INTRODUCTION  ❖  Editing a Fluid Text · Navigating the Typee Manuscript · Writing Typee · Scenes of Revision

Writing Typee

Until I was twenty-five, I had no development at all. From my twenty-fifth year I date my life. Three weeks have scarcely passed, at any time between then and now, that I have not unfolded within myself. But I feel that I am now come to the inmost leaf of the bulb, and that shortly the flower must fall to the mould.1

As a text, Melville’s Typee: A Peep at Polynesian Life represents two experiences: a four-week adventure in the South Seas, which we shall assume actually occurred, for the most part; and a several-month adventure in writing, which we know did occur. The first experience may not have happened exactly as Melville records it in Typee; as with any human event, reality fades the moment “happening” becomes “memory.” Melville’s more concrete experience was in his writing of the event, for he grew as his text grew, and the reality of that growth is recorded in the revisions of his manuscript.

From Sailor to Writer

On January 3, 1841, Melville set sail out of Fairhaven, Massachusetts, on the whaling ship Acushnet bound for the sperm whale ground in the South Pacific. In eighteen months of cruising along the Atlantic and Pacific coasts of South America, his ship entered port only twice, and on July 9, 1842, he and Richard Tobias Greene, two land-starved men in their early twenties, jumped ship as the Acushnet lay anchored in the bay of Nuku Hiva, one of the Marquesas islands situated about eight hundred miles northeast of Tahiti and a few degrees south of the equator.

Melville and Greene (known as Toby in Typee) made their way into the interior of the island, got lost, landed in the farthest reaches of Taipivai, the valley of the Taipi tribe, and resided there until each escaped individually during what scholars believe was little more than a three-week period. Melville and Greene were not reunited until after Typee was published, at which point Melville added “The Story of Toby” to his publication. Melville sailed from the Marquesas on another whaler, resisted authority again, but this time was punished for his resistance and spent a month in a Tahiti jail. From that island, he shipped out again on another whaler and landed in Hawai’i where in August, 1843, (a year after his sojourn in Taipivai) he enlisted in the navy on board the US frigate United States, at which point his mutinous days ceased. During these sailing years, Melville apparently did not keep a written journal. He was discharged from the navy in Boston on October 14, 1844, and made his way home to Lansingburgh, New York. There he was encouraged by friends and family to write down (or write up) his South Pacific adventures, and in the winter of 1845 he began writing in New York City.

What Melville called his “First Draught” of Typee may not have been much more than eighteen chapters long. And throughout the spring and summer of 1845, before and after he submitted a fair copy (no longer extant) to publishers, he augmented his text with additional materials and chapters. Rejected by the American publishing firm Harper & Brothers, he gave the text to his brother, Gansevoort Melville, newly appointed as a diplomat to London, to peddle the book in England in the summer of 1845. British publisher John Murray picked up the book in the fall of 1845 and Gansevoort saw the text through the press, correcting proofs for his brother during the winter of 1846. At that time Gansevoort also read selections in proof to Washington Irving (then visiting London), who recommended the book to his American publisher, Wiley & Putnam.

Typee first appeared as a book in England in February, 1846. A slightly altered and moderately expurgated version was published the following month in the United States. Almost immediately, on both sides of the Atlantic, this remarkable tale, remarkably written by what one somewhat skeptical reviewer called a “common sailor [but] no common man,” became if not a national best-seller then certainly Melville’s best-selling book.2 It would remain Melville’s most popular title throughout his life, eclipsing in reputation Moby-Dick.

Part of that popularity was linked to controversies over Melville’s strident critique of missionary abuses in the Pacific, whether the narrative was true, and even whether Melville had been the common sailor he claimed to be in that narrative. Reviewers, thinking the improbably named “Herman Melville”—is he Dutch? is he French?—was perhaps a seagoing gentleman taking liberties with facts, could winkingly accept the more romanticized Robinson Crusoe–like episodes in the book, but members of the religious press, especially in the United States, complained bitterly of Melville’s “moral obtuseness” with regard to the cause of missions in the South Pacific.3 And in response to criticisms, Melville issued a more fully expurgated version of his text in the fall of 1846 that reduced the book by almost one-fourth its original length.

Thus, in the course of six months, from February to August 1846, Typee went through three separate versions, and on the basis of this much variation alone, it could lay claim to being one of American literature’s most intriguing fluid texts. But the discovery in 1983 of the working draft manuscript of Typee revealed that Melville’s text was even more fluid than had previously been known, for the manuscript shows that Melville’s text traveled through at least three additional versions before it ever reached print. The extant manuscript itself is a portion of the original eighteen-chapter text that grew in its final form to over thirty chapters. But this three-chapter fragment is a sizable and central portion of Melville’s narrative, and it is laden with myriad revision sites, enough to indicate that at this earliest point in his publishing career, Melville was not only learning how to shape a narrative and hone his skills; he was also growing, as I argue in Melville Unfolding, in terms of his sexual identity and politics.

Herman Melville’s life changed the moment he began Typee. The simple act of inscription, initially just a matter of writing down some already well-rehearsed anecdotes, transformed him from sailor to writer. But the process was more than a trigger for a career shift; it became for Melville a way of knowing. He found himself trying to describe his past, dredging up memories and writing them down, but at the same time he found himself transforming memory into narrative and writing it up, making it all up. These inventions and romancings would in turn trigger deeper self-inspections and higher speculations. In a letter to Nathaniel Hawthorne, written five years after he began Typee and while he was composing Moby-Dick, Melville described what his writing process was like: he felt like the bulb or bud of a flower and found that from the moment he had begun writing at the age of twenty-five, scarcely three weeks had passed “that I have not unfolded within myself.” Each act of writing is the unfolding of the layerings of the self. For Melville, the unfoldings had not ceased with Moby-Dick, nor would they cease until Melville’s death in 1891. But what the readers of the letter to Hawthorne often fail to note is that Melville’s lifelong unfolding began, in his “twenty-fifth year,” with the revising of Typee.

