University of Virginia Library

In 1751 Thomas Edwards, who in The Canons of Criticism (1748) had mischievously exploded Warburton's pompous edition of Shakespeare (1747), was urged by his friend Philip Yorke to exert himself more constructively in the future—to turn editor himself, in fact, and rescue Spenser's works from their current state of dilapidation. As canny as he was unlovable, Edwards declined the invitation with thanks, and with a degree of prudence which one can certainly appreciate:

I regret as much as you do [he replied] that our Classic Authors have fallen into such unhallowed hands as they have of late been profaned by, but who am I that I can prevent the sacrilege? . . . Why, say you, publish a good Edition of [Spenser] yourself. Very well Sir—but to publish a good Edition of an Old Author is not, as we find by melancholy experience, as easy a matter as to poach eggs, nor is it to be done without a great deal of care and application, and some time too . . . .[1]
That last poignant sentence has the ring of a maxim that ought to be inscribed as a motto over the portals of the Center for Scholarly Editions—the editor's analogue to "Abandon hope, all you who enter here." But I expect we will agree that the business of editing, if a more exacting enterprise than poaching eggs, is no less wholesome, good authors being in their way no less nourishing.

Indeed, since they are likely to delight and instruct us the more according to how precisely we can ascertain what they wrote and what they meant, it is not too much to claim that an editor performs the fundamental act of criticism. As the work of Greg and Bowers has shown, editing in its primary aspect—that of establishing the text—has the advantage


Page 2
over other kinds of criticism in that its principles may be more surely formulated. There is, for instance, a certain compelling logic about the proposition that, in the absence of an author's manuscript, we can best approximate the linguistic texture of his work—the essential form in which he clothed his thought—by choosing as our copy-text the first edition, the version of the work that derives immediately from the manuscript. I have tried to demonstrate elsewhere[2] —by showing that the full sense of certain passages in Fielding's novels ultimately resides in his use of capitals, italics, type-sizes, etc.—that accuracy in this matter of rendering even the accidentals of a text is not at all the merely precious or pedantic consideration it is sometimes thought to be. Criticism as she is generally known is a disappointing mistress who has a way of changing her favorites: Richards gives place to Wittgenstein, Wittgenstein to Levi-Strauss, Levi-Strauss to Shklovsky, Shklovsky to Poulet, Poulet to Derrida, and Derrida, with any luck, will soon give way to Common Sense. The boulevards of Paris, as George Watson has noticed, are littered with ex-structuralists. But there is the permanency of sweet reason in Greg's rationale of copy-text.

Editors, however, have no comparable set of principles to guide them in that other, and I believe no less important operation, the annotation of the text. As far as I am aware, the most valuable attempt to supply such a guide is Arthur Friedman's essay, "Principles of Historical Annotation in Critical Editions of Modern Texts," published some thirty-five years ago.[3] Much briefer statements of principle sometimes occur incidentally in the prefaces to scholarly editions—as, for example, in Professor Friedman's Introduction to his edition of Goldsmith[4]—but because these statements rationalize the approach to annotation which a particular editor has found congenial, their general usefulness is limited. Most reviews of scholarly editions include, moreover, incidental criticism of the explanatory notes, but such criticism, being desultory as well as widely scattered through a hundred journals, is neither systematic nor very convenient; and, being founded on the personal preferences and prejudices of individual reviewers, it is rather too subjective and impressionistic. Every editor, it appears, is confident he knows what a proper note should do, but few readers are satisfied with the results. As James Thorpe observes,[5] the editors of the Yale edition of Milton's


Page 3
Prose Works are at one extreme, supplying an especially copious commentary; at the other is Herbert Davis's edition of Swift's Prose Works, which eschews explanatory annotation altogether. The spectrum runs from nothing to too much, and in between these extremes one finds every degree of amplitude or dearth. Indeed, the enterprise itself of explanatory annotation is often regarded as a necessary evil (and not always even as a necessary one), provoking in scholars feelings ranging from gratitude for the light a good note can shed on the obscure places of a text to dismay at the vanity and pedantry of editors who are thought to use such opportunities to primp and strut in public. To Charles Moorman, for example, "a comprehensive, carefully-prepared set of explanatory notes is, aside from the establishment of a hitherto unprocurable text, the greatest gift an editor can bestow upon a reader"; to Samuel Schoenbaum, writing of the editing of Shakespeare, the editor who goes beyond the glossing of words and phrases is not only professionally irresponsible but morally reprehensible, since he is "having a free ride at the expense of a captive audience that has paid its money for the plays."[6]

The state of the art, it appears, looks disconcertingly like anarchy; and it is time that those of us who are engaged in the exacting and, as our readers will assure us, expensive business of preparing "definitive" editions of major authors should follow Professor Friedman's lead in asking ourselves what, if anything, can be done to introduce a measure at least of order into the confusion—to mark out a terrain ample and firm enough to accommodate and support such a wide variety of edifices. The attempt is not likely to be entirely satisfactory, for, as I hope to make clear, much of the confusion, contradiction, and inconsistency that has characterized literary annotation from the time of Bentley and Warburton to the present is an inevitable function of the nature of the enterprise itself. This being so, I mean by the indefinite article in my title—which might otherwise seem to imply that I am here trying to do for the literary annotator what Greg has done for the textual editor—to emphasize that this will be a wholly tentative essay in which, on balance, I will be more concerned with defining the relative aspects of the problem than with proposing absolute solutions. Though I am persuaded that there is essentially one sound rationale for the treatment of copy-texts, I am no less convinced that there can be no single rationale of literary annotation that will prove universally practicable and appropriate. For, though they are colleagues in the community of textual


