University of Virginia Library

Practice, not Theory: Editing J. S. Mill's Newspaper Writings
by
John M. Robson [*]

Introduction

Recalling his father's educational practice (and theory), John Stuart Mill says: "I recollect . . . his indignation at my using the common expression that something was true in theory but required correction in practice; and how, after making me vainly stive to define the word theory, he explained its meaning, and shewed the fallacy of the vulgar form of speech which I had used; leaving me fully persuaded that in being unable to give a correct definition of Theory, and in speaking of it as something that might be at variance with practice, I had shewn unparalleled ignorance."[1] This lesson was administered in Mill's twelfth or thirteenth year, and, despite his general acceptance of James Mill's authority, he continued to tease out its implications for the rest of his life: it might be said, for example, that his great System of Logic is an extended analysis of theory and practice. And how comforting it would be to say that he resolved the problem, for which of us is not more at ease—in Zion and elsewhere— if our practice can be shown to fit into a sweet theoretical niche.

So, in editing Mill's works, from the beginning and it seems forever, we have sought that comfort. But it was and is not to be ours, as the following discussion will demonstrate, and I hope justify.

Mill properly qualifies as a polymath, and therefore presents special challenges to editors, not least because his works are of special concern to people of very different interests.[2] He also, if one may be forgiven

the word, is polygeneric, and therefore produces even more editorial problems.

Mill's longest works went through many revised editions, as did, with fewer changes, his shorter monographs. Even in these apparently benign cases, decisions about the proper treatment of lightly revised texts may be less than ideal for heavily revised ones. His major periodical essays were published in a collection that went from two to three volumes in his lifetime, and to four volumes posthumously; the revisions of these essays vary in scope and kind, almost as much as in subject matter, and there are manuscript texts for only two of them. Most of his essays, however, were not republished and, apart from those portions of some of them that were used in other works, have single versions. Furthermore, he wrote (and kept a record of) over 1700 despatches for the East India Company during his twenty-five years in the Examiner's Office; these are anomalous with any other of his writings, and there is no carefully thought-out guide to editing such materials. And, not to labour the matter, there are, of course, private and public letters, a few journals, and many speeches (debating, public, and parliamentary), all of which might be thought to call for special treatment. Finally, there are comparatively few manuscripts of any kind, except, of course, for the material he did not himself publish, such as letters, journals, and the India Office despatches (a few of which were printed in Blue Books). The mix is not as before—or ever again.

The mode moderne, however, calls for at least an appearance of consistency in bringing together a writer's works into a single edition, and so we in the Collected Works of J. S. Mill (twenty-five to date, with five more to come) have adopted principles and methods that we have applied as closely as seemed practicable and wise for each title.[3] Saving the newspaper writings for a late stage tested our will and our skill in ways that we did and did not foresee. In the four volumes those writings occupy, we have included everything of Mill's that appeared in a daily or weekly newspaper, with minor exceptions mentioned below. The articles span more than fifty years, from 1822 when Mill was sixteen years old, until 1873, the year of his death; the subjects range from abstract economics (with which he began) and practical economics (with which he ended), through French and British politics, reviews of music and theatre, Irish land reform, to domestic cruelty, with glances at what seems like everything else lit by the nineteenth-century sun or obscured by its fog.

Not questioning the importance of including these items in the Collected Works—though in some ways a heretic, as will become evident, I am not seeking martyrdom by preaching such a dangerous doctrine anywhere,

let alone in the holy land—I still can admit that individually they are comparatively unimportant. Mill did not reprint them (again with minor exceptions to be mentioned below), undoubtedly judging them to fall within the area of proscription he defines in his own collection of essays, Dissertations and Discussions. Those excluded from the volumes, he says, "were either of too little value at any time, or what value they might have was too exclusively temporary, or the thoughts they contained were inextricably mixed up with comments, now totally uninteresting, on passing events, or on some book not generally known; or lastly, any utility they may have possessed has since been superseded by other and more mature writings of the author."[4] While recognizing Mill's wisdom in many matters, we are not disposed to heed him here. At the very least, the bulk of these materials gives them very considerable significance.

