settings and tales for his fiction. The early manuscripts written between 1870 and 1876 conform to the conventional approach to reflective writing in which the writer attempts to build a bridge to his past by examining episodes out of his prior experience. In Clemens' s case, this meant creating sketches only several paragraphs long, a form with links to his talent for short, precise vignettes like those strung together in the chapters ofThe Innocents Abroad, Roughing It, A Tramp Abroad, and even The Adventures of Tom Sawyer. In these sketches, he attempts to reclaim a sense of the contrast between his present and his past. Longer autobiographical fragments—the Grant Dictations of 1885 and the dictations begun in Florence during 1904—were composed in the presence of a "shorthander" or stenographer. These later self-contained descriptions of Clemens' s experiences are precursors of a form of collaboration Clemens would pursue more energetically in his autobiographical dictations of 1906-1909.

Clemens used the dictating sessions of 1885 and 1904 to record immediate ideas and impressions. James Pond, Clemens' s lecture agent, worked with him on the Grant material; Isabel Lyon, recently hired to help carry some of the weight of the household, took down Clemens's 1904 Florence observations. Neither of these attempts at working with an amanuensis was Clemens' s first such attempt. The very notion of working with a stenographer appears quite early. During his 1872 trip through England, Clemens wrote to Olivia: "If I could take notes of all I hear said, I should make a most interesting book—but of course these things are interminable—only a shorthand reporter could seize them.’’ In 1873, he hired S. C. Thompson, a theological student, to accompany him to England and to keep notes and records of the trip; in 1883 he hired Roswell H. Phelps, a stenographer for the Continental Life Insurance Company of Hartford, to accompany him during his tour of the Mississippi River, although Phelps would not complete the trip. Clemens also dictated portions of The American Claimant and even toyed with the idea of dictating onto wax cylinders as a way to increase his output.

While these events demonstrate Clemens' s interest in dictating as one method of composing, they hardly suggest collaboration. The early experiences with Thompson, Phelps, and Pond do not suggest any attempt to share the work of composition. They were there to take down notes and ideas and reactions, not to take an active role in composing text. In Phelps's case, Clemens used dictated material much later when he wrote Life on the Mississippi. Even Clemens' s earlier—and rather quickly aborted— attempt at collaboration, a scheme to engage John Henry Riley to dispatch notes from South Africa so that Clemens could build a book on diamond mining by surrounding those notes with his own reactions and ideas, was more collaborative in spirit. Besides, Clemens certainly had neither wish nor intention to share either the composition or the credit with any other person. He shared bylines only with Charles Dudley Warner, Bret Harte, and William Dean Howells, but even then there was usually some tension between the collaborators.

During the 1870s and 1880s, Clemens was riding the crest of his own creativity. He was thus not interested in sharing the fruits of that success. He was, however, interested in using those around him to help him sharpen his storytelling. That interest led him to collaborate with family members by reading the day's work to them in order to gauge their reaction. Historically, critical confusion over Olivia's role as reader and censor has resulted in misinterpretations of Clemens's own descriptions of those evening readings. Clemens' s desire to use an audience of auditors was also behind his repeated pleas to Howells for editorial support. On one level, these are simply attempts to try his material out"on the dog,’’ attempts to get some notion of how the work was progressing and whether he should "edit'' material out. But what seemed either a game or a serious attempt at censorship, in various critical interpretations, was more likely Clemens's shrewd attempt to use a representative audience to see whether his work might have broader appeal. More specifically, Clemens used his wife and their daughters Susy, Clara, and Jean as a focus group to determine how his work would play to a wider audience. Later, he would look back at this period as among the more peaceful and creatively satisfying of his life. As the century came to a close, however, Clemens faced the loss of the support network that had sustained his work. He was geographically separated from Howells, Livy's illness took her further out of the circle of creative partners, and his daughters were maturing and looking to wean themselves from the family. Susy's death in 1896, Jean's increasingly serious epilepsy, and Clara's attempt to create her own life and singing career contributed to Clemens's physical and psychological isolation. With Livy's death in 1904, Clemens lost the creative compass that held him on course, and he began a methodical search for a new network that would sustain his creativity. He would spend the final decade of his life in a sometimes desperate attempt to recapture the muse that escaped him when the family circle was broken.

