[begin page 1]

Between 1870 and 1905 Mark Twain (Samuel L. Clemens) tried repeatedly, and at long intervals, to write (or dictate) his autobiography, always shelving the manuscript before he had made much progress. By 1905 he had accumulated some thirty or forty of these false starts—manuscripts that were essentially experiments, drafts of episodes and chapters; many of these have survived in the Mark Twain Papers and two other libraries. To some of these manuscripts he went so far as to assign chapter numbers that placed them early or late in a narrative which he never filled in, let alone completed. None dealt with more than brief snatches of his life story.

He broke this pattern in January 1906 when he began almost daily dictations to a stenographer. He soon decided that these Autobiographical Dictations would form the bulk of what he would call the Autobiography of Mark Twain. Within a few months he reviewed his accumulation of false starts and decided which to incorporate into the newer dictation series and which to leave unpublished. By the time he had created more than two hundred and fifty of these almost daily dictations (and written a final chapter in December 1909, about the recent death of his daughter Jean), he had compiled more than half a million words. He declared the work done, but insisted that it should not be published in its entirety until a hundred years after his death, which occurred less than four months later, on 21 April 1910.

This belated success with a project that had resisted completion for thirty-five years can be traced to two new conditions. First, he had at last found a skilled stenographer who was also a responsive audience—Josephine S. Hobby—which encouraged him to embrace dictation as the method of composition, something he had experimented with as early as 1885. Second, and just as important, dictating the text made it easier to follow a style of composition he had been drifting toward for at least twenty years. As he put it in June 1906, he had finally seen that the “right way to do an Autobiography” was to “start it at no particular time of your life; wander at your free will all over your life; talk only about the thing which interests you for the moment; drop it the moment its interest threatens to pale, and turn your talk upon the new and more interesting thing that has intruded itself into your mind meantime.”1

Combining dictation and discursiveness in this bold way was unexpectedly liberating, in large part because it produced not a conventional narrative marching inexorably toward the grave, but rather a series of spontaneous recollections and comments on the present as well as the past, arranged simply in the order of their creation. The problem of method had been solved. It was also liberating to insist on posthumous publication, but that idea had been around from [begin page 2] the start and was closely tied to Clemens’s ambition to tell the whole truth, without reservation. As he explained to an interviewer in 1899: “A book that is not to be published for a century gives the writer a freedom which he could secure in no other way. In these conditions you can draw a man without prejudice exactly as you knew him and yet have no fear of hurting his feelings or those of his sons or grandsons.” Posthumous publication was also supposed to make it easier for Clemens to confess even shameful parts of his own story, but that goal proved illusory. In that same 1899 interview he admitted that a “man cannot tell the whole truth about himself, even if convinced that what he wrote would never be seen by others.”2

But if delaying publication failed to make him into a confessional autobiographer, it did free him to express unconventional thoughts about religion, politics, and the damned human race, without fear of ostracism. In January 1908 he recalled that he had long had “the common habit, in private conversation with friends, of revealing every private opinion I possessed relating to religion, politics, and men”—adding that he would “never dream of printing one of them.”3 The need to defer publication of subversive ideas seemed obvious to him. “We suppress an unpopular opinion because we cannot afford the bitter cost of putting it forth,” he wrote in 1905. “None of us likes to be hated, none of us likes to be shunned.”4 So having the freedom to speak his mind (if not confess his sins) was still ample justification for delaying publication until after his death.

Seven months after he began the Autobiographical Dictations in 1906, however, Clemens did permit—indeed actively pursued—partial publication of what he had so far accumulated. He supervised the preparation of some twenty-five short extracts from his autobiographical manuscripts and dictations for publication in the North American Review, each selection deliberately tamed for that time and audience, and each prefaced by a notice: “No part of the autobiography will be published in book form during the lifetime of the author.”5 But not long after Clemens died, his instruction to delay publication for a hundred years began to be ignored—first in 1924 by Albert Bigelow Paine, Mark Twain’s official biographer and first literary executor, then in 1940 by Paine’s successor, Bernard DeVoto, and most recently by Charles Neider in 1959.

Each of these editors undertook to publish only a part of the text, and none ventured to do so in the way that Clemens actually wanted it published. Paine began his two-volume edition with all but a handful of the manuscripts and dictations carried out before 1906, as well as several texts that were probably never part of those early experiments. He arranged all of them “in accordance with the author’s wish . . . in the order in which they were written, regardless of the chronology of events.”6 It now seems clear that Paine’s understanding of “the author’s [begin page 3] wish” was mistaken: Clemens never intended to include all those false starts, let alone in chronological order; he intended only the dictations begun in 1906 to be published that way. But having chosen this course, Paine then had space for only a relative handful of the dictations. And on top of that, he felt obliged to suppress or even alter certain passages without notice to the reader. He eventually acknowledged that he had published only about one-third of what he regarded as the whole text.7

DeVoto was critical of Paine’s acceptance of “the arrangement Mark Twain originally gave” the dictations, “interspersed as they were with trivialities, irrelevancies, newspaper clippings, and unimportant letters—disconnected and without plan.” Instead he chose to print only passages that Paine had left unpublished, drawn from “the typescript in which everything that Mark wanted in his memoirs had been brought together” (that is, the Autobiographical Dictations begun in 1906). DeVoto then arranged the selections by topic, “omitting trivialities and joining together things that belonged together.” And he said with great satisfaction that he had “modernized the punctuation by deleting thousands of commas and dashes, and probably should have deleted hundreds more.” He was confident that he had “given the book a more coherent plan than Mark Twain’s” and he was unapologetic about having “left out” what seemed to him “uninteresting.”8

Neider, too, was unhappy with Paine’s acceptance of Mark Twain’s plan to publish the autobiography “not in chronological order but in the sequence in which it was written and dictated. What an extraordinary idea! As though the stream of composition time were in some mysterious way more revealing than that of autobiographical time!”9 Neider had permission from the Mark Twain Estate to combine some thirty thousand words from the unpublished dictations with what Paine and DeVoto had already published. Like DeVoto, he omitted what he disliked, and was also obliged to exclude portions that Clara Clemens Samossoud (Clemens’s daughter, by then in her eighties) disapproved of publishing. He then (figuratively) cut apart and rearranged the texts he had selected so that they approximated a conventional, chronological narrative—exactly the kind of autobiography Mark Twain had rejected.

The result of these several editorial plans has been that no text of the Autobiography so far published is even remotely complete, much less completely authorial. It is therefore the goal of the present edition to publish the complete text as nearly as possible in the way Mark Twain intended it to be published after his death. That goal has only recently become attainable, for the simple reason that no one knew which parts of the great mass of autobiographical manuscripts and typescripts Mark Twain intended to include. In fact, the assumption had long [begin page 4] prevailed that Mark Twain did not decide what to put in and what to leave out—that he left the enormous and very complicated manuscript incomplete and unfinished.

That assumption was wrong. Although Mark Twain left no specific instructions (not even documentation for the instructions that Paine professed to follow), hidden within the approximately ten file feet of autobiographical documents are more than enough clues to show that he had in fact decided on the final form of the Autobiography, and which of the preliminary experiments were to be included and which omitted. This newly discovered and unexpected insight into his intentions is itself a story worth telling, and it is told for the first time in this introduction.

Three printed volumes are planned for this edition, which will also be published in full at Mark Twain Project Online ( MTPO ). Exhaustive documentation of all textual decisions will only be published online.10 This first volume begins with the extant manuscripts and dictations that must now be regarded as Clemens’s preliminary efforts to write the autobiography and that he reviewed and rejected (but did not destroy) in June 1906. They are arranged arbitrarily in the order of their date of composition, solely because Clemens himself never specified any order. Some of these texts he explicitly labeled “autobiography,” and some are judged to be part of his early experiments on other grounds, always explained in the brief headnotes that introduce them. We include those preliminary texts for which the evidence is reasonably strong, without asserting that there were no others.

The Autobiography of Mark Twain proper begins on p. 201 in this volume, starting with the several prefaces Clemens created in June 1906 to frame the preliminary writings he had selected as opening texts, followed by his almost daily Autobiographical Dictations from 9 January through the end of March 1906—all that will fit into this volume. The dictations are arranged in the chronological order of their creation because that is how Clemens instructed his editors to publish them. The remaining volumes in this edition will include all the dictations he created between April 1906 and October 1909, likewise arranged chronologically, the whole concluding with the “Closing Words of My Autobiography,” a manuscript about the death of his youngest daughter, Jean.