There was, of course, a recognition in Mill's time by people as diverse as Newman, Carlyle, Bulwer-Lytton and Disraeli, that newspapers were becoming extremely important in determining public opinion, and that the determination of public opinion was increasing in importance with the growing power of democracy. Let me cite only Disraeli: "Opinion is now supreme, and Opinion speaks in print. The representation of the Press is far more complete than the representation of Parliament."[5] The campaign against the "taxes on knowledge" gained much of its force from the general belief that the "people" needed the information, political, scientific, economic, artistic, and so on, that could most easily and quickly be transmitted through the daily and weekly press. Indeed, from the sixteenth century to the present, one abiding belief has been the importance to civility and utility of freedom of access through the press to opinions and facts of all kinds. The great campaigns of the nineteenth century used the press effectively and constantly, and though newspapers' significance for scholars of many persuasions has come to be recognized only in the last quarter century, there has been since then an ever-increasing use of newspapers as historical sources for opinion as well as fact.

That Mill was as aware as anyone else in his time of the surge of importance and interest is demonstrated by his heavy participation in what could not fancifully be called the "newspaper movement."[6] Indeed, he identified his times as "this age of newspapers."[7] Not all is well in such an age (it is also an "age of transition" to something better), as Mill makes clear in several places. For example, praising his friend W. B. Adams ("Junius Redivivus"), he comments that, in the circumstances, "all he has written, perhaps all he will ever have the inclination or patience to write, will be ephemeral. . . ."[8] More vividly: "The Spartan in the story, who, for the crime of using two words where one would have

sufficed, was sentenced to read from beginning to end the history of Guicciardini, and at the end of a few pages begged to commute his punishment for the galleys, would have prayed to exchange it for death if he had been condemned to read a file of English newspapers five years old."[9]

Nonetheless, especially in his early years he thought the times were ripe for the kind of education in and stimulation of reform that he could encourage through the ephemeral press. The anonymity it provided gave him the opportunity, so often seized in British life, for authoritative judgments in a favourable context: that is, the authority for the judgments was attached to the newspaper itself, and the principal readers of any newspaper then were those who accepted its general political orientation. Mill was frequently, it should be mentioned, a member of a dedicated team: his leading articles and news reports, for instance, reflect his often intimate connections with the editors, particularly in the 1820s with John Black of the Morning Chronicle and in the 1830s with Albany Fonblanque of the Examiner. And his reviews are not seldom imbued with enthusiasm for his allies' endeavours. Although Mill's later journalism, while still polemical, in general shows more care in developing an authority based on argument than on statement (partly, one may guess, because the anonymity was less guarded, or even absent, because the tone of the newspapers in which he was writing had altered, and because he was both more skilful and more committed to a balanced tone), his opinions then gained from the increased prestige of the big dailies with large circulation. In any case, malgré their ephemerality, newspapers provided Mill early and late with a platform he considered worth mounting, as the bulk of our volumes will all too readily testify.

Description of materials

In the volumes will be found 427 items, from 27 newspapers, 17 of them daily, and 10 weekly; however, the greater number, 261, appeared in weeklies, the loading factor being Mill's dedication to the weekly Examiner, especially from 1830 through 1834, resulting in 235 contributions over his lifetime. In fact, after Mill's first few years of writing for newspapers (much of it consisting of letters to the editor of the Morning Chronicle, many of which were probably solicited, or at least guaranteed a place), contributions to weeklies dominate the record through the 1830s, Mill's busiest period as a journalist. Subsequently his contributions to dailies are more common, with leading articles for the Morning Chronicle (in which 114 items appeared) and a variety of letters to editors making up the bulk. In addition to the Examiner, the weeklies (in order of the frequency of Mill's contributions; and where these are equal in number, in the order of date of those contributions) include the Spectator

(12 times), the Black Dwarf (4), the Sunday Times (2), the Leader (2), the Republican (1), the Lancet (1), the Reasoner (1), the Weekly Despatch (1), and the Reader (1); the dailies, after the Morning Chronicle, are the Daily News (16), the Globe and Traveller (11), The Times (8), the Sun (2), the News Times (1), the British Traveller (1), Le Globe (1), Le National (1), the True Sun (1), the Guide (1), La Voix des Femmes (1), the Morning Post (1), the Penny Newsman (1), Our Daily Fare (1), and the New York Tribune (1).[10]