With these circumstances as background, we can better understand not only why Clemens began to show more interest in autobiography but also why he returned to dictation of his material. One of its primary attractions was the prospect of telling tales to a captive and appreciative audience. That notion began to take shape with Isabel Lyon in Florence as Clemens turned to the practice of dictating in the face of Olivia's deteriorating health. While the balance of the Florence materials is rather bland, with the exception of several nasty entries focusing on the Clemenses' landlady, it is clear that Clemens saw the work as important to his own peace of mind. Years earlier, he had retreated into work after Susy's death and produced his longest and most effective piece of autobiographical writing to date—"Ancestors.’’ With Livy's decline, he found a similar solace in the production of dictated text. In keeping with a practice that was particularly useful as he constructed Following the Equator, he attempted to infuse the later sections of the Florence dictations with an additional power and poignancy by returning to his notebooks to offer pieces of his earlier descriptions of Florence in 1892 so as to contrast with the 1904 experience. The mixture of dictation and notebook materials shows Clemens' s interest not only in his work with an amanuensis but also with the idea of placing his own past on display and of running that past head on into the present like, as he said later, "the contact of flint with steel.'' For a very brief period, that mixture fostered a collaboration with his own younger and happier self.

Livy's death on June 4, 1904, threw Clemens into personal and creative chaos. His family, in effect, became fractured. Clara entered a New York rest-cure where Clemens was not allowed to visit, Jean and Isabel Lyon spent a long stretch in the Berkshires, and Clemens himself arranged for new living quarters in New York. He continued to work, with ‘‘Eve's Diary’’ being published in 1905, ‘‘Adam's Diary’’ being revised, ‘‘3,000 Years Among the Microbes'' and a section of the Mysterious Stranger saga being begun. The prospect of gathering a circle of admiring listeners seems to have helped convince Clemens not only of the entertainment value of being interviewed by Paine for the biography but also of the creative potential offered by dictating material to a waiting and eager stenographer as grist for an autobiography that would both extend copyright on many of his earlier works by using portions as preface material and also keep Mark Twain before the reading public. Clemens adapted his notion of collaboration to fit the moment. For the biography, Paine would be able to use the autobiography that Clemens produced; Clemens himself would be able to sit and tell his stories to his audience and recapture some of the creative energy that emanated from the family gatherings at which he formerly read the day's work. Clemens would be free to weave a complicated quilt that would map his mind's wandering. He would survey his memories, introduce contemporary issues, even drop whole newspaper articles, letters, essays, and fiction into the massive collection. As early as March, 1906, he concocted an idea of an autobiography that would require an entire state to contain its volumes. The captive audience of Paine, Lyon, and Josephine Hobby, the first of several stenographers employed by Clemens, seemed perfectly to aid Clemens as he worked to recapture the creative momentum that was his while his family circle had been intact.

Paine describes the dictating sessions as taking on the aspect of performance: "We were watching one of the great literary creators of his time in the very process of his architecture. We constituted about the most select audience in the world enjoying what was, likely enough, its most remarkable entertainment." He records Clemens' s description of his return to dictating: '"With shorthand dictation one can talk as if he were at his own dinner-table— always a most inspiring place. I expect to dictate all the rest of my life, if you good people are willing to come and listen to it.'’’ Of course, they kept coming. The constellation of agendas guaranteed that each member of the little band preserved his or her own vested interest in continuing the sessions. Paine and Lyon, the two minor players in the intimate audience, were each marrying their futures to a connection with Clemens. Paine's interests lay in a projected biography, which made it imperative that he forge both professional and personal relationships with Clemens in order to assure continued access to source materials. As the biographical process continued, Paine's interest fo-cussed on maintaining the public persona, the public icon that Clemens had established for Mark Twain. On the other hand, Lyon's mixture of literary ambition—she was to be an editor of Clemens's correspondence—and personal interest—her personal affection for Clemens grew deeper as time passed—seem directed toward creating a safe environment for the aging writer. At the center of all the creative and emotional attention, Clemens immediately sensed the value of recording the re-creation of his past.

