University of Virginia Library

Editing the Letters of Letter-Writers
by
Robert Halsband [*]

DURING THE LAST FEW DECADES SO MANY NEW editions of letters have been produced that any editor contemplating still another must pause and take his bearings. In the edition of Lady Mary Wortley Montagu's letters which I am now preparing I hope to be guided by the virtues and warned by the faults of others before me. But if in my remarks here I raise questions which I do not answer with definite conviction, it is because my editorial resolutions are still somewhat flexible.

Certainly it is easy enough to criticize nineteenth-century editors of letters. We are often impatient with their methods—their casual attitude toward accuracy of text, their skimpy and superficial annotations, their partisan attitudes, and their diligence in excising any word or passage that might bring a blush to a maiden's cheek or a blot to a family escutcheon. Perhaps we can regard them more kindly if we recall what they conceived their function to be. They intended their editions to be enjoyed by a reading public interested in genuine literature but intolerant of pedantry and dullness. These editors did not require patronage or subsidies—from either well-endowed angels or universities; they prepared their handy volumes for a reading public who bought and read.

It is inconceivable, for example, that—aside from the indelicacy of publishing a living writer's complete letters—there could have been produced in the nineteenth century such a work as Professor Haight's recent seven-volume edition of the George Eliot letters. Let me say immediately that this edition is admirable as an editorial job; its only fault, reviewers have agreed, is that the letters are terribly dull. Although it is a useful work for the study of nineteenth-century literature, and will no doubt prove its value to students of George Eliot, of the novel, of literary economics, of German influence, of medical history,

its warmest defenders will hardly allege that it will be read by any one whose interests are unconfined by any of these special subjects.

Here, I believe, is the prime difference between letter-editing in the nineteenth century and today. In our view letters have become "documents"; and the editor, instead of presenting a literary work, is setting up an archive. The long shelves of books are no longer meant to be read but only to be referred to. Is this not corroborated by the proliferation of indexes in recent editions? Professor Haight has been temperate in allowing himself one uncomplicated index; what are we to think of Dr. Chapman's three-volume edition of Johnson's letters with its seven immense indexes? If index-learning turns no student pale, how fortunate is the student who holds the eel of science by seven tails!

Our editions differ from the nineteenth century's in another related way: our tendency to "monumentalize" editions of letters, to provide—as Dr. Chapman's reviewer in the Times Literary Supplement writes— "an edition which shall be, humanly speaking, definitive and final." The sceptic may well wonder how definitive and final this edition can be when each volume has a section of addenda, and there is even a last minute inclusion in the preliminary leaves of Volume I. Our concept of definitive editions—which nineteenth-century editors did not try to produce—is perhaps related to the advertisements one sees by beauty parlors which offer permanent waves "guaranteed for three months." Our desire to make these editions all inclusive leads us to throw into them every scrap of writing by, to, and about the writer, every collateral and associated reference however remote; and so annotations grow in length and appendixes in number.

Besides differing from us in their attitude toward readability and toward completeness, nineteenth-century editors necessarily reflected the genteel tone and taboos of their time. Thus in 1861 when Moy Thomas edited Lady Mary Wortley Montagu's letters he was confronted in one of her gossipy ones with a scandalous anecdote, and rather than discard it—for it makes a clever, vivid episode—he dropped it into a footnote and labeled it an excerpt from an anonymous friend's letter to her. Of course, following the editorial practice of his time, he corrected her grammar and usage like a fussy governess; and he emasculated or omitted her frank expressions and racy anecdotes. Today, fortunately, Victorian prudishness has given way to plain frankness. No word is too crude for the editors of Walpole's correspondence to transcribe or translate. The editor of Hart Crane's letters, a few years ago, published the full record of that poet's neurotic existence—only

twenty years after his death. There can be no difference of opinion as to the desirability of transcription without bowdlerizing; for in an age when four-letter words unabashedly litter our fiction, and Freudian theory has passed into everyday lingo, the editor need not be a verbal or moral censor.