In order to witness Melville’s revision process, we need to know where the manuscript fits in the stages of composition associated with Typee, and what kinds or phases of writing it exhibits.

Stages of Composition

The Growth of Typee

Some assume that Melville wrote Typee with an array of source books in front of him and composed each chapter pretty much in the order in which it appears.4 But manuscript evidence suggests otherwise. Most likely, Melville completed a shorter personal narrative to which he returned in discrete periods of revision, expanding the text with new chapters first based solely upon personal experience, and then based upon source books that he began to consult for the purposes of further expansion.5 That consultation of sources for factual information eventually aroused in Melville a desire to digress upon missionaries and Western interventionists, which swelled his text even more, and in directions that would lead to the expurgations of the American Revised edition. These broad developments are elucidated more fully in the following stages of composition. (See the Growth Chart [PDF, will open in a new window].)

Stage A: Performances and Rehearsals
After his three weeks in Taipivai and almost immediately upon his return to shipboard life, Melville was regaling his mates with anecdotes culled from his island experiences. He most certainly continued that practice for the benefit of friends and family upon his return home in October 1844. These oral performances were his rehearsal for the composition of Typee, and such polished episodes (in both manuscript and print) as the baked-baby feast and the fire-lighting scene ( chapters 12 and 14, respectively) are just a few of over a dozen likely candidates for written versions of these performed anecdotes. (See the Anecdote Chart [PDF, will open in a new window].) Also, during the fall and winter months of 1844–45, Melville may have prepared written outlines, notes, scribblings, even brief trial drafts, although none of these survives.
Stage I: “First Draught”
In its earliest complete version, Typee was probably a personal narrative of about eighteen chapters drawn entirely from the author’s personal experiences and anecdotes, including his shipboard life, his escape with Toby into the mountains and Taipivai, his growing attachment to Taipi culture and people (including a girl he first called Faaua), but also his apprehension over cannibalism and tattooing, and his escape. These sections (which are represented as Stage I, chapters 1–18; see the Growth Chart) include those chapters or parts of chapters presently numbered 2–14, 16–19, and 32–34. The Typee manuscript includes all but one page of the text of the three print chapters 12–14, which Melville originally numbered 10–12. Melville revised this document in various phases of writing up to the time he created his fair copy (Stage VI).
Stage II: Early Narratorial Expansions
Recognizing that he did not have enough material for a book-length publication, Melville probably made two manipulations to augment tensions within his narrative. He split up and expanded upon a hypothetical manuscript chapter 13 to create what would become print chapters 16 and 17, and did the same to a hypothetical manuscript chapter 15 to create print chapters 19 and 30. The effect of this splitting and augmentation of chapters, which may be one cause of the manuscript’s loose leaves, is to allow Tommo to elaborate more upon island harmony before he encounters the crucial problem of tattooing elaborated in print chapter 30. This gave Melville a total of twenty chapters.
Stage III: Early Non-Narratorial Expansions
Still less than two-thirds of the way toward his finished book, Melville probably added the two digressive chapters on breadfruit (print chapter 15) and on “Fishing Parties” and the eating of raw fish (print chapter 28). The insertion of the breadfruit chapter is another possible cause of the loose leaves in the manuscript. This gave Melville a total of twenty-two chapters.
Stage IV: Source Expansions I
Having added four chapters to the latter half of the book, Melville probably then turned to his opening chapters, which in their original form most likely comprised major portions of print chapters 2, 4, and 5. He augmented this section by two chapters’ worth of materials that help to establish his voice and add factual detail drawn from certain source books, in particular Charles Stewart’s Visit to the South Seas (see Melville Unfolding, chapters 13 and 14). However, in the process he began to develop a critical distance from Stewart’s missionary ideology and Western imperialism in general. A number of hypotheses can be developed to explain Melville’s possible sequence of revisions at this stage (see Melville Unfolding, chapter 9). At this stage, Typee grew to a chapter count of twenty-four.
Stage V: Source Expansions II
Reviewing Stewart inspired Melville to read more source books, in particular David Porter’s Journal of a Cruise. Returning to the latter half of his book, he would have added the Feast of Calabashes section (print chapters 22–26) and may have also augmented print chapter 30 on tattooing, by adding a section on taboo and an anecdote concerning Captain Vangs’s violations of taboo while bird hunting in the glen of Tior. These new and augmented chapters (echoing Tommo’s Stewart-related anecdote of Tior in chapter 4 added in Stage IV) contain a blending of direct source appropriations and Melville’s own anecdotes that focus mostly on island customs, including Porter-related materials on religion, social rank, governance, and burial (see Melville Unfolding, chapter 15). At this point, Melville may have also submitted his first draft to his brother Gansevoort for a vetting that resulted in a set of suggested revisions in pencil (discussed in Melville Unfolding, chapter 12). At this point, Melville had twenty-nine chapters.
Stage VI: Final Filler and Fair Copy
With a book-length work in hand, Melville probably submitted his Typee “draught” for fair-copying. During this period, Melville, family members, and other editors contributed a round of revisions that do not appear in the manuscript but must have occurred because Melville’s text in manuscript varies significantly from the text that appears in print. At the same time, Melville may have added two Porteresque chapters of “filler”: the chapter on island natural history (print chapter 29) and a chapter on additional island customs (print chapter 31). These polishings and filler may have been made before or after he submitted his fair copy (which has not been located) to Harper and Brothers, who rejected the book in May 1845. The chapter count now reaches thirty-one.
Stage VII: Source Expansions III
In July 1845, Melville entrusted his book to his brother Gansevoort, who successfully peddled the work to publisher John Murray. In the summer and fall months, Melville composed and sent by steamer (either on his own initiative, as Hershel Parker believes, or upon Murray’s request, according to Leon Howard) three new chapters: A “history of a day” in Typee (print chapter 20) including a salacious section on native dance (later removed at Murray’s request and added to Omoo), a chapter on island ruins (print chapter 21), and another new chapter on “The Social Condition and General Character of the Typees” ( chapter 27). The latter two draw heavily upon Porter. This brought Melville’s book to its present thirty-four chapters.