Page 4
criticism, the provinces of editor and annotator are different in kind. The editor is concerned with establishing the ideal form of the text, its physical reality on the printed page. The choices confronting him as he selects and arranges the linguistic signs which constitute the literary work as the author intended it will be essentially the same choices whether he is editing Shakespeare or Milton, Fielding or Hume. The element of uniformity in this operation makes it susceptible, therefore, of a kind of rational control and discipline that it is not too much to describe as scientific; and, like a scientist, the editor as he performs his task will be as objective and as unobtrusive as possible. The task of the annotator, on the other hand, is to mediate between the text which the editor has thus established and the reader who wishes to recover its meanings wherever they are obscure. Being concerned therefore with the mental, as it were, rather than the physical reality of the work, he will function not as scientist, but as historian and critic. He is necessarily obtrusive in this role, and he cannot be objective: every choice he makes as to when or when not to supply a note is subjectively determined, governed entirely by the quality of his own understanding of the author's intention and by his estimation of the reader's need to be enlightened. The element of uniformity which validates Greg's rationale of copy-text and which simplifies the task of the editor is lacking here, and without it, though we may suggest certain principles and guidelines for the annotator, these cannot amount to a theoretical system that will be applicable in all cases.

Let me briefly illustrate what I take to be the three chief variables affecting the annotation of any given literary work: (1) the character of the audience which the annotator supposes he is addressing; (2) the nature of the text he is annotating; and (3) the peculiar interests, competencies, and assumptions of the annotator himself.

The first of these variables is surely the most obvious. Before we can proceed very far in annotating a text we must have formed some notion of the interests and capacities of those who are likely to use the edition. Will they be professional scholars in the field or undergraduates in survey courses? Though in most cases the answer to this basic question will be immediately apparent and will determine our general orientation as we set about the task of annotation, yet its implications are today rather less clear than they would have been even a generation or two ago. I suppose it can be safely assumed that undergraduates are likely to be ignorant of most matters relating to the literary, social, and intellectual history of whatever period, to say nothing of the classics and the scriptures which not very long ago were the staples of every schoolboy's education. But can we as confidently assume that our colleagues will need


Page 5
much less enlightenment? We, too, are increasingly the products of an educational system that has abandoned the study of Latin and Greek and the Bible for subjects more "relevant" and secular; as a consequence most of us are ill at ease within the most important frames of reference shared by the community of literate men from the Middle Ages to the nineteenth century. In these times of narrowing disciplines and increasing specialization, furthermore, few of us are equally knowledgeable about all aspects even of the historical periods in which we spend our professional lives. The literary critic who knows Gulliver's Travels intimately may well need to be informed in some detail of the political circumstances referred to in The Conduct of the Allies, and it is more than likely that he has never read a sermon by Swift or anyone else. He will probably be unable to translate for himself a passage from the Aeneid, or to identify the work or to recall the context when it has been translated for him. Freud may have acquainted him with the myths of Narcissus and Electra, but will he be just as familiar with those of Hylas and Pasiphaë? With respect to many areas of knowledge, the distance between a bright philosophy major, say, who finds himself taking a course in Renaissance literature and the professor of history who wishes to read Spenser's Faerie Queene is probably shorter than we might suppose: for different reasons they may be grateful to know that in Una, Spenser allegorizes both the platonic ideal of the One and the character and destiny of the Church of England.

Though the annotator's conception of his audience will thus have a less radical effect on the process of selecting passages to gloss than might at first seem to be the case, yet it is certainly not irrelevant to this process. We may reasonably assume that the general knowledge of scholars is superior to that of undergraduates and that their command of the language is more comprehensive and more sophisticated. For this reason it has seemed to me supererogatory for the Twickenham editor of The Rape of the Lock to supply definitions of such terms as "Virago," "Termagant," and "Spark," which, one trusts, few scholars will need to be instructed in, and to gloss such other terms as "Treat," "Denizens," and "Resign'd," which, though applied by Pope in a somewhat unusual sense, may be easily found in the OED. Notes such as these in a definitive edition serve only to clutter the page and to betray the editor's pedantry, not to say his condescension toward his readers. Paradoxically, however, though scholars are generally better informed than college students, notes designed to enlighten the former will normally be fuller than those addressed to the latter. For scholars, who may be teaching the work in question or writing about it, and who are by and large a curious and skeptical lot, will not be content with a gloss that is superficial


Page 6
or perfunctory; they expect to be informed of the historical context that explains the author's use of a term or concept and whether his usage is conventional or novel, and before accepting an editor's assertions on trust they wish to be apprized of the evidence. Not only the decision itself to annotate or not to annotate will be affected, therefore, by our conception of an audience, but also our notion of the scope and thoroughness of any given note.

