University of Virginia Library

IV

What can be regarded as a contribution, in any field, is likely to be a different angle of vision, an unaccustomed way of perceiving something. Rarely is an entirely new idea formulated, but old ideas can usefully be given new emphases or be placed in new contexts. The recent dominance of textual instability as a subject for writers on textual criticism is valuable because it directs attention to an aspect of textuality that has not been adequately explored in the past. It has of course been recognized by everyone who has ever thought about texts: the changeability of texts over time is the basic fact that gave rise to such a field as textual criticism and has been the grounding for all editorial activity. But the fact of instability was taken as the starting point, as the essential condition within which one had to work, not as a particular focus of interest in its own right. There has always been an implicit understanding that individual versions are of interest, and this understanding has often taken explicit form in the construction of apparatuses and the publication of facsimiles; but the primary concern of textual scholars before the last few decades was authorial intention, and the primary editorial activity was presenting texts (however arrived at) that reflected authorial intention. Thus the recent interest in the process of textual metamorphosis, along with the emphasis on the value of every one of the myriad forms that texts of works take (however mixed — in many of those forms — are the intentions of authors with the intentions of others), is a most welcome development, throwing a spotlight on an area where it had not previously been directed.

Any such development, however, is apt to be accompanied by two kinds of problems. One is that enthusiastic advocacy for the new may involve an unfounded denigration, or an inaccurate characterization, of what went before. The recent writings on textual instability, the social construction of texts, and related ideas have amply demonstrated this point. Given the fact that the new position is characterized by an openness to all the forms of texts produced by the historical process, it is

surprising that there has not been more of an inclination to take intentionalist editing in stride as one of the social forces that produce alterations in texts, alterations that are validated by their origin in a social process. But the rhetorical need to discredit the staus quo persists, even though the merits of any new position are logically independent of whether or not an earlier position is flawed. In the case of textual theory at present, the older and the newer views simply concentrate on different parts of a complex process, and both are important in an attempt to understand the whole. If more people are talking about the newer one at present, fashion is not the only explanation, for a neglected topic by definition needs more attention. In the long run such imbalances continually shift; and in the unending process of sifting and winnowing, distortions and oversimplifications come to be recognized.

The other kind of problem often associated with new theoretical insights is that they may leave people at a loss to know how to proceed. Although it is no criticism of a theory to say that its practical implications are unclear, people are nevertheless bothered by theories that seem to lead to inactivity or to an impasse. Many of the essays on textual theory in recent years have lurking within them the implication that editing is an impossible, or unnecessary, activity. If all the variant texts that have existed are worthy of attention as the products of social forces and as the inspirations for readers' responses (and this point is clearly valid), then it may seem that there is nothing for editors to do. This feeling of helplessness is analogous to the aporia of deconstructionist readers, faced with words and texts of indeterminate meaning; and this similarity is not surprising, since recent textual theory springs from the same intellectual milieu that produced recent literary theory. Much of that theory is convincing in its own terms, and it frequently does help us to understand a little better what complex events are subsumed in the concepts of "text" and "reading." But whether or not acts of communication can theoretically be consummated, most of the time we behave as if we are not trapped in private prisons of language. Writing to express thoughts, as well as reading to receive them, will continue to be practiced, simply because they are activities that are congenial to the human mind. Reveling in or despairing over the impossibility of succeeding in these activities is also present in some minds, but it can coexist with the practicing of them. Similarly, scholarly editing will continue to exist because it, too, is a natural activity of mind. It is one of the forms of response to texts: of those readers who make public responses to what they have read, some write essays, some give lectures, and others produce editions. The existence, on some occasions, of a gap between what we do and how we theorize about it does not in any way suggest

that theory is useless. We could not force ourselves to stop theorizing in any case; but theory does lead to sensitive practice, for the process of thinking issues through on a theoretical level endows our practice with greater self-awareness and thoughtfulness.

Editing is the practical side of textual scholarship, and the recent emphasis in textual theory on the importance of versions calls increased attention to a long-recognized practical problem: how best to present texts and textual evidence, particularly in codex form. The concern with versions has international force behind it; in addition to the many discussions in English (from scholars in most of the English-speaking countries), German textual scholars have long concentrated on versions, and the French now have a flourishing school of critique génétique, emphasizing the pre-publication evolution of texts.[85] Much of this work has been practical, in that it has applied the theory of the significance of versions (set forth in a limited number of general treatments) to the textual history of specific writings; but the results have taken the form of essays as often as (probably more often than) of editions. The presentation in codex form of complete texts of versions has always posed difficult problems, both because of the space they take up (and the consequent unwieldiness of the resulting volumes) and because of the inconvenient process they entail for the detection and comparison of variants. For these reasons, nearly every editor in the past has presented only a single text (whether critical or diplomatic) in complete form and has recorded the variant readings from (and other information concerning) all the extant documents in abbreviated form in an apparatus — either incorporated into the single linear text (and identified with symbols) or appended to that text (at the foot of each page or at the end of the volume). Because the codex form forced this kind of compromise on editors, they have discussed endlessly the questions of selection, arrangement,