The remaining four stages involve the printed versions of Typee that appeared during or soon after Melville’s lifetime.6

Stage VIII: Murray’s British Edition
On Herman’s behalf, Gansevoort squired the fair-copy text of Typee through page proofs of this first edition published in February 1846; thus, he as well as Murray’s copy editor Henry Milton and Murray himself could have contributed changes to any of the more than five hundred additional revision sites created as early as the fair-copy Stage VI. Murray published this version throughout the rest of the century. (Gibbs’s 1850 pirated edition follows this version.)
Stage IX: Wiley’s American Edition
After reading sections of Typee in page proofs aloud to Washington Irving in London and with Irving’s enthusiastic recommendation, Gansevoort gained the acceptance of his brother’s book in the London office of the American publishing firm of Wiley and Putnam. Prepared from pages from the British edition, this version appeared in March 1846, and contains numerous accidental variations (misspellings and typos) and a handful of substantive changes, including line-length expurgations made by John Wiley.
Stage X: Wiley’s Revised American Edition
This revision of the American edition, appearing in August 1846, involves 129 expurgations ranging from single words to an entire chapter and the appendix. It also includes changes made on Melville’s behalf as well as the addition of “The Story of Toby.” Harpers took over publication in 1849. (Routledge’s 1850 pirated edition follows this version.)
Stage XI: Stedman’s Edition
Melville’s literary executor Arthur Stedman based his 1892 United States Book Company edition on the British version, but made numerous changes of his own, including a handful of requests made by Melville before his death.

While Melville’s participation in revising his working draft manuscript is abundantly evident in the numerous deletions and insertions in his hand on that document, we have more difficulty discerning whether the variants appearing in any of the subsequent print versions of Typee were initiated by Melville. A number of “readers” (including his brother Gansevoort, his British and American editors, and later, perhaps, even his wife) would have had the opportunity to revise or correct the text either on Melville’s behalf or on their own initiative. At the same time, meaningful variants in print might also be typos, the kind of accidental error and corruption endemic to the printing process. The problems to confront, then, are how to distinguish among correction, revision, and corruption, and how to assign responsibility for them. Because definitive answers are not likely to be found, the best alternative solutions to a given textual problem are shaped by discourses that take into account the probabilities and necessities inherent in the relationships among manuscript and print documents.

To begin with, the Typee working draft manuscript is not the fair-copy manuscript Melville sent to England with Gansevoort. Moreover, that fair copy, which Herman subsequently expanded, was also copyedited by Henry Milton and further revised by Gansevoort (NN Typee 282), so that its final text was significantly altered before it was submitted for publication on January 5, 1846. Proofs began to arrive at the end of the month, and Gansevoort would have read the British proof sheets against the augmented fair-copy document; however, neither fair copy nor proofs survive, so we can only guess at what changes to the British edition were made by whom. Murray’s printer was able to present Gansevoort with two complete sets of corrected proofs on February 3, at which time Gansevoort delivered one set to Putnam, and it is from these proofs that the American text would be set in New York. Gansevoort had only enough time to perform a cursory proofreading of the corrected proofs; therefore, it is highly unlikely that he (or his brother for that matter) had any additional input into the first American edition text (309). And if Gansevoort had been able to correct what was sent across the Atlantic, those corrections would have been made against the augmented and copyedited fair copy that no longer exists. The major changes to the first American edition—four sentence-long expurgations—were made by Melville’s American publisher John Wiley, who later recommended that the author himself make or agree to the additional and far more extensive expurgations found in the American revised edition, which appeared in the summer of 1846. Melville also added material including small changes (like “literal” to “liberal”) and large (“The Story of Toby”).