Secondly, the annotator's task will vary, and quite appreciably, according to the nature of the work in question—that is, according to whether its allusive and topical texture is complicated and dense or comparatively simple and straightforward. A poem like the Dunciad, for example, poses in this respect an editorial problem that is virtually insoluble. The conscientious editor who attempts an exhaustive commentary on this richly allusive work must either fail miserably or, by succeeding, earn for himself a niche in the poem (which I like to imagine Pope revising from time to time from his vantage point in the Elysian Fields). Such an editor would be obliged not only to retain Pope's and Warburton's own copious notes (which often require commentary in their own right), but also to supplement these by sharpening the identification of the hordes of obscure scribblers imprisoned in Pope's couplets; what is more, he would have the further formidable task of locating the sources of Pope's frequent plagiarisms and paraphrases, of explaining his innumerable puns, and spotting his countless allusions to various literary, religious, and political matters. By attempting this questionable feat the notes to the Twickenham edition virtually crowd the poetry off the page, yet, even so, admirable as Professor Sutherland's commentary is, they ignore scores of allusions and at best merely hint at the range and complexity of the ambiguities in which Pope veils his "deep intent." It is generally true—as, say, Roger Lonsdale's edition of Gray, Collins, and Goldsmith will attest[7] —that the poetry of a literate age which values the pleasures of imitation will have an allusive texture of uncommon density posing for the annotator a far more exacting problem than its fiction will do. One may hope, at least, to provide all the needful commentary for Moll Flanders or Emma, but to annotate Pope's poetry is to accept from the start the necessity of compromise. I am not, of course, proposing as an axiom that all novels present simpler challenges for the annotator than all poems: Defoe's Journal of the Plague Year and Jane Austen's Northanger Abbey, to say nothing of Sterne's Tristram Shandy and Fielding's Tom Jones, are plain examples to the contrary; and I can think of no poem of any period that affords a problem for the annotator as hopelessly demanding as that posed by Ulysses


Page 7
and Finnegans Wake. The point is that what we may reasonably expect an annotator to undertake to elucidate in any given work will vary according to a number of factors, such as the obscurity of the contexts to which it relates—literary, intellectual, political, biographical, etc.—and the density of its allusive texture.

A third variable preventing uniformity in the practice of annotation is the apparent (and, regrettably, perhaps inevitable) cause of that disconcerting subjectivity which is and has ever been the distinguishing characteristic of editorial practice and editorial reviewing. No two editors will annotate a text in the same way because each, according to his interests, competencies, and assumptions—according, indeed, to his temperament and sensibilities—will respond to the text in different ways: what to one may be obscure will be clear to another; what to one seems an allusion is to another the author's own turn of phrase; what to one seems significant about a passage may seem to another irrelevant. Such differences will obtain even among editors of comparable abilities and industry, and even if they should happen to agree on the audience they are addressing and on the nature of the annotation a specific text demands. According to his peculiar interests and competencies, an editor will see some things as happening in a text and be oblivious of others: he may be alert to every political, literary, and theatrical innuendo in the Dunciad, say, but be largely unaware of Pope's complex system of religious and philosophical allusion. In this respect the factor of subjectivity in annotation, though its distorting effects may be minimized by further study and research as the editor attempts to familiarize himself with the sources of his author's knowledge and thought, can never be eliminated and, consequently, the commentary even of a Mack or a Wasserman will be imperfect and incomplete.

Literary annotation being, then, the attempt of a particular editor to mediate between a particular text and a particular kind of reader, it follows that this process is a relative thing which in certain essential respects will vary from edition to edition. For this reason, as I have said, annotation more nearly resembles an art than a science; it cannot be completely regularized or reduced to a single set of invariable principles. Yet it need not be practised in quite the anarchic way it has been. Like any other art, this one has a purpose as well as certain strategies and techniques for achieving that purpose. We ought at least to be able to agree on what we are trying to do as annotators and on how we can best go about doing it. In the remainder of this essay, by drawing on Professor Friedman's early article and on my own experience editing Fielding's novels, I will try to bring us a little closer to a consensus on these basic questions, particularly as they apply to scholarly editions of


Page 8
eighteenth-century literary prose. Narrow as this focus may be with respect to the historical frame of reference and the nature of the text, the results of the inquiry should not be irrelevant to the problems of other editors concerned with other kinds of texts in other periods.

Though in some important respects it needs to be refined and qualified, Professor Friedman's statement of the purpose of annotation and of the two principal kinds of notes is useful. To paraphrase him, we may say that the editor in annotating a work intends to make the meaning of the text more intelligible to the reader, on the one hand by recovering for him certain information about specific persons, places, and events once known to the author's contemporaries but now obscure, and on the other hand by placing the author's ideas and expressions in the context of his own writings and those of his contemporaries. Professor Friedman calls these two kinds of notes "notes of recovery" and "explanatory notes," observing, however, that in actual practice the distinction between them often blurs. "Notes of recovery" are needed in Tom Jones, for example, when Fielding refers to acquaintances such as Richard Willoughby the lawyer (p. 458)[8] and Thomas King the coachman (p. 549); to places such as Prior Park (pp. 612-613) and Bridges Street (p. 375); and to events such as the wreck of the Victory (p. 593) and the riotous first night of Edward Moore's play, The Foundling (p. 729). Another sort of "note of recovery," implied though not specified in Professor Friedman's definition, provides information often crucial to the subtler play of meanings in a given passage: namely, the note identifying a literary allusion—such as Fielding's arch reference, interrupting his ostentatiously eloquent introduction of his heroine, to "the rude Answer which Lord Rochester once gave to a Man, who had seen many Things" (pp. 155-156); or his having Sophy reveal her distress at a scene in Southerne's Fatal Marriage which closely resembles her own unhappy situation (p. 796). And of course "notes of recovery" will also be required whenever it is likely that an author's language will be obscure to most readers, as when Fielding's surgeon wishes to subject Jones to "a Revulsion" (p. 411), or when Jones supposes that Partridge must be good "at capping Verses" (p. 414). Most editors will readily recognize the need to supply such information as this. Even so, as I will show presently, there is considerable disagreement about the fullness of treatment appropriate to certain kinds of "notes of recovery" whose function at first glance would seem to be quite straightforward.