The distribution over time is important: 42 of the items appeared in the 1820s, 246 in the 1830s, 99 in the 1840s, 20 in the 1850s, 11 in the 1860s, and 9 in the 1870s. Equally significant is the distribution of types: 182 are leading articles, 106 news reports, 72 letters, 47 reviews, 6 obituaries, and 14 miscellaneous articles.[11] Just why I call these important and significant cannot here be explicated, though I trust the reasons for the judgment will be obvious; I should emphasize merely that I am thinking not only of analysis of Mill's career and thought, but also of editing practices.

Simply referring to the contents of the volumes as "newspaper writings" disguises some problems. I said above that we have included all of Mill's writings that appeared in a daily or weekly newspaper, with some exceptions. This definition needs refinement in several ways. First, the volumes are part of a collected edition, and have been published towards the end of that edition. When Francis Mineka was preparing Mill's correspondence, he (following Mill) made a distinction that seemed perfectly appropriate at the time, and still has general validity, between "private" and "public" letters, the basis being the intended audience. He excluded "public" letters from his volumes, leaving them for later consideration. We decided to include them here, for all but three of them were designed for newspapers, though not all of them appeared.[12] This decision is based on our unwillingness to leave a few letters, more closely related in genre and form to what is in these volumes, for a miscellaneous volume that will be heterogenous enough without them. Similarly, we included those letters to the editor that failed to be published, because, though some of them are obviously drafts, they were intended for newspaper publication, and almost certainly were sent in fair copy.

Another problem arises with articles or parts of articles that were reprinted in newspapers from other of Mill's writings. If one sees these volumes as gathering together the total materials that revealed Mill to newspaper readers, it must be regretted that some very telling pieces are excluded as extracted reprints. But actually no one reader would have been able to see Mill the journalist whole, for most of his writings were

anonymous, and they were scattered over such a period of time and so many papers that the likelihood of anyone reading them all is so small as to be negligible. Furthermore, we cannot pretend that we have found all examples of such reprints: the newspapers of the day commonly made extracts of this kind (often with the intention of puffing), and Mill was a popular author. We collated those found with the copy-texts used in other volumes of our edition, though there is no evidence that Mill had anything to do with the text of the reprints; we have found nothing of textual value.

Finally, we were reluctant to reprint anything that appears elsewhere in the Collected Works, even though it could be argued that some items should have been saved for these volumes. In particular, and with great reluctance, we excluded letters already published in our volumes of correspondence, often in draft form; according to received theory, these manuscript versions have the greater authority (at least as to accidentals), but the printed versions have greater polish, suggesting a Millian finishing. This criterion has been put aside in a few cases: for instance, Mill used some of his leading articles on agriculture in an appendix to his Principles of Political Economy; these were collated for the text of the Principles, and the substantive variants are given there. But it seems appropriate to give the original versions in the Newspaper Writings, because they are part of a series, not all of which was used in the Principles, and because the rewriting altered the form of the argument, if not markedly its substance. And we reprinted two other items, Mill's obituary of Jeremy Bentham (1832) and a petition for free trade (1844), believing that we were mistaken in including them in appendices to earlier volumes, where they lose part of their appropriate contexts. These cases of exclusion and inclusion are, it will be seen, special to our situation.

Also special is the confidence with which we can identify these items as Mill's. How many of them could we have been able to call his were it not for the survival of a copy of the list he kept of his published writings?[13] Alexander Bain, Mill's closest disciple, whose detailed knowledge is shown in his slight but informative biography,[14] using the information in Mill's Autobiography, some now unrecoverable family memories, as well as his practised judgment, was able to identify the larger proportion of Mill's early writings in monthlies and quarterlies, but could do little to pin down his newspaper writings. Similarly, F. A. Hayek, to whom is attributable the revival of interest in Mill after World War II, and who had a good eye for Mill's work, dismissed the series, "The Prospects of France," signed "S," one of Mill's common identifications, as not being in his style, and failed to find another signalled by a reference in one of

Mill's letters.[15] The surviving private correspondence would lead with considerable certainty to a few items, and some others (especially letters to the editor in the latter part of his life) are signed either with his name or with initials or pseudonyms that—provided one somehow came across the items—would permit identification. But I much doubt whether we would have found more than 25% without his surer guidance.