As the dictations moved ahead, they became the locus for the self-mythologizing at which Clemens was now so expert. Paine found out the hard way that Clemens was using the dictations both to review and to re-create his past. At times, Paine discovered, Clemens became caught in the web of creative memory:

It was not for several weeks that I began to realize that those marvelous reminiscences bore only an atmospheric relation to history; that they were aspects of biography rather than its veritable narrative, and built largely—sometimes wholly—from an imagination that, with age, had dominated memory, creating details, even reversing them, yet with a perfect sincerity of purpose on the part of the narrator to set down the literal and unvarnished truth... His gift of dramatization had been exercised too long to be discarded now.

Paine came to believe that the material was infected by dramatization, a belief that drove a wedge between his work as a biographer and Clemens' s as autobiographer, between the work of historian and storyteller. Clemens, it seemed, was using the contrasting approaches to break free from the traditional demands of verisimilitude placed on biography in order to move back to the by-now-familiar, perhaps inevitable, blending of life and fiction that had been the foundation for his half-century of storytelling. As Paine came to realize the conflicting approaches, he drew a distinction between the materials related to Livy and Susy as being different from the other materials.

The things he told of Mrs. Clemens and of Susy were true—marvelously and beautifully true, in spirit and in aspect—and the actual detail of these mattered little in such a record. The rest was history only as Roughing It is history, or the Tramp Abroad; that is to say, it was fictional history, with fact as a starting-point.

This is, most likely, a distinction driven by Paine's own agenda to assure his readers of the quality and consistency of Clemens' s approach to the autobiography. It certainly points to a symbiotic relationship between his and Clemens' s work which makes them individual halves that need to be brought together to appreciate the life whole and entire. Paine's focus on truth ‘‘in spirit and in aspect,’’ in other words truth based on the consistency of the teller's voice and tone, is closely allied to Clemens' s concept of a truth created both by words and the impression left for the reader to find "between the lines,’’ one which leaves at issue the possibility of verifiable truth outside the reader/writer relationship. On a more practical level, it not only sets up the biography as a necessary historical check to Clemens' s creative wanderings but also lays the early groundwork for the two-volume autobiography that Paine would publish in 1924.

What is important here is the emphasis on performance and the way that Clemens intentionally and expertly places himself at the center of a creative process that ultimately recasts the collaborative relationship as one that—unlike his earlier episodes that begged for clear advice or at least clear vocal reaction from his family—uses a silent presence to stoke his talk. He is not looking for editorial advice but, rather, for appreciative listeners who will hang on his every word. In fact, what appear to be unrehearsed comments and reactions are often best related to Clemens' s rehearsed stage appearances. Contrary to the evidence of Clemens's and even Paine's descriptions of the dictating sessions, the formal sessions were not the only venue for Clemens' s use of audience to ignite his own creativity. Clemens, in fact, augmented the emotive power of the dictating sessions by frequent rehearsals with Isabel Lyon during the morning prior to the arrival of the stenographer and the beginning of the formal dictating session. Increasingly, Clemens needed an audience—even an audience of one—to prompt, but not interrupt, his talk. The audience—the listeners to his talk rather than the readers of his writing or the commentators on his writing—aided in the production of text. It was an effective, if somewhat unusual, use of his relatively new immediate social circle.

This immediate audience, however, was not only a boon to Clemens's work, enabling him to excavate stories and to amass text. It was also one of the primary influences on Clemens's approach to storytelling itself. That small audience, closed and comfortable and congenial, for the most part adoring, lent an element of informality to the enterprise, even though that informality did not extend to the possibility of their interrupting his monologue or diverting his dictation toward conversational give and take. Clemens' s analogy to the dinner table is key to our understanding of how he played off his loyal listeners not only because it places him at center stage but also because of the subtle suggestion that they are guests who may or may not be invited again. Still later, Clemens's habit of dictating from his bed, in effect a revision of his practice of welcoming visiting reporters to conduct their interviews while he lounged comfortably in bed, suggested an atmosphere steeped in confidence, safety, and confidentiality. In fact, that atmosphere seems another revision of yet another of Clemens' s storytelling experiences, a revised version of Clemens's own memories of the storytelling that echoed in the cabin at Angel's Camp—or around the real and metaphorical camp fires of the west—or within frontier kitchens and slave quarters. Each locus of storytelling enhances the potential for a variety of tales and emphasizes the creation of a community defined and restricted, and yet somehow inviting and welcoming, by shared stories. The mixed atmosphere allowed Clemens to offer a full range of tales and observations—some of which he would eventually hold back from publication because he felt that his comfort had allowed him to range far from the content and tone that would make for proper reading. Some of these, he would announce, were being held back in order to whet public appetite, as for example when, during his first six months of work, Clemens gave his agents instructions to withhold the material generated during twenty-six of the seventy-eight dictating sessions. This entire series of tales and recollections re-connected him to the base-line value of talk itself.

