G. Thomas Tanselle

Recent discussions of research libraries in an age of electronic texts have focused on the exchange of "information," with little attention to the artifacts that provide the means for the transmission of verbal texts. Indeed, these discussions frequently imply that book stacks can gradually be abandoned as texts are incorporated into machine-readable databases. Those who hold this view see book-centered librarianship as outmoded and special-collections libraries as an insignificant element in the structure of information retrieval. If, however, one understands the uses of artifacts, one will value more highly the function of those librarians who concern themselves with books as physical objects. The electronic future may even make special-collections libraries more important than ever.

Form and content depend on one another. All artifacts can be studied as physical objects for two major classes of historical information that can influence the interpretation of any visual or verbal symbols present on them. One class relates to their production history, the techniques of their manufacture; the other focuses on their post-production history, the implications of their appearance. Scholars pursuing the first type of information seek clues to how objects were made; with books, this pursuit is called analytical bibliography, and it can provide information about typesetting, proofreading, and presswork--essential data not only to printing and publishing history but also to textual history and criticism (assessing the genealogy and accuracy of texts). Scholarly editors have increasingly recognized analytical bibliography as a foundation of their work, with the associated understanding that books, like manuscripts, offer primary evidence for textual study and regularly present variant texts (since even copies from a single edition often contain variations).

The other major approach to artifacts concentrates on their sensuous-primarily visual--characteristics. Every object can be studied for aesthetic value, whether or not its producers intended it to have only a utilitarian function. The historically oriented form of this investigation not only attempts to show how the features of artifacts reflected cultural trends but also observes the effects those features produced on persons encountering the objects. In connection with books, this kind of study examines how elements of design (typography, layout, leaf size, and binding) emerged from contemporary tastes and affected reader responses. Such research is clearly relevant to the history of reading and of the spread of ideas--that is, to the broad field called histoire du livre, the history of the impact of books on society.

Of these two approaches, the first involves details that readers do not normally notice; the second considers features that readers were meant to notice and that do affect their interpretations. The first produces evidence for reconstructing the texts that authors (or others) intended; the second looks at the texts that actually appeared and their physical settings. (A discrepancy between intended and published texts--that is, works and documents--is always to be expected, since language is intangible, subject to distortion by any tangible representation of it.) The two approaches are thus complementary. Both illustrate how physical evidence is involved in interpretation; they show why the fullest and most rewarding kind of reading is a reading of the whole object conveying a text, not just the words.¹

This argument carries critical implications for research libraries. There are no books whose physical features are unrelated to the process of understanding their texts. Recognition of this point, one should note, does not impede an enthusiastic acceptance of new technology for the dissemination of previously published texts. Scholars who understand that microfilm and xerographic copies do not fully substitute for originals have gladly used them as interim tools. The availability of electronic texts is an advance greater in degree but not different in kind: It accomplishes in a more sophisticated way the same function fulfilled by xerography, making texts widely accessible at the price of removing them from their original contexts. Scholars should welcome the day when they can call up on their terminals an enormous array of texts without the cumbersome process of interlibrary loan or the ordering of photocopies. But they should also understand what evidence they are missing and why recourse to the originals, while inconvenient, can never be irrelevant. Many discussions of the future of libraries speak of access replacing ownership, but when it is understood that access to physical evidence is an essential kind of access, and that books must therefore be preserved in as many copies as possible, the questions of ownership and care remain significant.

The function of libraries as centers for information exchange was conceived long before the advent of computers and will undoubtedly loom larger in the future. What should logically follow is enlargement of every so-called special-collections department to encompass the library's entire book stack. (The resulting collection would be most appropriately named a "Department of Original Materials," for it would care for originals of all kinds, including computer disks and other forms yet to come.) This development would clarify the conceptual framework on which librarianship rests. In the past, books were treated differently in two parts of the library: a book in the general stacks was subject to rebinding or even replacement, since its physical form was not deemed important, but if it got transferred to the special- collections department, it would be carefully preserved. No satisfactory rationale has ever explained why books were treated sometimes as artifacts deserving preservation and sometimes as expendable vehicles for the transmission of texts. If all books were entrusted to experienced curators of artifacts, the incoherence would vanish. The dissemination of abstracted information would then be recognized as a distinct activity that has no bearing on the rationale for preserving artifacts.