Because the base version of this present edition covers only the range of text represented in the three-chapter fragment of the Typee manuscript, the edition reports only those print-text variants found in the corresponding texts of the three 1846 print editions. Since Gansevoort had little opportunity and his brother even less to revise or correct the text destined to become the first American edition, only a handful of the fifteen or so variants that fall within the three-chapter range (such as Melville’s “L-word” or other “oscillating variants” like luxurious/luxuriant and inefficacy/inefficiency) are discussed in the edition’s revision narratives. Minor variants that may be typographical—such as annoyance/annoyances and parent/parents—are listed in NN Typee 350. However, changes found in the English and American Revised editions, in which several individuals played identifiable roles, are treated fully here. Of course, with the working draft manuscript itself, there is no doubt that (with the exception of certain marginal pencilings) the revisions on the page and in different Phases of Writing are Melville’s and that they represent an intermediate stage of composition. [rev. 2009]

Melville’s “First Draught”: An Intermediate Stage of Composition

There is no doubt that the Typee manuscript fragment represents an early period in Melville’s highly creative compositional process, but though the inscription on Melville’s cover to the document states that it is the “First Draught of ‘Typee’—after which much was added & altered,” it probably does not represent the very first penning Melville made. Melville’s cover caption is certainly accurate in the one sense that “much was added and altered” to the final reading text of the manuscript, but the material evidence also suggests that Melville’s designation of “First” should not be taken to mean earlier, now lost, “pre-writing” or “rough draft” documents—notes, outlines, even full-length passages of prose— that would constitute the actual first draftings of Melville’s personal narrative, including trial versions of his oral presentations.

Properly speaking, the document represents a set of intermediate stages of composition between Melville’s actual first “draughts” (Stage A) and his fair copy (Stage VI), neither of which has been found. On this intermediate document, we find various distinct phases of writing, including passages of copying from the earlier materials as well as passages of new composition.

Two factors argue for this intermediate status. First, while the document is heavily revised, it does not bear the markings on all pages of the kind of incessant revising—false starts, interrupted thoughts or sentences, reworked phrases, and transposed passages—that we find in other, far rougher, working drafts such as The Confidence-Man fragments or certain poetry manuscripts.7 And while the handwriting is generally difficult to read or obscured by cancellations and insertions (suggesting that Melville was inventing, not copying), we also find pages that are far more legible than others, suggesting that Melville was attempting to put his handwriting on good behavior for the purposes of creating fair copy. In all likelihood, Melville was intending to make a concerted effort at creating a polished version of certain rudimentary materials already rehearsed, composed, and now lost to us, but he was also improvising new material as he wrote.

This impression is supported by word counts of each manuscript page. In all, the 32 pages of writing amount to 12,103 words, including all cancellations and insertions. If we list the pages in order of word count with the lowest and highest yields per page being 326 words (MS p. 34) and 453 words (MS p. 17), respectively, we find that the median word count of 378 words per page is almost the same as the average of 376 words per page (see Table 1). (The median represents the point at which there are an equal number of pages with lower word counts as there are pages with higher.) Taking the median and average word counts to represent the typical amount of revision per manuscript page, we can readily distinguish those pages with higher counts as having more revision relative to those with lower counts and less revision. If Melville were starting entirely from scratch and simply pre-writing in this document, we would expect a fairly uniform degree of revision on each page, and the average word count per page would be significantly higher and the highest and lowest word count leaves would range closer to the median. Instead, we find significantly more revision in the first twenty pages, which appear in the stitched booklet, and less revision in ten of the loose-leaf pages (see Table 2). Moreover, throughout the entire document, pages of comparatively less revision are interspersed with pages of more revision. In short, the intensity of revision seems to come in waves, beginning with a high tide at the beginning of the document that recedes toward the end. This distribution supports the idea that Melville would copy for a spell, then revise as he copied or compose afresh, then return to his copying, and so on. The word counts also show that, despite their rough physical appearance, the loose-leaf pages have comparatively less revision and seem to be in a more polished state than the bound booklet leaves.

MS Page No Word Count
34 326
6 327
21 344
10 345
28 348
26 350
27 355
29 361
31 363
1 367
22 368
14 369
16 373
2 374
33 374
20 375
9 377
18 383
3 384
19 386
4 389
15 389
13 391
11 392
25 393
8 397
32 400
5 406
7 409
30 415
12 420
17 453
23 Missing
24 Missing

Table 1. Ascending word counts. Here, manuscript pages are arranged by ascending word count; the range containing both mean and median counts (376 and 378, respectively) is shaded. Higher word counts (listed here below the shaded range) indicate a relatively higher amount of revision. (Pages 23–24 are missing.)

MS Page No Word Count
1 367
2 374
3 384
4 389
5 406
6 327
7 409
8 397
9 377
10 345
11 392
12 420
13 391
14 369
15 389
16 373
17 453
18 383
19 386
20 375
21 344
22 368
23 Missing
24 Missing
25 393
26 350
27 355
28 348
29 361
30 415
31 363
32 400
33 374
34 326

Table 2. Sequential word counts. With manuscript pages displayed in proper sequence and the comparatively higher word counts shaded in, the dispersal of the more heavily revised pages throughout the entire manuscript can be seen. The “booklet” pages (1–20) bear evidence of more revision than the loose-leaf pages (21–34). (Pages 23–24 are missing.)

Besides having been torn from a booklet, some of the loose leaves have tears in the upper and lower edges that obliterate bits of text. Chances are that a half dozen of these tears mark the spots where insertion slips (now lost) were pinned to the loose leaf and then torn off at some later date (probably during the creation of the fair copy), thus taking fragments of the manuscript with them.

A second factor arguing for the manuscript as an intermediate document is the existence of the six loose leaves (actually seven, if we count the lost loose Leaf 12). These pages indicate that at various stages of composition, Melville moved chunks of text by tearing already composed leaves away from his booklets either to reposition chapters or to make room for new materials. Common sense suggests that Melville would not have been so bold as to commit his first pennings to neatly bound booklets of good quality paper; instead he would have composed initial notes, outlines, and drafts on single sheets of cheaper foolscap.