Confirmation of the list, so far as the important early writings in the Examiner are concerned, is found in a bound set of that newspaper in the collection of books from Mill's library housed in Somerville College, Oxford.[16] On the front flyleaves of all but the 1830 volume Mill listed his own articles, and (for the volumes for 1831-33) enclosed the parts of the text by him in inked square brackets. Also he made some corrections in ink. For the most part these three sets of information confirm the other, independent list, but the Somerville material enabled us to add some items to Mill's corpus.

Other evidence, such as the correspondence, signatures, and internal connections with identified items, enabled us to add a few more. Also, Mill's list included only published writings; we have, as mentioned above, included the meagre number of unpublished letters intended for newspapers that remain in manuscript. We have scanned the likely sources for further items, especially for the years after 1865, because Mill's list contains a very disturbing if excusable note: "From this time [presumably after April 1865, the date of publication of the preceding item] no memorandum has been made of my letters which have appeared in print: numbers of my public or private letters having found their way into newspapers, of all of which (I believe) the original drafts have been retained in my possession."[17]

In total, we added only 27 items to Mill's list: 10 of these (mostly in Mill's later years) have signatures; 7 come from the Somerville College Examiner; 5 are unpublished holographs; 3 are identified by external evidence and 1 by internal; and 1 is an ambiguous entry in MacMinn's edition.[18] Therefore, while we are pleased to have added any to Mill's list, it remains the vital source. With all our experience, we should have greatly underestimated and underrepresented his journalism without it; it is, we say with full understanding of the implications, impossible to search carefully through the files of nineteenth-century British newspapers to find Mill's writings. We envy, as all editors must, the reliance of Dickens' scholars on the devoted and loving efforts of myriads of readers over the past century who have found, by accident as well as design, the trails of Dickens in tangled banks of Darwinian complexity—but Mill was not Dickens, and Mill scholars, avid as they are, do not have the devotion of Dickensians.

Treatment

1. Organization

Though the general principle governing the division of materials amongst the volumes of the Collected Works is thematic, there are strong generic marks that make the universal application of such a rule ridiculous: for the special purposes of anthologies it is sensible to isolate parts of Mill's System of Logic, for instance, but to dissect that tome and store its bits in bottles labelled "the syllogism," "abstract method," "ethics," "political theory," etc., would be quite irresponsible. So the rule has been applied only to materials, usually periodical essays and short monographs, that are of a length permitting combination within thematic volumes. Within these volumes, a chronological order is followed, although sub-themes could have been used to combine essays. In the Earlier and Later Letters chronology was followed without much hesitation, for the other main option, grouping by correspondent, is of value only when there is a large mass of letters and when both sides of correspondences are substantial; neither of these conditions is sufficiently filled in Mill's case.

I mention these obvious matters only to explain the background to our decision to organize the newspaper writings (and also the journals and speeches that will make up Volumes XXVI-XXVIII) in chronological order. Materials as heterogenous as these generally resist division into themes or subjects, though in Mill's case the newspaper writings provide, as indicated above, two major subjects: French politics in the early 1830s and Irish land in the late 1840s. However, so densely grouped chronologically are these two sets that they cohere even within an ordering by date. Furthermore, some other groupings would be objectionably arbitrary, and there would be a ragtail remnant.

This decision makes it difficult to break the material into coherent units that would serve as "chapters." Since few users of the volumes will read them through from beginning to end, in fact these aids to ease are not essential; nonetheless we introduced breaks into timely intervals that will at least allow the table of contents to be followed without strain; and indeed because of the relation of Mill's journalism to patterns in his life and to external events these "chapters" actually have thematic and sub-generic affinities that are explored in the Introduction.

