 | |  |
Notes
A manuscript or typescript may be reproduced photographically, but there is still only one original, and all the evidence is located in it. Typescripts produced from the same cards by a magnetic-card typewriter, however, would take on the characteristics of an edition.
"The Objects and Methods of Bibliographical Collations and Descriptions," Library, 2nd ser., 8 (1907), 193-217 (quotation from p. 210). He also said, "If a book is complete and perfect—i.e., in its original condition—or when we have ascertained how far it falls short of completeness and perfection, we have still to determine what relations it bears to other copies of the same work" (p. 207).
"Some Points in Bibliographical Descriptions," Transactions of the Bibliographical Society, 9 (1906-8), 31-52 (p. 41). There follows a discussion distinguishing "such differences as have been introduced by subsequent treatment, binding, mutilation, etc." from the "already important distinctions" with which copies left the press—the most important of which are the variant textual states of individual formes (pp. 41-42).
Transactions of the Bibliographical Society, 12 (1911-13), 213-318 (see "On variations in different copies of the same edition," pp. 282-289).
"A Formulary of Collation," Library, 4th ser., 14 (1933-34), 365-382 (p. 371); reprinted in his Collected Papers, ed. J. C. Maxwell (1966), pp. 298-313 (p. 303).
"Criteria for Classifying Hand-Printed Books as Issues and States," PBSA, 41 (1947), 271-292 (esp. p. 290).
"The Fallacy of the Ideal Copy," Library, 5th ser., 33 (1978), 108-118. This paper has been succinctly criticized (as one "which will no doubt cause considerable harm and confusion" by Rolf E. DuRietz in "'The Fallacy of the Ideal Copy,'" Text, 2, no. 1 (1978), 31-34. In his comments DuRietz hopes that I will publish a "refutation of all this nonsense as soon as possible." The present essay was written before DuRietz's remarks appeared and was not conceived of as a reply to Miss Pouncey. Insofar as her remarks will be taken seriously enough to require a reply, DuRietz's comments serve the purpose very well. And although my own essay is not organized as a systematic examination of the points Miss Pouncey makes, I believe that it does contain the arguments required to reply to those points.
In his important earlier article, "A Formulary of Collation" (see note 5 above), he refers to "the ideally perfect copy" as the subject of the collation, without at that point mentioning intention. But that intention is involved the context makes clear: cf. my further quotation from this piece in the discussion to follow.
Its character is certainly hypothetical; what differs from instance to instance is its likelihood of being true.
Although its original place as printed may also, when known, be made clear in the collation formula—as in "π1 (=M8)."
That is, they may be defined as anomalies on the basis of the evidence they present. Whether or not they are anomalies in a statistical sense cannot be known, since the evidence in the copies now lost cannot be recovered.
In his 1947 article (see note 6 above), he had defined it as "that form of the book which the printer or publisher wished to represent the most perfect and complete form to leave his shop" (p. 290). James G. McManaway used similar language in his contribution to the 1946-47 series of Rosenbach Fellowship lectures, where he defined ideal copy as "the book in the precise state in which the publisher intended it to be put on sale as complete and perfect" (Standards of Bibliographical Description [1949], pp. 70-71).
At a later point Bowers succinctly says, "the intention of the printer should take precedence over what happened to be the practice of various binders" (p. 122).
An issue, because it is a consciously planned publishing unit, can be thought of as having an ideal copy (or perhaps more than one); an issue or impression may in turn consist of states, but because they do not represent discrete publishing efforts, they are not given separate descriptions as distinct ideal copies. States are subdivisions of an issue, produced by the effort to make corrections or additions that are not to be called to the public's attention; they are not intended as identifiable publishing units. Cf. G. T. Tanselle, "The Bibliographical Concepts of Issue and State," PBSA, 69 (1975), 17-66.
One in which, as he says, "Nothing is invented. . . . Instead, all the evidence to be gained from the examination of numbers of copies is analyzed on the basis of the printing history of the edition, so far as it can be determined, in order to discover what was the actual most perfect form of the book achieved by the printer within an issue" (p. 113). The point is that this "form" is a composite of details from various copies and does not necessarily agree with any single actual copy.