Of course, we should not presume the young writer had common sense or cared at this point in his career about cost and convenience. In fact, never having written a book before, he may not have anticipated the kind of paper stock his new writing adventure would require. The use of neat composition booklets, only slightly more refined than the kind he and his siblings had used in school, may have been ill-advised for the composition of a novel from scratch, but not a prohibition for the novice. Perhaps, too, Melville did not feel he needed to do much pre-writing or rough-drafting. After all, with his oral rehearsals fresh in mind, he might have felt free to write his first book from scratch in neatly stitched booklets. While this is true, after only a few days of inscribing even the best-rehearsed bits of narrative, Melville might well have thought twice about composing from the very start in booklets. If so, he would have likely found himself tearing away and discarding leaves of botched writing from his booklets. Revision, he would quickly learn, was inevitable.

Phases of Writing

Keeping the larger stages of composition in mind, we can turn now to the more minute operations of Melville’s creative process, or what I call his phases of writing.8 These six phases consist of:

Although these phases seem to follow logically from one to the other in the sequence indicated here, the manuscript shows that at any given revision site Melville may have combined any of these phases in various ways. For instance, he may have copied for a stretch, making minor corrections as he copied, but then begun to compose fresh material as he copied and thus saved any proofreading for later, when he would revise once more as he proofed. Or he may have stopped periodically while copying or composing to proofread and revise before continuing with more copying or composition. Each revision site offers up its own set of sequencing problems, and since Melville did not use different inks when he wrote—which might have allowed us to distinguish, let’s say, revisions made while composing in one ink from revisions made in another while proofreading—we cannot readily discern in a given revision site whether a particular revision derives from the one phase or the other. Thus, an important function of the revision narrative, and indeed its principal challenge, is to articulate the more plausible sequencings of the phases of writing evident at the revision site in question. As we shall see, too, the meanings we can construct out of a revision site depend entirely upon the sequencing we find there.

Often, the only grounds for distinguishing among Melville’s phases of writing are the dictates of English grammar, syntax, and sentence structure. Other clues are the physical positioning of canceled words on, above, or below the baseline (indicating immediate, subsequent, and later revisions); or the fact that one insertion is squeezed between two other inserted words (indicating its subsequent addition); or the fact that we find a canceled word within a larger phrase that has also been canceled (indicating that the shorter cancellation preceded the longer). These and other features help us distinguish certain phases of writing and speculate on their probable sequencing in a given revision site. Only the insertion slips offer a higher degree of certainty about the sequence of a revision, for their existence on separate slips of paper attached to particular manuscript leaves clearly shows that this kind of proofreading revision or fresh composition occurred, in most cases, late in the manuscript revision process, well after the manuscript leaves on which the slips were pinned had been completed.

The six compositional phases are more than just conjectural. In copying, Melville would have had to his left any of his notes or rough-draft materials and to his right a set of blank twenty-page composition booklets. As he copied from his pre-writing materials into his booklets, he soon enough caught himself miswriting or skipping ahead, and therefore he corrected himself immediately on the baseline of the text he was writing. He also found himself rethinking certain wordings, and therefore he paused, reread what he had copied, and revised. He also found himself inventing new material, or squeezing in insertions between words and lines or down margins. In this kind of revision, he departed from the pre-writing materials to incorporate as much as a page or two of fresh composition. But more likely as not he also returned to his notes or outline to copy again. After some time, he would also proofread what he had written, and in proofing he found more reason to revise certain passages, some so severely that he rewrote their final texts on insertion slips. Melville may have proofed periodically throughout the process, not letting a paragraph go by without rereading it. But hard evidence shows that he also revised with all or most of the fully completed manuscript in hand, with the repositioning of certain loose leaves, and during a discernible moment in which he responded in ink to his brother’s penciled suggestions for revision. Melville also manipulated the text at some point when he added paragraph marks.9

Probably the most difficult issue in reading the manuscript is discerning copying from fresh composition. One is tempted to speculate that the relative neatness of a certain page suggests areas of copying, but let’s recall that Melville had rehearsed his “yarns” repeatedly, and there is every likelihood that he could have transposed those rehearsed tales directly onto paper without much revision simply because the words were already well-established in his mind. On the other hand, speaking is not writing and drafting words is tougher than spinning fireside yarns, so one cannot rely too heavily upon the idea of rehearsed tales yielding clean pages. Cleanliness is no certain measure of either copying or fresh composition.

Nor should one depend too heavily upon what scholars have, up until the discovery of the Typee manuscript fragment, determined was Melville's “typical” method of composition and manuscript preparation. We know that later in his life Melville typically relied upon his wife Elizabeth or his sister Augusta to make fair copies from his rough drafts, which he in turn revised, had fair-copied again, and revised again until his material was ready for a final fair-copying and submission to the publisher. Aside from the faint pencilings we may attribute to Gansevoort, there is no evidence of anyone other than Melville performing the six phases listed here. And one might thus argue that since he “typically” left his copying for others to do, the absence of other handwriting in manuscript suggests Melville was engaging more in fresh composition than copying.