For each item we provide a title, a headnote, the text, and footnotes. Since all decisions are questionable, some mention of the choice of titles is desirable. It might be thought that, like letters, newspaper writings need no titles; however, many of Mill's were entitled, whether by him or not, and the titles provide an index to the material, as well as making each piece easier to identify in short form. When a title appears on the

copy-text (or on another version of the text that Mill oversaw), it is used, even though there is a likelihood that a large number of them were not chosen by him. The guides to identification mentioned above, Mill's bibliography of his writings and the copy of the Examiner in Somerville College, provide many titles in his own wording or a modification of it. For example, in his bibliography Mill uses two main wordings in listing his news reports on French politics: "summary of French affairs" and "summary of French news"; in the Examiner he normally lists each of these same items as "Article on France": for convenience we adopted "French News" with a bracketed serial number for all of them. In the case of the series on Ireland, which he lists in his bibliography as being on "Irish affairs," we drew from the contents of the articles a more descriptive title, "Condition of Ireland," again with serial numbers to distinguish them one from another. In both these cases the serial numbers are editorially added; in a few cases ("The Spirit of the Age" being the best known) Mill or the newspaper provided numbers for series: to indicate the difference, we used roman numbers for those in the copy-text and arabic for those editorially supplied.

A few titles derive from references to the articles by Mill in letters, and finally some, in addition to the "Condition of Ireland" series, are editorially supplied as appropriate to the contents and genre. The reviews, for example, which are normally headed in the copy-text by bibliographical identifications, are headed in our text by the author's name and the short title of the work under review. The obituary notices are (in conformity to Mill's occasional usage) headed "Death of" the deceased.

Beneath the title is given the provenance and date of publication of the item, while the headnotes indicate, as briefly as is compatible with some obeisance to stylistic decorum and grace, the place of the item in relation to others in the volumes, and the minimal historical information needed as background (a broader scope is given in the Introduction, and more detail in the footnotes). Each headnote also gives the evidence that the item is by Mill and the reason for the choice of title. The context in the newspaper from which it is extracted is sketched (location within a section, general headings, for instance), and finally any information about the choice and treatment of the text (the general account of editing practices is given in a separate Textual Introduction).

Appendices include a speech Mill translated for the Examiner, the French version of the article intended for the National that appeared in the Monthly Repository, the English translation of a speech by Enfantin that Mill prepared for the Examiner, the English version of the unpublished letter intended for La Voix des Femmes, an obituary of Francis

Place that is dubiously attributed to Mill by George J. Holyoake, and lists of editorial interventions, as well as our Index to Persons and Works and an analytic index. Also we provide indexes to items according to their original provenance and to their signatures.

2. Copy-text

For most of the items in these volumes, the selection of copy-text is Hobson's choice: that or none. As indicated above, Mill did not reprint any of his newspaper writings in the three volumes of Dissertations and Discussions that appeared in his lifetime, and only a few appeared in other versions. In only 19 cases are there competing versions: 10 appeared in part in other writings of Mill's (3 of these in the posthumous fourth volume of Dissertations and Discussions, and 1 of these also in a pamphlet and a printed version of a lost manuscript), 5 appeared in more than one newspaper, 2 have surviving manuscript versions, and 2 exist in both English and French versions. These last we included in both versions; the others, almost all of them different in kind, are printed with notes indicating the variants, and with explanations of the choice of copy-text (almost invariably the newspaper version, for that is the volumes' mandate).

One matter troublesome to our consciences that will be a nightmare to editors of twentieth-century newspaper writings is that of editions. In the 1830s the Examiner (our principal source) often, but apparently not always, went through two editions, which are not clearly marked as such. We have tried to compare the two versions, without finding anything but occasionally different page numbers for Mill's articles. (One of the letters published in Earlier Letters from the Examiner had the signature "M" removed in the second edition.) But collections of nineteenth-century newspapers, even the indispensable British Library collection, seldom include different editions, and location lists are at best embryonic if not unconceived; it would be an immense task, well beyond our powers, to locate all possible existing editions of the issues in which Mill's writings appear.[19] And—once again let me show the traitorous flag—I wager that our main audience will not judge us wanting in this respect.

