How Reproductive is a Scholarly Edition?

Mats Dahlström

Lecturer and PhD student, Dept of Library and Information Science,University College of Borås and Gothenburg University.

E-mail: mad@adm.hb.se ; web site: http://www.adm.hb.se/personal/mad/index.htm

This article was first printed in:
Literary & Linguistic Computing. Vol 19 (2004) : 1, pp. 17-33.

The nature of a scholarly edition, as of any bibliographical tool, is determined by the historical, medial, social and rhetorical dimensions of the genre. This situatedness puts constrains of the force of scholarly editions: what they can and what they can not do. Claims have been made for the potent reproductive force of scholarly editions, as well as for the making of massive digital facsimile and transcription archives that can be used as platforms for producing new critical editions. This article questions the legitimacy of such assumptions when combined with idealist notions of documents, texts and editions. That the nature of editions is rhetorical rather than neutral, social rather than individualistic, and one of complex translation rather than simple transmission, for instance, suggests that the versatility and reproductivity of the edited material itself will be limited by significant factors. Recognizing this makes us better equipped at subjecting digital archives and editions, and the claims some of their surrounding discourses make, to critical inquiry.

Scholarly editions based on textual criticism have historically been developed in intimate relationship with particular script and print-based technologies and distribution logistics. In consequence, editorial theories and strategies are intertwined in their scope, rhetorics, and strategies with particular media materialities and epistemologies. This relationship was certainly there in the temporarily stabilized universe of print media, but was rarely discussed. It is now becoming so to an increasing degree. As publishing and distribution are considerably altered by new media ¹ and technologies for document management, these relations are affected, both between people and between media.

The making of scholarly editions - or SEs for short, as I will refer to them for the rest of the article - and archives using new media seems to open up new kinds of communication between academic and professional communities formerly more or less isolated from each other, e.g. between programmers and software designers on the one hand and textual critics and bibliographers on the other, creating grounds for new kinds of negotiation of competence and power. Editing as process and the resulting edition as document make use of many different technologies and media. These also stand in delicate relations to each other due to particular historical ecologies of media. The organisation and architecture of SEs as well as the task division between different media change as the ecology changes. Looking at, for instance, current Scandinavian national editing projects that publish both on the web, on discs, as e-books and in print, such as the Ibsen or Almqvist projects,² one sees the forming of a new division of labour between various display and distribution solutions, a changed balance between the variants of edition types. The web edition turns into a large resource archive and editorial laboratory, and even more often into a more or less temporary interface to a changing, dynamic digital archive. This affects the scope and function of the editorial material being printed. The printed version does not have to include the laboratory material of the editors (variants, alternative versions, minor paratexts, illustrations and so forth), but rather confines itself to a single, uniform reader's text with a minimum of editorial tools and paratexts. The digital cumulative archive on the other hand assumes the role of the primary, with or without a web interface, from which static spin-offs are secondarily launched in print, on CD, as e-books or on the web. The digital archive is thus able to play with various document forms as outputs (Svedjedal, 2000). A printed codex edition embodying one particular editorial theory ideal is therefore no longer the only possible output of the editing endeavour but rather one potential output from among many that at least in theory might satisfy several different and perhaps even rival theoretical ideals.

One of the questions we ask ourselves in the light of this development is whether the SE can and should continue to fulfil the same functions. To what extent, if any, might the logic and capabilities of new media affect the essence of scholarly editing? Do we need editions any longer, or should we rather invest our human, economical, and textual resources in massive, long-term digital archives? Any attempt at answering such questions will need to begin by reconsidering the nature of the SE, what forces it has and has not, what limits it has and what kind of factors determine its possibilities and limits.

This article attempts a tentative discussion of such forces and limits of the SE, and specifically looks at its supposedly representational and reproductive force. The aim is to identify poles of extreme positions in editorial discourse and thereby to map out the fields of tension and perhaps conflict that lie between them. Coming from the field of bibliography and library and information science, I will begin by making an argument for the bibliographical dimension of the SE.

1. Bibliographical Tools

There is a tendency in bibliography and the adjacent fields of library and information science, textual studies or historiography, particularly manifest in their textbooks and education, to treat bibliographical tools as more or less neutral instruments impeccably beyond the limitations of spatial, material, medial, historical, social, and ideological constraints, and free from the biases and tastes of any author. Any close reading of the tools as texts, however, reveals their situatedness: their dependency on particular historical media settings, their socio-cultural roles and functions, or their argumentative, rhetorical dimension.

For instance, the tools have historically been developed as solutions to particular historical media situations. The parameters of new media technologies and the logistics of distributive networks make us aware of such medial and technical constraints of the tools. Their shape and architecture at a given moment in history is not haphazard but a result of particular media settings. New media contain and distribute the genres and architectures of older media. Perhaps needless to say, they also impose constraints on what is both theoretically and pragmatically achievable with the bibliographic tools.³ Further, the tools are never genre neutral, but on the contrary steeped in certain genre assumptions and respective social functions (Andersen 2002).

