 | |  |
Some Principles for Editorial Apparatus
by
G. Thomas
Tanselle
Anyone who undertakes to edit a text must necessarily make some basic decisions about the kind of editorial apparatus that is to accompany that text. Sometimes these decisions are so thoughtless that they are hardly recognized as decisions at all — as when a publisher's editor selects a particular early edition to be photographically reproduced, without commentary of any kind, in a cheap paperback series. At other times they are the result of careful deliberation — as when a scholarly editor who has constructed a critical text sets forth in several lists the data he had at his disposal in examining variants and making emendations. In between, the spectrum includes a wide variety of kinds of apparatus: one edition may be entirely unannotated except for a prefatory note explaining the source of the text and perhaps generalizing about certain changes made in it; another may record variant readings (or the more important of them) in brackets within the text or at the foot of the page; another may limit its annotation, at the foot of the page or at the end of the volume, to definition of obscure words and identification of historical allusions; and still another may be principally concerned with citing previous critics' remarks about individual passages. But whatever form the apparatus takes — from the rudimentary to the elaborate — it represents some sort of thought about the extent to which the editor should make himself visible to the audience at which he is aiming.
Some kinds of apparatus, of course, are in part determined by the purposes of the edition: thus a variorum edition must by definition include apparatus which records variant readings. But an editor's basic job is to produce a text, and normally neither the exact form nor the extent of the apparatus is automatically determined by that job. A careful and reliable text obviously could be published without any accompanying apparatus, while an irresponsible text could be offered with extensive notes and tables. Such is not usually the case; but the
Just what apparatus is appropriate for a particular audience is a matter about which opinion naturally changes over the course of time, as scholarly techniques develop and bibliographical knowledge accumulates. In the nineteenth century, conventional practice for scholarly editions was to list variant readings (usually a selection of those considered most significant) in footnotes, with more discursive notes placed either at the end (as in William Aldis Wright's Cambridge Shakespeare, which began in 1863) or in a second set of footnotes, below the first (as in Horace Howard Furness's Variorum Shakespeare, which began in 1874). A new standard was set in 1904, when the first volume of R. B. McKerrow's edition of Nashe appeared; though in arrangement it was similar to earlier editions (variants in footnotes, with discursive notes in a supplementary volume), it marked a turning point
But the event which has been most important in influencing the form of the apparatus in many of the more recent scholarly editions was the publication in 1953 of the first volume of Fredson Bowers's Cambridge edition of Dekker. What Bowers did to make the apparatus more conveniently usable was to break down the listing of variants into several parts: only the substantive departures from the copy-text were recorded in footnotes; the departures in accidentals were then gathered into a separate list at the end of the text; and two more lists at the end dealt with press-variant formes and with substantive variants in other pre-1700 editions ("historical collation"). Since that time many editions have employed some variety of Bowers's plan, most notably the series of editions now in progress under the auspices of the MLA Center for Editions of American Authors. These editions maintain Bowers's distinction between lists of emendations and historical collation; but since they are "clear-text" editions, no apparatus appears on the pages of text, and all the emendations of the copy-text — both substantives and accidentals — are usually joined in one list. The adoption of this approach to editorial apparatus by the CEAA editions of Clemens, Stephen Crane, John Dewey, Hawthorne, Howells, Irving, Melville, and Simms[3] suggests that a new pattern for the treatment of
Although the form of the apparatus in a large number of scholarly editions from the last twenty years is therefore similar, it is by no means identical, even in those which follow the same basic plan. There is no reason, of course, to insist on such identity in outward appearance, so long as the approach to the material is sound, since different circumstances naturally entail somewhat different treatments. Nevertheless,
i. Arrangement
The first decision to be made about the arrangement of apparatus is its location. Are variant readings or editorial symbols to appear within the text itself? Or is the annotation to be provided at the foot of each page? Or at the end of the text? Or in some combination of these places? The tendency in recent years has been toward "clear text" — that is, no editorial intrusions of any kind on the pages of the text itself — and there are at least two important reasons for encouraging this practice. In the first place, an editor's primary responsibility is to establish a text; whether his goal is to reconstruct that form of the text which represents the author's final intention or some other form of the text, his essential task is to produce a reliable text according to some set of principles. Relegating all editorial matter to an appendix and allowing the text to stand by itself serves to emphasize the primacy of the text and permits the reader to confront the literary work without the distraction of editorial comment and to read the work with ease. A second advantage of a clear text is that it is easier to quote from or to reprint. Although no device can insure accuracy of quotation, the insertion of symbols (or even footnote numbers) into a text places additional difficulties in the way of the quoter. Furthermore, most quotations appear in contexts where symbols are inappropriate; thus when it is necessary to quote from a text which has not been kept clear of apparatus, the burden of producing a clear text of
Arguments can of course be advanced for inserting editorial apparatus into a text, and it is true that on certain occasions this arrangement is desirable. For instance, the nature of such materials as letters, notebooks, and journals — works never intended for publication — may not always be accurately reflected in clear text, which requires a choice among alternative readings that is often alien to the spirit of a private document. In these cases a text which includes editorial insertions and symbols recognizes that canceled readings and uncanceled variants are in fact integral parts of such works and comes as close as possible (short of a facsimile) to reproducing their essential character.[7] In
Notes at the foot of pages of text also can be defended at times. There is no denying the argument that the location is more convenient for reference than the end of the text, and editors may wish to place the most important apparatus there, even when they reserve the bulk of it for the end. Fredson Bowers, in his move to make apparatus less cumbersome, took the view that scholarly editions "should be made more attractive to the general user, first by removing all but the most immediately pertinent of the apparatus to appendices in the rear, thus freeing the text page from all information that is only of reference value and so of no immediate concern to the reader."[10] In his edition
If an editor decides to place his apparatus at the end rather than on the text pages, several questions of arrangement still remain. One of the first is whether, if a volume contains more than one work, the apparatus pertaining to a given work should come immediately after the text of that work or whether the entire apparatus should be gathered at the end of the volume. Practice has varied in the CEAA editions: the Ohio State Hawthorne and some volumes of the Virginia Crane have apparatus following each work, whereas the Southern Illinois Dewey has apparatus at the back of each volume. Of course,
A related question is the precise arrangement of the material at the end of the volume. Just as it seems easier to turn to the apparatus if it is all at one location, so is it simpler to refer to the information about a particular work if it is presented at one spot within that section of apparatus. Normally when a reader consults the apparatus he is studying a single work, and he finds the apparatus more convenient if all the lists pertaining to that work occur together, so that he does not have to turn back and forth from the relevant part of a long list of emendations for the whole volume to the relevant part of a full historical collation, and so on. One might argue, of course, that to have only one list of each type for the whole volume — as in the Dewey edition — consolidates the information more economically and is in conformity with the view of the volume as a single physical entity. The difficulty, however, is that the volume is still composed of individual works, each with its own textual history and each requiring separate editorial consideration; any apparatus which does not segregate the material relating to each work is likely to obscure the variations in copy-texts and in numbers of authoritative editions involved and to suggest a greater uniformity than in fact exists. Obviously this is not to say that the true situation cannot be perceived by a careful reader — after all, neither system, in the hands of responsible editors, conceals relevant information. The whole point is whether one system is clearer and easier to use than the other, and the advantage in this regard would seem to lie with the system which presents separate blocks of data for each individual work. The greater ease with which, under this system, one can move from one category of data to another and gain an overview of the kinds of editorial decisions involved (even without remembering which pages the work covers) surely outweighs the slightly greater expenditure of space entailed.[18]
Whether or not a volume contains more than one work or has apparatus arranged by work, another question which an editor must decide is the order in which the various parts of the apparatus are to be presented. Although the record of emendations in the copy-text generally precedes the historical collation, there is otherwise little consensus on this matter. Some editions place the discursive textual notes after the list of emendations, while others give them first; some put the list of line-end hyphenation last, while others insert it immediately after the record of emendations. The issue is not one of major importance: what is important is that these kinds of information be present, not that they be present in a particular order. Still, they have to appear in some order; and the editor, if he does not settle the question by flipping a coin, must have some rationale for selecting one order over another. If an advantage, however slight, does exist favoring one arrangement, that arrangement is preferable to a purely random order hit upon by an individual editor — preferable not only because there is some reason for it but also because its adoption would result in greater uniformity among editions. At any rate, the editor should be aware of the various considerations involved in the ordering of the lists.