The misconception that texts are easily extractable from books has contributed to policy decisions--all the more shocking for being deliberate--that will mark the present as an age of destruction on a scale beyond even that of the book burnings of the past. Since 1986 the Commission on Preservation and Access has been coordinating the microfilming of more than half a million volumes--a blessing for scholars, but hardly unalloyed, because the copies selected generally do not survive the process. The usual procedure is to chop off the spines ("guillotining") to leave a stack of loose leaves, since the books are usually too brittle to open flat without cracking. After microfilming, the mutilated fragments are thrown away. The term "Preservation" in the commission's title, and as practiced at many libraries, applies only to the extracted texts; for the artifacts, the project has been an agent of destruction. Though the commission's widely publicized film "Slow Fires" lamented the self-destruction of books printed on acidic paper, many books (including thousands in usable condition) have been doomed to a much swifter conflagration. Not only are the microfilmed copies discarded, but the existence of the microfilms causes some librarians to decide that other copies also can be disposed of.

The latest forms of book burning are not only wasteful but preventable. Both technological progress and wise policy decisions can save books. (Widespread use of the Gottschalk prismatic camera, for example, which can photograph the facing pages of a book opened at only a 60-degree angle, would drastically reduce the damage from microfilming.) Some segments of the library and scholarly community are beginning to recognize that microfilming and electronic scanning do not fully extract the useful information from books. It is encouraging, for instance, that the Modern Language Association in 1995 gave wide dissemination to a "Statement on the Significance of Primary Records"--with primary records being defined as the objects carrying texts, and not simply the texts themselves. If an organization as influential as the Commission on Preservation and Access were similarly to recommend that libraries keep microfilmed books, that action would be an important step in spreading the message that saving our verbal heritage means saving both texts and objects.

In 1847 Augustus De Morgan said, "The most worthless book of a bygone day is a record worthy of preservation." The study of our intellectual history since Gutenberg depends on our incorporating this point into our plans for the electronic age.

---

1. I have elsewhere dealt in more detail with the implications of the intangibility of language as a medium, and therefore with the distinction between works and documents and with the indispensability of physical evidence to reading: See, for example, A Rationale of Textual Criticism (1989). The application to library policy is particularly evident in "Reproductions and Scholarship," Studies in Bibliography 42 (1989), 25-54; Libraries, Museums, and Reading (1991), which also appears (without the footnotes) in Raritan 12.1 (Summer 1992), 63-82; "Books, Canons, and the Nature of Dispute," Common Knowledge 1.1 (Spring 1992), 78-91; and "The Future of Primary Records," Biblion 5.1 (Fall 1996), 4-32.

---

Related links:

Grolier Club of New York, a book collecting society

American Printing History Association

Rare Book and Manuscript Section, Association of College and Research Libraries

College Libraries Committee, Commission on Preservation and Access

Abbey Publications (nonprofit preservation advocacy group)

European Commission on Preservation and Access

David S. Bennahum, "The Future of Libraries," Meme

David Hudson, "Persistence of the Written Word," Rewired

Bibliographical Society, London

Bibliographical Society of America

Bibliographical Society of the University of Virginia

Library History Round Table

Society for the History of Authorship, Reading and Publishing

Resources for the history of books and printing, Daniel Traister, U. of Pennsylvania Special Collections

David L. Gants, "Toward a Rationale of Electronic Textual Criticism," ALLC-ACH conference, Paris, 1994

Encyclopedia of New Testament Textual Criticism, Robert Waltz (examples of textual scholarship)

---

G. THOMAS TANSELLE, Ph.D., is adjunct professor of English and comparative literature at Columbia and vice president of the John Simon Guggenheim Memorial Foundation. Earlier versions of this article appeared in Harvard Library Bulletin 4.1 (1993): 38-40 (based on a talk at the 50th anniversary symposium of the Houghton Library) and in Common Knowledge 2.3 (1993): 172-177.