In these circumstances, the editor's task is much lightened. But of course some emendations are called for in the interests of accuracy, consistency, and easy reading.[20] The texts are flawed in most of the ways typical of their genre: characters are dropped or broken, sorts are mixed or lacking, compositors are (by inference) inexperienced or careless, and Mill's hand has (again by inference) been misread. Also, some conventions of the genre and the period are not consistently followed in the originals, and if reproduced would be annoying to readers in the late

twentieth century. While we list the emendations in an appendix, they are so numerous and in the main so trivial that we cover many of them in general categories, and correct silently.[21] For example, except when the correction was indicated by Mill or when there is a possible ambiguity or when one such correction is contained within a more significant one, we do not list but only describe in the Textual Introduction the following:

• 1. Dropped and misplaced characters, including misplaced or absent word space (e.g., we do not list the corrections of "discharge sthe" to "discharges the"; "o fchildren" to "of children"; or "allthose" to "all those").
• 2. Missing or misplaced French accents, including those on proper names. This is a more contentious matter, but Mill's French was, though not perfect, very good, and undoubtedly better than that of most compositors, who, moreover, seem often not to have had the types (or enough of them) to hand. (In this context, one may mention that the habit of setting names in small caps meant that accents usually could not be indicated.) And there is inconsistency in nineteenth-century practice, which also differs in unpredictable but disturbing ways from twentieth-century usage.
• 3. French proper names. Once more Mill's knowledge suggests that at least many of the variant spellings were introduced by compositors, though one cannot know, and occasionally more than one spelling was acceptable. Our decision was to avoid the annoyance rather than keep the anomalies, so, for instance, we always give Louis Paul Courier (never Courrier), Casimir Périer (not Casimer or Pérrier), Jacques Laffitte (not Lafitte), Odilon-Barrot (not Odillon or Barrott), and (to illustrate what are more clearly compositors' errors) Cormenin (not Cormerin), and Cauchois-Lemaire (not Cauchors-Lemaire). We also cut through the hyphen knot in French forenames by printing them as separate names.
• 4. Initial majuscule / minuscule changes. These were made sparingly and only to make individual passages (not the volumes as a whole) consistent, on the grounds that Mill's hand is not infrequently ambiguous in this regard for some letters, and that the change in these specific words cannot be seen as emphatic.

Other emendations not signalled in the apparatus result only from the desire for easy reading without any implication of error in the copy-text. For example (and most of these apply throughout the edition, not merely in these volumes), monarchs are identified in the form "Louis XVI" rather than in any other way (e.g., "Louis the Sixteenth"), other ordinal abbreviations are regularized ("22d" becomes "22nd"), names in small caps are given in upper and lower case, italics are substituted

for small caps used for emphasis except when the small caps are themselves italicized (in which case they are retained in roman), and abbreviations for monetary units are always italicized ("50l." becomes 50l."). The styling of different newspapers is also not preserved, so, for instance, the salutation in letters to the editor is always given as "Sir,—", the square brackets sometimes given around such sub-headings as "[From a Correspondent]" are deleted, and the publishing information in the headnotes and headings is regularized.

3. Apparatus

In addition to the variant notes mentioned above, two kinds of footnote are appended to the items. Those from the copy-text, that is, Mill's own (with occasional notes by the editors of the newspapers) are signalled by the series *, +, etc., beginning anew in each item. Those editorially supplied (the great bulk of the notes) are signalled by separate series of arabic numbers for each item. In accordance with the practice throughout the edition, we attempted to identify in these notes all Mill's allusions to people and references to and quotations from written works and speeches, trying to specify where possible the edition he used or may be presumed to have used; to his notes we added (in square brackets) missing identifications and corrected mistaken ones. In the other volumes (excepting the correspondence) we avoided as assiduously as our consciences demanded and as our desires would permit adding any other information in notes, believing that Mill's texts were still almost as transparent as when first read. But newspaper writings are, like correspondence, much more time- and place-bound, and so we indulged our readers and ourselves (though still hounded by conscience) by giving explanatory notes of historical and biographical (as well as bibliographic) kind. The adequacy of these is of course for others to judge, but I should say we aimed a little higher than James Mill, whose confidence in his readers was considerably greater than ours; in one not untypical note he says: "See the writings of Kant and his followers, passim; see also Degerando, and others of his school, in various parts of their works."[22]