Before discussing specific questions of text and annotation, I should like to discriminate between what we may consider the letters written by writers who are noted for their other writings and the letters by writers noted mainly for their letters. This distinction, I believe, has some bearing on the form of the edition. For the eighteenth century—which I am primarily concerned with—we may consider as letter-writers Lord Chesterfield, whose letters were last edited by Bonamy Dobrée in 1932; Horace Walpole, whose correspondence under the urbane general-editorship of Mr. Lewis has been in progress since 1937; and Lady Mary Wortley Montagu, whose letters, last edited in 1861, still await complete publication. For these three writers, letters are their main claim to a place in literary history; hence they can be classified as letter-writers rather than mere writers of letters. The letter as a genre was exceptionally congenial to their talents, and they expended on it all their self-conscious, disciplined craft. In contrast, there are the mere writers of letters—Addison, edited in 1951; Steele, edited by Professor Blanchard in 1941; and Pope, whose correspondence has recently been edited by Professor Sherburn. Unquestionably these letters are important in literary history, but it is at least debatable whether they are themselves literature. To Addison and to Steele letters were a functional means of transacting business or sending messages. Addison's letters smell of the Public Record Office, where many of them indeed lie; and doesn't one become a little tired of the repetitive "Dear Prue" notes of Steele?

Between these two categories—letter-writers and writers of letters—fall Thomas Gray, edited in 1935, and Johnson, edited by Dr. Chapman in 1952. Neither Gray's nor Johnson's stature comes from their letters alone, but it is raised by them. Gray's scintillate with the lively intimacy repressed in his external life; and Johnson's display the drama of a combative many-sided personality. Many readers, no doubt, go to these letters not for any particular fact but for the enjoyment of reading.

And now to the editor's problems—to his "job of work." He must first decide what to include in his edition. This problem is, today, less obvious than it would seem. All of us agree that every letter by the writer should be included, whether it survives in manuscript or only

in a printed copy. The headaches that afflict editors of printed works in deciding on a correct copy-text are spared the letter editor: the manuscript has prime authority, and where it does not survive, then the printed text closest to it. In printing every letter, the editor may have to welcome brief and inconsequential notes. It is Robert Browning's recent editor, I believe, who was confronted by scores of postal cards on which the poet had merely accepted or declined dinner invitations. Fortunately eighteenth-century letter-writers did not bother to write letters unless they had something to say in them. Johnson remarked to an old friend, "I find myself very unwilling to take up a pen only to tell my friends that I am well." And where only brief notes survive, they are valuable for filling in sparsely documented biography.

If most editors print all the letters by their writers, there is some disagreement in the matter of printing letters to or even about them. Professor Haight calls his edition The George Eliot Letters, yet only two-thirds of the letters come from her pen. The editions of Gray and of Walpole, called correspondences, print letters to and from these writers; so does Professor Sherburn, although since he also prints letters by Pope's friends to each other, his edition may be called "Pope and His Circle." A more extreme example of this tendency is Theodore Besterman's edition of Voltaire's correspondence, now in progress; and that collection, one may observe, is not so much an edition of letters as a collection of materials toward Voltaire's biography.

The practice of giving both sides of a correspondence—as we see in the Walpole edition—has obvious advantages; it also has the disadvantage of swelling the volumes with a great mass of material which is dull and only of oblique literary interest. Horace Mann's letters to Walpole illustrate this; and his editors even print the long official documents which Mann occasionally enclosed. If this correspondence is intended to be a political history of the eighteenth century, the plan is sound; but it is difficult to see its literary justification. Mr. Lewis admits as much, when he writes (in Collector's Progress), "From the first the Yale Walpole has been planned as an encyclopedia of eighteenth century life and thought."

Any decision as to what to print rests ultimately on two factors: what is the edition's purpose and what materials are available. If I may discuss, for the moment, my own concern with Lady Mary's letters, this is how it works out: I wish to present, essentially, her own writings—not because they are a source of historical, literary, social, or biographical information, but because they have value as literature. (I need not take time now to say why they can be considered so; we can accept the

sanction of critical opinion since her own time.) In round numbers I have about 800 letters written by Lady Mary, more than twice the number printed by her last editor. Of these 800, about 650 survive in manuscript—which will give my text a solid and reliable basis.

What of the surviving letters written to her? Should they be printed in the text? I have about 90 from her husband, and about 60 drafts of letters sent to her by the Countess of Oxford. These two series are so dry and dull that they would, if printed, infect the book with a deadly tedium. I plan instead to quote from them, only where relevant, in my annotations. Lady Mary's correspondence with Pope presents a different problem. There are about 20 of his letters to her; and although they illustrate Pope's inflamed imagination and one of his epistolary styles, they tell us very little about Lady Mary's lost letters to him. Since Professor Sherburn has carefully printed from manuscript this series of letters from Pope, there is no point in my duplicating the text. The only surviving letters from Lady Mary to Pope are two very brief notes sent after his effusive friendship and correspondence had abated. These two will appear in my text, in accordance with my rule that all of her letters deserve a place there.

