The “Ashcroft-Lyon manuscript” is a 429-page diatribe written in 1909, one year before Samuel L. Clemens's (Mark Twain's) death. The text will appear in the third and final volume of his Autobiography, the current publishing endeavor of the Mark Twain Project.[1] The document is Twain’s rancorous attack on, and laborious account of, ways that his secretary, Isabel Lyon, and business manager, Ralph Ashcroft, allegedly embezzled money from him during the years 1907 to 1909, as well as manipulated Twain’s relationship with his youngest daughter, Jean, resulting in her needless residence in sanitariums for three years.[2] There have been monographs written about this manuscript as well as a number of articles.[3] The text itself has never been published but has been readily available in the Mark Twain Papers and is often cited as evidence of either Lyon’s and Ashcroft’s guilt or Twain’s overzealous persecution of them.

Putting aside the contextual life of this manuscript, editing the work itself has presented unique challenges for the editors, at least in part because it was written in the guise of an unsent letter to Twain’s good friend William Dean Howells. The editors argue that this device was an autobiographical strategy that Twain was experimenting with at the time, that is, writing a letter that would never be sent, in order to free himself from the constraints inherent in writing a letter (or anything else) that was intended for publication. While Twain was composing the final form of his autobiography, from roughly 1905 to 1909, he worked chiefly by dictating to a stenographer, a practice that he felt also allowed him a degree of freedom to say the (conventionally) unsayable, but after three years of dictating, this practice too seemed deficient.[4]

The letter form of the Ashcroft-Lyon manuscript provides us with an alternative approach to crafting an autobiographical narrative, but one the author did not continue to use after this composition. Twain did not intend for the manuscript to be a part of his autobiography, because he was meticulous about marking which dictations and manuscripts should be included in it. Also, it is noted in letters that he meant for the manuscript to be held privately and used only if Ashcroft and Lyon ever threatened legal or other action against the Clemenses.[5] Regardless, he did at times view the “letter” as a preferred way to relate a more honest account of an event from his life than he could achieve in his dictations. Therefore, including this text with the Autobiography required the editors to rethink how the text should be presented, compared to the presentation of the dictations. Since it is technically a letter, should it be transcribed in plain text as are the letters published by the Mark Twain Project? Or should it be printed in clear text as are the dictations of the Autobiography?

The editorial deliberation over these questions also raised, at least for me, a number of larger questions. How does this manuscript affect our understanding and definition of the “letter”? What do we understand letters to be when editing them for inclusion in a scholarly edition or elsewhere? Are they text of a conversation? Are they literature? Are they notes? Are they a dramatic performance? Are they private or public documents or both? Letters are sometimes all these things and more, and the expectation of what they accomplish as a textual medium changes depending on, among other things, what century of letter writing one is studying and the context of their composition. That said, they are a fascinating example of how a genre of text can evolve. So what can we learn from deliberations over the Ashcroft-Lyon manuscript, since the dynamic nature of letters is something not easy to pin down?

Within this frame of analysis, this article reviews the current thinking on the epistolary form and identifies where more nuanced thought is lacking when it comes to a critical textual study of letters being presented in different media formats. I will use Twain’s Ashcroft-Lyon manuscript and other examples to question various long-held assumptions about the “letter.”

The handwritten missive of past centuries, the “private conversation of distant friends,”[6] as Harriet Martineau described it, is typically a text that relies on the tension of being both a private and a public document. As Margaretta Jolly and Liz Stanley claim, “letters inhabit an interesting ontological as well as epistemological ‘space,’ situated as they are on the boundaries of personal [private] and impersonal [public].”[7] In fact, the argument about whether a letter is ever completely private or public has been the most pervasive and genre-defining in the study of correspondence, including attempts to theorize the so-called epistolarium.[8] Twain’s manipulation and use of the letter form is one of many examples, surely, of how it can be both. Regardless, theorists and editors often use the private-public binary as a point of engagement for discussing letters, and for choosing appropriate editorial treatments, most specifically, in deciding whether letters require a documentary or a critical approach.