Autobiography as a literary form had a special fascination for Mark Twain. Long before he had given serious thought to writing his own, he had published both journalism and fiction that were, in the most straightforward way, autobiographical. From the earliest juvenilia in his brother’s Hannibal, Missouri, newspaper (1851–53) to his personal brand of journalism in Nevada and California (1862–66), he played endlessly with putting himself at the center of [begin page 5] what he wrote. Twenty years and nine books later, in October 1886, he acknowledged (and oversimplified) the result: “Yes, the truth is, my books are simply autobiographies. I do not know that there is an incident in them which sets itself forth as having occurred in my personal experience which did not so occur. If the incidents were dated, they could be strung together in their due order, & the result would be an autobiography.”11 He was thinking of his travel books and personal narratives—The Innocents Abroad, Roughing It, A Tramp Abroad, and Life on the Mississippi—the only books up to that point in which he set forth anything “as having occurred” in his own experience. To be sure he also made extensive fictional use of that experience. The factual basis of characters and situations in works like The Gilded Age, The Adventures of Tom Sawyer, and Adventures of Huckleberry Finn has been thoroughly documented, and the autobiographical content is obvious in dozens of shorter works like “The Private History of a Campaign That Failed” and “My First Lie and How I Got Out of It,” even when they are not entirely factual.12

More germane to Clemens’s thinking about his own autobiography is his interest in fictional autobiography—that is, fictions in the shape and form of an autobiography. Mark Twain’s (Burlesque) Autobiography was written in late 1870 and published in pamphlet form in March 1871. Mark Twain tells us that his own parents were “neither very poor nor conspicuously honest,” and that almost all of his ancestors were born to be hanged—and for the most part were hanged. An even briefer “burlesque” called simply “An Autobiography” appeared in the Aldine magazine in April 1871: “I was born November 30th, 1835. I continue to live, just the same.”13 The whole sketch takes fewer than two hundred words and pointedly leaves the reader as ignorant of the facts as before.

Burlesque implies familiarity with genuine autobiographies, despite what Clemens told William Dean Howells in 1877 (“I didn’t know there were any but old Franklin’s & Benvenuto Cellini’s”). Benjamin Franklin’s didactic bent made him a lifelong target of Mark Twain’s ridicule. But he thought Cellini’s autobiography the “most entertaining of books,” and he admired the daring frankness of Jean-Jacques Rousseau’s Confessions and Giovanni Giacomo Casanova’s Mémoires, as well as Samuel Pepys’s Diary, which Paine said was the book Clemens “read and quoted most.”14

In 1871 he proposed writing “an Autobiography of Old Parr, the gentleman who lived to be 153 years old,” but apparently he never did so.15 In the summer of 1876 he wrote four hundred pages of a work he was then calling “Huck Finn’s Autobiography.” And in March 1877, he told Howells he was writing such a work about his own older brother: “I began Orion’s autobiography yesterday & am charmed with the work. I have started him at 18, [begin page 6] printer’s apprentice, soft & sappy, full of fine intentions & shifting religions & not aware that he is a shining ass.” He assigned various real incidents of Orion’s life and aspects of his character to an apprentice named Bolivar, and wrote more than a hundred pages before abandoning the project.16

In 1880, Orion’s decision to write a real autobiography prompted Clemens to suggest that he instead “write two books which it has long been my purpose to write, but I judge they are so far down on my docket that I shan’t get to them in this life. I think the subjects are perfectly new. One is ‘The Autobiography of a Coward,’ & the other ‘Confessions of a Life that was a Failure.’ ” The object here was not burlesque, but rather a kind of thought experiment to test the difficulty of telling the whole truth in an autobiographical narrative—in this case, by shielding it behind a deliberate fiction.

Another version of this scheme Clemens said was more difficult, to “tell the story of an abject coward who is unconscious that he is a coward,” and to do the same for “an unsuccessful man.”

The inspiration for both of these ideas was obviously two autobiographies that Clemens admired.

[begin page 7] Clemens himself seems not to have attempted what he urged Orion to try, but it is obvious he was thinking about the challenge of writing with the perfect frankness he admired in these writers. The question of how fully he could tell the truth about himself, and especially to what extent he could confess what he regarded as his own shameful behavior, occupied him off and on throughout work on the Autobiography.

Clemens’s plan to write his own autobiography is more or less distinct from these fictional uses of the form. The first indication that he had such a plan survives only in the report of a conversation that took place when he was forty. Mrs. James T. Fields and her husband were visiting the Clemenses in Hartford. She recorded in her diary that at lunch, on 28 April 1876, Clemens

This remarkable statement shows that Clemens was already committed to several ideas that would govern the autobiography he worked on over the next thirty-five years. The notion is already present that publication must be posthumous, a requirement linked to the ambition to have “the whole tale told as truly as I can tell it,” without censoring himself or allowing others to do it for him. He also plans to publish selections from the narrative while still alive, withholding the rest “to leave behind him.” He will not limit himself “as to space,” but will be as digressive and discursive as he likes, even ignoring chronology when it suits him. These cardinal points are clearly interrelated: absolute truth telling would be made easier by knowing that his own death would precede publication, and discursiveness (quite apart from his natural preference for it) would help to disarm his own impulse toward self-censorship. But it would take another thirty years to actually apply these various ideas to a real autobiography.

Just a year or so later, sometime in 1877, Clemens seems actually to have begun writing, prompted (as he recalled in 1904) by a conversation with his good friend John Milton Hay. Hay “asked if I had begun to write my autobiography, and I said I hadn’t. He said that I ought to begin at once” (since the time to begin was at age forty, and Clemens was already forty-two).

In late November 1877 Clemens listed “My Autobiography” among other projects in his notebook, reminding himself to “Publish scraps from my Autobiography occasionally.” He did indeed write an eleven-page manuscript at this time which he intended as the first chapter of an autobiography—very likely the “beginning” that in 1904 he remembered having thrown away. He titled it merely “Chapter 1,” but it is commonly known as “Early Years in Florida, Missouri,” the title Paine assigned it.20 It begins, “I was born the 30th of November, 1835”—the same way Clemens began his Aldine burlesque in 1871—and it goes on to reminisce briefly about his early memories of childhood in that “almost invisible village of Florida, Monroe county, Missouri.” Like “The Tennessee Land” (the only extant autobiographical fragment that was written earlier, in 1870) it ends somewhat abruptly, exactly as if the author’s interest had “melted away and disappeared.”

If Clemens did, as he says, make successive attempts to write the autobiography “every three or four years” after 1877, few are known to survive.21 What we have instead are such things as his advice in 1880 to Orion about his autobiography: “Keep in mind what I told you—when you recollect something which belonged in an earlier chapter, do not go back, but jam it in where you are. Discursiveness does not hurt an autobiography in the least.”22

Clemens took between three and seven years to complete almost all of his major books. He required that much time chiefly because he always encountered stretches during which he was unable to proceed, and composition came to a complete halt. Since at least 1871 he had found it necessary, when his “tank had run dry” in this way, to “pigeonhole” his manuscripts. And he learned to resume work on them only after the “tank” had been refilled by “unconscious and profitable cerebration.”23 But the time he spent on his earlier books is brief compared with the nearly four decades it took him to finish his autobiography. Its construction was certainly punctuated by long interruptions as well, but for somewhat different reasons. Until January [begin page 9] 1906, the tank seemed to “run dry” after relatively brief stints of writing, or dictating, because he grew dissatisfied with his method of composing the work, or with its overall plan, or both.

In the spring of 1885 Clemens made his first attempt at doing an autobiography for which more than a few pages survive. He had some previous experience with dictating letters and brief memoranda to a secretary, but he had never tried it for literary composition.24 Now he decided that it might be a good way to work on the autobiography. In late March he wrote in his notebook:

At about the same time, he realized that dictation might be of help to his friend Ulysses S. Grant. Grant had written several articles for the Century Magazine’s series on the Civil War. In the spring of 1885, when he was dying of throat cancer, Grant was close to completing the manuscript of the first volume of his two-volume Memoirs. Clemens had recently secured them for his own publishing house, Charles L. Webster and Co., confident they would earn large profits both for Grant’s family and for himself. As a frequent visitor to Grant’s New York house, Clemens knew that Grant feared dying before he could finish his book. He suggested that Grant hire a stenographer to ease his task. Grant at first demurred, but later hired a former secretary, Noble E. Dawson. On 29 April Clemens visited Grant on his first day of dictation and learned that it “was a thorough success.”26

No doubt encouraged by Grant’s experience, in early May Clemens asked his friend and former lecture manager James Redpath to serve as his stenographer. He liked and respected [begin page 10] Redpath, who had been a journalist and knew shorthand. On 4 May 1885 Redpath replied to Clemens’s proposal: “Now about the auto. When I do work by the week, I charge $100 a week for the best I can do. I have had a run of ill-luck lately but I found that that was what I averaged. It wd take you much less time than you think. I get you word for word & it takes a long time to write out.” Clemens accepted these terms and urged Redpath to come to Hartford soon. “I think we can make this thing blamed enjoyable.” It is clear that he was beginning to intuit the need for a responsive, human audience when dictating—something he articulated quite clearly six years later in a letter to Howells.27

The two men began working together sometime in mid-May and continued for several weeks. In the six dictations that survive, Clemens traced the history of his friendship with Grant, then talked about his own protégé, the young sculptor Karl Gerhardt, who had a commission to create a bust of Grant. In the longest of these dictations he launched into a detailed account of how he had acquired the right to publish Grant’s Memoirs, defending his tactics and countering newspaper insinuations that he had acted unethically.