The editor of a noncritical edition offers a closer parallel with the descriptive bibliographer; this point is discussed more fully below, in Part II.
This point is ironically demonstrated by the number of Rolf DuRietz's Text in which he comments on the Pouncey article (see note 7 above). At the center are two folds (eight pages) that are not counted in the pagination; they are blank except for the recto of the second leaf and the verso of the third. On the former of these is an announcement beginning, "These two inmost folds outside the pagination do not belong to the ideal copy of a complete bound volume 2 of TEXT. They are intended, most unbibliographically, to be removed and cancelled by the binder." DuRietz goes on to explain three functions that they serve (protecting the "inmost regular text fold" from the staples that hold the number together, protecting copying-machine glass from the staples, and providing space for listing errata) and adds information about the quality of the staples and the paper. The errata list, on the other printed page of these eight, is to be "reprinted on a regular page in a forthcoming issue of TEXT." However, despite the announced intention that these four leaves are to be excised and not to be considered a part of the ideal copy, history cannot be undone: the journal was in fact published with these leaves, which do contain text, and they are a part of the publication, whether its editor wishes them to be or not. In the future, copies lacking these leaves, far from being considered "ideal," will have to be regarded as defective.
Except to the extent that the present characteristics of individual copies must be recorded as documentation.
One cannot say simply "post-publication history of individual copies": if publication is taken to mean the time when the first copies are put on sale, then some post-publication changes can certainly be part of the production history of the book, in cases where the printer or publisher orders changes to be made in the copies not yet released. The distinction being made here relates to individual copies: what happened to each one before the time of its release as opposed to what happened to each one after that time.
When the publisher does call attention to a particular group of copies, as through a new title leaf or special binding, the result is a new issue of the book, which then has its own ideal copy. For more detailed discussion, see note 15 above and the article cited there.
Duplicated leaves are commonly regarded simply as a redundancy rather than a state. Yet if they become a feature of a copy during production rather than later, it is hard to see how logically they can be considered anything other than a state. Of course, one can always decide in certain cases that this one class of states will be relegated to the notes on individual copies and not be reported in the body of the description. Although this arrangement may seem best at times, in many instances it is bound to seem somewhat arbitrary and to raise more problems than it settles. If a substantial number of copies were found to contain the duplicated leaves, they would no doubt be regarded by most bibliographers as deserving of more attention. But when one is dealing with less than the full evidence, it is always difficult to say what relation a single copy with certain features bears to the whole issue. And in any case, if a feature is worthy of being recorded when it exists in large numbers, how can it be ruled out, when it is known only in a single copy, as not exemplifying one of the states in which the book was issued? The number of copies involved is likely to have an effect on one's thinking, but it cannot logically be allowed to determine what is reported as a state, if state is to be a term of historical classification, referring to what in fact took place. A cautionary example, involving what may be a cancel, is furnished by Jane Austen's Emma (1816). A copy was recently discovered in which M11 of volume 3 is a disjunct leaf pasted to a stub. Since this leaf corresponds exactly—textually and typographically—with the usual copies of M11, which are integral, one might be led to conclude that this copy is "made up"—that a missing M11 has been replaced by the corresponding leaf from another copy. However, a brief search has now turned up two more copies in which M11 is disjunct; the existence of three such copies makes one think that this supplied leaf may after all be a product of something that happened during the production of the book. The matter remains a puzzle: could the supplied leaf really be a cancel for a state of the text not yet located? The situation nevertheless illustrates that what may seem obvious when only one copy is known should not be confidently taken to be the correct explanation.