Due to the perceived “documentary” nature of letters (personal, historically significant), D. C. Greetham and Thomas Tanselle, two of the most prolific writers on textual theory, devote fewer pages to choosing an appropriate theory or method of editing letters than they do to other topics of textual inquiry. However, this is very likely because letters editions are not usually considered to consist of “critical” texts—that is, texts (such as literary fiction) that require a sophisticated practice for emending a copy-text from variant readings. Since letters most often consist of just one unique manuscript, there is usually no need to choose between variants. However, this is not always true, because sent letters may sometimes have preliminary drafts to them, as well as various public printings, all of which may bear on the establishment of a text for an edition.[9] Taking this into account, one could argue that letters are literary works and should not be treated strictly as historical documents, if indeed there is a difference, for as Tanselle has pointed out, “Any approach to editing that attempts to justify differing treatment for ‘literature’ and for other works is thus built on an insecure foundation, for the simple reason that ‘literature’ is not a fixed body of material and what constitutes ‘literature’ is a matter of judgment.”[10] Admittedly, in this deliberation, Tanselle was discussing literary fiction as compared to philosophical works and the like, but the statement stands, I believe, with regard to letters as well. Editors do and should engage in judgment in the practice of editing, because it is inescapable. Therefore, attempting to assign immutable treatments for genres of writing in an attempt to deny subjective judgments about the content is a futile practice.

In fact, discriminating between “literary” and “documentary” letters, and making their editorial treatment reflect that, is a distinction rejected by Tanselle: “[T]he real distinction is between writings of the kind normally intended for publication and those of the kind normally not so intended (critical editions often being most appropriate for the former, diplomatic transcriptions for the latter).”[11] Greetham admits that this is a difficult distinction to draw for pre-Gutenberg letters,[12] but it is also a difficult distinction to draw for a number of letters in the modern era. A letter may have been a private communication at its inception, but it became public when it entered the postal delivery service, was passed among friends, or when it was (perhaps) published in a public forum such as a periodical. Also, writers who are well-known figures might well expect their letters to be published one day, perhaps even having this in mind at the time of composition. Twain, for example, did expect this, and four years before he died he appointed specific people to be editors of his letters. Determining an editorial policy based on the public-private, literary-documentary binary may actually result in editors making compromises to construct a policy that incorporates aspects of both. Nonetheless, those ideas about the genre have influenced areas outside literary and textual scholarship that are concerned with the epistolary form.

Interestingly enough, some of the current critical work on epistolary theory and the formal aspects of the “letter” are in the fields of social science and history, and more specifically genre theory, in which scholars are establishing the characteristic parameters of the letter form. Social scientists and historians who study “life writing” (diaries, letters, etc.) find that establishing or sometimes deconstructing the formal and stylistic attributes of letters helps to codify the study of them and of the culture or persons they concern.[13] However, the conclusions most often drawn have not identified any kind of useful categorization, but rather have tended to highlight the “‘infinitely malleable features’ of letters, finding that letter writing is a social practice that morphs into and across genres and across historical periods in spite of (varyingly) prescriptive conventions.”[14] Then what becomes most useful for social scientists interested in just the content of letters and how the form supports or manipulates the writer’s life story is to decipher the “truth value” of an author’s writings—that is, how much of what an author describes in letters can be counted on as truthful after the reader takes into account, for example, the style, voice, and possible performance and manipulation of voice that the author uses. This type of study may be, also, how some literature scholars contribute with their interest in letters: as a way to unearth points of formation of an author’s creative ideas, perhaps, or the identification of influences in the professional lives of writers. The most common use of letters, though, has always been to construct the biography of a figure. Simply put, a biographer, historian, or social scientist will collect details from a primary source to serve a scholarly argument or construct a life story.