Clemens probably stopped dictating shortly before Grant died on 23 July 1885.28 In July and August (and possibly earlier) Clemens read over some of the typescripts that Redpath had created from his stenographic notes, adding his own corrections here and there but making few changes in wording. He found the result far from satisfactory, as he implied in a letter to Henry Ward Beecher:

Redpath’s work as an amanuensis was unskillful. None of his stenographic notes are known to survive, but his typescripts are manifestly ill-prepared—full of typing errors, struck-over characters, and extraneous marks—and his numerous penciled corrections create punctuation that is in no way characteristic of Clemens’s own habits.

No manuscripts for the autobiography written between 1885 and 1890 have survived, but the project was certainly not forgotten. In late 1886 as he worked on A Connecticut Yankee in King Arthur’s Court, Clemens wrote to Mary Mason Fairbanks: “I fully expect to write one other book besides this one; two others, in fact, if one’s autobiography may be called a book—in fact mine will be nearer a library.” His 1876 plan for a work not limited “as to space” was evidently alive and well. And in August 1887, two years after halting the Grant Dictations, [begin page 11] Clemens wrote to his nephew, “I want a perfect copy of Fred Grant’s letter, for my Autobiography. I was supposing I had about finished the detailed private history of the Grant Memoirs, but doubtless more than one offensive chapter must be added yet, if Fred Grant lives.” A few months earlier he told another correspondent, “No, I’ll leave those details in my autobiography when I die, but they won’t answer for a speech.”30

Then, in December 1887, Orion wrote to ask his brother’s permission to reveal “something of your boyhood” in an upcoming interview with a local journalist. He listed a few “points” he wanted to offer:

Clemens had already used a number of these “points” in published work. His making wicked fun of Jim Wolf’s pointless rescue of a wash-pan and broom from the threat of a fire next door was in fact his “first writing” for Orion’s Hannibal newspaper, “A Gallant Fireman” (1851).32 And in the first chapter of Tom Sawyer Aunt Polly (based on Jane Clemens) had mildly complained that Tom knew that if he could “make me laugh,” her anger toward him would disappear. Still, Clemens refused Orion’s request:

Clemens would make use of only a few of these “points” in the autobiography. But his stinginess about letting others reveal the raw materials of his history is certainly understandable, [begin page 12] and it may suggest that at this time in 1887 he still intended to write an autobiography that would include these anecdotes from his early life.

By the fall of 1890, Clemens had been investing money in the typesetting machine invented by James W. Paige for almost ten years (since 1881). It was, however, still not completed. The relevance of this project to his autobiography was inescapable, and in the “closing days” of that year he began to write “The Machine Episode,” an unsparing account of the way Paige had charmed and beguiled him into an enormous investment without having yet achieved a salable product. By the time Clemens added the second part to this self-revealing account, in the winter of 1893–94, Paige had still not perfected the machine but was about to sign a new, more satisfactory contract for it. Left in a rather unfinished state, the manuscript was very likely among those Clemens reviewed in 1906 before deciding to omit it from the final form. He did return to the subject in an Autobiographical Dictation of 2 June 1906.

Clemens’s hopes for the Paige typesetting machine were finally crushed in December 1894, and the bankruptcy of Webster and Company earlier that year had placed its debts solely on his shoulders. In the summer of 1895, in order to repay them, he, Olivia, and Clara undertook a lecture tour around the world (Susy and Jean stayed at home), which ended when they arrived in England on 31 July 1896. The family landed at Southampton and then traveled to Guildford, where they learned that Susy was ill in Hartford. “A fortnight later Mrs. Clemens and Clara sailed for home to nurse Susy,” Clemens recalled in 1906, and “found her in her coffin in her grandmother’s house.” Within weeks of this calamity Clemens wrote his friend Henry H. Rogers that he intended to “submerge myself & my troubles in work.” In the last week of September 1896 he reminded himself to “Write my autobiography in full & with remorseless attention to facts & proper names.”34 But he still needed to finish the book about his around-the-world lecture tour.35 The family spent the winter and spring of 1897 in London while Clemens wrote Following the Equator, which would be published in November.

In the summer of 1897 they retreated to Switzerland, and in late September they moved to Vienna. Two autobiographical manuscripts were begun that fall, “Travel-Scraps I” and a much longer sketch called “My Autobiography [Random Extracts from It].” “Travel-Scraps I” appears to be unfinished, or at least not quite ready for the typist, since Clemens made a tentative revision of its title, in pencil (“Travel-Scraps. from Autobiog”) and the manuscript itself still has two sets of page numbers (1–20 and 1–28). It was probably written soon after Clemens arrived in Vienna, for it is largely a complaint about London’s cab drivers and its postal service, things that would naturally have been on his mind since the spring.

On the evidence of the paper and ink used, “My Autobiography [Random Extracts from [begin page 13] It]” was begun about the same time, but probably not completed until 1898. Clemens identified the text as “From Chapter II.”36 (The first page of this manuscript is reproduced in facsimile in figure 1.) It begins as a history of the Clemens and Lampton relatives and ancestors and, more briefly, the despised Tennessee land. But it meanders, without apology, into an anecdote about an incident in Berlin in 1891, and it ends with an evocative description of Clemens’s idyllic summers on his uncle’s farm near Florida, Missouri. This typical combination of early memories and later experiences helps to make clear why Clemens would reject the idea of a completely chronological narrative: his preference for juxtaposing related events from different times deeply resisted that way of organizing his story. At the same time, labeling the sketch “From Chapter II” implied that most of what it contained would come early in the autobiography, as would befit a review of ancestors. The chapter number suggests that while he was not writing about his experiences in the order of their occurrence, he was still making an attempt to assign chapter numbers that respected chronology.

Before Clemens completed “Random Extracts” in 1898, he wrote several more sketches for the autobiography between February and June of that year, grouped here under the supplied title “Four Sketches about Vienna”: “Beauties of the German Language,” “Comment on Tautology and Grammar,” “A Group of Servants” (the only one that Paine did not include in his edition), and “A Viennese Procession.” These were not reminiscences but rather more like entries in a diary, with each piece prefaced by a date. None of these sketches would be included in his final plan, but he did eventually include another manuscript written at this time, “Dueling,” in the Autobiographical Dictation of 19 January 1906.

Two further sketches were written in the fall of 1898 and also later inserted into the final structure of autobiographical dictations. The first was “Wapping Alice,” a tale deemed unsuitable for magazine publication, which was based on an actual event. It joined a growing collection of manuscripts that Clemens would eventually draw on for what he called “fat”—“old pigeon-holed things, of the years gone by, which I or editors didn’t das’t to print”—that he would use to enlarge the bulk of the Autobiography. 37 More than a year after he began dictating his autobiography in 1906, he inserted “Wapping Alice” in the Autobiographical Dictation of 9 April 1907.

The second sketch was “My Debut as a Literary Person,” which he dated “October 1, 1898” and labeled “Chapter XIV.” The revision of this manuscript reflects a season of discouragement about the autobiography, a mood that shows up sporadically during the winter of 1898–99. Just below the title he first inserted a footnote: “This is Chapter XIV of my unfinished Autobiography and the way it is getting along it promises to remain an unfinished one.” Then he changed “unfinished” to “unpublished” and canceled the words following “Autobiography.” When the sketch appeared in the Century Magazine for November 1899, it omitted any reference to his autobiography. Still, it is the first “chapter” to be published in fulfillment of his long-held plan to publish selections from it.38

[begin page 14]

[begin page 15] Clemens’s unsettled attitude toward his “unfinished” autobiography is clear, but not readily explained. On 10 October 1898, even as he was preparing “My Debut” for magazine publication, he told Edward Bok, editor of the Ladies’ Home Journal,

A month later, in a more ambitious frame of mind, he wrote to Rogers that he now planned to “take up my uncompleted Autobiography & finish it, & let Bliss and Chatto each make $15,000 out of it for me next fall (as they did with the Equator-book).” But almost immediately he changed his mind about the need for money, and concluded that he would “never write the Autobiography till I’m in a hole. It is best for me to be in a hole sometimes, I reckon.” Then, just a few days later, he wrote again to Rogers: “I have resumed my Autobiography, and I suppose I shall have Vol. 1 done by spring time. I hope so I expect so.” And at last, in February 1899, still trying to find a magazine publisher for “My Debut,” he told Century editor Richard Watson Gilder: “I have abandoned my Autobiography, & am not going to finish it; but I took a reminiscent chapter out of it some time ago & & had it copyrighted & had it type-written, thinking it would make a readable magazine article.”40 So within the span of a few months he claimed that a “good deal” of the autobiography was written; that he would never finish it until he was “in a hole”; that he expected to have the first volume “done by spring time”; and that he had “abandoned” it altogether. He was obviously struggling with how, or even whether, to proceed with a work that had been in and out of the pigeonhole for twenty years.