Conclusion

These matters indicate that we have been thinking of our audience. I have speculated much on the curious mix of folk who are interested in John Stuart Mill, but I need not attempt to characterize them here, the relevant point being that they provide a differentia of our edition. More significantly for other editors, though even vaguer, is the consideration of future audiences. Scholarly editors, though they of course feel a special warmth for those who rush out to buy copies hot from the cold type, know that their volumes will sit quietly on shelves for most of their lives—that

is, the lives of the volumes, which should be much longer than those of their transitory midwives. (Which of us who habitually gives presents to family members thinks seriously of giving copies of our textually sound editions to them?) So we have tried to design our treatment for the fit few who will want to know all about Mill in ages to come.

What will they (or she or he) need to be told? It certainly is safe to assume that future readers will not know more about nineteenth-century England (or France, or Ireland) than do our contemporaries. And how much do the latter know? To make a worst-case assumption is not practicable, for it would entail more effort, time, and money than we can supply; it is also uncharitable, and we are essentially an eleemosynary profession. As indicated above, we have made what is almost a best-case assumption in the earlier volumes of the edition (excepting the letters), but here believe it unfair to well-intentioned readers. Somewhere in between lies the optimum case, which is related to our judgment of ourselves as representatives, even if not typical, of those to buy and to come; we are not adequate par nature, however, because in seeking for what we do not know but think should be known we modify necessarily our understanding of what we did know or thought we knew—here is a common but undiscussed analogy to the quantum uncertainty principle. Many hours were spent in discussion, decision, revision, as well as in search, and the results are certainly not perfect: like the strenuous Victorian strivers whom we emulate, we are seeking a personal best, not a perfect "10." What we have had to do is compromise, taking into account our resources, abilities, and the all-too-frequent inaccessibility of what we hoped to find.

All this is somewhat obvious, and might be taken for a counsel of despair. What I have earlier said about the text will suggest the same comment. But the despair will be such only to those who seek a comprehensive theory. Let me be confessedly open: in certain circumstances, theory is merely practice with the hard bits left out. It is wisely said that hard cases make bad laws; the history of law-making and -breaking in the last quarter century, when this maxim has often been scorned, indicates, at least to me, its validity. And editing is made up of hard cases.[23] James Mill's view, expressed in my opening quotation, puts the bald position that we repudiate. My confidence in our attitude may be explicated by an anecdote. Some years ago, Claude Bissell, then President of the University of Toronto, began to explain its long-standing college system; after a few moments, an Irishman stopped him, saying: "That's all very well in practice, but it won't work in theory!" Subsequently we had a President who tried to make it work in theory, and now it doesn't work at all. Even Jerome J. McGann, in his thought-provoking Critique of Modern Textual Criticism (1983), where he raises many of the issues

I am troubled about, though his is a literary corpus, feels compelled still to search for an encompassing theory, though his examples indicate quite clearly that such a theory would be empty of useful advice.

I am not saying, let it be clear, that one can manage without principles: these we have established, and have tried to follow for twenty-five years with all the diligence and care we can muster—but also with what commonsense and experience we can preserve. All editors worth their ink try to get it right. But what is "it"? And what if there is no "right"? I hold with Edmund Burke that circumstances alter cases, and—though he did not bother to say so—that in editing all cases are different.[24] Not all are significantly different, of course, and there is much that each of us can learn from others. My debts are very great, though my repayments may not seem worth the cost of collection.