One common arrangement is to place the list of emendations first (if it does not appear as footnotes), since it could be considered of most immediate importance, recording as it does the editorial alterations in the copy-text. Generally this list is followed (as in the Ohio State Scarlet Letter, the California Mark Twain, and the Dewey edition) by the discursive textual notes; because a common practice now (following Bowers's Dekker) is to mark with an asterisk those emendations which are discussed in the notes, the section of notes can in a sense be regarded as an appendage to the list of emendations. On the other hand, most editors find it necessary occasionally to comment on readings they have not emended, giving their reasons for not altering what might at first seem to be incorrect. Notes of this kind obviously do not relate to the list of emendations, since no emendations are recorded at these points. One way of solving this awkwardness is to insert asterisked entries in the list of emendations calling attention to
Once it is decided to make the textual notes the first of the four main divisions of the apparatus,[22] it is not difficult to settle on the list of emendations as the second. Of the standard lists, it is the one most directly connected with the edited text, since it records the editorial changes which that text embodies and enables the reader to reconstruct the basic document, the copy-text. The question that remains, then, is the order of the other two basic lists — the historical collation and the record of line-end hyphenation. The hyphenation list is often put in last position (as in the Dewey, Howells, and Irving editions, in certain volumes of The Mark Twain Papers, and in the first volume of the Hawthorne edition); apparently the reason is that hyphenation seems in some intangible sense less "significant" than the instances of substantive variation in the authorized editions. Yet if one of the reasons for giving precedence, among the lists, to the record of emendations is that it is concerned with editorial decisions, consistency would suggest that the hyphenation list should follow immediately, since it too records editorial decisions — that is, one half of it does, the half that lists line-end hyphens in the copy-text. And the other half — noting the established forms of compounds divided at line-ends in the edited text — is of direct use to any reader who wishes to make a quotation from the text. The historical collation, on the other hand, is simply a factual register of the variations (usually only the substantive ones) present in a given group of editions; although it is valuable in setting forth much of the textual evidence at the editor's disposal, its purpose is not primarily
A suggested standard order, therefore, for the four basic parts of the apparatus is as follows: (1) textual notes; (2) emendations; (3) line-end hyphenation; and (4) historical collation. Another essential part of the editorial matter is an essay describing the editorial principles followed in the edition, the textual history of the individual work, and any special problems emerging from the application of those principles to that particular historical situation. This kind of essay is frequently labeled as an "introduction" and placed at the beginning of the volume, preceding the text. Although such a location does not contradict the notion of clear text (since it obviously does not affect what appears on the text pages), it does result in the editor intruding himself at a very prominent place in the volume; what is probably more in keeping with the spirit of the clear-text principle and with the decision to place apparatus at the end is to think of the textual essay not as an introduction to the entire volume but only as an introduction to the apparatus — and thus to be placed at the end, though preceding the other parts of the editorial matter dealing with textual concerns (a plan followed by a number of editions, such as the Howells, Irving, Melville, and Simms).[24] In cases where related documents of textual interest exist (a manuscript fragment from an early draft, a published preface differing from the one in the copy-text edition, a map or other nontextual appurtenance of the copy-text edition, and the like), they can properly be printed or discussed as another section of the editorial apparatus — a section which should probably come last, since such material often is not keyed directly to the established text and in any case is only peripherally related to it. Finally, the editorial apparatus
ii. Symbols
Practically every edition makes some use of symbols or abbreviations; indeed, they are almost unavoidable unless one is dealing with a text so uncomplicated that scarcely any apparatus results. The primary motive behind most (but not all) symbols and abbreviations is economy, for if an editor is going to refer dozens (or even hundreds) of times to a particular impression, published at a given time by a given publisher and identifiable perhaps only by certain typographical peculiarities, it is merely common sense that he devise some concise way of making the reference. But he must also realize that, beyond a point, the interests of economy work in the opposite direction from those of clarity. In 1863 William Aldis Wright recognized (as every editor must) this dilemma: "We will now proceed," he said in the introduction to the Cambridge Shakespeare, "to explain the notation employed in the foot-notes, which, in some cases, the necessity of compressing
Perhaps a distinction should be made between symbols which stand for particular editions or impressions and those which stand for concepts. One cannot object to a multiplicity of symbols representing editions, if there happens to be a large number of editions involved, for the symbols are still easier to manipulate than cumbersome identifications of the editions in words; but further symbols to be used in conjunction with the edition-symbols for making comments about particular situations may easily proliferate to the point where they are less easy to follow than the same concepts expressed in words. Thus when McKerrow uses parentheses to indicate "a reading which is not identical with the one given but which is substantially the same in meaning or intention so far as the purpose of the note is concerned" (p. 82) and then inserts two parallel vertical lines within the parentheses "as a warning that, although the editions thus indicated support the reading in question, the context in which their reading occurs is not identical with that of the other texts" (p. 85), one may begin to feel that the goal of the apparatus has become compression rather than ease of
If it can be agreed, then, that the use of symbols is desirable for reference to editions and impressions, the practical question which arises is what system to use in establishing the symbols. There are two basic approaches: one is to arrange the editions in chronological order and then to assign them arbitrary sequential designations, such as the letters of the alphabet or numbers; the other is to construct each symbol so that it contains enough rational content connecting it with its referent to serve as a mnemonic. The choice of system depends on what kind of information is deemed most useful in connection with a given text, since each system, in order to make certain facts obvious, sacrifices other facts. In the first system, one can tell immediately that a particular reading is, for example, from the third edition (by means of a "C" or "3") but cannot tell (and may not be able to remember without checking) the year of that edition and whether it was English or American. The second system, conversely, might provide in the symbol the information that the edition was an 1856 American one (through some such symbol as "A1856" or "A56") but would not at the same time reveal its position in the sequence of editions. A variety of the first system has conventionally been used for pre-nineteenth-century books: a letter designating format (such as "F" for folio and "Q" for quarto), followed by numbers indicating the succession of editions within each format. Thus "Q4" would identify the fourth quarto but would not indicate the year of publication nor whether the edition came before or after the second folio. Attempts to combine the two approaches have not been successful because forcing too much significance into a symbol renders the symbol more cumbersome and to some extent defeats the purpose of establishing symbols as simple and easily recognizable designations. An edition reference like "3A56" is,
For earlier periods (before the beginnings of machine-printed books), the bibliographical and textual information conveyed by reference to format makes such symbols as F1, F2, Q1, etc., more revealing than reference to years of publication would be — and simpler as well, since the common situation in which more than one quarto appeared in a single year would have to be reflected in letters or other marks appended to the year designations. This system is one of the few well-established conventions in reference notation, and, with usefulness and simplicity on its side, there is little reason to oppose its popularity. For later books, however, format cannot always be determined and in any case is a less useful fact for incorporation in the symbol, since the variants to be reported are likely to be between impressions as well as editions. The most obvious adjustment would be merely to eliminate the format designation and use consecutive numbers (or letters) to refer to successive editions, with attached letters (or numbers) to indicate impressions within any edition. The Hawthorne edition assigns capital roman numerals to editions, with superscript lower-case letters for impressions (e.g., "IIIc"), while the Howells edition employs capital letters for editions, with arabic numerals for impressions (e.g., "B2").[27] Such a system is simple and neat; but, if a large number of editions and impressions are involved, it is difficult, even
Regardless of the variations in the basic system, an extremely useful convention which emerges is that a letter by itself stands for all impressions of an edition and a year is attached only when a particular impression is meant. But even this convention is best modified in certain situations: in the case of Irving's Mahomet there is only one English impression but nine American printings, all from the same publisher; the sensible way in which the Wisconsin edition assigns symbols here is to use "E" for the English impression and simple year designations without attached letters ("50" for "1850") for the Putnam impressions.[28] This arrangement is perfectly clear and is
In regard to symbols which stand for abstract concepts or relationships rather than concrete documents or impressions, the most prudent course of action is to keep their number as small as possible. Only two such symbols (both suggested in McKerrow's Prolegomena) have gained any currency in recent editions, and the reasons for their importance will suggest the kinds of circumstances in which symbols are desirable. Both symbols are used in reporting variants in punctuation: one, the centered tilde or wavy dash (˜), stands for the word previously cited, when the variant is not in the word but in the punctuation associated with it; the other, the caret (V), calls attention to the absence of punctuation at a given point. The justification for the first is not simply that it saves the effort of repeating the identical word, for the small amount of effort saved would be no justification at all if the repetition of the word would be clearer; the fact is, however, that the information is conveyed more clearly with the symbol than without it:
218.4 indefatigable,] A; indefatigable;
188.23 approaching,] 57; approaching
188.23 approaching,] 57; ˜
iii. Textual Notes
The section of discursive notes on textual matters is generally entitled "Textual Notes" but sometimes (as in the Irving and Melville editions) is called "Discussions of Adopted Readings." Both titles suggest an important point about the content of these notes: they are comments on individual problematical readings in the text which has been established, and any reading which appears there is the "adopted reading," whether it results from emendation or retention of the copy-text reading. The notes, in other words, discuss not simply emendations but also places where emendations might have been expected and yet, after careful consideration, have not been made. (The principal reason for putting the textual notes before the list of emendations, of course, is to emphasize the fact that they refer to the text itself and not to the emendations.) Since these notes deal with individual readings, it is important to keep general matters out of them and to limit them to cases which raise special problems. Any textual problem which recurs a number of times ought to be taken up in the textual essay, since it then constitutes a general textual problem and since a repetition of the same explanation several times in the notes or extensive cross-references between the notes would be more awkward and, indeed, less clear than one coherent discussion in which all the evidence is brought together.