Whether it is genre theory, social science, history, literary criticism, or biographical work, one area could stand more rigorous questioning of its perception of letters: that of scholarly editing, which more often than not provides other scholars with a text to study. The debate of public versus private documents alone leads to confusion, as we have seen. However, traditionally, a way for editors to determine policies for editions of any text has been to identify the definitional aspects of the text: what makes it a draft, manuscript, proofs, and so on, and how we can best, typographically, represent those aspects.[15] Then what were the author’s intentions for the manuscript, or on the other hand, what was the social text disseminated to the reading public? Various characteristics have been assigned to letters, such as: they have a writer and usually a recipient or at least an addressee, they are sent through the postal service or by some other carrier, they are thought of as covering distance and inhabiting a geographical space, they have attached to them an expectation of privacy (almost always met), they have reciprocity built in, and the writing of them has “referential aspects.”[16] However, in the end, many find that letters are still a transitionary form and a “particularly difficult text object to define: after all almost anything can be put in the form of a letter.”[17]

An editor may admit that any attempt to assign a formal or genre definition to a text is continually challenged by exceptions to that genre—an admission thrown into relief when trying to use established characteristics in order to formulate an editorial policy. Genre-defining can serve a useful purpose in some areas of critical study but with editing it can lead to quick judgments, misunderstandings, and confusion as to how a text should be published. An early attempt (within the current era of scholarly editing) at defining a methodology for editing correspondence was done by Robert Stephen Becker in 1984.[18] He surveyed the field of editions being produced at that time and looked for patterns of similar approaches to handling correspondence. In short, he found that “fifty percent of the canvassed editors would not recommend any edition of letters as a model. In general, they felt that individual problems created by an author’s correspondence are too great to be solved or even managed with a formula” (262). This survey and study grew from what at the time was a quickly moving current toward more diplomatic transcriptions of text. In editing, “diplomatic” can mean different things, but here I refer to a very general move away from heavily emended, truncated, and censored texts (e.g., in letters editions made in the nineteenth to mid-twentieth centuries), to transcriptions that showed in their printed surrogates the deletions, additions, and other marks a writer makes, for either deliberate or unconscious reasons.

This shift signaled a kind of crisis for letters editions, as touched on above in the literary versus documentary debate. Varying degrees of diplomatic transcription were a result, as was a more careful attention to the presentation of a holograph letter in a published form. Letters editions can still range from giving very literal, documentary texts, to giving clear texts with lists of textual alterations in the back of the book, and one might say that this latter approach preserved a kind of literary aura around letters editions that critical editions had held for years.

At this point it is important to remember that the resulting edited text is “already something other than what the writers in most cases intended them to be. . . . Editions of letters are so much a fixture of literate culture that it is easy to forget that in its published state a correspondence leads the second of two distinct generic lives.”[19] The recognition that what an editor creates is something different from the actual letter is important, and where an editor goes from there in the editing process matters; that is, how do the assumptions about letters, which I explained above, lead us astray in editorial decisions? Letters editions produced from approximately the late eighteenth through the nineteenth centuries were made under the theory that an edition should work toward creating one long narrative of the subject’s life, or highlight some aspect of it while suppressing others.[20] This coincidentally resulted in page after page[21] of blocks of texts, short excerpts, and all in all an aesthetic experience that appeared more like a literary tome or novel (consider the epistolary novel) than corresponding letters, which by their very nature are disjointed things: interrupted conversations, notes of reply in no context, fleeting invitations, utterances in transit sent to someone next door or across the world. In fact, most of the canvassed editors of Becker’s survey confirmed “that the regular form and spacing of print, its size on the page and the bound page itself, are all intrinsic elements of the printed medium that are foreign, even contrary, to the corresponding elements of the holograph.”[22] Could it be that the reason scholars edit volumes of letters is that they are influenced by the idea that “letters have an intrinsic value, which might indeed raise them to the level of literature”?[23] If so, as Robert Halsband surmised, “they intended their editions to be enjoyed by a reading public interested in genuine literature but intolerant of pedantry and dullness.”[24]

Fortunately, then, letters are never just gossip or informal conversation about commonplace domestic matters, nor finely crafted literary prose. They are complex textual documents that contemporary editors can only attempt to represent on a printed page or in a digital frame. It is not just a matter of representing typographically the cancellations and insertions in a letter but also a consideration of the purpose and aesthetics of them. Some might wonder why it matters, since we are only offering a surrogate to readers anyway. But what the editorial artifact gives us, through its very content, is an additional, metatextual framework of information, as well as a different aesthetic experience from the actual letter, which in turn manipulates our understanding of it and, I argue, distances us even further from what the letter accomplished decades and centuries ago. Our editorial policy and annotation policy should hope to rely on the reader's conception of the social and geographical space that letters inhabit and their varied purposes, intentions, and styles, whether it be in print or online.