Her correspondence with Lord Hervey raises still a different problem. For over twenty years they maintained a lively friendship, usually sympathetic but sometimes malicious. (Lady Mary is responsible for the epigram that there are three sexes: men, women, and Herveys.) After his death his son courteously sealed up all of Lady Mary's letters and returned them to her. Not a single one has come down to us. But of Hervey's letters to her, I have his manuscript copies of 43 which have never been printed. Many of them are direct replies to those she sent him, and so they tell us something of her lost letters; they are, besides, of great interest in their own right, displaying the copious talent and skill that make his Memoirs of the Reign of George II such a brilliant document. I plan to print them as part of my text; or, if I lose heart at thus flouting my own rule, I can assign them to an appendix.

Having decided, in general, that only Lady Mary's own letters belong in the text, and that those to her belong in the annotations, I must next decide how to arrange them: in a single chronological sequence or in separate correspondences. Mr. Lewis's edition of Walpole and the Nichol Smith-Cleanth Brooks edition of Bishop Percy's letters follow the latter arrangement. Mr. Lewis's defense of it for his edition seems to me persuasive for two special reasons: that both sides of the Walpole correspondences are very fully printed, and that Walpole channeled his four main interests among particular friends—so

that each series has an inherent unity. But editors of both letter-writers and writers of letters generally prefer the single sequence. The fascination of reading these collections comes from seeing the gradual unfolding of their writers' lives and ideas, as well as their coherent comments on topics which caught their interest.

If editors agree on the arrangement of their letters, they diverge sharply in the matter of how to present their texts. The root of the divergence is the fact that the copy-text is a manuscript. The editor must decide, essentially, whether he wishes to put the letter into the reader's hands as a manuscript or as a printed text. Dr. Chapman writes in his introduction: "I have aimed at reproducing my manuscripts as closely as typography admits." But does he? I notice, for example, that he does not reproduce Johnson's long s. I believe, perhaps naïvely, that the exact reproduction of a manuscript is impossible. Even if we used a photo-process, we should then begin to worry about the color of the ink, the quality of the paper, the manner of folding the sheet, and so on. When we decide to reproduce it by means of typography, we have made a great concession; and once having made it we need not be stingy as to its extent. No reader of the book will be fooled into thinking he has a manuscript in his hand.

Dr. Chapman, of course, is neither the innovator of this point of view, nor even its most extreme champion. The Toynbee-Whibley edition of Gray prints long s's, abbreviations, and raised letters; it is fussy and finicky about every detail of the manuscript. It does not indent, so that we sometimes come up against a tight-packed, full page of dense type without a crack in the text to allow us a breath. As a reference work this edition is excellent, but who reads it? At the other extreme of textual style—that of normalizing—stand the newest editions of Walpole and Chesterfield. Mr. Lewis's practice is to retain Walpole's spelling of proper names and (careful) punctuation, but to normalize other spellings and capitalization. Professor Dobrée also normalizes spelling and capitalization. His object, he writes in his preface, has been "to make the text as pleasurable as possible to the reader of today, while altering as little as possible." As a general principle this makes good sense.

It seems paradoxical that political and social historians—who, one would think, are sticklers for exactness—should prefer normalized texts, whereas literary historians strive for exact transcription. I can only conjecture why this is so: the former group want the letters set forth as factually exact as possible but without any distracting irrelevances because to them the facts are paramount; literary historians believe

that they will lose the nuances of style if the text is normalized, and that these small details convey the flavor of the writer's personality and milieu. Between these extremes there are other sensible styles. Nichol Smith writes in his edition of the Burnet-Ducket letters: "It has been found desirable to expand the simpler contractions and to add occasionally to the punctuation"; but in his Swift-Ford letters (twenty-one years later), although he expands contractions he retains raised letters for titles—a confusing procedure when footnote numerals are also raised. (One would think a title like Lord or Duchess simple enough to expand.) Professor Tinker in his 1924 edition of Boswell's letters has normalized except for spelling; and it is only necessary to glance at his page to see how much is gained and how little lost.

What kind of text—to come to my own job—do I plan to use for Lady Mary's letters? I am fairly certain that I do not want a facsimile manuscript reproduction. Nor do I want the letters to look as though they had appeared in this morning's Times. My solution, a compromise, comes from these factors: that the edition will be published by a university press, and so must be scholarly and reasonably complete; that as a great letter-writer Lady Mary will attract readers beyond the ranks of eighteenth-century students and scholars; and that her letters, as manuscripts, raise several peculiar problems.