It is difficult to be entirely sure, but Clemens seems to have become discouraged at least in part over his inability to be completely frank and self-revealing, after the fashion of Rousseau and Casanova. His solution was, at least temporarily, to recast the autobiography as a series of [begin page 16] thumbnail biographies of people he had met over the years. Several autobiographical manuscripts written in Vienna—“Horace Greeley,” “Lecture-Times,” and “Ralph Keeler”—are character sketches that were part of this reconception, one that he also relied on to some extent in 1904. The Vienna portraits recall men and women whom he knew in his days on the lyceum circuit in the early 1870s. The new plan probably owed something to the idea of a lecture he wrote back then called “Reminiscences of Some un-Commonplace Characters I Have Chanced to Meet.” He delivered this lecture, which he said covered his “whole acquaintance—kings, humorists, lunatics, idiots & all,” only twice. No text of it is known to survive, but in Vienna he evidently resurrected its premise.41 In an interview for the London Times in May 1899 the reporter explained:

Clemens even spelled out exactly why he had abandoned his original plan for an autobiography: “You cannot lay bare your private soul and look at it. You are too much ashamed of yourself. It is too disgusting. For that reason I confine myself to drawing the portraits of others.” And in an interview after he returned to London, he said again that the new idea had actually supplanted his earlier ideas for the autobiography:

If this switch to biographical portraits signaled frustration over the puzzle of how to tell even the shameful truths, his interest in it was still relatively brief. We have no indication that he wrote any further portraits until 1904, and by 1906 the character-sketch idea had fallen entirely out of favor. For about a year Clemens seems not to have added anything to his accumulation of autobiographical “chapters.” In the fall of 1899 he moved his family to London, and for about a year seems to have taken leave of the autobiography.

[begin page 17]

Clemens’s use of the terms “Scraps” and “Extracts” (as well as “Random”) in 1897–98 suggests that he was looking for a way to label “chapters” which, while not themselves strictly chronological, might still have been parts of some coherent narrative sequence. In the fall of 1900 he used the term “Scraps” in the titles of three more sketches for the autobiography: “Travel-Scraps II,” “Scraps from My Autobiography. Private History of a Manuscript That Came to Grief,” and “Scraps from My Autobiography. From Chapter IX.” Only one of these made it into the final form: “Travel-Scraps II” continued the 1897 recital of grievances about London’s telephones and postal system and was ultimately inserted in the Autobiographical Dictation for 27 February 1907. “Scraps from My Autobiography. Private History of a Manuscript That Came to Grief” was much longer. It concerned a recent experience with T. Douglas Murray, an amateur historian, who had invited Clemens to write an introduction for an English translation of Joan of Arc’s trial records. Clemens submitted his draft, and wrote Murray: “When I send the Introduction, I must get you to do two things for me—knock the lies out of it & purify the grammar (which I think stinks, in one place.)”43 Murray took this invitation all too literally and proceeded to revise the text extensively, making the language more formal, even pretentious. Enraged by this tampering, Clemens proceeded to draft a reply in the shape of a scathing letter to Murray, which of course he never sent, preparing it instead for the autobiography.44 The third manuscript, also excluded from the final form, nevertheless illustrates a rather different dynamic, namely the persistent reluctance or inability to break entirely free from the chronological structure of conventional autobiography. The manuscript was titled (as revised) “Selections Scraps from my Autobiography. Passages f From Chapter IX.” Paine thought it was written “about 1898” but it was in fact written in 1900, as one reference in the text makes clear. The assignment of a chapter number is something that it shares with only a handful of other manuscripts, summarized in the following list.

Chap I	Written in 1877. Describes Clemens’s home until aged 4, when the family moved to Hannibal.
Chap II	Written in 1897–98. “My Autobiography [Random Extracts from It].” Clemens aged 7 to 12.
Chap IV	Written in 1903. “Scraps from My Autobiography. From Chapter IV.” Transcribed in ADs, 1 and 2 Dec 1906. Clemens aged 14.
Chap IX	Written in 1900. “Scraps from My Autobiography. From Chapter IX.” Clemens aged 14, 38, and 61.
Chap XII	Not found, but written after 1898 since it refers to Vienna. Clemens aged 62. Mentioned in chapter XVII.
Chap XIV	Written in 1898. “My Debut as a Literary Person.” Clemens aged 30.
Chap XVII	Written in 1903. “From Chapter XVII.” Transcribed in AD, 3 Dec 1906. Clemens aged 62.

[begin page 18] Although some additional numbered chapters may have been written and subsequently lost or destroyed, it is highly unlikely that in 1903, when Clemens labeled a text “Chapter XVII,” he had actually written seventeen chapters. But the numbers assigned to the chapters that do survive correspond roughly to the chronology of their topics, even though they do not accurately reflect the lapse of time: Clemens was fourteen in both Chapter IV and Chapter IX, but between Chapter IX and Chapter XIV he aged from fourteen to thirty, and then to age sixty-two by Chapter XVII. Still, this rough approximation is exactly what one would expect if the chapter numbers were only estimates, intended to place the chapters in approximate chronological order. Together they again suggest that although he was not writing about his life in the order of its occurrences, he was still trying to maintain an overall chronology, even as late as 1903.

The text numbered “Chapter IX” (“Scraps from My Autobiography,” written in 1900) is suggestive in a related way. The chapter number would place it relatively early in his life. It recounts two stories from Clemens’s youth, when he was fourteen (1849–50), but it concludes each story with much later events—the first in Calcutta in 1896 when he was sixty-one, and the second in London in 1873 when he was thirty-eight. In both cases it seems that to follow the stories to what Clemens regarded as their natural conclusion, it was necessary to skip over several decades of his life. So whatever else “Chapter IX” was in 1900, it was not a purely chronological account—even though the chapter number placed it toward the beginning of the narrative.

A similar tension occurs in the two manuscripts with chapter numbers written in 1903 but revised in 1906 after Clemens had settled on discursiveness as the principle for the whole autobiography: both were inserted into the dictations for 1, 2, and 3 December 1906. In 1903 he titled the first one “Scraps from My Autobiography. From Chapter IV,” and began it with a marginal date (“1849–51”). It concerns his youthful encounters with mesmerism in Hannibal. The second 1903 manuscript, paginated separately but probably written at the same time (they share the same ink and paper), brought this story to its conclusion. It is another story about mesmerism, in which a haughty aristocrat is embarrassed by being hypnotized and ordered to undress, in retaliation for his incredulity. Clemens originally titled it “From Chapter XVII.” But when he decided to use the manuscripts in the December dictations, he removed all reference to chapter numbers. So the first mesmerism story was originally assigned to Chapter IV, and its natural conclusion to Chapter XVII, separated by some twelve putative chapters. Their revision shows that in 1903 Clemens was still wrestling with the compulsion to maintain some semblance of his life’s chronology, while in 1906, when he made the manuscripts into one continuous narrative, he had clearly shed that compulsion.

On 15 October 1900 the family arrived in New York City, where they soon rented a house at 14 West 10th Street. “Jean is learning to type-write,” Clemens wrote a friend, “& presently I’ll dictate & thereby save some scraps of time.”45 Jean’s new skill may have prompted Clemens [begin page 19] to think again of dictating, rather than writing, the autobiography. There were other temptations as well. The president of Harper and Brothers, George Harvey, was clearly interested in the prestige that would flow from having the rights to publish the autobiography, even though it would not actually appear until long after both men had died. On 17 October 1900 Harvey proposed to Rogers (who was acting as Clemens’s agent) to “publish the memoirs in the year 2000” and suggested that Clemens “insert a clause in his will to the effect that the memoirs shall be sealed without reading by his executors, and deposited with a trust company.”

Harvey was in fact eager to make Harper and Brothers into Clemens’s exclusive American publisher, and on 14 November, after much discussion, he proposed a rate of twenty cents a word for the exclusive serial rights to anything he might write in the next year, as well as the exclusive right to publish all of his books in the same period. One week later Clemens wrote to Harvey, “Let us add the 100-year book to the arrangements again, & make it definite; for I am going to dictate that book to my daughter, with the certainty that as I go along I shall grind out chapters which will be good for magazine & book to-day, & not need to wait a century.” Nothing dictated to Jean at this time has been found, but Clemens soon agreed to Harvey’s “proposal regarding the publication of my memoirs 100 years hence,” although no formal contract for the autobiography was signed at this time.47

In August 1902, Olivia’s health grew alarmingly worse. Despite temporary improvements, it continued to decline, and in 1903, on the recommendation of her doctors, Clemens decided to take the family to Italy. In early November they settled into the Villa di Quarto near Florence. In addition to Clemens himself, the travelers included Olivia, Clara, and Jean. Three employees were also with them: longtime family servant Katy Leary, a nurse for Olivia, and Isabel V. Lyon, who had been hired in 1902 as Olivia’s secretary but had since assumed more general duties.