Something close to this point is implied by Philip Gaskell in his discussion of ideal copy in A New Introduction to Bibliography (1972). He says that a description of an ideal copy would note all leaves "which belonged to the most perfect copy of the work as originally completed by its printer and first put on sale by its publisher. This is the basic ideal form; and the description of ideal copy is completed by the addition of notes of any subsequent changes made by the printer or publisher . . . and of any unintentional alterations to its form" (p. 315). Despite the reference to a "basic ideal form," Gaskell seems in this passage to be moving toward the view I am setting forth here, suggesting that the description of idcal copy contains within it in chronological order the various states that historically existed. If, however, intended and unintended differences in copies put on sale later than the time when the first ones were put on sale are to be recorded, it is difficult to see why the "most perfect copy" as "first put on sale" should be the starting point, since there may have been other less "perfect" copies also offered for sale at that initial time and thus also deserving of a place in the historical record. When Gaskell repeats his definition of ideal copy a few pages later (p. 321), he introduces a significant difference: "It will be recalled," he says, "that the basic form of ideal copy is the most perfect state of a work as originally intended by its printer or publisher following the completion of all intentional changes." Of course, he had earlier, and more appropriately, used the phrase "as originally completed" instead of "as originally intended."
Bowers agrees that the "complete bibliographical description . . . apprises the reader of all variant states in which the described issue of the book is found" (p. 123). Where my approach is somewhat different from his is in suggesting that these states may be presented chronologically at the appropriate points in the main description, rather than that some of them be appended to a description of a single ideal copy. The information set forth is the same either way; but the rather arbitrary choices sometimes required to construct a single ideal copy may occasionally produce a misleading effect, one that can be avoided by treating the various states at the points where the elements they concern come up within the framework of the description.
In addition to the alternatives occasioned by the correction or the occurrence of error, there are of course planned differences among copies (as when copies are issued in various colors of casing cloth) and differences resulting from the nature of the publication plan (as when different original prints or other illustrations are tipped in to different copies). There are also those publications that allow for the rearrangement of their parts (such as boxes containing loose pieces, as in Marcel Duchamp's 1934 publication of facsimiles of 93 examples of his notes); in these cases there may or may not be any one arrangement of the pieces that can be described as standard or "ideal." Allan Stevenson, in his long introduction to the second volume of the Catalogue of Botanical Books in the Collection of Rachel McMasters Miller Hunt (1961), is bothered by such variability among copies: finding that the positions of the plates in different copies of botanical books vary greatly, he calls ideal copy for these books a "will-o'-the-wisp" (p. clxxii). Yet ideal copy, by definition, must exist, regardless of the number of variations. If the publisher is responsible for the binding of copies, and thus for the insertion of the plates, then all the variations originally present become part of the description of ideal copy; if the plates were inserted by binders employed by individual owners of copies, then the variations are of little significance compared to any binders' directions that were issued or any indications on the plates themselves announcing their intended locations.
For a discussion of the probabilities involved in drawing conclusions from a limited number of copies of an issue, see David Shaw, "A Sampling Theory for Bibliographical Research," Library, 5th ser., 27 (1972), 310-319. The problems of interpreting bibliographical evidence, including that arising from the existence of variant formes, are thoroughly analyzed by Fredson Bowers in Bibliography and Textual Criticism (1964).
An example from Greg's "Bibliography—An Apologia," Library, 4th ser., 13 (1932-33), 113-143: "what the bibliographer is concerned with is pieces of paper or parchment covered with certain written or printed signs. With these signs he is concerned merely as arbitrary marks; their meaning is no business of his" (pp. 121-122). Reprinted in his Collected Papers (see note 5 above), pp. 239-266 (p. 247).
This study is of course most fully set forth in The Printing and Proof-reading of the First Folio of Shakespeare (1963).
For this reason, Bowers recommends the choice of uncorrected formes generally for facsimile editions (and they are the formes that would normally supply copy-text for critical editors). This point and related matters regarding press variants are carefully explored in his "The Problem of the Variant Forme in a Facsimile Edition," Library, 5th ser., 7 (1952), 262-272. He describes "the ideal photographic facsimile" as "a collection of formes from any number of copies, these formes being chosen first according to the principle of their textual state, and second according to clarity and fidelity of the inking" (p. 263).
Or to those in a given manuscript text; but the focus here, in the context of ideal copy, is of course on printed texts.