A number of examples can be found among letters written throughout the centuries that challenge the conception of the so-called traditional letter, or that do not fit into the genre categories mentioned above. I by no means claim that Twain’s letters are the only letters that present challenges to the traditional definitions; rather, I imagine many letters by other writers would show the same kind of playful genre-bending that his sometimes do.

The Ashcroft-Lyon manuscript gives us an opportunity to unpack and rethink what the letter is. To begin with editorial policy: the dictations included in the three volumes of the Autobiography, with few exceptions, are printed in a clear, critical text.[25] This means the editors collated different versions of the dictations (such as one or more typed copies made by the stenographer, sometimes holograph manuscript, sometimes published text, and often one or more typescripts with Twain’s holograph revisions) and used these to establish a text, and a textual apparatus was constructed listing variants and emendations. The goal was to get as close as possible to the text Twain wanted published after his death. He indeed intended all of it to be published, some during his lifetime in modified form in periodicals, and some only long after his death.[26] In other words, the autobiography was “finished” before he died, in that he was done giving dictation, had completed some revisions to the typescripts produced, and had written what he declared to be the end of the work. That being the case, one’s initial thought could be that the Ashcroft-Lyon manuscript should be handled the same way as the dictations and published as clear text. However, it is not part of the dictations, and was written after most of the dictations were completed. It was crafted as a narrative of events in his life, but there is where its similarities to the dictations end. First, Twain wrote it himself, in manuscript. It is in more or less a rambling informal prose style, which is not uncommon for him, but the manuscript does give less appearance than the dictation transcripts do of being revised for a purpose beyond drafting. Second, he never intended for the manuscript to be formally published (which I will address further on), but he did expect unknown others to read it. Here it begins to inhabit this nebulous space of being both a private and public document. Third, it is in the form of a letter that he never intended to send. He never used this literary device in his dictations, but this form has surfaced elsewhere in his writings. So if it is a letter, what kind of a letter is it, and are we justified even calling it that (as Twain himself did) when it is first and foremost, by content, autobiographical and not “correspondence” in the traditional sense?[27] Should content or form or both decide our methodology for choosing our policy?

In the end, we opted for plain text rather than clear text for all the reasons given above as to how it differed from the autobiographical dictations. However, we soon ran into problems when trying to adhere to the specifications of plain text transcription. Plain text at the Mark Twain Project is a transcription method that was devised by general editor Robert H. Hirst for the editing and printing of Twain’s letters and other texts. Through certain typographical elements and symbols (e.g., strikethrough, slash, caret), it represents on the page the cancellations, insertions, and other substantive markings that constitute the sent letter. But it is also a faithful transcription, that is, faithful to what is literally on the page but also faithful to the intentions of the author. This means that in plain text, editors will also use their best editorial judgment not to transcribe words or phrases in a way that will confuse the reader, but rather offer them a transcription that benefits a comprehensive reading:

We assume that the purpose of publishing letters is to make them easier to read than they are in the original documents. On that assumption, a successful transcription must include enough of the text to enable someone to rely on it, rather than the original, and it must exclude enough to make the transcribed text easier to read (or at least not more difficult to read) than the original. Thus, when the documents originally sent are intact and available, we transcribe them as fully and precisely as is compatible with a highly inclusive critical text—not a literal, or all-inclusive one, but a typographical transcription that is optimally legible and, at the same time, maximally faithful to the text that Clemens himself transmitted.[28]