During his eight-month stay in Florence Clemens made unusual progress on the autobiography, in large part because of a renewed enthusiasm for dictation as a method of composition. He had experimented with mechanical methods of transferring words to paper ever since the [begin page 20] dictations to Redpath in 1885. In 1888 he tried (and failed) to get access to one of Thomas Edison’s recording phonographs.48 Then in 1891 he suffered an attack of rheumatism in his right arm and, compelled by the necessity of working on his current book (The American Claimant), he did briefly experiment with the phonograph. “I feel sure I can dictate the book into a phonograph if I don’t have to yell. I write 2,000 words a day; I think I can dictate twice as many,” he wrote to Howells on 28 February. But by 4 April he had concluded that the machine “is good enough for mere letter-writing” but

His expectation in December 1900 of relying on Jean to type up dictated autobiography at last became a reality in January 1904, when he tried dictating once more, but not to a machine. According to Isabel Lyon,

Lyon did not know shorthand and so took down Clemens’s words in full, then gave Jean her record to be typed. Shortly after he had begun to dictate, Clemens wrote to Howells on 16 January:

It seems that he recognized Lyon’s lack of shorthand as an advantage, for he went on to urge Howells to try this method, but “with a long-hand scribe, not with a stenographer. At least not at first. Not until you get your hand in, I should say. There’s a good deal of waiting, of course, but that is no matter; soon you do not mind it.” More important even than the leisurely pace was the scribe’s role as audience: “Miss Lyons does the scribing, & is an inspiration, because she takes so much interest in it. I dictate from 10. 30 till noon. The result is about 1500 words. Then I am a free man & can read & smoke the rest of the day, for there’s not a correction to be made.”

Dictation proved so congenial, in fact, that his opinion of the drafts and experiments he had written over the years now began to change. He continued to Howells:

Howells replied to this letter on 14 February, shrewdly raising a familiar issue (clearly not for the first time)—the difficulty of telling the whole truth:

Clemens had of course already reached the same skeptical conclusion. He answered Howells:

What those “safeguards” were remains unknown, since no copy of the “first day’s dictating” has survived. The most one can say is that Clemens seems to have moved on from his despair at not being able to tell “the black heart’s-truth,” rationalizing that that truth would emerge anyway, in spite of all his attempts to suppress it. In a dictation made in late January 1904 he hinted at the disinhibiting nature of talk:

Two years later, in mid-June 1906, he would look back on this time in 1904 as the moment he discovered free-wheeling, spoken narrative as “the right way to do an Autobiography.”54

Only six Florentine Dictations are known to survive. Three of them are portraits of friends or acquaintances—“John Hay,” “Robert Louis Stevenson and Thomas Bailey Aldrich,” and “Henry H. Rogers”—presumably products of the “portrait gallery” concept. Two are reminiscences: “Notes on ‘Innocents Abroad’ ” and a sketch (untitled) recalling his first use of the typewriter. The sixth is a complaint about the Villa di Quarto, the family’s current residence near Florence.55 It is the longest and the least polished, an extended diatribe about the rented villa and especially its hated owner, the Countess Massiglia. Clemens concluded it by inserting an 1892 manuscript about the Villa Viviani, where the Clemenses had lived during an earlier, more enjoyable stay in Florence. Despite that moderating addition, the 1904 dictation is replete with fiery insults to the countess—so much so that when, in May and June 1906, Clemens considered publishing selections of autobiography with S. S. McClure, he marked in blue pencil the offending passages on Jean’s typescript and wrote (on the verso of page 2), “Leave out that blue-penciled passage (& all blue-penciled passages) in the first edition,” and added, “Restore them in later editions.”56

It is clear that there were Florentine Dictations that have not survived, at least not as originally dictated. In August 1906 Clemens said that he had created more than a dozen “little biographies,” of which we have almost none.

Only one such Florentine Dictation is known to survive: Clemens certainly “flayed” the Countess Massiglia in “Villa di Quarto.” We can only guess who the “deserving” men were by considering other evidence. For example, in a letter of 29 January 1904 Clemens vented his anger toward Henry A. Butters, head of the American Plasmon Company, whom he held responsible for his investment losses: “As soon as I get back we will pull Butters into Court, & I guess we can jail him. . . . He occupies space enough in my Autobiography to pay back all he & his pimp have robbed me of.”58 But no such text from 1904 or earlier survives, and Butters is mentioned only in passing in the later Autobiographical Dictations. (On 31 October 1908, for example, Clemens described him as “easily the meanest white man, and the most degraded in spirit and contemptible in character I have ever known.”) There are others who might have received harsh treatment in now-lost dictations from 1904 whose portraits were then “done over” between January and August 1906. Clemens scattered a few sarcastic remarks about Charles L. Webster in the early 1906 dictations, then excoriated him at length in the one for 29 May 1906. Other candidates include Daniel Whitford, Clemens’s attorney; James W. Paige, inventor of the failed typesetter; and of course Bret Harte.59

When Olivia died suddenly on 5 June 1904, Clemens’s interest in his autobiography quite naturally evaporated, and in the next year and a half he wrote only one short sketch for it, “Anecdote of Jean.” But before Olivia died, his new enthusiasm for dictating the autobiography gave rise to a scheme to provide income for the family he thought would survive him (he was then sixty-eight). In the same January letter to Howells in which he enthused about dictation he described this new idea: “If I live two years this Auto will cover many volumes, but they will not be published independently, but only as notes (copyrightable) to my existing books. Their purpose is, to add 28 years to the life of the existing books. I think the notes will add 50% of matter to each book, & be some shades more readable than the book itself.”60

This notion was still alive almost a year after Clemens had begun his dictations to Josephine [begin page 24] Hobby in January 1906. In December of that year he spoke of it to a reporter, who then summarized it in the New York Times:

It is difficult to know just how serious Clemens was about this scheme because it was never put to the test. On 24 December 1909, in “Closing Words of My Autobiography,” he explained that the

Because of that change in the law, he evidently told Clara much the same thing. She protested, and he replied on 23 February 1909:

Still, it is clear that extending his copyrights had never been his primary motive for creating the Autobiography.

[begin page 25]

Clemens, his two daughters, Katy Leary, and Isabel Lyon accompanied Olivia’s body back to America in July 1904. They buried her at Quarry Farm, and Clara entered a rest home in New York City while Clemens, Jean, and Lyon spent the rest of the month in a rented summer home in Lee, Massachusetts. On 10 August Clemens went to New York and soon signed a three-year lease for a town house at 21 Fifth Avenue, which he and Jean occupied in December 1904 after it was renovated and furnished.

One year later, in January 1906, work on the autobiography had been at a standstill for eighteen months. A catalyst was needed to revive the enthusiasm of 1904, and on the night of 3 January it arrived in the form of Albert Bigelow Paine. Paine was an experienced writer and editor who in 1904 had published a biography of Thomas Nast which Clemens admired. On that January evening Paine attended a dinner in honor of Clemens, hosted by The Players club, and happened to be seated “nearly facing” him. Three days later he called at 21 Fifth Avenue and asked to write Mark Twain’s biography. After brief preliminaries Clemens turned to him and said: “When would you like to begin?” Paine went on to suggest that a stenographer be hired to take notes of what Clemens said in response to the biographer’s questions. Clemens said that he thought he would “enjoy dictating to a stenographer, with some one to prompt me and to act as audience,” and he offered Paine office room and access to “a trunkful or two” of his manuscripts, notes, and letters. “Whatever you need will be brought to you. We can have the dictation here in the morning, and you can put in the rest of the day to suit yourself. You can have a key and come and go as you please.”63

It was agreed that work should begin on 9 January. On that Saturday morning, Paine arrived accompanied by a stenographer, Josephine S. Hobby. The procedure for working on the biography was promptly decided, but Clemens

Josephine Hobby was an experienced stenographer and an excellent typist, known to Paine for about eight years. She had previously worked for Charles Dudley Warner and Mary Mapes Dodge and was currently employed by the Century Company, which, since 1899, had also [begin page 26] employed Paine as an editor of St. Nicholas, a magazine for young people. Hobby charged one dollar per hour of dictation and five cents per hundred words of typescript. She began immediately with a transcription of the morning’s conversation. “We will try this,” Clemens told Paine, “see whether it is dull or interesting, or whether it will bore us and we will want to commit suicide. I hate to get at it. I hate to begin, but I imagine that if you are here to make suggestions from time to time, we can make it go along, instead of having it drag.” He proposed a schedule of four or five days a week, for roughly two hours each morning.65 Clemens talked while Hobby took him down in shorthand and Paine listened appreciatively. For these early sessions, Paine recalled, Clemens usually dictated from bed, “clad in a handsome silk dressing-gown of rich Persian pattern, propped against great snowy pillows.”66

Before Clemens was done dictating in 1909, he and Hobby, along with three other typists, generated more than five thousand pages of typescript. That enormous body of material has, since Clemens’s death, constituted the largest part of the manuscript known as the “Autobiography.” But probably since DeVoto’s time as editor of the Mark Twain Papers, anyone who consulted that file was likely to be puzzled by two things. First, most of the Autobiographical Dictations between January and August 1906 were filed in folders—one per dictation—containing between two and four separate, distinct typed copies of essentially the same text. No one understood the purpose of the duplicates. Second, the differences (if any) between these various “duplicates” were not obvious or readily intelligible: pagination differed, seemingly without pattern; some contained handwritten authorial revisions, while others were unmarked; and many were extensively marked by at least half a dozen different (mostly unidentified) hands, in addition to the author’s. These documents constituted the central puzzle confronting anyone who set out to publish the Autobiography of Mark Twain.