I have therefore decided on the following style. I shall reproduce her exact spelling and capitalization. The range of her spelling—extending over more than fifty years of her activity as a letter-writer—illustrates the gradual evolution of her writing from that of a girl haphazardly and self-educated to that of a poised woman of letters. Her spelling of family and place names is historically useful. Her use of capitals for nouns, though unsystematic, was often her way of achieving emphasis. These differences (in spelling and capitalization) from twentieth-century norms do not impede the reader, for he finds them in the printed texts of the eighteenth century. In regard to capitalization, Mr. Lewis states that it is often not possible to be sure when the writer intended a capital; but an editor, who is presumably best qualified to read the manuscript, is least likely to make this error in transcription.

With all abbreviations (including the ampersand) and with raised letters I intend to be pitiless: the contractions and abbreviations will be expanded in full, and the raised letters lowered to the plane where they belong. (Proper names will be expanded within square brackets.) These niceties are of no more importance, it seems to me, than an extra flourish of the pen or the placing of a date on the left or right

hand side of the manuscript page. It may be objected that abbreviations are significant in showing the degree of familiarity or of formality in a letter, for letter-writers tended to avoid abbreviations in their formal notes. But surely we judge the tone of a letter by its actual words and ideas and not by its flourishes and shorthand. As the reader picks his way through stenographic symbols like D s for Duchess or y m for them, he may forget the thought. Even Dr. Chapman, so faithful to his manuscripts, does not reproduce Johnson's "displayed" conclusions because he believes they would waste space and offend the eye. Professor Sherburn, with more precision, uses vertical lines (as in the bibliographical description of title pages) to set forth the conclusions of Pope's letters; but it is difficult to see any advantage beyond mere exactness. Nor do I see any benefit in indicating the pages of the manuscript letter, as the editors of Percy's letters do by printing marginal glosses. As a result their page is disfigured, and the reader is burdened with useless information about the physical condition of the manuscript.

How should the editor treat the writer's erasures (assuming he can read them) and inadvertent errors like repeated or omitted small words and syllables? Dr. Chapman neglects Johnson's erasures but preserves his inadvertences "partly because they furnish some indication of his state of health or his state of mind, partly because they show the sort of error to which he was prone. . . ." As good or even a better case can be made for preserving the one and neglecting the other; for the erasures show us the writer's literal intention, whereas his inadvertences provide an incalculable number of speculative possibilities. When Johnson repeats a syllable in carrying a divided word over to a new page, what does that signify? Perhaps merely that a fly alighted on his nose and distracted his attention. For both types of lapses the editor need not set himself any rigid rules. In many instances he can avail himself of square brackets to insert what the reader does need, and common sense to omit what the reader clearly does not.

My most radical treatment of the text is this: I shall add punctuation where the sense demands it, and I shall capitalize the first word of every sentence. Lady Mary's custom in these two respects is rather cavalier; its exact reproduction would be meaningless. I have noticed that when I read a text exactly reproduced from manuscript, a sentence ending without a period or a sentence beginning with a lower-case letter stops me in my tracks, and my eye swings forward and back to catch hold of the sense. Careful analysis of the clauses usually determines the sentence sense; and if the editor does this first, then the reader can read without interruption.

The exact typographical copy of a manuscript has this ironic effect (as Mr. Lewis shrewdly remarks); it "imparts an air of quaintness to a text which was not apparent to the correspondents themselves." For their letters printed in their own day do not challenge the modern eye, as we can see in any eighteenth-century collection. The text-style I have chosen, then, is a compromise, but I believe a sensible one—something between eighteenth-century printed and manuscript styles and at the same time hospitable to the modern reader. So much for my text. These matters of spelling, capitalization, abbreviations, and punctuation are relatively trivial compared to the fact that every word written by Lady Mary in her letters will be clearly reproduced.

A manuscript letter contains more than simply the writer's message; and the editor must decide how to deal with this other material. The recipient's name and the date of the letter, although they may not appear on the letter (and are hence editorial intrusions) are obviously a necessary adjunct to the letter if it is to be intelligible. The postmark, if any, should be recorded, for it is evidence that the letter was actually sent, and may have some bearing on the letter's date of composition and of transmission. But I see no reason why it should be with the text when it can be dropped into a footnote. The endorsement, since it is also part of the letter, should also appear on the page—under the text, if by the writer; as a footnote, if by the recipient. Most of Lady Mary's letters to her husband were endorsed by him with a date and a summary of their contents, and this information is often clarifying.