For the transcription of letters, this makes good sense, because as we have found, one of the hallmarks of letters is their idiosyncrasy, the nature of a text in the midst of being crafted but not polished, and this deserves a critical treatment. For the Ashcroft-Lyon manuscript, the plain text retained the integrity of the document: that of a drafted letter, loosely composed. Twain’s strikeouts, additions, many inserted documents, and tireless repetition of events and facts display quite literally his frustration and mounting anger over the period of five months when he wrote the manuscript. Also, Twain stipulates in the manuscript itself that “This original Manuscript will be locked up & put away, & no copy of it made.” Copies would have introduced other witnesses to the text and perhaps made it necessary to collate and emend a copy-text, but with a unique manuscript, plain text is all the more called for. However, using the plain text rules that were established for editing letters did not always allow us to transcribe the manuscript satisfactorily. Much of this challenge came from the extraordinarily diverse components of the text. It is a long manuscript “letter,” but it also includes newspaper clippings, testimonies written by the author’s daughters, Clara and Jean, legal contracts, household bills and accounts, and typed documents. Are we justified in treating these as plain text items as well? Do we retain the common misspellings that occur in newspaper articles where clearly the typesetter was inept? Do we treat Clara’s and Jean’s manuscripts as we treat Twain’s? This is where it became difficult to treat this “letter” like other letters, relying on preconceived notions. What resulted in this case was using plain text but also making allowances for the other types of text included so as to produce a faithful text, a historically accurate text, and one that does not obscure the author’s intentions.[29]

Twain’s purpose behind writing this document was the other guiding force for the editors in deciding how to treat it. Why did Twain choose to unburden himself of all his accusations against Ashcroft and Lyon in an “unsent letter” to Howells? His intention was not to publish it, but rather to keep it so that the family or Twain’s lawyer would have it at hand in case Ashcroft or Lyon ever tried to take legal action against the family. That way, even after his death, Twain’s daughter Clara would be protected with the document. But why the letter form? He had been attracted for many years to the catharsis of the “unsent letter.” It was a way of unburdening himself of a topic that might not be suitable for others to read. He expresses this quite well in an actual sent letter to Howells written just a few days before he started composing the Ashcroft-Lyon manuscript:

My mind’s present scheme is a good one; I could not like it better if I had invented it myself. It is this: to write letters to friends & not send them. I will now try it on you as a beginning, & see how it works. Dictating Autobiography has certain irremovable drawbacks. . . . Talking to the same, same, same old one-character stenographer all the time, is talking to the vague—with [unsent letters] there’s no definite target for each letter, & you can choose the target that’s going to be the most sympathetic for what you are hungering and thirsting to say . . . you can talk with a quite unallowable frankness & freedom, because you are not going to send the letter. . . . When you are on fire with a good thing that’s indecent you’ll save it for Howells, who will love it. As he will never see it, you can make it really indecenter than he could stand; & so no harm is done, yet a vast advantage is gained.[30]

Twain considered the Ashcroft-Lyon affair to be indecent because in his own words, directed toward Lyon but written in a letter to his daughter Clara, it was about someone who was “a liar, a forger, a thief, a hypocrite, a drunkard, a sneak, a humbug, a traitor, a conspirator, a filthy-minded & salacious slut pining for seduction & always getting disappointed.”[31] He chooses this “freer” and more honest form of the letter, but throughout the manuscript his language toward Ashcroft and Lyon becomes increasingly violent and exaggerated, and he begins to utter not a number of honest statements but rather half-truths and sometimes complete falsehoods.[32] He claims to be his most honest self in this text, but instead he uses it to give voice to his version of truth in whatever language he chooses, which could admittedly be another kind of honesty. This in turn lends itself to the “truth value” study that social scientists are concerned with in life writing, which also is of concern to editors. The interesting part of Twain’s use of the letter form in other instances is that he often bends the truth in order to effect a response from the reader, a literary device found in several of his writings.

