoll “excrescencies” he cited in the beginning of his essay are but the death-induced evidence of the continuance of organic life. Watson is not, of course, the only one to notice that the cycle is of birth, procreation, death. Poets—writers—have always paid rather more attention to the middle third of that particular cycle than they have to the other two. Not so Pym, the author of this narrative. Aside from the suggestive comments about oriental consumers of the biche de mere found on Tsalal, sex is not present at all in The Narrative of Arthur Gordon Pym. There are no female characters in the narrative except for the three ships upon which Pym sails. Both the Ariel and the Grampus are referred to by the feminine pronoun, as is, of course, the Jane Guy, with its feminine (albeit ambiguous) name. Poe's granting femininity to the three ships is conventional, of course, but the natives of Tsalal come closer to the organicism of Watson in their belief that the Jane Guy is a “living creature.” The chief goes so far as to indicate his sympathy “in what he considered the sufferings of the schooner, patting and smoothing” a gash where the cook had struck the deck with an axe (169). Pym makes the doubly ironic comment concerning this anecdote (doubly ironic, considering the final fate of the ship) that their behaviour showed “a degree of ignorance for which we were not prepared, and for my part I could not help thinking some of it affected.”

One might argue that all the aspects of femininity are present in the various kinds of ships Pym sails in or encounters. The Ariel is woman as sister-companion of youth, promising the freedom of the imagination that its name alludes to. The Grampus (“She was an old hulk” [65]) holds Pym in her maternal womb to term and beyond, but nurtures the shipwrecked mariners with her stores when their need is great. The two ships which approach the shipwrecked group, on the other hand, behave meretriciously indeed, lure with false hope and prospects, and then disappoint utterly, the Dutch hermaphroditic brig evidently the vessel for some horrid disease, the second ship by metaphorically “turning its tail” to the temporarily enlivened group. Finally, the Jane Guy is wife and helpmeet to Pym's southerly quest, but as its double and ambiguous name suggests, is only a “plain Jane” and “guys” him as well as it directs him like a “guy” wire or cable. The final episode's canoe is too obviously reminiscent of the many surrogate-mother embarkations undergone by questing heroes of myth to need further description. So if we are to conclude that the female sex is absent from The Narrative of Arthur Gordon Pym, we must essentially close down our connotative decoding stack.

Something of the same is true about the sex act in the narrative. Again, there are no females and no untoward suggestions of those activities which often, it is said, accompany “the life of a seaman” even in “his more terrible moments of suffering and despair” (65)—that is, buggery in the fo’c’s’le. Indeed, the all-male cast of characters represents a wondrous variety of sexually ambiguous yin and yang and possibilities for all sorts of S & M activities without ever crossing even a modern line of decorum, let alone violating nineteenth-century standards. Yet it is not hard to understand its appeal to the French Decadents, to name just one ambiguously epicene group which assured the transmission of The Narrative to post-modern readers. Pym's original quest, for “shipwreck and famine; of death or captivity among barbarian hordes; of a lifetime dragged out in sorrow and tears, upon some grey and desolate rock, in an ocean unapproachable and unknown” (65), is one that would cheer the cold hearts of de Sade and Sacher-Masoch, and speak, in opposite ways, to both Jean Genet and J. K. Huysmans.

Pym's frequent falls in his narrative are no more orgasmic ejaculations than they are falls from grace, but to the connotative code stack their repetition requires decoding as such. Pym falls from the “grace” of his control over his environment by reason into rescue at the hands of a benevolent providence under such various names as Tiger, Augustus, and Dirk Peters. In the same way, each of his “falls” is a “little death,” an orgasmic escape of the self by bestowing itself, giving itself up to the Great Nonself. That these “falls” are always felicitous suggests a sexual decoding no less than a theological one as well. In the organic universe described by Richard Watson, although the mechanics of sexuality might persist in mystery despite scientific probing, the nature of the act is universal, metonymic. Watson comments that the disputes among biologists “whether every animal be produced ab ovo femellae, or a vermiculo in semine maris, are exactly similar to those among botanists concerning the manner in which the farina foecundans contributes to the rendering the seed prolific.”⁸ Doubly, trebly fortunate may be Pym's falls—into life, into grace, into … a new encoding.

Indeed, the connotative coding of The Narrative is no more conclusive as a sub-program of investigation than any of the others, but the questions it raises do aid in some aspects of the proairetic and hermeneutic decoding. To see The Narrative as a cycle of cycles is to understand the conclusion as a “rescue” analogous to the one performed on Pym and Augustus by the Penguin in chapter one and all of the other “fortunate falls” intervening between the two, but this time into a new and somehow “higher” state. The recursiveness of the multiplication of cycles produces the paradox of the generation of a new dimension of meaning.⁹ That new dimension has something to do with the relationship between reason and faith on the one hand and sexuality and the organic creation on the other. The “death” of Pym is followed by the literary “birth” of the author of the Note, whose literally hermeneutic analysis of the hieroglyphics of the final episodes is obviously intended to mediate between the coding of the text and the further encoding by the reader. That encoding must be multi-dimensional, since the “new voice” of the author of the Note is specifically exegetic about the “text” of Tsalal—the island, the text. The reader is forced by the Note to an overview of the text as text-world. The interpretation of the hieroglyphics forces the reader to merge the creation of Pym with the larger creation of the (Symmesian and Watsonian) earth. The text is the world. It is/is not Pym/Poe who has “graven it within the hills … [and] upon the dust within the rock.” The final act of falling is the merging of the author with the reader in hermeneutic copulation as the creative maleness of Pym merges with mother Earth to close the cycle of the text. Like the solitary spermatozoon uniting with the egg after its long night swim, Pym rushes into the self-immolating embrace of the shrouded figure and completes one cycle—the cycle of text—with the creation of the next—a cycle of meaning.¹⁰