Does the editor have to record which passages in a letter have been struck out by the recipient or subsequent owners of the manuscript? Here his own judgment takes precedence over definite rules. Of course he will always decipher and print the passages. Any mutilation of the manuscript likely to mislead a later user can be noted at the foot of the page. This editorial problem does not exist in Lady Mary's letters, for her heirs, instead of using scissors or india ink, more effectively censored the letters which they disapproved of by burning them. Fanny Burney censored her father's letters, Mr. James Osborn tells us; there is good literary justification for an editor's noting her editorial treatment. The recent owners of Boswell's manuscripts censored before selling; and to record their attempts at suppression may illustrate Victorian prudery, but nothing is thus added to Boswell's letters. Mr. Lewis's edition of Walpole points out what has been omitted in previous printed editions. Although this unwittingly focusses the reader's attention on the scandalous and off-color anecdotes, it adds nothing to

the letters. Any reader interested in what earlier editions have omitted can simply collate the printed texts.

With a manuscript letter in his hand the editor can disregard a previously printed copy. But if he cannot find the manuscript and must depend on a printed text, he must be grateful—and very cautious. In my own work on Lady Mary's letters I have had a variety of illuminating experiences. Moy Thomas printed a letter from Lady Mary to her daughter, but he could not find the original among the Wortley Manuscripts, and so he took his text from Dallaway, Lady Mary's first acknowledged editor (in 1803). The letter contains an elaborate anecdote about William Pitt's ministry, and Thomas footnotes it thus: "It is remarkable that Lady Mary herself appears to have appropriated the passage from a letter of her friend, Miss Tichborne"; and he then quotes the identical passage from the friend's (manuscript) letter. But the truth of the matter is that Lady Mary's letter—which exists among the Wortley Manuscripts, though Thomas could not find it—does not contain the anecdote. Dallaway had simply transferred it from the friend's to Lady Mary's letter. Thomas should have been on guard, particularly since in his preface he accuses Dallaway of flagrant tampering.

Of greater importance than this kind of conflation—which an editor can discover by carefully checking his manuscripts—are those letters which exist only in printed form and without clear authority. The editor of Lady Mary's letters must deal with this situation: her Embassy Letters were published posthumously and without her family's permission, but their authenticity is beyond doubt since her own manuscript still exists. Four years later the same publisher issued an Additional Volume containing six letters, of which one had been published long before (in 1719). The other five, I do not doubt, are spurious, yet they have since then been printed among Lady Mary's letters. Moy Thomas put them in his text, labeled as of doubtful authenticity, but that has not prevented careless literary historians and anthologists from using them as examples of Lady Mary's epistolary art. How shall I treat them in my edition? Probably consign them to the purgatory of an appendix.

What of each letter's provenience as a manuscript and its bibliographical history? Most modern editors put these facts at the head of the letters. I believe that this information, although interesting and perhaps important, is primarily antiquarian; and so instead of cluttering the text to the reader's distraction, it can be neatly dropped into the footnotes.

Having set up the text of his letters the editor must then wrestle with his annotations. In this bout he will find even less agreement than in textual style. "It is impossible," wrote Dr. Johnson, "for an expositor not to write too little for some, and too much for others." Mr. Lewis's Walpole annotations are considered by many to err on the side of generosity; and in England one hears quips about the full genealogies attached to everyone mentioned in the text. This tendency to annotate Walpole widely and deeply has become more pronounced in the Mann correspondence, so that the volumes are becoming not a source for political history but a veritable reference collection. Whether or not one agrees with this objection, there can be no doubt as to the balance, tact, and efficient accuracy of these annotations.

Although the editor may set up formulas to determine what requires elucidation, he must actually judge each instance individually. When Lady Mary writes of Gulliver's Travels: ". . . and very wonderful it is, God knows!" should the reader be referred to the Bible passim or to Sir Harold Williams' bibliographical study? But if she refers to an obscure politician, the reader should be told his identity and referred to the DNB. If you will recall my initial premise—that the reader wishes to read the letters—then the annotations must remain subservient. They must not be tantalizingly brief lest they generate their own puzzles. Nor should the editor disregard a name or quotation or allusion in a letter because—he may rationalize—his main job was to present the text.