The First Typescript (TS1)

A first step in solving the puzzle was to find reliable ways of distinguishing between the several, nearly identical typed copies in any given folder. The paper used, specific characteristics of the typewriter and habits of the typist, and of course the unexplained differences in pagination proved to be essential pieces of evidence. The very first typescript Hobby created from her stenographic notes was eventually isolated and identified in this way, and it is called hereafter TS1 (for typescript 1). Each of the other typescripts (sometimes with carbon copies as well) were similarly identified, and are herein referred to by number (TS2, TS3, and TS4). Once they were physically distinguishable in this way, it became possible to see that the (understandable) fashion in which daily dictations were filed had in fact long obscured why there were different typed copies.67 It was in turn possible to decide, on the basis of meticulous collation, [begin page 27] which was copied from which, and to begin to make some sense of the various differences between them.

By 18 January Hobby had settled on a standard format for each session: she recorded the time spent on dictation and a word count at the top left of the first page, with the page number centered, and the date of dictation at the right, usually followed by a summary of the contents. Hobby marked in pencil any errors she had failed to correct on the machine. The typescript then went to Clemens for correction and revision. TS1 would total roughly twenty-six hundred consecutively numbered pages, beginning at page 1 with the dictation of 9 January 1906 and ending with the dictation of 14 July 1908, which Hobby completed shortly before leaving Clemens’s employment. Two later stenographer-typists produced another hundred or so pages of typescript, which did not continue the TS1 pagination sequence.68

Hobby soon agreed to give up her job with the Century Company in order to work for Clemens exclusively during the summer and possibly longer. On 13 March Lyon commented in her journal, “Mr. Clemens finds her entirely to his liking & he says ‘it is a case of established competency’ which is saying a great deal—for she is a good audience, is sympathetic & very appreciative.”69 By 8 April Hobby had transcribed her notes through the end of the 28 March 1906 dictation.

Despite the risk of somehow losing TS1, a unique copy (there was no carbon copy at this point), Clemens allowed Clara to carry away and read about five hundred pages, through the end of the 16 March dictation. He also lent Twichell three days’ dictation, probably those for 23, 26, and 27 March. On 8 April, Lyon recorded that “Mr. Twichell is here—Mrs., too—& Mr. T. thinks the auto. MS. is absorbingly interesting.” He presumably returned the three dictations at that time. The same day Clemens took the entire 28 March dictation (Orion’s misadventure with Dr. Meredith’s “old-maid sisters”) up to lunch with Howells at the Hotel Regent. Howells returned the pages by the next day’s mail:

Lyon noted that Clara had written “enthusiastically about it too,” but her letter has not been found. Clemens replied to her in Atlantic City, two days before she was supposed to return the pages to him:

On 11 April Clemens took a break from his dictating routine, and did not resume until 21 May, after he, Jean, and Lyon were settled for the summer in the isolated Upton House near the village of Dublin, New Hampshire. Clemens arrived there on 15 May; Paine and Hobby arrived a few days later and found quarters nearby. Paine reported, “We began in his bedroom, as before, but the feeling there was depressing.” He described the view from the verandah of the Upton House as “one of the most beautiful landscape visions on the planet,” and reported that Clemens soon saw the opportunity it presented: “I think we’ll do our dictating out here hereafter. It ought to be an inspiring place.” Lyon occasionally recorded Clemens’s (and her own) impatience with the “old-maidish whims” and slowness of the “Hobby horse,” but for the most part the morning sessions seem to have been remarkably amicable.71 Lyon described one of them in her journal:

In late May Clemens also began in earnest the job of reading and correcting the four months’ accumulation of TS1, which by then consisted of over seven hundred pages (through the dictation for 11 April). He revised the typescript in black ink, only rarely in pencil, making relatively few changes in wording—at least after the first ten dictations—and he consistently made a smattering of changes or corrections to spelling, punctuation, and paragraphing. For the most part Hobby seems to have learned, directly or indirectly from Clemens, how he preferred to spell and punctuate. Inevitably there were mistakes, especially in proper names (“Katie” instead of “Katy,” “Susie” instead of the preferred “Susy,” “Twitchell” instead of “Twichell”). And with a backlog of hundreds of pages typed before he began his review, Clemens inevitably found repetitive errors. Lyon noted in her journal:

He had found (and corrected) over two dozen instances of this trivial but irritating error. One can only hope that he alerted Hobby to her mistake in a less ferocious mood.

S. S. McClure and Syndication

Clemens’s correction and revision of TS1 was given a special impetus by S. S. McClure, founder of McClure’s Syndicate and McClure’s Magazine, who offered to pay Clemens a dollar per word for the right to syndicate fifty thousand words from the autobiography. As Clemens reviewed Hobby’s typescripts and some of his pre-1906 texts, he began noting likely candidates for McClure’s proposed syndication. For instance, the blue-penciled notation “Mc” is written at the top of the first page of the 1904 “Villa di Quarto” typescript, and the same or a similar notation can be seen on several other pages of TS1 prepared between January and March 1906. Still other dictations were marked “Not for MC.”

Clemens was not, however, free to accept McClure’s bid. He had signed an exclusive contract with Harper and Brothers on 22 October 1903, which prohibited his publishing with another firm any of his “books, writings or works now existing or which may hereafter be created.” It further stipulated that all “miscellaneous articles accepted for magazines or periodicals shall be paid for at the rate of thirty cents a word.” Even so, on 27 May Clemens kept Lyon busy for two hours taking notes while he outlined “a course of action to be followed out in his scheme of breaking away from the Harper Contract & selling 50000 words to McClure for $50000.00 to be syndicated.” His interest in publishing with McClure was more than financial, however. He had an express desire to see the selections “go into the papers—even into the Hearst papers—to reach his ‘submerged clientele.’ ”74 He further explained to Rogers,

Rogers, however, did not favor accepting McClure’s offer. Believing that the Harper contract was “so valuable that they would seize the opportunity of breaking the arrangement if it were possible,” he urged Clemens not to “think of anything that will vitiate” it.75

[begin page 30]

On 10 June Clemens wrote to a friend, “I’ve stopped dictating—tired of it. I’ve stopped reading autobiography & admiring it—tired to death of it!”76 Clemens’s lack of enthusiasm must have been merely a passing mood, however, since he did not in fact stop dictating or “reading autobiography.” From 11 to 14 June he dictated every day (as well as on six more days before the end of the month), while he continued to read and revise TS1, working his way through the backlog of over nine hundred pages by 21 June. And it was also at this time that he decided to return to a task that he had begun the previous winter but suspended in May: reviewing his earlier manuscripts, including his preliminary attempts at autobiography. On 8 June he sent Paine to the house at 21 Fifth Avenue to fetch the cache of manuscripts that he had gathered together for use in the biography as well as for his copyright scheme.77 Many years later, Lyon annotated a copy of Paine’s edition where “The Tennessee Land” began, explaining that “in the winter of 1905–6” Clemens pursued his idea of using “autobiographical notes to be added to each volume on its copyright expiration, thus creating a new volume with its new copyright to be extended for 14 years. . . . He asked me for the notes he wrote in 1870 & later—& here is the beginning.” It is also likely that at about the same time she heard Clemens read the manuscript of what Paine titled “Early Years in Florida, Missouri.” She noted in Paine’s edition, “Mr Clemens called for this MS. which he read aloud to me; often deeply moved by memories his voice momentarily lost in emotion.”78

Paine arrived back in Dublin on the thirteenth with a “small steamer trunk” of manuscripts. On 22 June Lyon wrote in her journal:

Clemens may have been considering which of his early reminiscences he liked well enough to add to the autobiography, if only to enlarge its bulk. A few days earlier (17 June) he had written a long letter to Howells in which he referred to yet another way to expand his text, this one taking advantage of posthumous publication:

And, in a postscript, he added: “I’ve written a short Preface. I like the title of it: ‘Spoken from the Grave.’ It will prepare the reader for the solemnities within.”80

The manuscript of this preface (whose subtitle is “As from the Grave”) and a draft of the title page survive in the Mark Twain Papers, as does a typed copy of the title page, on which Clemens drafted a series of notes specifying restrictions and conditions for publishing the autobiography. He then decided to add to his short “Preface” by enlarging on these notes. Addressing his “editors, heirs and assigns,” he dwelt at facetious length on how successive editions could include more and more of his (supposedly shocking) “sound and sane expressions of opinion.”