Much more problematical are the decisions relating to what Professor Friedman calls "explanatory notes." It is not always easy to distinguish


Page 9
in a text those ideas and expressions which require commentary from those which do not, and here, especially, an editor's judgment and powers of discrimination are tested. In editing Tom Jones, for example, I glossed rather fully Fielding's use of such concepts as "Good-nature" (p. 39) and "Prudence" (p. 36), because I am personally convinced that a modern reader cannot adequately appreciate Fielding's characterization and moral purpose in that novel unless he is apprized of the personal and historical contexts which deepen our sense of the significance of these ideas to Fielding and his first readers, and which enable us to estimate the degree of originality apparent in Fielding's treatment of them. Another editor, unconvinced of the importance of the ideas, would leave them unexplained, or treat them in a cursory way. The question of scope—of how full our commentary should be and of what kinds of information it should contain—must be faced in any sort of annotation; but the vaguer, more subjective question of relevance is particularly vexing as the editor tries to determine whether or not a note of the "explanatory" kind is needed. With respect to frequency and fullness, the character of the notes in any edition will be determined by how the editor answers these two questions. Indeed, the all too apparent anarchy of modern annotative practice—the aspect of contemporary scholarly editing in which, with reference to the commentary, individual editions range themselves (as we have remarked) along a broad spectrum defined by the antithetical practice of the Yale Milton and Herbert Davis's Swift—is chiefly to be accounted for in this way.

Must this confusion continue? Is the business of annotation really such a relative and subjective enterprise that any editor's notions of scope and relevancy are as valid as any other's? I think not. The trouble has been that, though most editors can agree that the purpose of annotation is "to make the meaning of the text intelligible to the reader," individually they have construed that definition in very different ways. In practice we have been unable to agree on what constitutes "the meaning of the text" or on the procedure required to make that meaning "intelligible to the reader." This disagreement is curious since, as linguists from Richards to the present have demonstrated, the meaning of a text—and particularly of a literary text—is not limited to the strict denotative or referential signification of its language, but includes as well the connotative values of that language, the full range of associations which the words had for the author and his first readers. Though it is true that we can never hope wholly to recover the meanings of a text in this sense, it is also true that an editor who has given some years of his life to studying his author and the historical context in which he wrote is most likely to be aware of them, and that, whenever in his judgment the text may be


Page 10
obscure, he has a responsibility to share his knowledge and understanding with the reader by providing whatever information may be necessary to make the author's meaning intelligible.

In this view, even a "note of recovery"—the neater and more straight-forward of the two kinds Professor Friedman has defined—will often require fuller treatment than we might suppose. It will of course serve the basic function of identifying person, place, event, or literary allusion; but it must frequently do more than this. In Tom Jones, for example, Fielding several times compliments Ralph Allen, referring to his friend's benevolence (p. 4), to his being both esteemed and condescended to by Pope (pp. 6, 404), to his having built hospitals (p. 38), to his intelligence, wealth, business activities, as well as his taste, hospitality, and personal integrity (pp. 403-404), to his estate (pp. 612-613). Clearly, no note which in a sentence records his dates and characterizes him as a philanthropist and patron of letters will adequately identify Allen for the reader of the novel; the editor must also illuminate the specific aspects of Allen's life and character which comprise the substance of Fielding's various compliments, for that is Allen's meaning in the text. Similarly, when Fielding mentions "the well-wooded Forest of Hampshire" and troubles to add a rare footnote of his own implying that it has been stripped of its trees (p. 259 and n.), the editor will oblige the reader by identifying the place as the New Forest; but he will oblige him more by explaining the particular historical circumstance to which Fielding refers, and still more by supplying the biographical information which in all probability accounts for the occurrence and prominence of the allusion in the novel: namely, that the Duke of Bedford, Fielding's patron and Warden of the New Forest, had appointed him its High Steward in 1748.[9] What the reader wishes to know, in other words, is the author's use of an allusion, why he chose to make the reference and how it works in the text.

To supply the reader with the information he requires in order to know the text this intimately is the function of any note, but the problem of recovery is often especially complex in annotating works of fiction and poetry, in which literary allusions are generally a principal device of the author for complicating, deepening, and extending the denotative sense of a passage. A proper note on such an allusion will not only identify the author and the work quoted, but will briefly point out those features of the context and circumstances of the passage quoted that serve to clarify whatever analogy or irony may be latent in the allusion


Page 11
Consider, for example, the following allusion occurring in Fielding's celebrated discussion of conservation of character, an aspect of the principle of Probability he is recommending to fellow novelists:
It is admirably remarked by a most excellent Writer, That Zeal can no more hurry a Man to act in direct Opposition to itself, than a rapid Stream can carry a Boat against its own Current. I will venture to say, that for a Man to act in direct Contradiction to the Dictates of his Nature, is, if not impossible, as improbable and as miraculous as any Thing which can well be conceived. Should the best Parts of the Story of Marcus Antoninus be ascribed to Nero, or should the worst Incidents of Nero's Life be imputed to Antoninus, what would be more shocking to Belief than either Instance. . . . (pp. 405-06)
The editor annotating the allusion in the opening sentence of this passage will of course state that Fielding is here recalling a remark of George Lyttelton's in his Observations on the Conversion and Apostleship of St. Paul (1747); and, since Fielding is paraphrasing rather than giving Lyttelton's exact words, he should probably quote the passage so as to enable the reader to judge for himself how Fielding alters the original. Having done this much, the editor yet has not helped the reader to grasp how the allusion works; in itself the passage Fielding paraphrases is a simple generalization about human nature. In context, however, the aphorism summarizes Lyttelton's crucial argument to prove that St. Paul's sudden conversion, from being the persecutor of Christians to being an apostle of Christ, was a true and genuine miracle, not to be explained by natural causes alone. By clarifying this context, the editor will enable the reader to respond to the specifically religious resonances of the word "miraculous" in Fielding's second sentence and of the phrase "shocking to Belief" in his third; and he will enable him to see that, implicit in the allusion, is a specific example of Fielding's point: though the authors of divine history may concern themselves with the Marvellous, the province of the novelist is the Probable. In his own life St. Paul was a greater contradiction than what Fielding next imagines would occur if an historian confused the characters of Nero and Marcus Aurelius. The allusion, moreover, works in a more personal way as well, serving as another of Fielding's compliments to Lyttelton, his friend and patron, to whom he dedicated Tom Jones and who stood with Ralph Allen as a model for Squire Allworthy. Since this biographical context will have been described in an earlier note, the editor may by a cross-reference direct the reader to this information. But, while enhancing the sense we already have of Lyttelton's goodness as a man (he is, we are reminded, the pious author of Christian apologetics), the present allusion contributes a new dimension to Fielding's


