University of Virginia Library

I

The distinction between catalogues and bibliographies is an elementary one; yet it is not always kept clearly in mind by those engaged in bibliographical activity. Of course, one person can call his work a "catalogue" and mean something entirely different from what someone else means by "catalogue." Neither person is necessarily confused in his thinking; each is merely using a different definition of "catalogue." Such differences are always possible where matters of definition are involved. But I am not speaking here of the definitions of particular terms; rather, I am concerned with two different concepts, regardless of what they are labeled. One kind of record of books[6] —which it is convenient to call a "catalogue"—is concerned with the particular copies of books that happen to be in a given collection (a private library, an institutional library or special collection within it, a dealer's stock, and the like) or that constitute a specifically defined assemblage (items brought together for an exhibition or an auction, for instance).[7] Another kind of record of books—which it is convenient to call a "bibliography"—is concerned with books which are related in some way, but not with specific copies of those books. In other words, an entry in a catalogue refers to a particular copy of a book; an entry in a bibliography refers to any copy of that book. This distinction can best be illustrated by noting that the goal of descriptive bibliography is the description of an "ideal copy" of each book, a term carefully defined by Fredson Bowers (Principles, pp. 113-123) to refer to the complete state that the printer or publisher "considered to represent the final and most perfect state of the book." An ideal copy is not necessarily free of textual errors, but it is free of those physical deficiencies which would prevent its representing a standard form of the book as published. It is therefore an abstraction, for conceivably all existing copies of a book might be defective in one way or another, but a description of an ideal copy could still be constructed by combining details observable in the defective copies. A description of an ideal copy sets a standard against which individual copies can be measured;

a catalogue entry describes or records an individual copy with all its peculiarities.[8]

Simple as this distinction essentially is, it raises some complicated questions which have been much discussed. Sometimes, there is a tendency to think that the amount of descriptive detail is a part of the distinction.[9] But it should be clear that both catalogues and bibliographies can run the gamut from the sparse to the elaborate. Some catalogues, like Allan Stevenson's volume of the Hunt catalogue (1961), include fuller physical descriptions than are found in a great many descriptive bibliographies. And some catalogues, like William A. Jackson's Pforzheimer catalogue (1940), involve comparisons between copies in the collection and other copies, so that the precise nature of the copies in the collection can be more clearly specified; in this way a catalogue can actually record the characteristics of an ideal copy, but so long as the object of description is one specific copy the work remains a catalogue. Naturally, as the quantity of detail declines, the differences between the entries for a given book in a catalogue and a bibliography are likely to become slighter—or nonexistent. If an entry consists of nothing more than a simple listing of author, title, and date, such an entry for a book in a catalogue could not be distinguished from the same style of entry for that book in a bibliography

(or "checklist"). But the indistinguishability of the two entries would not alter the fact that the purposes of the two lists, and thus the significance of the two entries, were different. In the first case, the entry refers to one specific copy of the book; in the second, it refers to any copy. A catalogue, then, is not merely a less detailed bibliography; the extent of detail is irrelevant to its classification as a catalogue. Furthermore, the account of an ideal copy in a descriptive bibliography, however much or little detail is given, requires a great deal more research (the examining of many copies) than is normally expended on a catalogue entry; but the catalogue entry can—within its limits (the description of a specific copy)—be equally authoritative. Of course, in describing a single copy without the benefit of a published bibliography or the examination of other copies, one cannot always know exactly which features may be of special significance; for this reason the entries in catalogues are in practice rarely as informative, even for the specific copy, as entries in bibliographies. Nevertheless, it is not the quantity of detail or the extent of research which distinguishes the two kinds of entry but solely the nature of the copy which each aims at recording.[10]

Another question which complicates one's thinking about catalogues and bibliographies is the determination of what is meant by "book"—of what, in fact, is being recorded. The word "book" is sometimes used to refer to a physical object (or a group of physical objects, such as all those comprising one edition); at other times it refers to the work (the verbal construction) embodied in the pages of the physical book. Clearly this distinction is basic, and the nature of a bibliographical record is determined by whether that record is principally concerned with books or with works. It might at first be supposed that a catalogue inevitably deals with books, not works, since it lists specific copies. Unquestionably a catalogue must involve this element; but many catalogues of books are used as guides to the works on a given subject, and their compilers sometimes have this function in mind and provide annotation which emphasizes it. Seymour Lubetzky, in Principles of Cataloging (1969), has offered a careful analysis of the book-vs.-work distinction in the context of library cataloguing (pp. 1-17).[11] First he sums up the perennial debate over whether a library catalogue should be a "finding list" or a "reference tool," whether it should merely locate certain books for its users or provide a guide to the works incorporated in those books.[12] Later he