As he warmed to the return to oral storytelling, Clemens used his tales to introduce a variety of voices, not only those of actual people he had known but those he himself invented, as he orchestrated the autobiography. His voices are many, and they are complex. Unlike the rather simple juxtaposition of present and past that Clemens used in ‘‘Early Years in Florida, Missouri’’ or in the Florence dictations, the dictations of 1906-1909 present a broader and more complex picture of Clemens as he approaches the tales of his own past from the perspective of seventy years and a full experience with life and death, success and disappointment at his disposal. He adjusts his own voice as he moves among the tales: he is calm and reflective when he speaks of Susy and Livy; angry and defiant when he blasts his various editors; uncompromising when he issues judgments on politicians and contemporary political scandals; shrill when he addresses the practices of an unjust deity; remorseful as he considers his own role in the financial problems that haunted his family.

Of the voices that Clemens adopted, and perhaps the most useful in establishing a relation between him and his audience (both immediate and future), was his guise as a dead man. He announced that he would play this role as early as 1904:

In this Autobiography I shall keep in mind the fact that I am speaking from the grave. I am literally speaking from the grave, because I shall be dead when the book is issued from the press.

This idea of grave speech (perhaps a genre unto itself) regulated Clemens's approach to the dictations. It is one of the primary motifs that run through the typescripts. The pose—itself a mixture of indifferent, disconnected voices—gave Clemens the opportunity to integrate his present and future audiences and to use this composite to gauge his success at creating a serious and effective story. As he wove what he hoped would eventually become a tale renowned for its truthfulness (whether that truth came out directly in his words or indirectly in the impressions that would be left to be discovered between the lines), he was aided by both his actual and imagined audience as he drew a shade of intimacy around these confidantes who were sharing the wanderings and workings of his mind. It is, of course, one more instance of Clemens collaborating with silent listeners to tease out what he felt were his most private thoughts.

Each day found Clemens ready. He would often slide into new topics in order to explore a particularly nagging question or issue. This mixture does not easily allow for easy characterization. In fact, the broad expanse of the dictations in raw form is intimidating for any reader, although their length is not as daunting as the shifts among topics. For example, from the initial burst on January 6, 1906, that begins in his thirtieth year and focuses on his Nevada experiences, Clemens detours into contemporary events. The look at his present interests keeps him solidly interested until January 19, when he returns to the topic of his experiences in Nevada. The loudest voice in this series may be that of the tour guide who is pleased to meander along wherever the paths he opens might lead. At one point, Clemens claims, "In this autobiography it my purpose to wander whenever I please and come back when I am ready.''

That self-assured and relaxed voice is at the heart of the dictations. It leads us through a series of day-long and month-long episodes that have their own particular focus. Still, only when reading the full collection of materials does a reader come away with a sense of the intellectual territory that Clemens was attempting to map. The final scheme is best described as a collection of linked tales. Clemens' s survey extends from a host of dictations in February, 1906, that deal broadly with family concerns through a series of commentaries devoted to his career, accounts that speed through April and May and June of 1906 and on into a brief series devoted to the relationship between man and God. The collection of topics and themes continues through 1907 and 1908 as Clemens offers his readers a tour not only through the events of his life but of the very process of storytelling itself. As the dictations move ahead, in fact, it is quite reasonable to see the broad sweep as an extended experiment in practical storytelling rather than as a deep analysis of a life. Throughout the dictations, we never lose sight of the ever-present audience. Neither does Clemens.