emphasis, and form that the construction of an apparatus poses. Such questions may appear to be entirely practical, but they continually intrude themselves into theoretical discussions, since the compromises they deal with would not be considered compromises if they did not depart from what editors regard as theoretically desirable. The early 1990s have been no exception; a number of the theoretical articles from this period touch on matters of apparatus, and there have continued to be other articles dealing exclusively with the subject.[86]

Jo Ann Boydston's "In Praise of Apparatus" is particularly eloquent in its account of how the reading of an apparatus is "a stimulating and highly productive intellectual adventure" (p. 10). Her attitude echoes that of a great many other editors who have produced editions containing single eclectic texts with accompanying apparatus: their presentation, in each edition, of only one text in full does not mean that they fail to see the importance of other versions. Some of the recent critics of eclectic editing have assumed that conventional editions of the past reflect a theoretical belief in the primacy of a single "ideal" text for every work (or for most works). But a more realistic explanation for the kind of presentation given to texts and variants in past editions is the constraints of the codex form. Given the effort that most editors have expended on apparatus and the stress they have habitually laid on the importance of lists of variants for understanding the textual history of a work, it is unreasonable to think that they have not understood the significance of versions as part of the process of experiencing a literary

work. In most cases, they would have been delighted by the opportunity to offer readers multiple texts in full, but the physical limitations and economic realities of codex publication precluded doing so.

It is true, of course, that the single texts chosen for full presentation by editors (having been forced to make such a choice) do reflect personal judgments as to what will be most useful to most readers. The judgments inevitably involve some mixture of practical and theoretical considerations, but they cannot be assumed to arise from any lack of openness, on theoretical grounds, to multiple texts. At the same time, they may very well signal a belief, on theoretical grounds, in the value of critically constructed (that is, eclectic) texts. Those recent writers who have attributed to intentionalist editors the notion that there is a single ideal text for each work have blurred two separate issues — singleness and ideality. The use of the word "ideal" is itself part of the problem. Sometimes it seems to mean something like "best" or "perfect"; at other times it means "not real" in a physical sense — that is, not extant in a document. The first sense (not often used by critical editors) implies singleness; the second is fully compatible with textual instability. A conviction of the importance of attempting to reconstruct nonextant texts does not entail any concept of the sufficiency or finality of a single text; indeed, critical editors are well aware not only of the lack of finality attaching to any product of critical judgment but also of the fact that any number of past moments (not just the one judged to represent final authorial intention) can be selected for critical reconstruction. Critical editors of the past clearly believed in the importance of critical texts. But their presentation of single texts did not necessarily mean (and in fact was not likely to have meant) that they believed only one text was valid or desirable; it only meant that the option of presenting more texts was not open to them.

Technological developments have now made that option feasible, and editors are rightly excited by the possibilities that electronic presentation offers. The capabilities of word-processing and hypertext programs for textual study have already been the subject of a considerable literature, ranging from (to name two works from the early 1990s) the general theoretical treatment of George P. Landow's Hypertext: The Convergence of Contemporary Critical Theory and Technology (1992) to the detailed practical overview of Charles B. Faulhaber's "Textual Criticism in the 21st Century."[87] In a hypertext edition, one can have

as many full texts as one wishes, regardless of the length of the work involved, and one can easily switch from a given word in one text to the variant at that point in another text, having the whole context available in each case. The imaging capabilities of this technology, along with the availability of space for multiple texts, means that one can now have, for every variant text of a work, not only a newly keyboarded rendition (searchable for any word) but also a facsimile that shows the original typography or handwriting, lineation, and layout. Indeed, the first requirement of an electronic edition should be that it contain both forms of every text.[88] The result accomplishes what we always wished that printed apparatuses could do, for now we are able immediately to see variants in their verbal and visual contexts.

This is not the only requirement that we should have for electronic editions, however. The elimination of space constraints takes care of the problem of the "single" text that was associated with the codex form, but it has nothing to do with the kinds of multiple texts that get presented. Writers on hypertext editions frequently think of the goal as a kind of "archive" limited to documentary texts. The dissemination in this way of manuscript and printed texts surviving from the past is extremely valuable, obviously, just as the publication of documentary editions has always been valuable. But hypertext editions offer great advantages for the presentation of critical texts as well, and one can argue that the potential of the electronic form is not being very fully exploited unless editors' critical reconstructions are included along with documentary texts. Critical texts may be out of favor among many theorists at present, but there will always be scholars and other readers who understand the need for reconstructions of additional historical moments besides those represented in surviving documents. And for such persons, electronic editions offer the possibility of multiple (and linked) critical texts, attempting to show different intentions (those of a publisher, say, as well as those of an author) as they existed at different times.