The most convenient form which the textual notes can take is a citation of page and line number (or act, scene, and line),[31] then the reading under discussion (shortened by an indication of ellipsis if
iv. List of Emendations
The purpose of the list of emendations is to provide a convenient record of all the changes of textual interest[34] — both substantive and
In the light of this summary of the general rationale behind the idea of a separate list of emendations, two common variations in the basic form outlined above are worth examining. One is the segregation of substantives and accidentals into two different lists. This system is used in Bowers's Dekker and the Cambridge Beaumont and Fletcher (where emendations in substantives are listed at the foot of the page and emendations in accidentals at the end of the text), as well as in the Virginia edition of Crane's Maggie and the Wesleyan edition of Fielding's Joseph Andrews (where both lists come at the end of the text). Since the purpose of a list of emendations is to make the whole range of emendations easier to examine and analyze, it follows that under certain circumstances — particularly when there is an especially large number of emendations in accidentals — the separation of substantives and accidentals will make such examination easier still. In other words, if the total number of emendations is small or even moderate, little is gained by exchanging the simplicity of one list for the complication of two; but when there are a great many emendations, with the possible result that the emendations in substantives would be obscured by being included in the same list as a large number of emendations in accidentals, the data may be much easier to use if the two categories are listed separately. To do so is only to extend the principle of convenience and clarity on which the whole list is founded in the first place. And, by a further extension, certain large categories of automatic alterations within the list of accidentals itself may be separated so as not to overwhelm the other individual alterations of probably greater significance. For example, in the Northwestern-Newberry Typee 224 words ending in "-our" in the British copy-text are changed to the American "-or"; once the policy of making this category of changes is adopted, the changes themselves are automatic, and to list all 224 instances in a list of emendations would place an unnecessary impediment in the way of using the list to trace the more important alterations. Yet it is unwise to make any textual changes silently;[36] so these
A second variation from the most basic form of an emendations list is the inclusion of at least some of the further history of the rejected copy-text readings. That is, instead of providing simply the copy-text reading at those points where the copy-text has been emended, the entry includes the sigla for certain other editions which agree with the copy-text and sometimes includes the full history of the reading in the collated editions. One often-used plan,[38] following Bowers's Dekker, is to trace, at each point of emendation, the readings of all collated editions (that is, all which might contain textual authority) down to the earliest which can serve as the source of the emendation. Thus in the entry
The question of distinguishing between substantives and accidentals is often not an easy one, because some alterations of punctuation, for instance, do have an effect on meaning; but unless the substantives and accidentals are to be placed in separate lists, the question does not arise in constructing the record of emendations. Two other basic
One problem in the specification of copy-text is raised by the existence of variations within an impression. Such variations may be caused by stop-press corrections or by type which slipped or shifted during the course of printing. The precise definition of copy-text in terms of particular states of the variations obviously determines which of these readings qualify for inclusion in the list of emendations: thus if uncorrected formes are taken as copy-text, the only press-variants which would turn up in the emendations would be those adopted from corrected formes; and if the correct spacings at points where letters shift around are regarded as characteristics of the copy-text, the only variants of this kind which would be reported in the emendations list would be those for which no copy with correct spacings had been found. The decision as to whether correct or incorrect states are taken as copy-text may vary in individual circumstances, but the point is that the copy-text must be defined in terms of the specific variants within the impression which embodies it; for this abstract "text" must have one and only one reading at any given point,[44] and to define a copy-text merely in terms of an impression is not sufficiently rigorous, since more than one reading may exist at many points within various copies of that impression.[45] Because sheets embodying corrected states of some formes (or correct spacings of letters) will be bound with sheets embodying uncorrected states of other formes (or incorrect spacings of other letters), it is unlikely — when more than a few press-variants occur — that any single physical copy can be found which contains the entire copy-text.[46] Emendations in the copy-text, therefore, are not simply emendations in the text of a particular copy; and the copy-text remains an authoritative documentary form of the text, even though no one existing physical entity (or even no one physical entity that ever existed)
There are some variations among copies of a given impression which are nontextual and need not be reported, any more than differences in design between the copy-text print and the critical edition
Finally, a few minor points about form should be noted. (1) First, the list will be clearer in the end if each lemma consists simply of the word or words which constitute the emendation, without any of the surrounding words.[49] Occasionally an editor will feel that it would be helpful to the reader to have an additional word or two of the context, to enable him to see more clearly the nature of the alteration involved, while he is looking at the list. It is difficult to say, however, what would be sufficient context for this purpose, but generally a few words would not be enough; and as the cited readings become longer, the actual
v. List of Ambiguous Line-End Hyphenation
Until Fredson Bowers called attention to the matter in 1962,[53] no consideration (to my knowledge) had been given to the editorial problems raised by possible compound words hyphenated at the ends of lines. Such hyphenation clearly presents problems in two ways: first, when a possible compound is hyphenated at the end of a line in the copy-text, the editor must decide whether to print the word in his edition as a hyphenated word or as a single unhyphenated word; second, when a possible compound is hyphenated at the end of a line in a scholarly critical edition, the editor must have some means for informing his readers whether this word should be reproduced, in any quotation from the text, as a hyphenated word or as a single unhyphenated word. As a result, the necessity of including two hyphenation lists in the apparatus of critical editions cannot be denied. The first of these lists, recording line-end hyphenation in the copy-text, is essential to complete the record of editorial decisions. The editor's decision whether or not to retain a line-end hyphen in a given word can be more difficult than some of his decisions reported in the list of emendations. Yet it does not really produce an emendation, for if he prints the hyphen he is only retaining what, after all, is already present in the copy-text; and if he eliminates the hyphen he is only treating it as the printer's convention for marking a run-over word. Obviously some line-end hyphens present no problems: those simply breaking a word which cannot possibly be a compound (as "criti-|cism"), where the hyphen is only a typographical convention, not to be retained when the word will fit within a line; and those dividing compounds in which the second element is capitalized (as "Do-|Nothing"), where the hyphen is to be retained whenever the word is printed. But in between is a large area of possible compounds where no automatic answers can be given; the treatment of these hyphens depends on various factors (the author's characteristic usage, the conventions of the time, and the like), and the editor is not providing readers with a full record of his textual decisions unless he specifies these cases. The second list, recording line-end hyphenation in the editor's own text, is necessary if the editor is to complete his task of establishing a text — for if there are places in a text where a reader does not know precisely what reading
Because the Center for Editions of American Authors has required editions prepared under its auspices to include the two hyphenation lists (as specified in its 1967 Statement of Editorial Principles), the value and importance of these lists are becoming more widely recognized. Among the CEAA editions themselves, however, there are some variations in form, arrangement, and approach; and a glance at the principal variations will suggest some of the factors which need to be considered in setting up these lists. Probably the most noticeable difference among editions is in the order of the two lists. One may feel that it makes little difference about the order, so long as the two lists are there; but if an editor is trying to follow some consistent rationale in the overall arrangement of the entire apparatus, then surely one arrangement of the hyphenation lists fits that scheme better than another. Several editions (the Crane, Fielding, Hawthorne, and Simms and The Mark Twain Papers) place the critical-text list before the copy-text list, while several others (the Dewey, Howells, Irving, and Melville) reverse this order. The general rationale outlined above suggests placing nearest the text those parts of the apparatus taking up decisions affecting the edited text. Following this plan, the copy-text list should precede the critical-text list, for the copy-text list does record editorial decisions and in this sense is an appendage to the list of emendations (the immediately preceding section, according to this arrangement); the critical-text list, on the other hand, does not involve editorial decisions in establishing the text[55] but only printer's decisions