But again, the Melville who composed Typee was a freshman when it came to the processes of literary creation. Except for two juvenile publications, he had never, to our knowledge, attempted so involved a literary work; nor did he have at the time a wife who would dutifully fair-copy for him. Chances are he felt obliged to “do it all himself” at least until the very end, when he may have used the services of an amanuensis or willing sister at home in Lansingburgh to transform his difficult handwriting into something readable. The absence of other hands (except for Gansevoort’s faint marginal pencilings), then, does not necessarily have any special relevance here in resolving the question of copying versus composing . It does indicate, however, that Melville was in full charge of this particular document. In short, the Typee manuscript does not necessarily conform to Melville’s putatively typical practice. Moreover, since the document predates the evidence previously used to determine that practice, and since the Typee manuscript represents the earliest extant evidence of his fiction writing, we are obliged to consider it coming from a point in Melville's career in which nothing had yet become “typical.”

False Starts

Evidence of copying or fresh composition is best found in seemingly insignificant details. Take the “false start,” for instance. This is a word that the author considers in mind, begins to write on paper but never completes, and cancels in mid-execution. A kind of sprite dancing between mental and physical states, the false start may be a word, partial word, single letter, or even a half letter such as an uncrossed t. If it can be decoded—that is, if a full, intended wording can be determined from its partial execution—a false start can provide a glimpse at the author’s more volatile mental twists and turns: the decisions and indecisions at the moment of creation. On the one hand, it might indicate that at a certain point the writer, rather than simply copying his words mechanically from an earlier draft, momentarily considered alternative words or whole new sentence structures as he copied. (A partially inscribed w for the word “which” may be all we need to know that Melville had considered converting an independent clause to a dependent clause.) On the other hand, the false start might indicate the interruption of a flow of freshly composed words for the elaboration of a newly invented image; it may, as well, indicate the frustration of trying to get an image out, as with “blo{ody} recent sacrifice,” a false start discussed in “Editing a Fluid Text.”

Some false starts, however, help identify fresh composition. A set of revision sites (RS10ms145-146) shows Melville interrupting one genetic moment to elaborate a new point he has generated as he writes. In speaking of what he first calls a “wild” but then “musical recitative” of the natives, Melville came to something of a block, as the following revision sequence shows. He first wrote:

1. the musical recitative, which for all I knew might

Given the phrasing that Melville finally used later on in his next paragraph, he probably intended to complete this false start with something like “which for all I knew might {have been taken for a throng of the devotees of Ceres}.” But Melville interrupts this thought at the word “might” and cancels “for all I knew might.” With the idea of first evoking the sound of the natives’ singing (rather than the classical reference to Ceres), he then considers something like

2. the musical recitative, which […] they s{ang}

but interrupts himself again and cancels the false start “s” before he can complete the word “sang.” He then tries his third and final option:

3. the musical recitative, which […] with various alternations they continued until we arrived at the place of our destination.

Here, the full range of revisions strongly argues that Melville is making things up as he writes. With the false start “they s{ang},” Melville’s first impulse was to change the structure of his nested clauses; that is, to connect his “which for all I knew might” clause to a simpler “which they sang” clause. But the interruption of “which they sang” suggests he wanted some intervening phrase between “which” and “they” to replace the rhythms of “for all I knew.” He chose to add “with various alternations,” and continued on, going back to “they” to add “continued” in place of the partially inscribed “s{ang}.” At this point, Melville then composed his next new paragraph describing the “picturesque procession” of chanting natives, which delivers us to Melville’s usage of “a throng of the devotees of Ceres.” In short, the false start “they s{ang}” is part of a process of fresh composition in which Melville first conceives then delays the image of devotees in order to record, on the spot, newly invented phrasings concerning the “picturesque procession.” The relevance of the revision sequence is that we can “see” the sudden appearance of a classical allusion and how it triggers related thoughts. Moreover, the fact that Melville or an editor cut the Ceres image from the first British edition (see RS10e119) suggests that this aspect of the crafting of Tommo’s voice surely carried with it a certain rhetorical anxiety.

False starts can be determinants of fresh composition or revision occurring during copying. No less concrete but just as hard to pin down in time are the myriad other cancellations and insertions that, despite our difficulty in sequencing them, may be associated with specific versions of the manuscript. A canceled word, for instance, that is replaced by a word immediately following it on the baseline generally indicates that the revision occurred spontaneously while Melville was copying out or composing down an empty line, and this distinguishes that revision from later revisions typically inserted above the baseline when Melville in a different frame of mind returned to his fully inscribed lines to proofread. Immediate baseline revisions imply fresh composition or invention and may be linked to what I call Melville’s transcription version of Typee. However, insertions replacing canceled words and inserted material without cancellations generally indicate revisions made in a later proofreading phase and may be features of the transformation and translation versions. All three versions are discussed next in “Versions of Typee.”

Versions of Typee: A Narrative of Revision

Given its classification as an intermediate-stage composition exhibiting several phases of writing, the revision text of Typee does not exist in several distinct, physically separable versions; instead, it consists of a layering of “inferred versions,” or modes of revision. To distinguish these versions, we need first to link them to a broader understanding of the interrelationship of the manuscript’s various revision sites. That is to say, we need to construct a revision narrative of Typee that allows us to connect its seemingly disparate, utterly localized sites into a global understanding of Melville’s revision process.