It would be easy, but not really necessary, to quote examples of faulty annotations. Let me instead give a couple of samples of how I intend to deal with this. In a letter from Avignon, Lady Mary writes to Lady Oxford in October 1744 that she supposes the slowness of the mails is "occasioned by the cessation of correspondence between Dover and Calais." My footnote reads: "War between England and France had broken out on 15 March 1744." Merely that. It is not necessary to give the causes or battles of the war, the organization of the post office, or a survey of postal routes in 1744. In the same letter Lady Mary sends her opinion of Pope's will; and in my footnote I quote the few sentences from Lady Oxford's letter which had told Lady Mary the news. I doubt that more is required, for this is not a biography of Pope or a study of his reputation; it is a letter by Lady Mary.

Should passages in foreign languages be translated in the footnotes? Most editors have conceded that the reader of even scholarly editions needs to have Latin translated for him. (Professor Sherburn's edition of Pope does not make this concession.) The relatively few Latin quotations in Lady Mary's letters will be translated in the footnotes. And

like all these other editors I do not intend to translate French passages. To the reader unfamiliar with that language, about twenty newly discovered letters in French by Lady Mary will have to remain buried in the obscurity of a modern language.

I hope Dr. Chapman's extensive use of abbreviations in his letterheads and footnotes does not win many imitators. (So far, Miss Norton, in her edition of Gibbon's letters, has imitated.) The reader must fight his way through an alphabetical thicket, for he is forced to expand initials into names so frequently that reading gives way to mnemonics. The elaboration of short cuts can reach a point of diminishing returns. I also confess myself irritated by Dr. Chapman's abbreviated year dates —'74 for 1774. Only one space is saved, and again the reader's attention is distracted by having to fill in what has been omitted. The editor's job, after all, is to remove obstacles, not to create new ones. For my own abbreviations: since I have such mouth- and line-filling names for Lady Mary and for her husband, I designate them as M and as W. But there I draw the line. Where sensible and useful, the editor can use short cuts. The letters, for example, can be assigned numbers, and these can then be used for cross references. If the letter-numbers are printed as shoulder-notes on each page, they can easily be found by the reader.

The freedom and elasticity of footnotes are a great temptation to a vivacious editor. He can fill them with materials for his autobiography as a scholar, telling us how and where he has searched, whom solicited, why failed, when succeeded, and so on. His acknowledgments belong in the prefatory material; after page one the reader is concerned not with the editor but with the letter-writer. A simple rule for avoiding editorial exhibitionism is to avoid the first person pronoun. Besides cozy footnote chats, the editor will be tempted into sly quips and oblique jokes; and this should also be resisted. The editor, in short, should practise the most extreme self-effacement in favor of the letter-writer.

I hope I do not seem to be in favor of erasing the editor. I wish, on the contrary, to restore to him the function stated in Dr. Johnson's definition: "he that revises or prepares any work for publication." The editor, as I see him, should be discriminating as to what goes into the edition and also as to what belongs in the text and in the annotations. He should prepare his text in such a way as to make it faithful to his manuscript and at the same time serviceable to his readers; he should arrange his annotations and apparatus in a simple, lucid form.

Earlier in this paper I made a distinction between letter-writers and writers of letters. Perhaps this gap can be narrowed if we print

more letters as though they were worth reading for their own sake. Unless we print letters in such a form as to allow them to be read as widely as possible and with ease and pleasure we are in danger of creating a coterie scholarship, when we will only read each other's footnotes; and the ranks of monumental editions will, in truth, be only monuments.

If I seem reactionary in advocating a return to the nineteenth-century concept of readable editions, I hope it is still progress—on a spiral, so to speak; for we can also make our texts accurate and our annotations full and scholarly.

Is this possible? you may well ask. I hold here the first volume of the new edition, in the Pléiade series, of Madame de Sévigné's letters. In several respects it is more radical than I have outlined: the notes, of exemplary brevity and directness, are in the back of the volume, and the spelling, punctuation, and capitalization are normalized. (The text is for the most part based not on the writer's own manuscripts, which do not survive, but on a later transcript.) While it is both scholarly and complete, it is at the same time attractively made and invitingly readable.

"Mon cousin," she once wrote to Bussy-Rabutin, "s'il vous prend fantaisie un jour de publier mes lettres, je vous prie de les corriger." He answered, "Ma cousine, on ne corrige pas le Titien." (But he did.) If today we do not believe in correcting our Titians, we as editors should at least clean, frame, and light them for all to see and enjoy.

Notes