Our tour through the text of The Narrative of Arthur Gordon Pym completed, we are now ready to consider the question that opened this exercise. The dispute between the Hoaxers and the Freudian Fryers dissolves into nothing once the text has undergone Barthesian decoding. Perhaps alone among the great writers of the nineteenth century, Poe was capable of engaging in freeplay with serious content, hoaxing his readers at the same time that he was encoding culturally, connotatively, and symbolically ideas about which he could be deadly serious in other contexts.¹¹ Pym's universe is the organic universe of Eureka, but Pym himself is only a blind and comical seeker probing the crannies of some of the private parts of that world. Indeed, the greatness of Pym resides in that very lightness and playfulness of the text, compared to which the profundity of Eureka, admired as it was by its author, sinks into bathos of its own specific gravity. Poe recognized the problem, trying to lighten the omelette with the soufflé excerpts from “Mellonta Tauta” and asking that Eureka be read as a poem, but no sensitive reader of Poe prefers Eureka to Pym. The continuing appeal of Pym through unsympathetic periods of Victorian repression and unsmiling realism argues for its success as a work of art. Now that we have reached an age which values writers like Borges, Barth, and Nabokov, Pym must be appreciated for what it is, a metafictional classic.

Notes

1. So much has been written about The Narrative of Arthur Gordon Pym that it would be fatuous to attempt a listing of studies here. Frederick S. Frank, “Polarized Gothic: An Annotated Bibliography of Poe's Narrative of Arthur Gordon Pym,” Bulletin of Bibliography, 38 (1981), 117-27, gives all the important studies (and others, too) to that date in the third part. Burton R. Pollin's Introduction to the text of Pym in The Imaginary Voyages volume of Collected Writings of Edgar Allan Poe (Boston: Twayne, 1981) perhaps best illustrates the hoaxer position (and some of the interpretive difficulties that follow from it), and Harold Beaver's Introduction to the Penguin edition of The Narrative (1975) gives about a page of lip-service to the possibility of a hoax and then goes on to a rather complete summary of the various psychological, sociological, and phenomenological readings that the text has generated. Structuralist and deconstructionist essays by Asselineau, Ricardou, Mourier, and Levy are listed in Frank; they are vastly different from this essay.

2. S/Z (Paris: Editions du Seuil, 1970). My quotations from Barthes are drawn from the translation of S/Z by Richard Miller (New York: Hill and Wang, 1974). Citations from Pym are drawn from Pollin's edition and are given by page number in his text.

3. S/Z, p. 19.

4. (60). See Pollin, p. 221 note, for a detailed analysis of the physical impossibility of the event and references to some early proairetic responses.

5. “Cant.”

6. See my “Usher's Madness and Poe's Organicism: a Source” in American Literature, [AL] 39 (1967), 379-89; the edition of Watson cited in that article and here is Chemical Essays (London, 1787), V, 126-71.

7. AL, pp. 384-85; Watson, pp. 167-69.

8. AL, p. 383; Watson, p. 146.

9. For a consideration of the paradoxes involved in this kind of event, I can hardly improve on the general discussion (interpreted analogically) in Douglas R. Hofstadter, Godel, Escher, Bach: an Eternal Golden Braid (New York: Basic Books, 1979), passim. For a suggestion of the mathematical underpinnings of such a reading, see Alexander Woodcock and Monte Davis, Catastrophe Theory (New York: Dutton, 1978), passim.

10. See, among many other possible intertexts, Plutarch, Moralia, V, trans. Frank Cole Babbitt (London: Heinemann; Cambridge, Mass.: Harvard University Press, 1936), 351, where an alternative to the island of Tsalal is offered as a location for the earth's omphalos; P. H. Gosse, Omphalos (London: J. Van Voorst, 1857) and Edmund Goose, Father and Son (New York: Scribner's, 1907), for the evidentiary nature of the etiology of the belly-button and the nature of creation on the one hand, and the trouble that etiology can cause on the other; Lewis Thomas, The Lives of a Cell (Toronto: Bantam, 1979), p. 4, for a modern comparison perhaps even more shocking than Plutarch-Poe-Gosse's.

11. Melville is the only other writer of English one would mention in this regard, and The Confidence-Man the one work that compares with Pym in this quality.