concisely states these two questions which are involved in cataloguing the "records of man's thought": It is obvious that a catalogue will be useful to more people if it performs both functions, and users of institutional libraries now regularly expect to find such catalogues of the holdings of those libraries. As a result, library cataloguing is generally divided into two activities, descriptive cataloguing (dealing with author or title entry and with the physical characteristics of the books) and subject cataloguing or classification (dealing with the content of the works). Lubetzky's principal point is that in neither activity has the distinction between book and work been clearly focused on in the past; therefore, he says, it has not been sufficiently understood that physical description is basic to all cataloguing, since works exist in a library only as embodied in individual books. If the record of particular copies in catalogues can thus serve to supply information about works, there is no question that bibliographies or checklists (recording ideal copies) can do so too. Indeed, the commonest form of checklist is that in which the primary emphasis is on the content of the works named. The border line separating details relating to the book from those relating to the work is not always sharp (some details serve both purposes), but one cannot think clearly about catalogues and bibliographies without keeping this division in mind and recognizing the extent of mixture of the two approaches in any given listing. Both catalogues and bibliographies can vary in the degree to which they lean toward providing information either on books or on works; but where they stand in that respect does not affect the crucial distinction between catalogues and bibliographies, based on the difference between specific and ideal copies.

The division between books and works is analogous to Lloyd Hibberd's separation of the field into "physical bibliography" and

"reference bibliography."[14] His useful essay recognizes that the amount of detail which a bibliographical record contains is less significant for classifying it than whether it is concerned with the physical form or with the content of the items recorded. His survey of the confusing array of terms in use and his suggestions for terminology which more accurately shows the relationships among bibliographical activities have been generally well received; but Rolf Du Rietz in a thoughtful essay[15] recently complained that Hibberd's "well-meant" proposal "has unfortunately further contributed to separate the two supposedly widely different 'kinds' of bibliography from each other instead of bringing them closer together" (p. 22). According to Du Rietz, the belief that reference bibliography is concerned only with the content of books leads to such lax standards in the inevitable allusion to physical detail in reference bibliographies that physical bibliographers are bound to have a low opinion of it. His principal point is that all bibliographical lists are to some extent physical (e.g., pp. 15-16, 22), because lists referring only to works and not to the books embodying the works would simply have entries like "Hitler's Mein Kampf" and "do not deserve the name of bibliographies, since they do not list books at all" (p. 24); therefore, he says, reference bibliographers must have training in physical bibliography, so that their lists will offer responsible treatment of the physical details which they cannot avoid. This warning is salutary and, in noting the physical element in book lists, calls attention to a fact not sufficiently recognized. But Du Rietz goes too far, it seems to me, in the direction of blurring a useful distinction when he is led to conclude that there is "no such thing as 'a' physical bibliography, or 'a' reference bibliography" (p. 24). It is true that all bibliographies in one sense involve a mixture of both physical and reference elements, but that does not prevent the principal emphasis or concern of a given listing from being on one or the other. Reference lists, for instance, frequently cite the city, publisher, and date of the first printing of a work without implying that the reader is necessarily being directed to the first printing in preference to a later printing or edition. The facts of publication are offered as historical annotation, not as physical details, even though these same details would of course be a part of a physical bibliography as well. Such listings are similar to references sometimes found in the body of a literary discussion: "Moby-Dick (New York: Harper, 1851)" may not mean anything more than "Moby-Dick," except that more historical details are provided; indeed, page references might be given to a later and more accessible edition—but again without any implication that the reader should not turn to still another edition, more convenient for him, to locate the cited passages. Even a catalogue can emphasize reference bibliography, if the interest is more in what works are represented by the books in a collection than in what particular books are there. To be sure, a catalogue cannot avoid physical implications, since it is based on a specific gathering of books; but the purpose of a catalogue can be, as its annotation would make clear, to show what works (regardless of edition) are available in that collection.[16]

Reference bibliography can simply be regarded as primarily concerned with works, physical bibliography as primarily concerned with books. The approach in each case will determine what details are reported and how they are treated; but it should not be surprising that some of the same details will turn up in both kinds of bibliographies, since the two approaches are complementary. I take it that Hibberd is making the same point when he says, "And though divergent in purpose and scope, the two divisions start from the common basis of systematic compilation and end in reciprocal fructification" (p. 133). Du Rietz, too, wishes to show the intimate relationship between the two, but in stressing the physical elements in reference bibliography he makes reference bibliography in effect a preliminary step leading toward, or a less thorough form of, physical bibliography.[17] He is unwilling to let the word "bibliography" move beyond its

etymology and encompass a concern for works as well as for books; the result is that for him reference and physical bibliography together form one camp and "information science" the other. Although he is reluctant to link "the immensely powerful information science" with "the rather humble and unsophisticated kind of activity of reference bibliography," he admits that a bibliographical list could conceivably be regarded as involving both physical bibliography and information science (p. 26). The issue is thus a question of terminology, for the dichotomy in any case is between books and works. It makes little difference whether "reference bibliography" is salvaged as a term, so long as we know when we are thinking about works rather than books.