Indeed, the functions of the two lists are so different that it is some-what artificial to place them side by side; only the superficial fact that both deal with hyphenation has caused them to be grouped together. The copy-text list fits logically into the textual apparatus because it is historically oriented: that is, it records certain words in a historical document about which the editor of a critical text has to make decisions. But the critical-text list is merely a guide to the proper interpretation of certain fortuitous typographical features (hyphens) of a given edition of that critical text; its usefulness is not in studying textual problems but simply in reading the edited text. In other words, the edited text is not really complete without the critical-text list, for without it certain hyphens in that text would be ambiguous. The other parts of the apparatus are important to certain audiences, but the edited text could of course be printed without them; the critical-text list, on the other hand, is essential to all audiences, and the edited text should never be printed without it. If, for example, a publisher leases a CEAA text and reproduces it photographically, he should include the critical-text list, whether or not he is including any other apparatus; if, instead, he sets the CEAA text in type anew, he should prepare a new critical-text list which applies to his own edition. It is extremely unfortunate that the copy-editors' convention for indicating to the printer which hyphens are to be retained (one hyphen above another, resembling an equals sign) has never become a generally accepted convention for use on the printed page; if an editor could utilize such a double system of line-end hyphens, the printed form of his edited text would be self-contained, without any typographical ambiguity requiring a separate list to elucidate.[56] As matters stand, however, to do so would violate the notion of clear text, since the double hyphen would strike the reader as an unfamiliar symbol. It will not be possible, therefore, in the foreseeable future to eliminate the critical-text list, and yet it presents something of an anomaly in the textual apparatus. Logically it should be separated from the rest of the apparatus and placed as an independent entity immediately following the text. Yet it is unrealistic to think that the easy grouping together of all matters connected with hyphens will be readily superseded; and
Some editions contain more than two lists in the section on line-end hyphenation. For instance, a third list that sometimes appears (as in the Crane, Dewey, Fielding, Hawthorne, and Simms editions) is a short one recording those instances in which a line-end hyphen occurs in a possible compound in the critical text at the very point where a line-end hyphen also falls in the copy-text. The function of a separate list of these words is to show that the established forms in these cases result from editorial decisions. Nevertheless, these words do not logically constitute a third category; they merely belong to both the preceding categories. A simpler arrangement, therefore, would be to have only the two lists — the copy-text list and the critical-text list — with certain words appearing in both. The introductory note to the critical-text list could not then say — as these notes do in some editions — that the words occurred with hyphens (or without hyphens) in the middle of lines in the copy-text; it would have to say that for each word the "established copy-text form" is listed. If the reader wishes to know which forms were established through editorial decision, he can quickly check the appropriate spot in the copy-text list to see if the word also turns up there.[57] Still another hyphenation list which has been employed (as in The House of the Seven Gables and The Marble Faun) records line-end hyphens in the critical text which are true emendations (that is, hyphens at points where none are present in the copy-text). Again, such words do not form a separate category but, rather, readings that belong in two categories — in this case the critical-text hyphenation list and the list of emendations. The simplicity of an arrangement which keeps the number of word-division lists down to the basic two is not merely an advantage to the bewildered reader who may never have encountered any hyphenation lists before; it also dramatizes the logical division between the two functions which hyphenation lists serve.[58] Furthermore, it sets as few obstacles as
The matter of deciding just which line-end hyphens are to appear in these lists can be approached in two ways. One method is to list all compound words and all words which might be regarded as compound, if they are hyphenated at a line-end, recording the forms they should take when they fall within the line; such a list would contain both hyphenated and unhyphenated words. Another method is to list only those words whose line-end hyphens are to be retained when the words come within a line and to say that all other line-end hyphens can be ignored as compositorial word-division; such a list would contain only hyphenated words.[59] Each method has its advantages and disadvantages. The first system has the advantage of being explicit (listing all words about which a question might arise), whereas the second proceeds by implication (making the absence of a word assume positive significance); on the other hand, the second system has the advantage of covering in condensed fashion — through its combination of direct statement and implication — every instance of line-end hyphenation in an entire work, whereas the first may result in an extremely long list and still omit words that some readers would consider "possible compounds." Presumably one could infer, even in the first type of list, that omitted instances of line-end hyphens are not significant (that is, that those hyphens should not be retained in transcription), but the fact remains that the actual content of the list is not precisely defined, since the question of what constitutes a "possible compound" is a subjective one. It might never occur to one person to think that the line-end hyphens in "cup-|board" or "inter-|view," for example, should be retained, while another person might expect to find them in the list for explicit guidance. The first kind of list, in other words, is somewhat inefficient, because for all its length it may always fail to note words considered "possible compounds" by some people; the second type of list, in contrast, can in shorter space be positively complete, because
This second type of approach, then, might seem preferable for the hyphenation lists in a scholarly edition, were it not for two further considerations. In the first place, this approach, for full effectiveness, requires that one have at hand the edition referred to. That is, if an editor says that all line-end hyphens, other than those listed, are merely compositorial, the reader who wishes to look over those allegedly compositorial hyphens must consult the edition under discussion and run his eye down the right margin of the pages. Furthermore, if the policy of an apparatus is to record all the editor's textual decisions, those instances in which a line-end hyphen in a possible compound has been dropped are just as significant for inclusion as those instances in which it has been retained; to define the first category by a process of elimination (as what remains after the second category is specified)[60] is as unfair to the reader as to make silent emendations, for it requires him to search through a text himself to locate the individual instances. It becomes obvious, therefore, that one of these methods is more appropriate for one of the hyphenation lists, and the other method is more appropriate for the other list. The copy-text list should follow the method of noting all possible line-end compounds and showing the editorially established form of each, with or without hyphens — for this list refers to a document outside the volume which the reader has in his hands at the moment, and it records editorial decisions necessary for the reader to know about in evaluating the editorial process or in reconstructing the copy-text. The critical-text list, on the other hand, more appropriately follows the system which notes only those line-end hyphens to be retained in transcription — for this list refers to the printed form of the text in the volume already in the reader's hands, and it has nothing to do with editorial decisions. In other words, the more explicit system is necessary for a full recording of editorial decisions, whereas the more concise system is preferable for elucidating purely typographical ambiguities of the new edition. Once again, the differences in the purposes of the two lists are reflected in differences in method. If the hyphenation lists are set up in this way, and if their introductory comment[61] and their form[62] are kept as simple as possible,
vi. Historical Collation
The remaining principal division of the apparatus is the one which records the variant readings that have occurred in significant editions of the text. Its emphasis is historical, as distinguished from the list of emendations, where the emphasis is on the changes made by the present editor in the basic text he is following. Some of those changes were probably adopted from (or noted in) other editions, but the primary function of the entries in the list of emendations is not to provide the history of the variant readings at those points;[63] such history, as well as the history of other variants (where no emendation of the copy-text occurred), is reserved for this "historical collation," as it is often called. Two limitations are normally imposed on the historical collation. In the first place, it does not usually survey (at least in the case of nineteenth- and twentieth-century works) every edition of the text which has ever appeared, but only those of possible textual significance; thus all authorized editions which were published during the author's lifetime are included (since any changes present in them could have resulted from his revision), as well as any posthumous editions which purport to utilize newly available authoritative documents or which could conceivably have utilized such evidence.[64]
Some editions (such as the Howells and the Irving) limit the historical collation in one further way: by entitling it "Rejected Substantives" and listing in it only those substantive variants which are not adopted as emendations in the copy-text. Under the basic form of this system, each entry in the list of emendations must provide the full history of the readings at that spot, because none of these entries will reappear later in the historical collation. In effect, the historical collation is split into two lists, one containing entries involving emendation of the copy-text and another covering the remaining substantive variants, where no emendation is involved. (In another version of this system, any agreements with the rejected copy-text reading in editions later than the one from which the emendation is drawn — or any additional post-copy-text readings — would appear in the list of rejected substantives, and thus in these cases the list of emendations would not provide the entire history of the variants.)[66] The obvious motive for this arrangement is economy, and there is no doubt that in many cases the apparatus can be considerably shortened by the procedure; how much it is shortened depends on the number of substantive emendations (exclusive of those initiated by the editor at points where no
The form of the entry in a historical collation is essentially the same as in a list of emendations, except that the sources of the rejected readings must be specified (whereas in a list of emendations the rejected readings are by definition from the copy-text and thus do not have to be individually identified as such). In addition, since the reading from the edited text provides the lemma in each case, there is strictly speaking no necessity to identify its source, since if it is not from the copy-text its source is recorded in the list of emendations.