Miss Pouncey asks, "Can any Anglo-American analytical bibliographer be persuaded that the literary works which are the content of printed books are the products of the creative imagination, and that the creative imagination disallows mechanistic studies of its productions?" (p. 115). This question obviously confuses the works themselves with their means of transmission. The workings of the creative imagination are not mechanical; but so long as the resulting creations are transmitted by physical means, a "mechanistic" study of those means is clearly necessary for an informed approach to the content of the works thus transmitted. She had earlier asked, in similar vein, "to what avail is knowledge of the printing process when the editor must make a choice based on intuition?" (p. 109). But a detailed knowledge of the printing history of a book may settle some textual points conclusively and is thus necessary for establishing the limits within which editorial judgment (or "intuition") can operate. No matter how far editors may finally move beyond the physical evidence in their attempt to define the products of creative imagination, they must begin with that evidence and cannot in their speculations contradict it.
The latter two examples are recorded in the Northwestern-Newberry Edition of Mardi (1970), p. 706, and White-Jacket (1970), p. 484; the former two will be reported in the forthcoming editions of The Piazza Tales and Moby-Dick.
For example, Fredson Bowers, in "Purposes of Descriptive Bibliography, with Some Remarks on Methods," Library, 5th ser., 8 (1953), 1-22, says, "I do not see how, ordinarily, a descriptive bibliographer can be responsible for collating the texts" (p. 15), though he adds that the bibliographer should be able to "guarantee that every page of every listed copy is in the same setting and imposition unless specific exceptions have been made" (p. 16). Reprinted in his Essays in Bibliography, Text, and Editing [1975], pp. 111-134 (pp. 127, 128). And I have myself said that "collations are theoretically part of the necessary background for every descriptive bibliography, whereas from a practical point of view it is unreasonable to demand them"—"The Use of Type Damage as Evidence in Bibliographical Description," Library, 5th ser., 23 (1968), 328-351 (p. 351).
A recent example is James L. W. West III, William Styron: A Descriptive Bibliography (1977), which includes tables of variant readings.
A convenient list of these editions appears in the Modern Language Association's The Center for Scholarly Editions: An Introductory Statement (1977), pp. 5-6; this Statement is also printed in PMLA, 92 (1977), 586-597.
These and other networks are identified, among other places, in an extremely useful introductory discussion of some of the issues that bibliographers should be aware of in relation to computer cataloguing: Terry Belanger and Stephen P. Davis, "Rare Book Cataloguing & Computers," AB Bookman's Weekly, 63 (5 February 1979), 955-966.
The Independent Research Libraries Association has established a committee (which first met in Washington on 13-14 March 1979), under the chairmanship of Marcus A. McCorison, to consider ways of supplementing AACR2 (the second edition of the Anglo-American Cataloguing Rules) and ISBD(A) (see the following note) to accommodate the interests of those concerned with the physical aspects of books. No library can be required to include extensive records of physical details in its cataloguing; but it is important to reserve spaces now that will allow for the recording of such details and to provide a standard framework for them so that any libraries wishing to note certain details will be doing it in the same way. It would be a mistake not to plan in this fashion for the future orderly accumulation of bibliographical details in a system that can make them easily retrievable.
The 1978 draft of ISBD(A)—that is, the International Standard Bibliographic Description designed to be a counterpart, for antiquarian (A) books, of ISBD(M), which is primarily concerned with current monographs (M)—states specifically that it deals with the special problems of "older books." Its scope is defined as "requirements for the description and identification of non-serial printed items issued before the year 1801, as well as for later works when the items are produced by hand or by methods continuing the tradition of the hand-produced book" (0.1.1.). Machine-produced books, however, are also of interest as artifacts, and recognition should be given to the fact that details about their production can be equally valuable to have on record. Thus when the introduction to ISBD(A) says, "for older books all leaves are important, even blanks," the opening prepositional phrase is actually superfluous and imposes an unnecessary limitation. The particular approach to books which ISBD(A) attempts to accommodate is in fact not limited to hand-produced books: what distinguishes that approach is a focus on books as physical objects, a focus which obviously is relevant to any books of any period.
| | |