But Clemens could not sustain the full weight of the dictations himself: not even his powerful and compelling voice would support the structure he was attempting to build. And he knew it. One way to turn away from a dependence on his solitary voice was to import tales from other sources. Clemens's sensitivity to the voice in his fiction—a voice like that which resonates through Huckleberry Finn, Hank Morgan, Roxana, and Mark Twain—sent him off to scour the autobiographical materials he had accumulated as well as his own journals and notebooks. More than that, however, it pushed him to search other (and external) sources: from newspaper accounts to letters, from the morning's post to the biography of him that Susy wrote when she was thirteen years old, to the journals that Ralph Ashcroft kept during the 1907 visit to England. Clemens used all of these materials: he would carefully and scrupulously juxtapose their stories with a flurry of his own commentaries and observations. Newspaper clippings appear throughout the dictations and provide Clemens with grist for his disillusionment with American politics and policy (he being particularly fond of blasting Theodore Roosevelt). In later dictations, he turns his attention to tales he finds in letters that he receives. Susy's biography appears in twenty-six of the dictations that Clemens completed during 1906; his responses to her writing formed the most effective and most emotional recurring thread within the first year's dictalions. Ralph Ashcroft's notes begin to appear in August, 1907, and play an important supporting role in the series of dictations during July, August, September, and October, 1907, that describe Clemens's trip to Oxford. During 1908, Clemens incorporates letters and writing samples from Dorothy Quick, the model for his aquarium of Angel Fish, and follows his established pattern of excerpt followed by his own commentary.

My main point, however, is not merely the specific trajectories of Clemens's interest but rather certain patterns of dependence that Clemens establishes early in the dictating process. At its most basic, the need for external aids was a part of his career as a writer: a host of his books from The Innocents Abroad to Roughing It to Life on the Mississippi to Following the Equator relied heavily on gathered materials. With the turn to autobiography, however, his reliance on notebooks, journals, and correspondence, as well as on the physical presence of a select audience, became even more profound. His need to identify, include, and expand on external sources and influences increased as he dove into the final series of dictations. In fact, the use of bits and pieces of external materials—from clippings to full literary works—acted as a stimulant. The chorus, perhaps the chaos, of voices helped Clemens adjust the tone and vibration of his commentary. He could act as a foil for the honest and innocent and cutting observations of his children; he could adopt the swagger of celebrity to intimidate any number of literary or business adversaries. The external voices could move him to deep sentiment or mild amusement or spitting anger. At its most mundane, his collection of manuscripts simply gave him something to talk about when his creative well was low. At its most inspired, it energized his attempt to escape the constraints of time.

All of this collecting and excerpting and commenting is supplementary to Clemens's well-established dependence upon his small and captive audience. He uses both to enhance the creative process, a process that he seems genuinely incapable of engaging without the support of these collaborators. Of course, the primary voice and perspective throughout is Clemens's; however, he clearly understands and makes allowances for the difficulty of the task he has set for himself. By the time Clemens began the final series of dictations in 1906, he had moved away from writing books, except for the rare exceptions of Adam's Diary (1904), What is Man? (published in 1906 but written in 1898), Christian Science (1907) and ‘‘Extract from Captain Stormfield's Visit to Heaven’’ (1908), which were not particularly successful either financially or artistically. The dictations—a project that he originally envisioned to require some 600,000 words—was far beyond any project he had ever attempted during his prime. His ability to begin and complete the project was assured by his ability to depend so completely on collaboration.

The collaborative work on Clemens's autobiography, however, did not end with his final scratch of pen on paper for his description of Jean Clemens' s death on December 24, 1909, or with his own death on April 10, 1910. Clemens did plan and oversee the publication of a portion of his autobiographical manuscripts in the North American Review of 1906-1907. Subsequently, his materials have been presented to the public in several competing versions. The competition to define Clemens' s autobiography began in 1924 when Albert Bigelow Paine published his two-volume Mark Twain's Autobiography. Paine offered both written and dictated material and arranged the pieces in the order of composition, a plan that Clemens would have approved. Unfortunately, Paine not only presented a mere handful of dictations, material that extends from January through April, 1906, but also edited sections of the pre-1906 manuscripts that would seem to taint Clemens' s status as an American icon. A radically different version was published by Bernard DeVoto as Mark Twain in Eruption in 1940. While including material that was not part of Paine's edition, DeVoto arranges the material in thematic cluster—a useful compilation but one that altered Clemens' s own scheme. Charles Neider's revised version The Autobiography of Mark Twain (1959) takes the autobiographical manuscripts and fashions them into a conventional cradle-to-grave chronology, also violating Clemens's instructions. Clemens' s own version of his autobiography that he published as "Chapters from My Autobiography'' in the North American Review were issued as a single text in 1990 as Mark Twain's Own Autobiography: The Chapters from the "North American Review.'' Each edition is, in fact, a collaborative work between Clemens' s text—if not Clemens himself—and an individual editor. No version presents the whole of the autobiographical manuscripts; such an edition is still in the planning stages by the Mark Twain Project at the University of California at Berkeley.