Page 12
praise of his friend, who is now complimented specifically for his excellence as a writer. By adducing similar compliments from Fielding's other works, the editor may also wish to assure us of Fielding's sincerity in admiring this less obvious virtue in his friend. And finally, in another cross-reference, he may wish to direct us to a later instance in the novel in which Fielding again alludes to the work quoted here.[10]

In short, the meaning of an allusion such as this—the full range of its associations and resonances—is complex. To render that meaning intelligible to the reader, the editor must recover for him the several elements which constitute it in all its complexity. Those elements will certainly include the identity of the author and work alluded to, but they will also include a knowledge of context, of historical and biographical circumstances, and of the occurrence of the same allusion or of similar ones elsewhere in the author's writings; for only when the reader is apprized of these latter kinds of information will he understand what the allusion meant to the author, why he chose to introduce it at the particular place in the text where it occurs, and how it works there. Unless in such exceptional cases as I discussed earlier—cases, such as the Dunciad, where the extreme density of a work's allusive texture obliges us to compromise with this ideal—the literary editor who undertakes to annotate a text in a scholarly edition must accept his responsibility to recover all these elements for the reader.

This responsibility, let me hasten to add, should not be construed as a license to display everything he knows about the topic; a note should not be, as George Sherburn once remarked, a sort of wastebasket into which the editor may conveniently dump the accumulated debris of his filing cabinet. This self-indulgent practice—the impulse "to tell all that they have learned rather than what readers need to know"—is what James Thorpe (p. 199) rightly deplores as the "occupational disease of


Page 13
editors." The ideal note is that which supplies essential information only and in the briefest compass possible. It does not of course follow that such a note will be short: since meaning can be complex, to supply the reader with even the essential elements that will enable him to recover it in all its complexity will frequently require a certain amplitude (in the Wesleyan Edition the note to the passage analyzed above takes up fourteen lines of print). But it will be as brief as the editor can make it, its brevity being measured by how well he has distilled the essential matter from the mass of potentially relevant information at his disposal. Far from encouraging self-indulgence, the making of a proper note requires of the editor a discipline and restraint that can be painful, for it demands of him the nicest discrimination in selecting his material and the most rigorous polishing of his own exposition.

Nor should a note, ideally, offer the editor an opportunity for self-indulgence in another sense. Though he is responsible for supplying essential information, he should strive to avoid imposing on the reader his own interpretation of a passage. His aim is to make the act of criticism possible, not to perform it. The editor's lack of restraint in this regard is what Professor Schoenbaum deplores as "having a free ride at the expense of a captive audience . . . ." Regrettably, however, by making available the elements out of which the reader may construct his own interpretation of the text, the editor to some extent will unavoidably control his reader's understanding of the text. For to identify the pieces of a puzzle and to arrange them in order, as rational discourse requires, is not only to define the nature of the puzzle but to suggest how the pieces should be assembled. One distinguished reviewer of the Wesleyan Tom Jones lamented this consequence of the rationale of literary annotation I am here proposing and which I have tried to apply in my own practice. The reviewer, Professor Middendorf, observed that, though the aim of the Wesleyan editors is not to interpret their texts, yet "in giving necessary background information and explaining 18th-century—and Fielding's—political and religious preoccupations in their introductions and notes, [the] editors open up their texts in ways that suggest and even occasionally establish judgments and evaluations." He continues in this "gloomy" vein by asking: "Once quotations are traced and names, places, events identified, how does one decide what deserves a note, what doesn't? Obviously the editor's interests and previous knowledge come into play, and though these may be, as here, of the most responsible and wide-ranging sort, may they not also, especially in an imposing edition like this, forestall fresh responses and invite an end to debate?"[11] This is not, certainly, a captious criticism of the difficulty—indeed,


Page 14
I would agree with Professor Middendorf in calling it the "impossibility"—of perfectly reconciling the editor's responsibility to supply essential information with his wish to be as objective and unobtrusive as possible in doing so. As I have observed, the process of annotation is the process by which an individual scholar with certain personal interests and certain personal kinds of competencies mediates between a particular text and a particular kind of reader. That he should be entirely objective and self-effacing in performing this function is a happiness unobtainable by the very nature of the art. Try as he may not to do so, the editor cannot help governing, to some degree, the reader's response to the text.