A Revision Narrative of Typee

A “revision narrative” is a detailed story of how a particular agent revises words in a particular sequence at a particular revision site. But to tell the story of the complex of strategies that can account for all or even a representative selection of those localized revision sites, we need a more comprehensive narrative of revision, one that elucidates the interpenetrations of writer and cultural discourse evident in a range of revision practices and intentions. In writing Typee, Melville’s unfoldings were deeply personal yet manifestly public; they were driven by sexual anxieties, aesthetic quandaries, political rage, and cultural bafflement. And his manuscript revisions are the material manifestation of these forces. Each revision site can be transcribed, and each site’s revision sequence can be presented as an array of texts, but as such they simply cannot be “read” in any coherent way until we begin to disclose their relation to other localized revision sites and sequences, and such relations cannot be addressed without our construction of a global narrative that can explain them in concert. Because the construction of this comprehensive narrative of revision articulates speculations on personal and social causes of textual evolution, it permits a more usable form of cultural analysis than found in the interpretation of single texts. For when we watch Melville unfold from revision to revision and version to version, we have a more concrete way to witness how this artist negotiated language and the world.

What follows is first a brief explication of the narrative of revision developed in Melville Unfolding and then an articulation of that narrative in terms of the three discernible versions I have called transcription, transformation, and translation.

Sexuality and Self-Colonization

The revision narrative developed in Melville Unfolding stresses politics, family, and sexuality, and follows a thesis both private and cultural through the stages of composition already discussed. In all likelihood, Melville began writing Typee with the idea of supplying readers with a personal narrative of his escape from the whaling ship Acushnet into the valley of Taipivai, home of a fierce, reputedly cannibal tribe, from which he soon enough also escaped. He began by relating his experiences in oral anecdotes rehearsed before fellow sailors, friends, and family (Stage A). Eventually, he came to write those anecdotes down, thus rendering a world from his past that he first called “Tipii” and a world of his imagination as well, eventually to be called Typee (Stage I). To be sure, the necessarily partial recollection of the past required an active imagination to fill in holes, if fullness of a narrative eventually to be called Typee was Melville’s professional goal. Inevitably, the writing process taxed both his memory and self-awareness. As he composed, he found himself laboring to recollect even mundane details: Was there one building on the hoolah hoolah ground, or two? Getting thoughts and words to correspond to memory was a primary concern; making his words vivid and communicative came as a natural adjunct to accuracy but invariably led to fictionalizations. In describing people and places, Melville found himself creating characters, setting scenes, and providing an imaginative cast to things if only to familiarize them for Western readers; thus, a gathering of islanders becomes an Irvingesque assemblage about the “the tavern door of a village” (a phrase he much later corrected to “door of a village tavern”). And these easy, perhaps automatic recastings of native life to fit the ideologies of amiable humor led to further, more complex cultural transformations: for example, natives bearing comestibles become in revision a procession of devotees to Ceres.

In due course, such cultural manipulations led to still deeper and more personal self-transformations. The twenty-five-year-old was writing in order to discover what it was the twenty-two-year-old had experienced, and what it all meant then, and what it was all beginning to mean now. Former island acquaintances and lovers were now characters playing roles in the dramatic unfolding of the writer’s being. In revision, Toby, Fayaway, Kory-Kory, and Marnoo began to serve as triggers of the young Melville’s sexual anxiety. Each became a kind of colonial extension of Melville’s previously unquestioned, imperial self. And just as an emergent power tests its national strength by sending its pirates, religionists, speculators, and conquistadors abroad to colonize new worlds, Melville used his pen to venture into new psychological territories. And like other colonizers, who, expecting to dominate, nevertheless find themselves changed, assimilated, and dominated by cultural, social, and sexual doubts triggered by the new world conquered, so, too, did Melville find that his self-colonizations—portraits of himself as lover of mother, brother, and other; and lover of women, and of men—challenged his sense of self and startled him toward a retrenchment of his sexual identity as safely heterosexual and marriageable.

Ironically, though, Melville’s reactionary self-colonizing revisions ran counter to a more liberal pattern of expansion. Once the writer had “finished” with his first working draft of Typee and discovered he had not quite enough material to make a book, he taxed his memory further, this time elaborating more personal anecdotes and reflections upon the islands and native life, which he inserted in portions sizable enough to make him tear pages out of his manuscript booklets to make room for the expansions and to force him to renumber chapters (Stages II–III). Requiring more material, he resorted to source books by Charles Stewart and David Porter and appropriated from them passages, sometimes word for word (Stages IV–VII). But the sources he plagiarized also angered Melville. These avatars of imperialism, despite their liberal attraction to, even acceptance of, Polynesian life, displayed the intolerance, arrogance, and brutality of Western power, and they became insufferable to the writer. In response, Melville expanded his narrative first by adding digressions against missionaries like Stewart (IV) and then against power merchants like Porter (V–VII). The digressions proved fateful; they were largely cut in the American revised edition (Stage X). But the deeper irony of this stage of revision is that while Melville was revising his text, on the one hand, to assert his personal, imperial dominance over the various colonized sexual identities he projected onto Typee, he was also revising, on the other hand, so as to castigate imperialism in the Pacific, and in particular his imperialistic predecessors.

A final element of this global narrative concerns Melville’s family. From the beginning, he had the support of mother, brothers, and sisters in his writing project, but he also had his future wife and in-laws in mind. Manuscript evidence indicates that perhaps as early as Stage V, Melville submitted the work to his brother Gansevoort, who penciled in suggested revisions of various passages, some of a sexual nature. Later, Gansevoort played a larger role in helping to place Typee with John Murray and making revisions in proof (Stage VIII). Melville’s future family also had a presence in the making of the book. Typee is dedicated first “affectionately” to his father’s best friend Lemuel Shaw, and then “gratefully” to the same man, his soon to be father-in-law. Quite possibly, several of Melville’s own changes to John Wiley’s expurgated edition (Stage X) suggest his toning down of randy implications out of respect for the sensitivities of the author’s future wife, and women readers in general. These revisions in a familial context give an added dimension to the more private political and sexual revisions manifested in earlier stages of composition.