All of this has about it a bit of the child's game "telephone.'' It is a simple game, but it presents us with an analogy that helps to explain the difficulty that is part of any exploration of Clemens' s autobiographical work. To play this game, children establish a sequence, e.g., a line or a circle. The first child begins by whispering a sentence to the next in line. That child whispers what he or she has heard to the next. And so on. And so on. The last child in line announces the sentence to the group after it has made the rounds. Invariably, by the time the last child speaks, the simple sentence has been transformed: "The fox ran to the tree ahead of the dogs'' becomes ‘‘The dog rested his head on the box.’’ In Clemens' s version of the game, he plays the role of several children. His first version of a tale is rehearsed before the arrival of stenographer and biographer, the dictation is recorded, Clemens edits the typescript, the tale is redone. Then editors come along and choose, edit, and rearrange the material. While individual episodes may remain intact, the whole of the autobiography may be distorted as editors place their own stamp upon and infuse their own voice into the text; voice, after all, is affected by the order and arrangement of material which often helps to set the tone for a work. Paine, DeVoto, and Neider imposed their ideas upon the materials that Clemens composed and arranged. Each offered his version of what he thought was the appropriate form for the autobiography; each started with Clemens' s text but then moved quite far afield. The line that runs from Clemens through Neider is very much a game of "telephone."

Ultimately, it is important to place Clemens back at the center of his creative process and reinforce his voice as the authoritative voice. The collaborative work that remains should now focus on expanding our understanding of Clemens's creative and autobiographical process. Examining the public text of Clemens' s life is one way to illuminate his thinking about autobiography; examining the public text against the original materials sheds even more light on Clemens' s creative choices as well as on his abilities, both strong and weak, to consider audience. We should turn away from the seductive prospect of retelling his story by adjusting his words and return to the original materials to understand the complex process in which Clemens was engaged. That approach to collaboration holds considerable scholarly and intellectual promise.

For Clemens, collaboration became a way to meet a personal and creative need. Companionship and human connection were central to the entire process of his autobiographical dictations. Dictating became a social activity, a way to keep people with him, a way to recall the intimacy of evenings surrounded by family. The very act of talking out his story brought him back to the most basic of his experiences—oral storytelling and the sharing of tales in comfortably domestic surroundings. It also brought the potential for connecting to a readership in posterity that would assure his reputation and standing as an author. It was a way to bring order and constancy to his long and often troubled life.

In the end, the complexity of the dictations suggests that Clemens actively sought order as he moved ahead with the project; the presence of a physical audience made it more likely that he would contain his wanderlust and control bundles of tales by staying within admittedly loose chronological and thematic sequences. The collaboration introduced a structure, at times a discipline. One result is that Clemens is at once navigator and guide, architect and engineer. He steers the story along the current, and he ties up at a network of interconnecting lattices. Yet neither drift nor design offers a single strand that runs from end to end. Stories connect at well-timed junctures and lead both writer and audience to an insight into the writer's process and personality with each new turn and dip. Clemens aims at painting the full portrait using an effective blend of language and silence. The silence of quiet listeners as well as the silence that inhabits the space between the lines of story becomes an effective collaborator as Clemens creates his life. Returning to this basic tale—and to the context in which that tale was composed—will allow us to look behind the myth of Mark Twain, and it will free the original tale from the editorial round-robin that has so confused our image of Clemens.

Source: Michael J. Kiskis, ‘‘Mark Twain and Collaborative Autobiography,'' in Studies in the Literary Imagination, Vol. XXIX, No. 2, Fall 1996, pp. 27–40.