But the gloom this dilemma inspires need not be quite so oppressive as it seems to Professor Middendorf. As a reader myself of scholarly editions, far from resenting I have been grateful for the opportunity a good set of notes provides to share the editor's insights into the text and his knowledge of its contexts and circumstances—to share, that is, his informed personal response to the text based on a consideration of its meanings more prolonged and intensive than my own could be. Though it is true, furthermore, that the interpretations suggested in the notes to a definitive edition carry the stamp of authority, need we really be quite so apprehensive that the readers of such an edition, who are for the most part teachers and scholars and critics, cannot be trusted to judge for themselves the validity and usefulness of these interpretations? A scholarly edition is addressed to the most sophisticated audience imaginable; its readers may be supposed, I think, to accept or reject what is offered them according to their sense of its accuracy and cogency. Anxious as he understandably is to keep to a minimum editorial interference in the reading process, Middendorf, I suspect, would prefer Friedman's spare and extremely conservative manner of annotating Goldsmith to Donald Bond's more liberal practice in his edition of The Spectator; and surely Friedman's modesty and self-restraint as an editor are admirable. But knowing how brilliantly, after so many years of study, he might have illuminated his author for us, many have thought these virtues achieved at too great a sacrifice.

Though I have so far illustrated the purpose and nature of literary annotation chiefly by reference to "notes of recovery," these same principles apply equally to notes of the "explanatory" kind, which must also supply the reader with essential information only and in the briefest compass possible. Here, however, as I have remarked, the task of the editor defines itself less neatly: it is generally speaking simple enough to recognize those allusions to persons, places, events, or literary works that need clarifying; it is not so simple to decide which of an author's


Page 15
ideas or expressions need to be rendered more intelligible by placing them in context. Such decisions will of course be easier to make the better an editor knows his author and the historical and intellectual milieu that influenced his thought, for he will then be better able to tell when the author is using concepts that carried a special significance for him and his first readers of which a modern audience may be unaware. Ideas such as "Prudence" (p. 36) or "Good-nature" (p. 39), the dramatic unities (p. 209) or the chain of being (p. 481), the double standard in sexual matters (p. 755) or the reputation of the Dutch for cowardice (p. 513) had meanings for Fielding and his contemporaries which in the course of time either have been lost entirely, or have been modified to the point where some explanation is required to restore them to their original contexts. Notes such as these are particularly difficult to write, however, since the topics in question—often implying the basic assumptions and representative ideas of an entire culture—invite a copiousness of treatment that the editor must try to resist. Essays, indeed monographs, have been written on such subjects; but the editor, whose medium might be thought the sonnet form of scholarship, must distill an adequate sense of them as efficiently as possible. Here especially his powers of discrimination and his sense of proportion will be tested.

Having discussed the general principles governing the purpose of literary annotation, let us consider more practical matters of procedure—a subject on which Professor Friedman's pioneering article is particularly judicious and helpful. With some slight qualification Friedman's insistence that illustrative evidence used in annotation must not be "drawn from writings later than the one being edited" (p. 119) should be regarded as axiomatic. The editor in a note is trying to reconstruct for the reader those original elements which informed the author's intention, who, as he wrote the passage in question, cannot have been acquainted with the fruits of modern scholarship or, for that matter, with any sources of information published later than the period in which he composed the work. Friedman's formulation of this primary axiom of annotation is, however, rather too strict. The meanings a given allusion or idea has for an author do not cease to exist at the moment his work is published; they may continue to live in his thoughts and therefore may find expression in his later works in ways that can illuminate the passage we are annotating. This same persistence of meaning, moreover, is just as obviously a feature of the historical context in which the work was written. In attempting to recreate meanings for his reader the editor should actually confine himself to considering the useful life of the idea in question. All things being equal, he will prefer to adduce illustrative evidence from earlier sources (not much earlier, of course:


Page 16
since, as the very purpose of annotation suggests, meanings change over the course of time, we will generally wish to annotate an author by reference to writings and ideas more or less contemporary with him). Lacking such earlier sources—or, for that matter, lacking earlier sources which will illuminate a passage as effectively as some later source—the editor may reasonably look ahead to find his evidence. The closer such evidence is to the time of composition of the passage, the more persuasive it will be; and in any event we must not range beyond the immediate historical and personal contexts which define the life of the idea. Some latitude in this matter is admissible. In annotating the passage in The Champion (12 February 1739/40) in which Fielding commends Samuel Boyse's poem, Deity, for example, the editor of that journal will certainly wish to cite the almost identical compliment in Tom Jones (VII. i), published nine years later. Or in glossing Fielding's attitude toward duelling implicit in Jones's reluctance to fight Northerton (pp. 383-384), we will not fail to mention the much fuller and more explicit condemnation of this practice in Amelia (1751) and The Covent-Garden Journal (14 January 1752). Indeed, the best sources I could find to illustrate the specific nature of Fielding's comments on The Bull's Head Inn at Meriden (p. 574) or on "Beau" Nash's reputation as self-appointed guardian of the young ladies at Bath (p. 585) were books published, respectively, in 1757 and 1761, several years after the novelist's death.