10.31 whom] who 50-60,E [historical collation]
127.4 moan] 45-57; moon 37-42; mean 60-70
Variants within impressions raise special problems for the historical collation, just as they do for the list of emendations. Since a knowledge of such variants in the copy-text edition is necessary for the precise specification of copy-text, they should certainly be recorded (at least those which involve more than variations in inking or slight type damage); but since these variations are likely to be in accidentals as well as substantives, not all of them would be appropriate for recording in the historical collation (if, as usual, that collation is limited to substantives). It seems sensible, therefore, to set up a special list to record such variants[72] (examples are the lists of press-variants in the Dekker and Beaumont-Fletcher editions and the lists of variants within the first and within the second editions in the Ohio State Scarlet Letter); alternatively — or additionally — these variants can be discussed in the textual essay as part of the definition of copy-text or of the bibliographical comment on other editions (as in the Melville edition). If any of the variants do turn out to be substantives, they should also be reported in the regular historical collation, since they form a part of the full history of the readings at these points. But determining which ones are substantives sometimes turns — as information about variants within impressions necessarily turns — on the particular group of copies collated or examined. In Chapter 70 of Melville's White-Jacket, for instance, the American edition (copy-text) reads "President" at a point where many copies of the English edition read "[]resident"; the space suggests that a "P" failed to print, but what did print — "resident" — is a different word, and, if no copies of the English edition could be found with the "P," the word would technically be a substantive variant. Copies reading "President" were eventually located, however, and this variant — though it deserves mention in the textual essay (or in a special list) — need not be entered in the historical collation. Once again, the intimate connection between descriptive bibliography and editing is evident: the greater the number of copies which are examined, the more reliable the evidence on which the edition is based.
The idea of separating certain categories of historical information for
presentation in special lists can be applied to other situations as well. Two
kinds of special lists may result. One kind merely repeats data present in
the full historical collation — data which the reader may find useful
to
have brought together in one spot. In the Hawthorne edition, for example,
there are sometimes (as in The
Any consideration of editorial apparatus is misguided if it loses sight of the convenience of the reader. For some audiences, the apparatus may be irrelevant and need not accompany the edited text; but for most scholarly audiences an edition without apparatus resembles any other work that lacks documentation — it may be brilliantly done, but it provides no aids for facilitating the scholar's independent investigation of the evidence. The apparatus (as the word itself suggests) is a tool for expediting further study, and a tool, to be effective, must be as simple and as easy to use as the circumstances allow. Fredson Bowers — through his connections with editions of Dekker, Beaumont and Fletcher, Fielding, Hawthorne, Crane, and Dewey — has done more than anyone else to set the course of modern apparatus along these lines. As a result of his efforts, there is now not only a widespread acceptance of an efficient basic approach to apparatus but also an increased awareness of the significance of apparatus. Though just a tool appended to a text, the apparatus may well be the only part of an edition that can meaningfully be called "definitive": there may legitimately be differences of opinion about certain emendations which an editor makes, but a responsible apparatus is a definitive statement of the textual situation (within the limits of the copies examined). What constitutes an apparatus responsible in both form and content is therefore a matter worth serious consideration. Only by being fully cognizant of the issues and problems involved in setting up an apparatus can an editor make those decisions which will establish his apparatus as a lasting contribution to literary study.
Notes
"The Rationale of Copy-Text," SB, 3 (1950-51), 19-36; reprinted in Greg's Collected Papers, ed. J. C. Maxwell (1966), pp. 374-91.
The ground rules might be different—if this were what Fredson Bowers calls a "practical edition"—so that less research would be expected. But the care devoted to establishing the text on the basis of the available evidence would be no less, for this is the editor's essential task.
Informal citations of these editions throughout this essay refer to the following: The Mark Twain Papers, ed. Frederick Anderson et al. (Berkeley: University of California Press, 1966- ), esp. Hannibal, Huck & Tom, ed. Walter Blair (1969), and Mysterious Stranger Manuscripts, ed. William M. Gibson (1969); The University of Virginia Edition of the Works of Stephen Crane, ed. Fredson Bowers (Charlottesville: University Press of Virginia, 1969- ); The Early Works of John Dewey, 1882-1898, ed. Jo Ann Boydston, in consultation with Fredson Bowers (Carbondale: Southern Illinois University Press, 1969- ); The Centenary Edition of the Works of Nathaniel Hawthorne, text ed. Fredson Bowers, with Matthew J. Bruccoli (Columbus: Ohio State University Press, 1962- ); A Selected Edition of W. D. Howells, ed. Edwin H. Cady, Don Cook, Ronald Gottesman, David J. Nordloh et al. (Bloomington: Indiana University Press, 1968- ); The Complete Works of Washington Irving, ed. Henry A. Pochmann et al. (Madison: University of Wisconsin Press, 1969- ), esp. Mahomet and His Successors (1970); The Writings of Herman Melville: The Northwestern-Newberry Edition, ed. Harrison Hayford, Hershel Parker, and G. Thomas Tanselle, with Richard Colles Johnson (Evanston and Chicago: Northwestern University Press and The Newberry Library, 1968- ); The Writings of William Gilmore Simms: Centennial Edition, ed. John Caldwell Guilds and James B. Meriwether (Columbia: University of South Carolina Press, 1969- ). In addition, the following are referred to allusively throughout: The Dramatic Works in the Beaumont and Fletcher Canon, ed. Fredson Bowers et al. (Cambridge: Cambridge University Press, 1966- ); The Dramatic Works of Thomas Dekker, ed. Fredson Bowers (Cambridge: Cambridge University Press, 1953-61); The Journals and Miscellaneous Notebooks of Ralph Waldo Emerson, ed. William H. Gilman et al. (Cambridge: Belknap Press of Harvard University Press, 1960- ); The Wesleyan Edition of the Works of Henry Fielding, ed. W. B. Coley, Fredson Bowers, et al. (Middletown, Conn.: Wesleyan University Press, 1967- ), esp. Joseph Andrews, ed. Martin C. Battestin (1967).