It is now clear that by the time Clemens read aloud the “44 typewritten pages . . . telling of his boyhood days, & the farm” on 22 June, he had already decided to use that sketch to begin the Autobiography. He wrote a one-page preface called “An Early Attempt” to introduce it, then followed that with a single page instruction: “Here insert the 44 old type-written pages.” This “old” typescript has been lost, but we now know that it was a typed copy of the manuscript he called “My Autobiography [Random Extracts from It],” written in Vienna in 1897–98. It is not known when this (now missing) typescript was prepared, but it was probably no later than 1900.81

At about the same time he also decided to further illustrate the evolution of his ideas about autobiography by including some of the dictations produced in Florence in 1904. To frame these he wrote a matching preface called “The Latest Attempt,” characterizing them as examples of “the right way” to do an autobiography. And he made one more change, adding “The Final (and Right) Plan” and an epigraph (“What a wee little part of a person’s life . . . ”).82 The result was a three-part preface, concluding with the “Preface. As from the Grave” (divided into three sections), followed by the introductory note “Here begin the Florentine Dictations.”

The present edition prints this extensive front matter, complete and in the sequence that [begin page 32] Clemens intended, for the first time. All of the material was known to Paine (his penciled page numbers are on the manuscript pages). But he apparently realized that it interfered with his own plan for the autobiography: a sequence of early sketches and the Florentine Dictations in the order of their composition, followed by a selection of the Autobiographical Dictations from January through April 1906. He included the epigraph and the first section of “As from the Grave” at the beginning of his first volume, placing “The Latest Attempt” before the Florentine Dictations but calling it “Author’s Note.” He omitted entirely “An Early Attempt” and the second and third sections of “Preface. As from the Grave.”83 The prefatory pages, all in the Mark Twain Papers, are shown in sequence on the facing page and reproduced in facsimile in figures 2–13.

Since the “44 old type-written pages” are admittedly lost, how can we be sure that they were in fact a copy of “My Autobiography [Random Extracts from It]”? And how can we tell which of the six surviving Florentine Dictations were intended to follow “As from the Grave”?

The multiple typescripts of the January–August 1906 dictations hold the answer to both questions.

Two More Typescripts: TS2 and TS4

In his postscript to the 17 June letter to Howells, Clemens had said: “I think Miss Lyon told you the reason we couldn’t send you the Autobiography—there’s only one typed copy, & we had to have it for reference, to guard against repetitions. The making of a second copy is now begun; & so, we can presently begin to mail batches of it to you.”84 TS1 had been begun without any provision for a carbon copy. But Howells’s and Twichell’s interest in seeing the text earlier in April, and McClure’s interest in late May, made it increasingly clear that duplicates were vitally needed—hence the decision to begin a carbon copy of TS1 from that point, certainly no later than 11 June.85 But that still left more than eight hundred pages of dictation in a unique copy, much of which had been revised.

Clemens’s postscript shows that by 17 June “a second copy” had been commissioned. In fact, not one but two typed copies of TS1 were begun in mid- to late June, soon after the various prefaces had been created: the first typed by Hobby (TS2) and the second by an unidentified typist (TS4). These sequences are distinguishable by their differences in pagination and by minute differences in their typists’ styles. Collation demonstrates that TS2 and TS4 were both copied independently from the recently revised TS1, not one from the other. Both TS2 and TS4 originally began with the “Random Extracts” text, but both omit the “Early Attempt” preface written for it. TS4 includes the other three-part preface and four of the Florentine Dictations (“John Hay,” “Notes on ‘Innocents Abroad,’ ” “Robert Louis Stevenson and Thomas Bailey Aldrich,” and “Villa di Quarto”), and TS2 originally did so as well. But only parts of TS2 for these early texts survive: gaps in it (shown by missing page numbers) cannot always be certainly reconstructed, but all surviving evidence shows that the missing pages were identical in content to continued after plates

[begin page 33] [begin page 34] [begin page 35] [begin page 36] [begin page 37] [begin page 38] [begin page 39] [begin page 40] [begin page 41] [begin page 42] [begin page 43] [begin page 44] [begin page 45]

[begin page 46] those of TS4, which is the only complete record of these initial elements in Clemens’s plan. This conjecture explains why the page numbers for the January–August 1906 dictations in TS2 and TS4 are different from each other and consistently higher than the page numbers of TS1 for the corresponding dictation. TS1 begins with the Autobiographical Dictation of 9 January, having been started before Clemens decided to include any of the early material.87

We now understand why there are often two, three, or even four nearly identical typescripts for the January through August 1906 Autobiographical Dictations. The resolution of this first part of the textual mystery shows, among other things, that TS1 is the primary source for the text of those dictations, and that when parts of TS1 are lost, the missing text can be reliably restored from either TS2 or TS4, because they were created by copying TS1 before the losses occurred. Our understanding of the typescripts also helps to explain the multiple inscriptions on so many of their pages: they are the traces left behind by the editors and typists who collaborated with Clemens in 1906–9, and by the editors who published parts of the autobiography after his death, from Paine to DeVoto. The four typescript pages reproduced in facsimile in figures 14–17 illustrate some of the many hands that had to be identified and, above all, distinguished from Clemens’s own hand.

To recapitulate: by 21 June Clemens had read through and corrected all of TS1 that Hobby had so far typed (over nine hundred pages, probably through the dictation for 20 June 1906).88 He had reviewed his earlier manuscripts and selected at least those he wanted to begin with (he would later select several more, inserting them in later dictations). And he had written the title page and the several prefaces to frame those early pieces and introduce the 1906 dictations. Hobby began to create TS2, and an unidentified typist started TS4, probably as soon as Hobby made the revised TS1 available.

With all that in train, Clemens left Dublin on 26 June to be away for a month, in Boston and New York City, occasionally visiting Henry Rogers at his home in Fairhaven, Massachusetts, and joining him on his yacht, the Kanawha. Following Rogers’s advice, he met several times with the Harper executives and lawyers in order to resolve their mutual disagreements about the recent republication of Mark Twain’s Library of Humor. While in New York he also met with S. S. McClure and left with him some pages from the dictations about Susy—probably those of 2–7 February. McClure wrote Clemens about them on 2 July:

[continued after plates]

[begin page 47] [begin page 48] [begin page 49] [begin page 50]

Finally, with Harvey’s return from England in mid-July, the troublesome Library of Humor problem was resolved and Clemens was able to return to Dublin on 25 July.89

Clemens had already published several essays in Harvey’s North American Review: “To the Person Sitting in Darkness” and “To My Missionary Critics” (1901) as well as his satirical commentary on Christian Science (1902–3). His first impulse had been to sell Harvey selections from the autobiography to go into Harper’s Weekly (which Harvey also edited), a much more widely read journal than the Review. But when Harvey finally made the twice-postponed visit to Dublin, arriving late on 31 July, he had big plans for the Review. 90 He promptly immersed himself in the autobiography, and by the time he left Dublin on 4 August he and Clemens had agreed on what would become a sixteen-month series in the Review. “I like this arrangement,” Clemens confided to his friend Mary Rogers (Henry’s daughter-in-law), even as Harvey departed, “& so will Mr. Rogers; but he didn’t much like the idea of McClure’s newspaper syndicate, & I ceased to like it myself & stopped the negociations before I left New York.”91 As Clemens wrote to Clara on 3 August, he was impressed by Harvey’s “great plan: to turn the North American Review into a fortnightly the 1st of Sept, introduce into it a purely literary section, of high class, & in other ways make a great & valuable periodical of it.” He was also clearly flattered by Harvey’s response to his text:

[begin page 52] Not only did Harvey “carefully read close upon a hundred thousand words” of the autobiography’s 250,000, but by the time he left Dublin on 4 August he had read the remaining 150,000 words, encompassing the pre-1906 material and all the dictations through the end of June.93

A Composite Typescript: TS3

Harvey’s “wonderful editing” was exactly as Clemens characterized it: he selected excerpts and patched them together to create five installments, essentially without “altering a word.” Domestic anecdotes were a principal theme. He used the moving description of Olivia, and of Susy and her death, with nearly all the excerpts from Susy’s biography of her father that were quoted in the February and March dictations. Other favored topics included Clemens’s amusing misdeeds, such as his swearing about the missing shirt buttons; Susy’s charming eccentricities of spelling; Clemens’s puzzlement over the “spoon-shaped drive”; and the challenges of the burglar alarm at the Hartford house. Harvey also included the recollections of Hannibal, such as the story of “playing bear” from “Scraps from My Autobiography. From Chapter IX” (which Clemens had omitted from his plan for the Autobiography), the dreadful anticholera “Pain-Killer,” and Orion’s 3:00 a.m. call on a young lady. Harvey did not share Howells’s squeamishness about Orion climbing (by mistake) into bed with the “middle-aged maiden sisters” (as they were called in the North American Review text): he included that episode without apparent concern. He rejected material with less broad appeal, such as the death of Patrick McAleer (the Clemenses’ beloved coachman), and omitted Clemens’s excursions into political commentary—the massacre of the “rebellious” Moro people in the Philippines and the treatment of Mrs. Morris, who had been thrown out of the White House. Harvey’s penciled notations on the typescripts, such as “Begin,” “End,” and “Continue,” show that Clemens’s own participation in the selection process was probably much smaller than has previously been supposed.94