One of the ways that traditional printed critical editions can be faulted is that, whereas the apparatus was used to record the variants in documentary texts, there was no attempt to use it to show what editorial emendations should be made to produce other critical texts besides the one presented as a full reading text. (This criticism has of course not

been made by recent theorists, who are likely to believe that even one critical text is too many.) Although critical editors have recognized textual instability as reflected in documents and have understood that any critical texts they offered were the attempted reconstructions of specific past moments and intentions, they still, strangely enough, did not see that critical texts representing other moments and intentions could be presented in an apparatus as efficiently (that is, with as little cost in number of pages) as documentary texts could be. In any case, the availability of electronic space for multiple texts, plus the fact that electronic linking makes bulky quantities of material logistically manageable,[89] means that in the future there is every reason to expect electronic editions to include a wide array of critical texts as well as a generous selection of documentary texts, all presented in full.[90]

Even an electronic edition of the kind envisaged here, with images of primary records, newly keyboarded texts of them, and a range of critical texts, would not satisfy some recent theorists, for two reasons: the reader is still removed from the physical objects that originally conveyed the documentary texts, and the reader is dependent upon the subjective reconstructions of a single editor (or series of editors, no matter how numerous). Neither of these points can be denied, but whether they are grounds for complaint is an interesting question because it involves the most fundamental characteristics of editing as a practical undertaking. The first fact that one must confront in thinking about it is that not all readers are interested in history. There is no reason why they should be, if they do not wish to be or if (whether or not they have considered the matter) they are not temperamentally inclined to give any thought to the past. Such readers can respond to the linguistic and design features of the document (whether paper or electronic) that they have in front of them, and how those features compare with the ones presented to past readers of what might be called the same work is quite properly of no concern to them. Scholarly editing, when it is conceived of as an activity oriented toward historical recovery, is irrelevant to their purposes.

So, in turn, are most of the theoretical debates, for most of them presuppose

a concern with the past. Yet there is a fatal wavering in some of them between a belief that all previous readers' responses to the texts they encountered are valid, because they occurred, and a feeling that readers today and in the future should not be presented with texts that inflict upon them the views of present-day historical or critical editors. Yet contemporary scholarly editions will in the future be regarded as a product of their time and as an influence on the responses of the readers who used them, just as editions from the past are now so regarded. A critical edition from the past is a documentary edition to us; and a critical edition produced now will be a documentary edition to future readers. Whether arguments can be made for or against particular editions is a separate matter from the point that they were thought worth producing by certain people at certain times. The view that all past editions are acceptable because they occurred does not fit very well with the position that some approaches are objectionable because they constitute interpositions between history and readers.

Either we are talking about historical approaches to literature, or we are not. If not, then the arguments of theorists that scholarly editing (or any kind of editing, for that matter) creates a barrier to historical discovery and understanding is irrelevant. But if we do wish to be concerned with literature in its historical setting, there is still good reason to believe that recent theoretical arguments critical of scholarly editing are misstated. They do not always say explicitly that there is no future for editing, but the idea is implicit in them: for if all the physical details of documentary artifacts are essential to the historical experience, then facsimile editions are not adequate; and if all documentary wording must be directly encountered by readers, without the intervention of editors' subjective emendations, then editions with critical texts are inappropriate. Each of these points can, however, be stated in a more understanding way; and doing so leads to the conclusion that both kinds of editions are not only inevitable but are desirable and necessary.

Of the importance of artifactual details in the historical understanding of texts inscribed or printed on physical objects, there can be no doubt. Whether or not authors or readers in any given instance regard physical characteristics as part of the text, those characteristics do reveal information about the production of the objects, the social milieu of the text, and the bases for readers' responses. Every visual and tactile detail is relevant, and no attempted reproduction can possibly carry the same historical suggestiveness as the object that survives from the time in the past that is the subject of one's interest.[91] But editors have always

known this. All editors have recognized that their own experience was richer, as a result of working with primary records, than that of the readers of their editions, who were generally limited to what was presented in those editions. To criticize editors of facsimiles (printed or electronic) for providing derivative forms of artifacts, or to object to the production of facsimiles for distorting history, misses the mark, since editors have always understood that what they offered in facsimiles did not entirely substitute for the originals. They were simply responding to the inescapable fact that most readers do not have access to the originals, nor do they have any realistic expectation of ever seeing them. Facsimiles must, by default, serve as partial substitutes for originals. And when they are accompanied by various aids, such as transcriptions of handwritten texts, physical analyses of the original artifacts, and records of press variants, they provide information not available in the originals. Landmark printed facsimiles like Charlton Hinman's of the Shakespeare First Folio (1968), Michael Warren's of King Lear (1989), and David Vander Meulen's of Pope's 1728 Dunciad (1991) provide extraordinary assistance of this kind to the reader;[92] and electronic facsimiles, like those being produced at the humanities computing centers of the University of Virginia and the British Library, enable readers to magnify at will particular areas of the text pages for detailed examination. None of these can — nor are they intended to — substitute fully for the originals; but even those persons with access to the originals will find them useful. And for other people, facsimiles are a necessary — and indeed a productive — compromise.