One example that displays Twain’s use of a letter to compose a type of fiction is one he wrote in 1874 to, again, William Dean Howells.[33] It gives the reader a multilayered address through the “correspondence” of a letter addressed to his wife (“Dear Livy”) but sent to Howells and dated 1935. It is a “prophecy,” as Twain later termed it, a fictional letter to Livy written when Twain would have been one hundred, Livy ninety, and Howells ninety-eight. He sent this letter as an exercise in “forecasting the Monarchy and imagining what the country would be like when the Monarchy should replace the Republic,” revealing his cynical view of how one powerful political party in America essentially functioned as a monarchy (Autobiographical Dictation for July 16 and September 12, 1908, forthcoming in volume 3). He imagines that the Irish have taken over politically on the East Coast and have renamed some of the major cities, Boston being named “Limerick.” He and his close friend the Reverend Joseph Twichell are visiting Howells in New York and Twain reports on the trip to Livy. Most of what he says involves the technological advances that will be at their disposal in 1935, such as “air-ships,” and even more telling, he gives his thoughts on advances in communication. It is in a sense a monologue on what he finds to be the benefits of the “old” written word, transmitted through letters and telegraphs, versus the new vogue in 1935 of “thought transference,” when people will telepathically transfer their thoughts to one another instead of writing:

The gentlemen-in-waiting stare to see me sit here telegraphing this letter to you, & no doubt they are smiling in their sleeves. But let them! The slow old fashions are good enough for me, thank God, & I will none other. When I see one of these modern fools sit absorbed, holding the end of a telegraph wire in his hand, & reflect that a thousand miles away there is another fool hitched to the other end of it, it makes me frantic with rage; & then am I more than implacably fixed & resolved than ever, to continue taking twenty minutes to telegraph you what I might communicate in thre ten seconds by the new way if I would so debase myself. And when I see a whole silent, solemn drawing-room full of idiots sitting with their hands on each other’s foreheads “communing,” I tug the white hairs from my head & curse till my asthma brings me the blessed relief of suffocation. In our old days such a gathering talked ^pure^ drivel & “rot,” mostly, but better that, a thousand times, than these dreary conversational funerals that oppress our spirits in this mad generation.[34]

This letter defies many genre staples. It serves as a communication to three or more addressees, and is not private to Livy. It was sent to Howells, and he in turn “passed it around” to Thomas Bailey Aldrich and James R. Osgood.[35] Twain then publishes the letter himself by including it in his Autobiography, in one of his dictations, thirty-four years later. It was first a creative exercise, in 1874, and then a platform from which to divulge his views on the country’s political system, in 1908. It was never just a letter to Livy; rather, it displayed Twain’s early fascination with the letter as a malleable literary device.

One other instance of Twain’s rumination over the unsent letter is a humorous sketch he wrote but never published, titled “Unmailed Letters,”[36] advice to a young letter writer. It explains his idea that there can be two versions of letters: an honest, bilious unmailed version and a truncated, dishonest sent version. Writing these two versions was a practice Twain saved for recipients such as reporters, reviewers, and overzealous fans who sent him endless obtrusive requests.

When you get an exasperating letter, what happens? If you are young, you answer it, promptly, instantly—& mail the thing you have written. At forty what do you do? By that time you have found out that a letter written in a passion is a mistake, in 99 cases out of a hundred, that it usually wrongs two persons, & always wrongs one—yourself.

Instead, he gives the young writer advice to lay out all his passionate outrage in a letter and stick it in a pigeonhole instead of mailing it, and then write a letter void of offense and mail that to the recipient. This advice is followed by amusing examples of such versions. Yes, this is a humorous exercise, but it speaks to a larger point, that the so-called Ur-text of a letter can have various limbs, expressions, and versions, and be used for various purposes.

By 1909 and the writing of the Ashcroft-Lyon manuscript, Twain is very interested in the so-called truthful form of letters, and he engages in a conscious manipulation of the form when he writes the manuscript to Howells. His preface to the letter (note that the use of a preface blurs the lines between its being a literary manuscript or a letter) reads:

I have set this history down in the form of a Letter—a letter to an old & sympathetic friend, a friend of thirty-five years’ standing, the novelist William Dean Howells. This was to give me freedom, utter freedom, limitless freedom, liberty to talk right out of my heart, without reserve. I could not talk like that to the general public, I could not strip myself naked before company.