Since Harvey drew each Review installment from several different daily dictations, Hobby was now charged with creating yet another typescript to serve as printer’s copy, working “under high pressure” to get it ready in time for the first chapter of the series, in the issue slated for 1 September. This composite typescript, called TS3, was typed from the start with a carbon copy, and was paginated independently of the three other ongoing sequences.95 Hobby managed to complete two batches of TS3, of some two dozen pages each, in time for Clemens to revise them lightly and give them to Harvey when he left Dublin on 4 August. She immediately set to work typing a third large batch of TS3 (sixty-three pages), which Clemens agreed to send off “as soon as finished.” These three batches of TS3 were intended for installments 1–5 in the Review. Because TS2 had at that time been completed only through 12 February, Harvey also [begin page 53] asked that the “copy of complete dictation beginning with Feb. 13 as it proceeds” be forwarded to him.96

Clemens soon revised and returned the proof for the first installment, first to Clara for her approval, and then to the Review editors. But no sooner had he done so than a new decision was made: an excerpt from the first half of “My Autobiography [Random Extracts from It],” the text that he had just recently chosen to begin the Autobiography, would now lead off the series. It has not been discovered who made this late change, or why, but by the end of August Clemens had read galley proofs for this “Virginia-Clemens” installment, as he called it in a letter to Mary Rogers.97

The first issue of “Chapters from My Autobiography” appeared on 7 September, accompanied by a statement that would be repeated before each installment:

In the “Editor’s Diary” section of the same issue Harvey “let go all holts,” as Clemens might say, in an announcement of the upcoming series:

[begin page 54] Reading the above in proof, Clemens facetiously professed himself “troubled” and suggested to Mary Rogers that she write a letter of protest to Harvey, even providing her with a text. She was to say that Harvey’s “prodigal, even extravagant” praises “sounded cold & indifferent” to him. “He is almost morbidly fond of compliments, & he realizes that these are good ones, but thinks they are over-cautious & thin. When we of the family butter him we do not do it with a knife, we use a trowel.”99

Harvey’s first round of selections, the second through sixth installments, appeared in the Review between 21 September and 16 November. According to Clemens, during his August visit he had actually earmarked a total of twenty-four selections—“a year’s lot”—drawing on the dictations of January, March, and April 1906, “John Hay,” and the second part of the “Random Extracts” sketch for installments 7–8 and 10–13, published through 1 March 1907.100 Later material, from the dictations of October 1906 through February 1907, began to appear in installments 14 and 15, published on 15 March and 5 April 1907. Hobby made only one additional batch of TS3, for installment 16, published 19 April 1907. Apart from the dictation of 21 May 1906 (“My experiences as an author . . . ”), which had been used in installment 2, no material from the dictations of May through August 1906 was published in the Review. 101 In those months Clemens dictated some rather stringent comments about religion, business, and various of his associates—comments he had no intention of publishing during his lifetime. Besides, there were soon so many excerpts stockpiled that the making of further selections could be safely postponed.

This basic work flow remained in place for the next sixteen months. Clemens continued to revise his typescripts, censoring or “softening” them as needed. For example, he deleted the phrase “ ‘Stud’ Williams was his society name”; altered “Plasmon thieves” to “Plasmon buccaneers”; changed his description of “Dr. Meredith’s two ripe old-maid sisters” to “Dr. G.’s two middle-aged maiden sisters”; altered “a dying parishioner” to “a fictitious ailing parishioner”; and deleted “God forgive me.”102 The revised typescripts were sent to the Review and marked for house style by editor David A. Munro; after typesetting, the typescripts were returned to Clemens along with the galley proofs ready for correction. The editorial relationship was easygoing: Munro and Harvey offered very few substantive revisions, and Clemens often took a joking tone in his responses. He occasionally addressed remarks to Munro in the margins of the printer’s copy or the galley proofs; pressing matters might be handled by mail and telegram and, conceivably, by telephone.103

[begin page 55]

Between September and December 1906 brief excerpts from the early North American Review “Chapters from My Autobiography” appeared in several newspapers, accompanied by a scattering of complimentary remarks. The New York Times, for example, reprinted the passage from the first installment (excerpted from “My Autobiography [Random Extracts from It]”) in which Clemens identified James Lampton as the real Colonel Sellers, commenting that the passage was “noteworthy” as an example of “honest self-revelation.” Two weeks later the Times remarked that the second installment, the “story of how G. W. Carleton refused Mr. Clemens’s first book and twenty years afterward called himself for so doing ‘the prize ass of the nineteenth century,’ ” was “a good story.”104 Other reprinted passages included the Florentine Dictation about Robert Louis Stevenson; the emotional descriptions of Olivia and Susy; the humorous episode about the burglar alarm in the Hartford house; and the essay about dueling.105 A reviewer in the Louisville, Kentucky, Courier-Journal (possibly its editor, Clemens’s lifelong friend and distant relative Henry Watterson) called the autobiography “delightful,” and while conceding that Clemens did not claim to be “strictly speaking a historian,” went on to correct the inaccuracies in his account of Jeremiah and Sherrard Clemens.106 The Washington Post characterized the installments as “filled with his gentle humor,” and an editor of Pearson’s Magazine noted:

None of the notices of the autobiography found in contemporary newspapers and journals, however, offered any substantial critical commentary or analysis, and after the early months of 1907, the installments received little attention.

For all Clemens’s insistence on publishing the Autobiography only long after his death, the excerpts in the North American Review were surprisingly important to him. Just how important became clear only when he realized how few readers had actually seen the magazine text. On 30 July 1907, nine months after the installments began, Lyon made the following entry in her journal:

One result of this problem was that between 27 October 1907 and 27 September 1908 the North American Review chapters were reprinted, with newly commissioned illustrations, as a series in the weekly “Sunday Magazine,” a supplement that was syndicated in many large-circulation newspapers. Harvey himself proposed the syndication, but he implied that greater circulation was his only concern, not how much money it made. Lyon noted on 6 September 1907 that Harvey

Clemens hoped that this syndication would expose his work to a very different class of readers. He was undoubtedly pleased, and he even praised his portrait by F. Luis Mora that accompanied the first installment reprinted in “Sunday Magazine.”110 Many years later, on 28 October 1941, the artist John Thomson Willing, who was the art editor of the Associated Sunday Magazines at the time of the syndication, wrote to an unidentified correspondent in response to a “request for Mark Twain stories”:

Not surprisingly, in 1907 and 1908 the intensity of Clemens’s interest in adding to the autobiography gradually abated. In each successive year the number of dictations declined by half, [begin page 57] they became briefer, and the proportion of inserted clippings and other documents grew larger. By 1908 much of what he produced for the autobiography was actually original manuscript that he labeled as dictation. When on 24 December 1909 he wrote that because of Jean’s death “this Autobiography closes here,” he had in fact produced fewer than twelve new pages of typescript in the previous eight months.

In 1910, after Clemens’s death, Howells reported in My Mark Twain that at some point Clemens had “suddenly” told him he was no longer working on the autobiography, although Howells was unclear whether Clemens “had finished it or merely dropped it; I never asked.” He also recalled that at the outset of his work Clemens had intended the autobiography to be “a perfectly veracious record of his life and period,” but he now admitted that “as to veracity it was a failure; he had begun to lie, and that if no man ever yet told the truth about himself it was because no man ever could.”112 Of course, by 1904 Clemens had already convinced himself, by experiment, that an autobiography “consists mainly of extinctions of the truth,” even if “the remorseless truth is there, between the lines.” And in April 1906 he had said in one of his dictations,

We have seen that in 1898 Clemens had been so discouraged by this insight that he (temporarily) decided to change the very nature of his autobiography. But there is good reason to suppose that by the time of his death he had reached a more enlightened understanding of what his or anyone else’s autobiography could accomplish. In mid-1909 he was asked whether his remarks about the Tennessee land as published in the North American Review were true. “Yes,” he replied, “literarily they are true, that is to say they are a product of my impressions—recollections. As sworn testimony they are not worth anything; they are merely literature.”114

A hundred years have now passed since Clemens’s death. It certainly seems fitting that his plan for publishing the Autobiography of Mark Twain in its entirety should just now be recovered from his vast accumulation of papers, and that the Autobiography’s standing and value as “literature” be at last recognized. This edition, prepared by his editors (if not his “heirs and assigns”), relies on the eloquent evidence of historical documents to understand and carry out his wishes for this, his last major literary work. His long-standing plan to speak as truthfully as possible “from the grave” is no longer just a plan. And as Colonel Harvey predicted more [begin page 58] than a hundred years ago, the Autobiography is being published both as printed volumes and “by electrical method,” a fact that would no doubt have appealed to Mark Twain’s “vivid imagination.”115

Harriet Elinor Smith

Mark Twain Project, Berkeley