Friedman's second axiom concerns the focussing of a note: "the editor," he asserts, "should annotate only what his author has to say about a subject, not the whole subject and everything connected with it" (p. 119). As a general principle this, of course, is perfectly sound, since the purpose of any note, as we have been insisting all along, is not to indulge the editor in displaying the full range of his erudition, but to clarify the author's specific meaning and intention. When annotating Fielding's representation of conditions in Newgate Prison in Amelia (I. iii), for example, we should focus on those particular features of the subject which illuminate the corresponding circumstances of the narrative; we should not rehearse the entire history of the prison or stray into a sociological survey of English penal reform. In actual practice, however, this axiom also requires a certain latitude of interpretation that Friedman, if we consider the conservative character of his notes to the Goldsmith edition, seems not to have intended. For, as I earlier remarked, what an author has to say about a subject can often be adequately understood only if the reader is apprized of the historical contexts and circumstances affecting it. Implicit in what Fielding in Tom


Page 17
Jones has to say—often in rather elliptical and oblique ways—about such subjects as "Prudence" (p. 36 and passim), the aesthetic principle of contrast (p. 212), or the polity of the gypsies (p. 666) are a whole range of complex historical associations which, once we are aware of them, significantly deepen our sense of the author's quality of thought, of how conventional or original he may be in his treatment of a subject.

Friedman's next two caveats are, one hopes, too patently just to require further comment. The editor, he insists, must be precise in distinguishing from among several possible contexts the particular frame of reference implied in his author's treatment of an idea; and he must be scrupulous enough not to mislead the reader by setting up "false parallels" between the passage he is annotating and those adduced to illustrate it (pp. 120-121). Errors of discrimination this fundamental are obviously irresponsible; indeed, they are in a sense unethical, since the editor who commits them is abusing the reader's trust.

From Friedman's analysis two further axioms may be drawn respecting the kinds of cogency to be achieved in explanatory annotations. I have some slight reservations about the general usefulness of the first of these, which seems to confuse the advantage of identifying the specific source of an idea or expression with what seems to me the greater and rather different virtue of illuminating the author's thought; the second principle, however, constitutes a cardinal rule for effective annotation. Friedman thus offers the following definition:

We may, I think, set it down as a principle that the most convincing explanatory notes are those in which unmistakable plagiarism from earlier writings is shown, for in such cases we have the unique elements out of which the author has constructed parts of his text. If this is true, it follows that other notes (at least notes that are intended to show similarity, not difference—that is, that reveal the author following a tradition of thought or expression, not departing from it) will be convincing and enlightening to the extent that the parallels pointed out approach plagiarism (pp. 121-122).
No doubt the most "convincing" note is one which demonstrates an author's verbatim borrowing from a particular source, since in this instance the reader can have no suspicion that the correspondence proposed is the product of mere coincidence or of the editor's fancy. And there is always a heady thrill in catching a writer in the act of pilfering. Few authors, however, are such irrepressible plagiarists as Goldsmith, on whose example Friedman chiefly bases his analysis. For the most part, an editor cannot expect to find a specific source for his author's ideas; indeed, to set about the task of annotation assuming that the only proper and successful note is one that demonstrates a plagiarism—as Friedman,


Page 18
following this axiom, seems to have done in editing Goldsmith—is to achieve cogency at the expense of other, no less useful kinds of illumination. Rather, the editor must be alert to more general correspondencies, to echoes from various works that approximate the phrasing of the text. He must have a good memory, certainly, so that he "hears" the text as he conducts his research; but he must also rely on his judgment to distinguish the viable context of his author's thought. The note that supplies this context, though it may be somewhat less "convincing," will often be more "enlightening" than one which merely records a plagiarism; for it will inform the reader not only of what the author read, but of how he read—how he was influenced, whether positively or negatively, by the vital currents of thought of his time.

For this reason Friedman's next axiom, which addresses itself to the usefulness of this latter kind of note and to the most effective method of proceeding in such annotation, is especially valuable. Friedman here rightly insists that when attempting to reconstruct the viable contexts of an author's thought, the editor will be more convincing when the parallels he adduces are drawn from sources more or less contemporary with the work being annotated. Correspondences between an author's ideas and those expressed in works published, say, a hundred years earlier are likely to seem merely fortuitous, whereas the same correspondences occurring in several works published within a few years of each other probably derive from a common current of thought. Thus, in Friedman's model, Louis Landa shed much light on Swift's "Sermon upon the Martyrdom of K. Charles I" by demonstrating numerous parallels between Swift's topics and rhetorical strategies and those of many other thirtieth-of-January sermons. The point, of course, will not be that the author actually read all or indeed any of these contemporary documents, but that, in Friedman's words, the work "is in many ways explained by being thus related to its historical background" (p. 123). More than the identification of outright plagiarisms—which is seldom possible—the definition of such relationships between text and context is one of the most valuable services an editor can perform. I would, however, again insist on a somewhat more liberal construction of the principle than Friedman's statement allows. What constitutes the contemporaneity of a parallel passage—and therefore its potential validity as an element in the editor's commentary on the text—is defined only in part by the proximity of its actual date of publication to the period of the text's composition. Fundamentally, of course, the contemporaneity of such a passage is determined by its currency during the period of composition. In this sense Hamlet and Paradise Lost, say, are no less contemporary


Page 19
with Tom Jones than Moore's The Foundling or Boyse's Deity; and, as Fielding himself makes clear, the sermons of the seventeenth-century divines Isaac Barrow and John Tillotson influenced his thought to an even greater degree than the works of Bishop Hoadly.