Here and elsewhere he believes in the ultimate freedom found in letter writing. In 1905, when Twain and Lyon were still close, he spoke to her about the kind of honesty that can only be given in letters, and he added the necessary component of an addressee or an interlocutor to his theory. A recipient, or at least the concept of a recipient, provided Twain with the accepting audience he required in order for his communication to be genuine. In her diary entry for June 2, 1905, Lyon recorded that Twain said the following:

People think things—know things, but they don’t dare to say things. If a man pokes his head up over the crowd & dares tell the truth, he has potatoes hurled at him by people who know that what he says is true. . . . [but] he never dared to burst out with the things that he could put into a letter—a letter to a man like Twichell—for Twichell is sympathetic—They wouldn’t be things you could print, except to weave into a volume of letters. Letters would be the best kind of biography any way.[37]

Consider this along with the fact that writing a letter is, yes, to (sometimes) correspond with someone else, but it is also a performance for an interlocutor. For, as Bruce Redford has pointed out, the letter is a text feigning a conversation, thus a performance of conversation and rhetoric.[38]

This is key, because more than anything else it is this feigned interaction that attracted Twain to the letter form. Consider what he was doing with his autobiographical dictations; he was performing for and interacting with his different stenographers, whom he prized for their skills as sounding boards for him. (This list includes his first biographer, Albert Bigelow Paine, and the later-accused Isabel Lyon, to whom, before their falling out, he would read aloud sections from the dictations.) Lyon wrote in her 1906 diary, “Mr. Clemens was saying that Mr. Paine and Miss Hobby [the stenographer] make a good audience, and quite enough too.”[39]

They provided him with listeners and acted, in a way, as an audience. However, he still found these “audience members” as lacking in certain respects. In the previously quoted April 17 letter to Howells, Twain wrote, “A stenographer is a lecture-audience; you are always conscious of him; he is a restraint . . . you are not talking to yourself; you are not thinking aloud—processes which insure a free & unembarrassed delivery—for that petrified audience person is always there, to block that game.” With the Ashcroft-Lyon manuscript, however, the correspondent (Howells) is not there sitting in front of him, and this fact gives him the best of both worlds, a person he can still perform for in the text (in the form of asides he writes to Howells throughout) but who isn’t there reacting in person and implicitly encouraging Twain to censor himself. Howells, or the concept of Howells, is an audience that will “beguile me into dwelling leisurely & lovingly upon it, & enjoying the taste of it in my mouth” (ibid.).

Through the Ashcroft-Lyon “letter,” he gives us a bridge from the traditional letter to that of prose autobiography. Twain’s manuscript does not fit into the parameters that current scholarship attempts to assign to letters: that the sender writes with the thought that the letter will travel through a specific geographical space, and that there is an expectation of privacy with one recipient. Twain has both these characteristics in mind and works toward a hybrid form. First, he takes away the idea of distance and keeps it local, writing to someone who is real but present only in his imagination. Second, there is an expectation of privacy to this letter but not completely: he is not, albeit imaginatively, just corresponding with Howells, because his daughters (Clara and Jean) are also involved in the composition of the manuscript by giving him their accounts of events. It is a collaborative, performing correspondence that he wants to be kept private for the time being, but that he knows one day will be read by others. Consider that his preface to the manuscript is addressed “To the Unborn Reader,” for he thinks that one day the text will be read by a stranger, be it in a court of law or by the general public. Suffice to say, with this statement he is introducing yet another correspondent, an unknown “other” that he is likely also performing to. Therefore, we have a three-layered addressee, similar to his 1874 letter to Livy and Howells.

So, to bring us back to where we started, policy: with all of these observations and considerations in mind, the editors of the Ashcroft-Lyon manuscript used plain text because it was the truest option for us to represent this document, but when we make these decisions we are weighing ontological arguments about what form is and how it is considered against the significance of the content. This exercise of editing the Ashcroft-Lyon manuscript helped me to more closely evaluate what a letter is and what it may accomplish in different authors’ hands, and consequently how an editor’s decisions could be affected by these conceptions. Editors will, one hopes, continue to weigh considerations with regard to form, content, and aesthetics when producing editions of letters, whether they be in print or online. Mark Twain as letter writer is only one example of an author producing unorthodox “letters”; therefore, there must be a considered regard in editorial scholarship for how different kinds of letters look and how they are read. One assumes that more options will be available in terms of rendering letters digitally as technology advances, but efforts in this vein should begin with a closer look at what letters accomplish beyond our preconceived notions.