Standard brief descriptions of this kind of apparatus appear in two essays by Fredson Bowers: "Textual Criticism," in The Aims and Methods of Scholarship in Modern Languages and Literatures, ed. James Thorpe (1963; rev. ed., 1970), esp. pp. 53-54; and "Some Principles for Scholarly Editions of Nineteenth-Century American Authors," SB, 17 (1964), esp. 227-28. See also Statement of Editorial Principles: A Working Manual for Editing Nineteenth Century American Texts (CEAA, 1967), pp. 9-10. During the same years there have, of course, been proponents of other approaches to editing and to apparatus. Among the best known are Edmund Wilson (represented by his essays in the New York Review of Books on 26 Sept. and 10 Oct. 1968, reprinted the same year in pamphlet form as The Fruits of the MLA) and F. W. Bateson (represented by his editorial plan for "Longman's Annotated English Poets" and reflected in his letter in the TLS on 1 Jan. 1971, pp. 14-15). Some discussion of Wilson's views can be found in Professional Standards and American Editions: A Response to Edmund Wilson (MLA, 1969) and of Bateson's in Thomas Clayton's letter in the TLS on 18 Dec. 1970, p. 1493.
The term "practical edition" is used here in the sense established by Fredson Bowers in "Practical Texts and Definitive Editions," in Two Lectures on Editing (1969), pp. 21-70. Further comment on the relation between definitive editions and widely disseminated reading editions appears in his "The New Look in Editing," South Atlantic Bulletin, 35 (1970), 3-10.
Encouraging such reproduction does not imply that only one reliable text of a work can exist. Obviously more than one text can be prepared following sound scholarly procedures, for there may legitimately be differences of opinion about certain emendations which rest on critical evaluation. (For discussion of this point, see G. T. Tanselle, "Textual Study and Literary Judgment," PBSA, 2nd Quarter 1971.) The point is that a practical edition should embody some reliable text, and, if such a text exists, the publisher of a practical edition should be encouraged to lease it rather than reprint, with no rationale, whatever previous text comes most readily to hand.
Some works of this kind may be of such importance that they will be frequently quoted; in these cases it may be more convenient to have a clear text (with variant readings recorded at the end), even at the sacrifice of the basic texture of the original. This sort of decision, involving a weighing of what is gained against what is lost, has to be made separately for each individual case. For further comment on this problem, see G. T. Tanselle, "The Editorial Problem of Final Authorial Intention," forthcoming in Bibliographia.
I use the term "finished piece of writing" rather than "literary work" in order to include historical, technical, and scientific writings or any other work completed for publication or of a type usually intended for publication.
Another use of symbols in the text, convenient in certain situations in practical editions, is to draw the student's attention to important revised passages. In the Signet Typee (ed. Harrison Hayford, 1964), passages which Melville deleted in the revised American edition of 1846 are enclosed in square brackets, and those which he revised are both bracketed and numbered, with the revised wording given at the end of the volume according to the reference numbers.
"Old-Spelling Editions of Dramatic Texts," in Studies in Honor of T. W. Baldwin, ed. D. C. Allen (1958), p. 14.
The question takes a slightly different form for practical editions, since it may be felt that a classroom edition with little apparatus except explanatory notes should offer those notes as easily accessible footnotes. It is perhaps true that more students will read them as footnotes, but the price paid for this attention is a high one: not simply the distraction from the text (which is after all more important for the students to read), but the cumulative psychological effect of always (or nearly always) encountering classic works encased in an obtrusive editorial framework which sets them apart from other books read outside of class. Sometimes it is objected that references to line numbers are awkward and inconvenient when side-numbers counting the lines do not appear on the text pages. Side-numbers have been so widely used in connection with poetry that they probably constitute little distraction there (and thus do not prevent a poetic text from being "clear"); but their presence on a page of prose remains an intrusion and lends the page a "textbook" air. The psychological advantages of clear text, therefore, can be said to compensate for the minor inconvenience of having to count lines.
The use of parallel texts is often a more sensible way of exhibiting complicated revisions than to present one established text with the revisions recorded in apparatus; besides, some complex revisions result in what amounts to a different work, so that both forms of the work deserve to be presented as texts in their own right. Placing two texts in parallel columns or on facing pages is in itself a kind of apparatus; but, except for that, the comments made here about the texts of other editions would also apply to the individual texts of a parallel-text edition.
Indeed, sometimes the apparatus appended to a text is more important to scholars than the text itself, since, if it is well done, it provides the evidence on which other editions can be constructed by editors who do not agree with the interpretation of the evidence represented in that particular edition. A good example of an apparatus presented as a piece of research in its own right is Matthew J. Bruccoli's "Material for a Centenary Edition of Tender is the Night," SB, 17 (1964), 177-93.
Even a facsimile or a diplomatic edition of one particular impression of a work or one particular copy of an impression is based on a decision to present a given text and cannot be approached with the attitude that "it makes very little difference what text is printed in extenso."
Since the basic goal of an edition is to establish a text rather than to present an apparatus, the effect which the text makes would apparently—if it comes to a choice—be given somewhat more weight than the convenience with which the apparatus can be located.
Both because one knows less readily where to turn to find the apparatus and because comparative study involving several works requires more extensive page-turning. In addition, the editor may have practical reasons for preferring a single block of apparatus at the end, since it enables him to key all his apparatus to page proof at one time. If sections of apparatus are scattered through a volume, the process is inevitably less efficient; for if the editor gets galleys first, he must wait to key the apparatus for his second text until the apparatus for his first has been made into pages, so that the pagination of the second text is known, and so on through the volume; and if the editor receives pages directly, then the apparatus must be set up with blank references ("00.00") and all the figures later altered.
Considerations of the ease with which individual texts and apparatus can be reproduced photographically have little relevance here, for even when the apparatus immediately follows the work the pagination would be appropriate for separate issue only for the first work in the volume; and when pagination must be altered in any case, there would be no additional problem in taking the apparatus from the end of the volume and altering its page numbers also. A real problem might arise, of course, if the apparatus for a given work were not presented as a unit and if apparatus pertaining to other works appeared on some of the same pages; but there is no reason why apparatus at the end of a volume cannot be so arranged as to avoid this problem. (See the following paragraph and footnote 18.)
Some extra space is required, of course, for the additional headings which would be needed. Further space would generally be used if one always began the section for a given work on a new page, in an attempt to facilitate photoreproduction of an individual text with its apparatus (see footnote 17). But if only a few long texts were involved, little space would be wasted in this way, while for shorter texts, such as essays, stories, and poems (where more space might be wasted, since more of these works could be included in a volume), there would be less reason to accommodate photoreproduction, because less demand exists for separate reprints of individual short works. There would be little reason, in other words, for beginning the apparatus to each work on a new page except when a volume contains only two or three long works.
Bowers's Dekker employs this system, though the situation is somewhat different since the record of substantive emendations appears as footnotes. It is also used by Matthew J. Bruccoli in "Material for a Centenary Edition of Tender is the Night," SB, 17 (1964), 177-93, but again there are special circumstances since here the apparatus is presented independently of the text. And the Virginia Crane edition uses the "stet" system even though the list of emendations follows the textual notes.
Still another arrangement is employed in the first volume of the Simms edition, where the list of emendations and the textual notes are merged: that is, whenever a reading requires comment, the comment is inserted at the appropriate point in the list of emendations. The advantage, of course, is a reduction in the number of separate sections of apparatus, so that the reader is involved in less cross reference between sections. But, as usually happens when notes are tied to a list of emendations, some entries for unemended readings have to be included in the list. Furthermore, there is the danger that the insertion of blocks of discursive material into the list will make the list less easy to follow; in the case of the first Simms volume, the number of notes is small enough that this difficulty does not arise to any significant extent, but it remains a possibility when there is a considerable number of notes—and when the number is small, the notes may actually prove less readily accessible if imbedded in a list. Finally, difficulties of design in joining paragraphs and lists (as when lengthy notes must be accommodated to a double-column page designed primarily for listed items) provide an additional argument against using this system under ordinary circumstances.