As I have said, perhaps arrogantly, the differences between works are obviously relevant to the procedures used in editing them; certainly arrogantly I am implying that this obviousness has not always been obvious enough. In brief, and using categories I have employed elsewhere in discussing rhetoric, one should take into account (though not equally or in all cases) the particularities of the author, the audience, the purpose, thesis, and subject, the genre, the occasion (time and place), the "resistance" to communication, and the argument (in the broad sense)—and should recognize, as relevant, the fact that each and all of these may be, in fact likely are, plural, and that in many cases the ostensible and the actual are not identical. Of these, the most significant in our case are those of subject and audience. Mill is seldom "literary" in the narrow sense, and when he is, writes as a critic, not as an author of imaginative works. Consequently, his audiences are seldom, in their primary interests, literary. What matters for them is not what matters for close students of Elizabethan drama or seventeenth-century lyric poetry or nineteenth-century fiction. Philosophers, political scientists, economists, historians: all should care about the accuracy of what they are reading, and usually they can be persuaded that it is important to know just what basis a text has. But their span of attention for such matters is very short, and they insist on pressing on to matters that concern them more. They will, within reason (their definition of reason) humour the textual editor, but they will not offer reverence. So, trying to serve them, one must take into account their reasonable expectations (our definition of reasonable). And one must recognize the different value that minor variants have in non-imaginative prose, compared with those in Shakespeare's plays and Byron's poetry.

Let me mention just one other of these rhetorical considerations, that of single authorship. Quite apart from any undetected surgery by the

editors of the newspapers, there is an unresolved problem about the part played by Harriet Taylor (later Mill) in several of the leading articles and letters to the editor dealing with justice and domestic cruelty. The entry in Mill's list of his published writings says of a leader dated 10 February 1846: "a joint production—very little of which was mine" (MacMinn, p. 59). Thirteen subsequent entries for newspaper writings of the 1840s and 50s contain similar comments, one of which (for 28 August 1851) is upsetting: "This, like all my newspaper articles on similar subjects, and most of my articles on all subjects, was a joint production with my wife."[25] Several items in these volumes that are not specifically described as "joint productions" certainly can be inferred to be covered by this blanket of vague dimensions, and a few, including some of the manuscript texts, even though they are in Mill's hand, almost certainly have two authors.

Even to mention the applicability of the other rhetorical variables to our practice is here impracticable. My main purpose in alluding to them is to indicate that we accept the sad inevitability that our edition cannot be either definitive or a sure guide for others. It may be that ours will be the last (as it is the first) comprehensive edition of Mill's writings, though of course some of his works will probably be reedited if more evidence comes to light and when fashions change. But the market is not bullish: even some of our most popular volumes are now out of print, though the edition is not yet completed, and I cannot foresee a time when it will be thought useful to begin again de novo on a parallel endeavour. (I am not overlooking such possibilities as electronic and microform editions.)

We are also guilty of not paying sufficient attention to the considerations that D. F. McKenzie has been insisting on in recent years. There can be no doubt that the form in which a work appears affects its received meaning. His argument is always telling, and it applies with special force to newspaper writings, for newsprint affects all the rhetorical variables. Certainly a case could be made for facsimile reproduction of texts, and I for one would be happier at least part of the time if we had taken this route. But I would not be happy all of the time, for many of the texts would be illegible, and the impossibility of reproducing each of the defining qualities (the feel of the paper, the textual "surround," and so on), make the ideal unattainable; and if it were practicable, it would in many respects be annoying to most readers. Furthermore, we are not the original audience, and cannot realize fully, even with the exact artifacts in front of us, that audience's sensibilities and expectations (which of course were not uniform). Nonetheless, one should make some attempt

to make the reader constantly aware of the original form through comment and reproduction of the textual context—and as many illustrations as the price will admit.

As I have said, even if our edition, or let us say more specifically our volumes of Mill's newspaper writings, were definitive (taking out of the account the mistakes that we have made), I do not think it should serve as a model for others. The conditions that I have described make it necessarily different. Of course our practice will, if reasonably "right," offer suggestions to others, just as the practice of others has been of help to us. Our main effort, however, has been, and has properly been, to make Mill's writings available in a reliable form, and a form clearly explained, to those who wish to read him. That we have contributed little to the theory of editing will, we hope, be at least partly offset by our contributing something to its practice (sins and silliness included), but if finally and fully offset, it will be by our fulfilling our primary intention of satisfying our real and ideal scholarly audiences' desire for a clean and wholesome Mill.

Notes