Three Versions: Transcription, Transformation, and Translation

This global revision narrative of Typee also manifests itself in three modes of revision, which for the purposes of this edition shall stand as three inferred versions of Typee. These versions are transcription, transformation, and translation.

In his first version of Typee, Melville set out to transcribe events from memory, but this process inevitably led to (indeed bled into) acts of revision in which Melville essentially transformed remembered events into dramatized scenes. Thus, what began as a personal narrative version (transcription) evolved into something approaching a fiction or romance version (transformation), and one that would prove to be something of an “arrested romance.” But even before going into print, and especially with the publication of the revised edition, Melville reconceived his personal narrative/romance as a text more directly engaging readers. In this version, Melville essentially translated his radical critique of imperialism into an idiom most readers would be able to accept. This kind of negotiation constitutes a third version (translation) of Typee that involves both manuscript manipulations as well as later print revisions and expurgations.

While distinct evidence of Melville’s transcription, transformation, and translation is located on the “physical version” of the Typee manuscript (that is, the document itself), these modes of revision are, of course, “inferred versions,” or hypothetical designations useful in distinguishing different aspects of the revisionary process. They cannot be seen as rigid separations; indeed, a particular physical revision site might be the locus for two or even all three modes. That is, an early revision correcting an inaccuracy of memory might also involve later, artful transformations of a character or scene as well; and a late revision at the same site might also transcribe remembered events more clearly or add to the narrative’s artistic unity. Despite such obvious overlapping, the three versions correspond to the larger movements of Melville’s growing reconception of Typee. A more detailed understanding of Melville’s revisions through these versions can be achieved when we attend to various kinds of narrative expansions that are evident on the physical document itself.

Next: Scenes of Revision
Previous: Navigating the Typee Manuscript

Notes

1. Herman Melville, Correspondence, ed. Lynn Horth, Harrison Hayford, Hershel Parker, and G. Thomas Tanselle (Evanston and Chicago: Northwestern University Press and The Newberry Library, 1993), p. 193.

2. Review in the London Critic 3 (7 March 1846): 219–22; rpt. in Brian Higgins and Hershel Parker, eds., Herman Melville: The Contemporary Reviews (Cambridge and New York: Cambridge University Press, 1995), p. 9; NN Typee, “Historical Note,” p. 298.

3. Review in the New Haven New Englander 4 (July 1846): 449–50; rpt. in Higgins and Parker, Contemporary Reviews, p. 51.

4. Laurie Robertson-Lorant flatly assumes that Melville set up shop in his mother’s attic in Lansingburgh and surrounded himself with source books to compose Typee (Melville: A Biography [New York: Clarkson Potter, 1996], pp. 135–36). In Herman Melville: A Biography (Baltimore: Johns Hopkins University Press, 1996), Hershel Parker places Melville in New York City but more deftly asserts that Melville was sufficiently well-versed in the available Marquesan source books (Langsdorf and others) to be “confident, from the start of his work on Typee, that he would not have to rely on his own unaided memory” (p. 357).

5. This is in fact a pattern Melville followed throughout his writing career and is the assumed pattern scholars have used to account for the composition of other works, in particular Mardi, Moby-Dick, The Confidence-Man, and Billy Budd. See Watson Branch, “The Etiology of Melville’s Mardi,Philological Quarterly 64 (Summer 1985): 317–36; Watson Branch, “The Genesis, Composition, and Structure of The Confidence-Man,Nineteenth-Century Fiction 7 (March 1973): 424–48; and Harrison Hayford and Merton M. Sealts, Jr., eds., Billy Budd, Sailor (Chicago: University of Chicago Press, 1962). Several studies theorize on the genesis of Moby-Dick: George R. Stewart, “The Two Moby-Dicks,” American Literature 25 (January 1954): 417–48; James Barbour, “The Composition of Moby-Dick,American Literature 47 (November 1975): 343–60; and Harrison Hayford, “Unnecessary Duplicates: A Key to the Writing of Moby-Dick,” in New Perspectives on Melville, ed. Faith Pullin (Kent, Ohio, and Edinburgh: Kent State University Press and Edinburgh University Press, 1978), pp. 128–61.

6. For further details, see NN Typee, pp. 306–14, from which these stages are drawn.

7. Readers can access these documents in person at Harvard University’s Houghton Library, or in photo-reproductions in the NN The Confidence-Man as well as in my fluid-text edition of one fragment in the Random House edition of The Confidence-Man (2003). The NN edition of the published poetry (with manuscript transcriptions) and the NN edition of the late manuscripts (including the unpublished poetry and Billy Budd) are forthcoming. My fluid-text editions of selected poems in manuscript appear in Melville’s Tales, Poems, and Other Writings (Random House, 2001).

8. I call these kinds of writing “phases” in order to distinguish them from stages of composition and modes of revision, or versions, both of which involve a specific time sequencing. That is, stage III follows II, and the translation version follows transformation; however, a proofreading phase can occur at any time in the process, after initial composition or much later after a series of revisions. The word “phase” allows for this useful distinction because it implies patterns of writing behavior within a creative process but does not, in this case, insist upon a determinate sequencing.

9. Parker believes “Melville inserted clear paragraph symbols as he went” (Herman Melville, p. 361), but the squeezing of some paragraph symbols beside insertions suggests that Melville added the marks after the manuscript was completed.