There are as well social and historical dimensions in the tools of bibliography, as instruments performing on various social arenas, mediating between communities. The tools are also always to some extent hermeneutical documents, subjective interpretations, in two senses: they carry with them a history of ideology and a hermeneutical heritage, and they also exert an interpretative influence over the objects they are designed to manage.

2. The SE as Bibliographical Tool

We are currently experiencing not one but several parallel introductions of new media and technologies, exhibiting radically different logistics and parameters for document production and distribution than print-based technologies do. This has had and continues to have tremendous effects on the nature, construction, and purposes of bibliographical tools. The conditions for both the tools themselves and the documents they are designed to manage are dramatically altered. For instance, new media and web distribution promise to vastly enhance the spatial confines of SEs, or even to annihilate them altogether. It is then of little surprise that the concept and epistemological assumptions of SEs are discussed and reconsidered:⁴ what is a scholarly edition, really? What does it do, what should it do, and why? Who edits what for whom and for what purposes? Where is the line between archive and edition? What kind of power, exhaustiveness and concentration do we want SEs to exert? What changes are we witnessing in the division of labour between the people involved in scholarly editing, the tools they use and between the various media outputs from such endeavours?

The SE is, and has been for a long time, a complex and diverse family of document types in the sense that many technologies, professional practices and academic areas converge in it, resulting in a spectrum of variant types ranging from facsimile, diplomatic, synoptic, genetic, critical (also known as historical-eclectic), variorum editions to large-scale digital archives on compact discs or mounted on the web.⁵ There is little general agreement as to the classifications. The division of critical versus non-critical, for instance, might render the impression that the latter type, whereby diplomatic and transcription editions are usually designated, somehow escapes implementing the scrutiny of textual criticism or of critical inquiry. More to the point, the labelling of the results of scholarly editing in new media environments seems, as has been pointed out (Robinson, 2002, pp. 45 ff.; Vanhoutte, 2003), to further blur some of these classifications. To even talk about digital editions as one particular type of edition is for instance debatable. Current discussions on digital editions tend to talk about the genre as based on media form and publishing technology, whereas traditional discussions in editorial theory rather identify the genre as based on its epistemological foundation and theoretically based strategy. Discussions therefore end up mixing apples and pears: digital editions versus, say, eclectic editions. This presupposes one predefined function and theoretical base for the digital editions to counter the ones identified in printed editions, when in fact all kinds of editorial approaches - both traditional and innovative - are being tried out in the realm of new media.

There is little room or intention in this article to elaborate further on the classifications of SEs. Perhaps we can at least agree on their quality as tools and results of scholarly inquiry, enabling us to refer to them as scholarly editions. I would argue that there is also another common denominator for SEs: their nature as bibliographical instruments.

There is obviously an historical bond between, on the one hand bibliographic activity, and on the other scholarly editing based on textual criticism, from Alexandria and onwards. Particular branches of bibliography have collaborated closely with scholarly editing, such as textual and analytical bibliography. But there are also deeper epistemological bonds. Both bibliography and a critical edition are occupied with clustering a set of documents and their texts around an abstract work notion. This is why it is central to both bibliography and editing to understand and define the concepts of works, texts and documents. This is also why concept relations and conceptual analyses are crucial ingredients in the emerging theory development within both fields.

A work can be defined as the principle of abstracted and complex ⁶ relations between clusters of documents. Bibliography tries to exert order and category among such clusters by applying the concept of work. So does the SE, making it as it were an instrument for bibliographic activity. A critical edition is a statement as to the extent and confinements of a particular work.

Scholarly editing for instance undertakes the mapping of relationships between versions of a work as manifested in documents, the perhaps most famous map illustrations being the stemma and the choice of copy text or texts along with variant apparatuses. In fact the entire editorial task whether to present the edited work as a facsimile, an eclectic, a variorum, a synoptic, or a web-based hypermedia edition, is largely a matter of identifying the work with one or several chosen relationships. There is an intellectual link between this relationship mapping and the kind of bibliographic relations categories that bibliographic theorists have identified.⁷ The way the SE manages work-version-document relations is analogous to the way a catalogue manages bibliographic relations or the way IFLAs Functional Requirements for Bibliographic Records talk about them (IFLA, 1998).

3. The SE as Icon

There is a related outcome of the way bibliography, primarily enumerative bibliography, and editing based on textual criticism are similar ⁸ activities, or in effect two variants of the same activity: iconicity.⁹ This is one of the chief objectives of both activities, which is to produce surrogates by iconic representation. As Ross Atkinson pointed out in his stimulating 1980 article, bibliographic records, catalogue posts, and text-critical editions all function as simile representations, ranging from the single catalogue entry, via full-text records in databases, via facsimile editions, transcriptions, critical editions, variorum and synoptic editions, over to full-scale exhaustive databases or digital archives. There is obviously a considerable scale of exhaustiveness and completeness, but nevertheless a commonality in iconicity.