Another category of dissatisfaction that recent theory has found with editorial activity relates to the production of critical texts. The principal objections are that a critical edition conceals the fact of textual instability by presenting a single "ideal" text and that such a text offers an unhistorical conflation reflecting an editor's subjective judgments about authorial intention. In order to focus on the real issue here, we can

immediately eliminate two superficial elements in these objections: the points about single texts and authorial intention. It should be understood by now, as I suggested above, that the offering of single texts is a product not of editors' theories but of the limitations of the codex form; and it should be equally evident that there is nothing tying the idea of critical texts to the concept of authorial intention, simply because such intention was focused on in the days of single-text editions. (Editorial judgment can obviously be applied to the reconstruction of texts intended by anyone.)[93] We are thus left with the basic issue: what is the value of constructing texts that differ from those surviving in physical objects? In most fields of historical inquiry, it is taken for granted that artifacts (with all their contingencies, symbolized by their random survival) must be supplemented by informed attempts at reconstructing past events. But artifacts carrying verbal texts are frequently thought about in a different way, presumably because the presence of such texts on them suggests that they can speak more directly than other artifacts. Yet texts of documents do not necessarily represent in every respect the intentions of any of the persons involved in their production, and those intentions are historical events, even if they never took physical form. Mental events are ultimately the essence of history, as John Searle understands in The Construction of Social Reality (1995), where his argument recognizes that "social reality" — the social structure of conventions and interchanges — depends for its existence on mental and physical reality. The acts of constructing texts and works are social events, as many textual theorists have been telling us; but we are not going as far as we can toward understanding those events if we limit ourselves to surviving objects and exclude from our deliberations the mental events that are a fundamental part of the textual process.

Many writers have commented, as Shelley did, on the difference between the idea for a work and the executed forms of the work. Virginia Woolf put it this way:

The "gulf" that "words can't cross" marks the beginning point for critical editors' activity. No text exists on the "far side"; but as soon as thoughts are "pulled through" and made into words, there is a verbal event that editors can attempt to reconstruct.[95] Woolf speaks of works being "written"; but of course the words were in her mind before they were written, and she may not always have written down the mental text that she intended to write. An editor who tries to recover that mental text is not venturing to the "far side," for the gulf does not divide the intangible from the tangible but rather separates unarticulated thoughts and feelings from groups of words arranged in a particular order. And those groups of words are mental facts first, before they develop into any other kind of fact. These points can be applied to every stage in the history of a work: authors have mental texts of successive versions as well as of initial versions, and compositors and publishers have mental texts before they produce texts in tangible form. The effort to reconstruct such mental texts is like the efforts made in all other historical fields to build up a fuller sense of the past than is provided by artifacts alone. These activities are the natural consequence of recognizing that the artifactual report is always incomplete, both because some of it has been lost and because artifacts do not in any case record everything that happened.

Accepting the necessity of critical extrapolation from artifacts still leaves open the question of whether the results should be published. That in effect is the question being answered in the negative by those recent theorists who say that readers should decide for themselves how (or whether) they wish to alter documentary texts. Readers will indeed finally make this determination, but they are not all equally qualified to engage in historical reconstruction, which involves knowledge as well as imagination, and they may wish to have the results of specialists' critical thinking. No one would be likely to claim that historians and literary scholars should not publish essays on the grounds that readers should not be told what to think. Similarly, one can scarcely claim that critical editions (which, like historical essays, are the products of systematic efforts to interpret the past) are objectionable because they inflict

particular editors' views on readers. Readers will decide how much or how little they wish to rely on the historical activities of readers who preceded them; but those activities — which include the publishing of essays and editions — are inevitable by-products of the ongoing process by which the human mind struggles to make sense of the hints offered by artifacts. Recent studies of the unstable nature of texts are bound to add to our cumulative sophistication in textual matters, but they cannot change the concept of what editorial work is. Technology will improve the ways in which the results of editorial thinking are presented; but the ingredients that make up such thinking are set by the unalterable disparity between the limitations of artifacts and the insatiability of our interest in the past.