The Cambridge Dekker and Beaumont-Fletcher editions also place the textual notes immediately after the text, but in effect these notes follow the substantive emendations, since the substantive emendations are recorded in footnotes on text pages.
The positions of various other special lists which may be required are commented on below, at the points where these lists are discussed.
Though incidentally it does list the substantive ones, since each entry has to be keyed to the reading which appears in the edited text.
In the Dewey edition, both a textual introduction and a historical introduction precede the text; but the pages on which they appear are numbered with small roman numerals, and this sequence of pagination is resumed at the back of the volume for the remainder of the apparatus. This arrangement thus makes it simple, in other printings or photographically reproduced sub-editions, to bring all the editorial matter together.
In the Wesleyan Fielding, though the textual apparatus is at the end of the volume, the explanatory notes are placed at the bottom of the text pages (keyed to footnote numbers in the text).
McKerrow was clearly aware of this feeling and admits that compression was his principal consideration: "I may say here that the conventions, which at first sight may appear somewhat complicated and even perverse, have only been adopted after careful thought and experiment, and actually do—at least in my deliberate opinion—make it possible to give all necessary facts in the minimum of space" (p. 77).
In both cases, however, some of the symbols do not follow the same system. In The Scarlet Letter, the first two editions are designated not by roman numerals but by "18501" and "18502"; and in Their Wedding Journey the serial publications are referred to as "S1" and "S2" (chronologically S1 would precede A, and S2 would follow C). Both editions use the mnemonic symbol "MS" for manuscript texts, and the Hawthorne uses "E" to distinguish English editions. In the Wesleyan Joseph Andrews, the five editions published during Fielding's lifetime are designated by simple arabic numerals, 1 through 5.
The system referred to in Mahomet, p. 584, note 17, as the standard system for the Irving edition uses such symbols as "1A1, "1A2," etc. It is difficult to decide whether the mnemonic value of these symbols is greater than that of symbols incorporating references to years; but under most circumstances of any complexity, symbols which employ the last two digits of the year (instead of the second figure here) and initials of publishers (when more than one publisher in a given country is involved) are probably easier to remember.
Of course, if a variant in punctuation were not involved, the punctuation would not be included in the reading at all; but the fact that it is included does not rule out the possibility that a spelling variant also exists at this point.
As stated above, editions of manuscript material which attempt to show stages of composition may require more elaborate sets of symbols (including perhaps angle brackets for canceled matter and vertical arrows for insertions, both of which have been fairly widely used); but the general principles for evaluating symbols outlined here would still apply. (Useful examples of symbols for editing manuscripts are found in the Hayford and Sealts Billy Budd and in the Emerson and the Irving journals.) It should be noted that complicated alterations in a manuscript can also be set forth in verbal descriptions, without any symbols at all, as in the Ohio State House of the Seven Gables and in the California Hannibal, Huck & Tom and Mysterious Stranger Manuscripts. Fredson Bowers, in his review of the New York Public Library edition of Walt Whitman's Blue Book (ed. Arthur Golden), makes some comments on the relative merits of the two approaches and finds Golden's method an uneconomical mixture of the two—see JEGP, 68 (1969), 316-20.
A sensible convention which has become well established is to use periods to separate the elements of these reference numbers (e.g., "240.17" or "III.ii.75"). All references in the apparatus should obviously be to the edited text (though in the textual essay a discussion of type damage or defective inking in an early edition might well involve page-line references to that edition).
A space, for example, would suffice; or, as in the Howells edition, the reading itself need not be cited, since the discussion can be constructed so as to make clear what word or words in the cited line are in question.
It might not literally waste a particular reader's time if a given fact, though easily ascertainable in the dictionary, were not already known to him. But an editor cannot pitch his annotation at his text's least informed reader, even if he could discover who that is; some minimum level must be recognized, and it seems reasonable to say that spellings or usages readily discoverable in standard dictionaries fall below that level. (Explanatory, as opposed to textual, annotation is of course a different matter; in an explanatory note it may well be useful to have a brief identification of a historical figure, even though he is listed in the basic biographical reference works. The essential difference is that historical allusions, however numerous, are manageable in number and affect one's understanding of the meaning of the text, whereas the kind of textual notes ruled out here might logically involve half or more of the individual words of a text and by definition would not raise such special problems as interpretation or meaning.)
As opposed to those which may be classed, for one reason or another, as nontextual—about which more is said below.
Often each of these entries is placed on a separate line, but sometimes, to save space, they are run on in paragraph form (a form which makes individual entries somewhat less easy to locate).
Nontextual changes—those affecting the design of the document embodying the copy-text but not the text itself—may of course be made silently. But to make textual changes silently, even though the categories of such changes are announced and discussed in the textual essay, is to deny the principle that it is risky to allow the absence of a positive designation to be significant (cf. the comments on the caret above). Thus, in the Melville example, merely informing the reader that "-our" spellings are changed to "-or" does not allow him to reconstruct the copy-text with the certainty he would have if he could follow an actual list of changes; for he could not be sure that every "-or" word in the edited text was originally "-our," whereas with a list he would know explicitly just where the changes were made. Furthermore, the specific instances of any type of textual alteration, however trivial they may seem, may be of particular concern to some linguistic, literary, or historical scholar, and the burden of locating these instances should not be placed on the user of the text but is rather the editor's responsibility.
The same principle is followed in the Ohio State Fanshawe and the California Mysterious Stranger Manuscripts, where a number of groupings of identical changes in accidentals are made; here, however, the references are cited in paragraph form in the list of emendations at the point of first occurrence of each type. To some extent this arrangement disrupts the smooth sequential flow of the list of emendations and makes it somewhat less easy to follow; but there is no doubt that it is an advantage to the reader to have these groups of identical emendations brought together somewhere.
Only if the source of the emendation were the editor himself or the last of the collated editions could the reader know that no additional information would be found in the historical collation.
In the Howells edition, when editions later than the one from which an emendation is taken agree with that edition, the complete history is implied in the entry. But when a later edition reverts to the reading of the copy-text, that further history is not offered in the emendations list. See footnote 66.
Of course, editors of facsimile or diplomatic editions are necessarily concerned with formal and typographical matters and must take them into account.
Nontextual details often play a great role in the bibliographical analysis which leads to the establishment of the text (as when wrong-font types allow a bibliographer to learn something about the timing of the distribution of type from preceding formes or about the order of formes through the press), but they are nevertheless not a part of the text.
With the exception that, when all variants are manifestly incorrect and their order is indeterminate (as in the example cited in footnote 48), designating only one of them as the copy-text reading becomes a pointless exercise. (In the case of a manuscript copy-text, of course, alternative uncanceled readings may well exist at individual points.)
Thus Bowers, in the general textual introduction to the Cambridge Beaumont and Fletcher, defines the copy-texts as embodying the readings of corrected formes: "The normal assumption is that the present edited text reproduces the corrected readings when press-variation is present if no contrary record is made" (I,xix).
Since a photographic reproduction of a single copy is often used as printer's copy for a critical edition, it follows that not every textual alteration marked on that copy is an emendation in the copy-text, for some may bring the printer's copy into conformity with the copy-text.
Some further discussion of this point appears in G. T. Tanselle, "The Use of Type Damage as Evidence in Bibliographical Description," Library, 5th ser., 23 (1968), esp. 347-48.
When neither form is correct, as in "t obelieve"/"tobelieve," it makes little difference which is considered the copy-text form, since an emendation is required in either case; in such instances, especially when the order of the variants is not clear, there is no point in choosing among incorrect forms, and both readings might as well be listed as the rejected copy-text readings.
Similarly, punctuation following (or preceding) a word need not be cited when only the word, and not the punctuation, is at issue.
William Hill Brown, The Power of Sympathy, ed. William S. Kable (Columbus: Ohio State University Press, 1969).