The relationships I am describing can perhaps be clarified by a diagram:

What this diagram attempts to suggest is, first, that there are two basic kinds of finished product resulting from bibliographical activity: the catalogue, dealing with specific copies, and the bibliography, concerned with ideal copies. The catalogue may refer to copies outside a given collection or to accounts of ideal copies, but its primary function is to refer to particular copies; the bibliography may cite the peculiarities of individual copies or offer a census of surviving copies, but its primary function is to refer to standard copies, free from the deficiencies which may happen to occur in any one copy. Both catalogues and bibliographies may take the form of essays rather than lists, but their essential function remains unchanged. Second, the arrangement of the diagram suggests that both catalogues and bibliographies partake of both reference bibliography, in which the subject matter is the works embodied in books, and physical bibliography, in which the subject matter is the books as physical objects. However, their interest in these two approaches is rarely equal, and they may move in one direction or the other, stressing either reference bibliography or physical bibliography. Finally, both catalogues and bibliographies, whether emphasizing reference or physical bibliography, can present a great deal of detail or very little detail. If the emphasis is reference, that detail will take the form of annotation suggestive of the nature or value (or both) of the works included; if the emphasis is physical, that detail will take the form of description of the physical makeup of the books included.[18] As the detail in a reference or a physical bibliography becomes less, the entries in the two come to resemble each other more and more, and for that reason I have employed the same word, "enumerative," to refer to lack of detail in either case.[19] But the fact that the entries are stripped to the information basic to both approaches does not mean that the functions of reference and physical bibliography have become blurred; the entries may even be identical, but their significance is different depending on the context in which they occur. And the context is determined by two factors: whether the emphasis is on reference or physical bibliography and whether the product is a catalogue or a bibliography. For example, the city and year of publication reported in a catalogue entry are to some extent physical details because a particular copy is being referred to; yet the general approach, as revealed in a preface or in other notes attached to entries, may be to regard the listing as primarily useful for its record of works, not books, and in this case the city and year are not essentially physical details. The cataloguer has a right to take this approach if he wishes to; the trouble comes only if his practice in recording what are partly physical details is positively misleading to anyone familiar with the way the same details would be handled in a catalogue stressing physical bibliography. The problem arising from the fact that any catalogue or bibliography can move toward the physical or toward the reference end of the scale is not simply a matter of how many details are included but rather of how the included details are treated.[20]

There is no question that the differing approaches of reference and physical bibliography have frequently produced incompatible results in the past. Du Rietz has said that bibliography and information science are "notoriously at loggerheads in all matters terminological" and that "the libraries will apparently remain an unavoidable battle-field for the combatants until some modus vivendi may be achieved" (pp. 26-27). The libraries are at the center of this debate precisely because they attempt, through catalogues and indexes in whatever form, to offer a guide both to the books in their holdings and to the works contained in those books. Of course, any cataloguer or bibliographer confronts this issue to some degree in his own work, but institutional libraries, because they process large numbers of books, naturally become the most prominent illustration of the problem. The real point of contact between the two approaches (or the

"battle-field" where one can see the conflict in progress) is in the pages or cards of bibliographies and catalogues. Certainly the concerns of information retrieval can lead one far from the physical book; but since information must be recorded in some concrete form and since different physical embodiments of the same work may contain variations in text which affect the "information" conveyed, the two approaches are ultimately inseparable. Physical and reference bibliography—or whatever we choose to call them—are tied together (as my diagram tries to show) in every catalogue or bibliography that is produced. Since they move in different directions, however, a catalogue or bibliography which is primarily concerned with reference bibliography may have only a small area which overlaps the concerns of the descriptive bibliographer, and vice versa—but they inevitably do overlap. It is in that overlapping area where the methods of the two approaches must be compatible; if they are not, catalogues and bibliographies will be less efficient tools, and scholarship will suffer. In preparing, using, and evaluating catalogues and bibliographies, one must keep firmly in mind the various relationships among the three sets of paired concepts discussed here: works vs. books, reference bibliography vs. physical bibliography, enumeration vs. detail. One will then realize that it is pointless to criticize a catalogue for being insufficiently descriptive of physical details, if it has set out to perform a different service; but one can legitimately complain if the physical details included are presented ambiguously or misleadingly or in a manner which is in actual conflict with the way those details would be presented in a catalogue stressing physical bibliography.[21] Reference bibliography and physical bibliography are complementary, and those who are seriously interested in contributing to either field must approach their individual task in a spirit of cooperation with those who are working in another branch of what is finally a single undertaking.