At the risk of belaboring a point I have already insisted on, I wish to conclude this review of Friedman's useful article by heartily endorsing a principle which, though applied in his final remarks to the making of "notes of recovery" only, should in fact be applied to the procedure of annotation in general: "namely, that as far as possible [an editor's notes] should be derived from contemporary sources rather than from modern reference books" (p. 125). The editor who takes this advice stands a far better chance of being able to illustrate the precise nuances of meaning a particular topic had for the author and his first readers. If it were not so apparent that many editors are unaware of the most obvious of these contemporary sources—namely, the journalism of the period—it would seem impertinent to add that the first step toward recovering information of this kind should be the conscientious perusal of at least those newspapers and magazines published during the period of the work's composition. By studying not only the "leaders," but also the news items and advertisements contained in such documents, the editor will come as close as possible to recreating in his own mind the daily circumstances in which his author wrote and to which he sometimes alludes in his writings. By analyzing the advertisements and monthly catalogues of works published during this period, moreover, and by tracking down and examining as many of these works as possible, the editor will often discover in even the fugitive literature of the time patterns of thought and allusion that illuminate his text in unexpected ways. From such hints couched in contemporary news items and advertisements, for example, I was able to deduce the personal circumstances underlying the episode of the puppet show in Tom Jones (XII. v—vi), and to identify Fielding's cryptic allusion in Amelia (I. ii) to "the celebrated Writer of three Letters" as a reference to Bolingbroke.

Let me conclude by proposing, in summary form, the following definitions of the purpose and nature of literary annotation, together with some recommendations concerning basic procedures for the annotator. I have in mind specifically, of course, as I trust will be clear by now, the typical case of a scholarly edition of an eighteenth-century literary text of unexceptional allusive density; but perhaps other editors in other fields will find at least some of these principles helpful in their own tasks.

The purpose of literary annotation, whether of "explanatory notes" or "notes of recovery," is to recover for the reader, as briefly and objectively


Page 20
as possible, all essential information (and only essential information) necessary to render the author's meaning wholly intelligible, the "author's meaning" being understood as not only the primary denotative significance of a passage but also, when appropriate, its full range of implicit associations, whether biographical, historical, or literary. To achieve this purpose an effective note will serve one or more of four functions: (1) it may define obscure terms or provide translations of words and passages in a foreign language; (2) it may identify persons, places, events, and literary allusions, supplying the reader, when appropriate, with such additional contextual information as he needs to appreciate how the reference "works" in the text; (3) it may illuminate the author's ideas or expressions either by citing specific sources for them or by adducing parallel passages from contemporary writings; and (4) it will record, whenever they seem significant, parallel or contradictory passages from the author's other works, as well as indicating such passages by cross-references as they occur elsewhere in the text itself. Since the aim of annotation is to reconstruct what a passage meant to the author and his first readers, all such information should be drawn as far as possible from contemporary sources rather than from modern reference books.

Assuming of course that he already has a sound general knowledge of his subject—that he is well acquainted with his author's life, works, and times and has carefully read the text in question—I recommend that the scholarly editor who undertakes to annotate a literary work according to these principles proceed as follows:

(1) He should first establish the text by collation of all editions published during the author's lifetime, observing as he does such unusual typographical features in the first edition especially as may signal the occurrence of a topical reference or literary allusion and noticing as well those substantive revisions which may require explanation.

(2) On the basis of external and internal evidence he should define the period of composition of the work, for by so doing he will define the chronological scope of his most intensive research in contemporary sources.

(3) Having defined the period of composition, he should inform himself in depth of the circumstances in which the author wrote—of his activities, residences, and acquaintances, and of the historical and intellectual events that may have mattered to him. As the surest means to this end, he will of course read through the author's correspondence and other personal papers when they are available. And he should conscientiously examine the newspapers and magazines issued during this period,


Page 21
thus familiarizing himself with its daily history and current topics of interest, and enabling him to identify books, poems, and pamphlets which his author may have read.

(4) He should also extend these investigations into the period immediately following the publication of the work in order to acquaint himself with its early reception as reflected in reviews, letters, etc. In this way not only will he be able to judge the initial general response to the work, but he will also often recover some of the specific meanings the work had for the author's contemporaries.

(5) With the text in all its detail firmly in mind, he should reread the author's other writings, keeping alert for parallel or contradictory passages. Since this review of the canon will often result in the identification of obscure references in the text which the author has treated elsewhere more explicitly, it should be undertaken at an early stage of the editor's research, though not before he has performed the basic task of establishing the text.

(6) He should then conduct his remaining research by consulting contemporary sources whenever possible, so as to acquaint himself with the specific meanings and connotations an allusion or an idea may have had for the author and his first readers.

(7)Finally, since the identification of some references will generally elude even the most conscientious research, he may wish as a last resort to appeal for help to the community of scholars by publishing a list of queries in appropriate places. (Almost, it may seem, as a vindication of Fielding's belief in a providential universe, three such queries which I published in the TLS were answered by three letters, two from German classicists and one from a Canadian scholar, each addressing himself to and identifying a separate allusion!)

The editor who observes these procedures conscientiously will have done what he can to fulfill his responsibility as a mediator between the text and his reader. The thought and spirit of an era being evanescent things and scholars being mortal, he will not of course succeed in illuminating all the dark corners of the work. Too often after the effort to track down an allusion he will find himself having to confess his uncertainty, introducing notes that can claim to be nothing more than guesswork with those, to him, most pitiable of adverbs, "probably" and "possibly"; or declaring his utter helplessness in that still more humiliating phrase—"Not identified." Annotating a topical and complex work from another time can be as arduous a task as writing a proper scholarly book of one's own, and there are no convenient dodges available to an editor, who cannot control his subject as he pleases but must answer, as it were,


Page 22
every question on the examination. In this business the demands and frustrations will sometimes be enough to tempt him to apply to his author what Harry Thunder declares of Rover in the play: "I don't know a pleasanter fellow, except when he gets to his abominable habit of quotation."[12] Even so, the editor has his compensation; for if his author is worth knowing, few will have come to know him better.