Editorial comments of this kind should of course be enclosed in square brackets to show that they are not actual readings; italicizing them is usually not sufficient, since italic words could appear in the text.
In other words, the letters do not stand for general stages of revision or alteration but refer only to the sequence at a given point. Thus there is no reason to suppose that one reading labeled "A55b" occurred at the same time or in the same process of revision as another with the same label; all that the symbol implies is that these are the second readings at each of these points.
In the Ohio State edition of The Scarlet Letter and in his paper before the South Atlantic Modern Language Association (cited in its 1964 published form in footnote 4 above).
And obviously, if the reader is accurately to reconstruct the copy-text from the critical text, he must have this information for interpreting the critical text.
If a possible compound coincidentally hyphenated at the end of a line in both the copy-text and the critical text is not recorded in a separate list, then it would appear in both these lists, and to that extent words involving editorial decisions might appear in the critical-text list. But their presence there has nothing to do with the fact that their established forms result from editorial decisions; they are there only because they are hyphenated at the ends of lines in the critical text. The fact that editorial decisions are involved can be learned only by noting the reappearance of the same words in the copy-text list.
Of course, an editor could insist that the lines of the text be reset until no hyphen which should be retained in quotation fell at the end of a line; in practice, this approach is often prohibitively expensive and, in some cases, virtually impossible of achievement.
And if the editor feels that it is of some help to the reader to have such words noted, a symbol can be placed beside those words which appear in both lists. (The Melville edition uses a dagger for this purpose.)
And emphasizing this division helps to make clear—as removing the critical-text list to another location would make still clearer—why some words turn up in two lists: since the critical-text list has nothing to do with editorial decisions, any word in it which in fact results from an editorial decision must naturally be found also in one of the two lists which record editorial decisions—the list of emendations or the copy-text hyphenation list.
It should be clear that the opposite possibility (employed in the first volume of the Simms edition)—that is, recording only those instances of possible compounds hyphenated at line-ends in the critical text which should be transcribed as single unhyphenated words—leaves ambiguities unresolved, for the reader still has to distinguish between purely compositorial hyphens, dividing unhyphenated words at line-ends, and the hyphens which should in fact be retained. (Of course, listing every line-end hyphen which should be eliminated in transcription—the only way to make this approach unambiguous—would be foolishly inefficient, since the majority of line-end hyphens in any printed work normally fall into this category, and the list would be extended inordinately.)
Of course, what remains is actually made up of two categories: possible compounds which, by editorial decision, should not contain hyphens, and words which are not possible compounds and which naturally do not contain hyphens.
Because the functions of these lists are not always grasped at first by the general reader, it is important that the headnote to each list not make the lists sound more complicated than they are. For the copy-text list, nothing more is needed than a statement of this kind: "The following are the editorially established forms of possible compounds which were hyphenated at the ends of lines in the copy-text." And for the critical-text list: "In quotations from the present edition, no line-end hyphens are to be retained except the following."
The simplest form is merely to list, following the appropriate page-line number, the word in its established form. Since the place where line-end division occurred is obvious in most cases, there is usually no need to mark it with a vertical line. Of course, when the point of division is not obvious—as in a compound with three elements and two hyphens—a vertical line can be used; but even then the vertical line is useful only in the copy-text list, not in the critical-text list. (The Dewey edition, in the critical-text list, gives the word first as a lemma, showing the line-ending with a vertical stroke, and then the established hyphenated form; such repetition does not make the function of the list clearer and indeed would seem to add a needless complication.) Furthermore, in the critical-text list, where every page-line citation would technically contain two line numbers (since each cited word runs over a line-end in the critical text, to which all citations are keyed), the awkwardness of the double-line reference serves no real purpose, and each page-line citation might as well refer simply to the line on which the word begins.
Sometimes certain of these entries do in fact provide histories of the readings involved, but that is not their primary function.
Certain other editions which, because of their wide popularity or impressive scholarship, have been influential in the history and study of the text may also be included in the historical collation, even though the variant readings present in them can carry no authorial sanction; indeed, editions of Elizabethan works often include practically every previous edition in their historical collations and thus provide the complete history of the treatment of the text with regard to substantives.
This plan is followed in the Howells edition (note the entries in Their Wedding Journey for 102.28 or in Literary Friends and Acquaintance for 223.23). As a result of the overlapping function of the two lists under this plan, the reader cannot know, when looking at the list of emendations, whether or not any given substantive entry contains the complete history of the variants at that point and must turn to the list of rejected substantives to see if any additional history is recorded there.
However, impressions of an edition need not be specified when there are no variants in them. Thus if "A" stands for the only American edition, that symbol alone could signify all the collated impressions of the American edition. But if a variant first shows up in, say, "A1847" (or "A47"), its history will be represented more clearly by "A47-76" than by defining "A47" to include all subsequent collated impressions. The use of inclusive notation, of course, does not result in the appearance of every siglum in each entry, but, when the symbols include mnemonic allusions to years or sequences, the grouping which would include any given siglum is obvious. For convenient reference, a list of all collated editions (or impressions) with their sigla should be included in the headnote to the historical collation (as well as in the headnote to any other sections in which the sigla are used).
A related symbol of McKerrow's, the plus-and-minus sign (±), is put to good use in Bowers's Dekker to stand for a general but not exact agreement among several editions, where the minor variations are irrelevant to the main fact which the entry is recording The same method could be applied to the specification of individual editions by enclosing in parentheses those sigla which refer to editions containing the slightly variant readings. (Such a practice would conform to McKerrow's use of parentheses, referred to above, to indicate "a reading which is not identical with one which is given but which is substantially the same in meaning or intention so far as the purpose of the note is concerned" [p. 82].) Sometimes earlier editions went too far in multiplying symbols of this sort: in the opening volume of the Variorum Shakespeare (1874), for example, Furness employs "&c," "et cet.," and "the rest" to stand for different groups of editions. Another symbol relevant to the matter of inclusive notation is the dollar sign, which has been borrowed from descriptive bibliography and introduced into textual apparatus by Bowers in the fifth volume of the Crane edition; it is used there to mean "all" or "every" when attached to symbols which subsume a number of documents (such as "N," the syndicated newspaper texts of a given work). The symbol is useful in "$N" to emphasize the fact that all the examined N texts agree and "$N (—N4)" to reinforce the statement that all but one agree; but since "N" is already a generic symbol, defined as all the examined newspaper texts, the dollar sign is essentially a device for adding emphasis rather than for condensing the statement.
Even in such a case, a variant at a point of emendation is somewhat awkward, since the reading in the edited text must be cited as the key for the entry, and it is different in these instances from the first-column (copy-text) reading.
It would be highly undesirable to have a situation in which a reading from a third edition had to be placed in a third column, even though it agreed with the reading in one of the other two editions; such an arrangement would make it more difficult for the reader to note agreements among editions and would open up more possibilities for typographical errors in the list.
Sigla in these lists would refer to particular copies of books, not just to particular impressions.
These lists do not record the complete history of the variants listed, for their function is only to note that the variants were present in a particular edition and are not adopted in the critical text. (Strictly speaking, therefore, no sigla at all would be required in such lists.)
When only a brief manuscript fragment survives, it can be treated either in a separate list (as "The Ohio State University Leaf" in The Blithedale Romance) or in a complete transcription with accompanying apparatus (as in the Northwestern-Newberry Typee, Mardi, and White-Jacket).
Placing all such lists immediately after the basic historical collation helps to make clear that they are parts of the historical record, appendixes in a sense to the historical collation. (The attempt to make the list of emendations serve as a partial historical collation is not an extension of the principle that certain categories within the historical collation can be conveniently separated, for it mixes the functions of the lists; all these special supplementary lists are purely historical in function.) Sometimes certain of these lists—especially those dealing with variants within an impression—are placed first in the apparatus, since they often deal with material which chronologically precedes that taken up in other lists; but chronology is not the general basis for the organization of the apparatus as a whole, and readers can probably find their way around in an extensive apparatus more easily if the arrangement is based on the distinct functions of the several lists.
| | |