"An enumerative bibliography," Atkinson (1980) writes, "reproduces its Object in microcosm; it is a reflection, a picture of its Object. As such, the relationship between sign and referent in enumerative bibliography is one of similarity and may consequently be designated iconic." The same really goes for descriptive bibliography, he continues, and interestingly enough also for the way textual criticism is a reflection, a picture of the edited work as perceived and constructed by the editor in one or several documents. "[T]he document (in its various conditions)," Atkinson (1980) goes on, "is approached as a set of representable characteristics -- a raw material -- from which a product, the description, is to be created". The difference between reference and descriptive bibliography as activities is quantitative. If you take into account the consequences and suppose an exhaustiveness degree at its fullest, what you end up with is textual criticism. Again, there is theoretically only a scale of exhaustiveness. From this angle, textual criticism is a natural extension of bibliography, and Atkinson in fact posits a constellation of iconicity as EDT (Enumerative bibliography, Descriptive bibliography, and Textual criticism).¹⁰ Atkinson's is one of the few argumentations I've seen for the epistemological denominator of iconicity between textual criticism and bibliographical activities, and deserves merit for this quite simple but important observation.

4. The SE as Media Translation

But the potency of iconicity can certainly be subjected to inquiry. Iconic representations are bridges between documents (as interfaces to the works at hand), striving to maximize the degree of similarity when transporting the perceived work contents between them - but how potent are our tools at achieving such similarity? Representation is in this case an instance of the activity of copying, reproduction, or what I will refer to as media translation.¹¹

Using documents, we hope to be able to repeatedly have access to some of the originally intended qualities of a work, and we also intend for these original qualities to be repeated in more or less the same manner every time we access the document.¹² If 'the same manner' is repeated we are satisfied that we have had access to the work. If however the manner deviates far enough, we sense that it no longer conveys the same work. But as long as each document manifestation is more or less adequate, a tangible, readable, accessible instance of the work is presented to the world. In other words, with documents we hope to be able to keep the work alive, or rather to keep the memory of the acted work alive.

Needless to say, perhaps, documents are also media and matter, and so documents are more or less subjected to the natural decay of all matter. They crumble away and die. If the material instantiation of a work dies, the work it contains dies with it unless we keep it alive in the internal memories of people or in external memories, i.e. re-instantiating the work in a new document or set of documents.

Media translation goes from a departure document to a target document. It entails many phases, e.g. scrutinizing a document, trying to establish what particulars in the document that are substantive elements of the work we suppose the document contains, and then using a new document (from the same type of medium as the departure document or from a different type of medium) into which we try to carve text and other signs in order to manufacture a target document that purports to be a remake of the departure document and, to some extent, of the work the latter contained. But it is vital to recognize that the target document is always derivative to the departure document.¹³

There are many types of media translation at use now and in history: monastic hand-copying, micro-filming, or digitization such as scanning are all examples of media translations using departure and target documents. Translation brings about transmissional noise. Although rather unproblematic in an abundance of genres, such noise however tends to become considerably awkward in cultural heritage works and other material that particularly call for the critical inquiry of human subjectivity. Textual criticism is an historical solution to come to terms with such noise. There are vast numbers of potential parameters introducing noise and constraining the target document and its text:¹⁴

Each medium as well as each document type produced within and for that medium brings to the text a semiotic system of its own.¹⁵ In the translation process, certain features of the work are preserved that can be carved into the flesh of the new medium and be expressed by its architecture and the language of its web of signs, while others are treated as noise, obscuring the substantive signals. If translation is successful (in the sense that a human agent accepts the target document as representing the same work as the departure document) we feel the work has been kept alive for yet a little time, namely the time span of the new document instantiation. Then the work is translated again and again, perhaps even across centuries and millennia. At the same time, it is being reinterpreted by new readers and users and thus brought to new life, each new manifestation mirroring particular contemporary medial, social or cognitive settings. But when no more translations take place, no more new documents refresh the work and the old documents finally die, the externally memorized work has ceased to be. And this is precisely how we have lost the vast majority of the works produced in history. At the same time media translation is, alongside the preservation of the original document, a crucial instrument for bringing external memories of past works between generations.

5. The SE as Scientific Tool

Scholarly editing is an important instrument in such media translation processes, and the SE is consequently subjected to the constraints discussed above. These are recognized by much editorial theory, but far from always explicitly acknowledged in the SEs themselves. Particularly in idealistic editing discourse, SEs have often been presented as neutral "scientific" instruments.¹⁶ An SE does contain introductory essays, editorial principles statements, and reports of the methods that were implemented in the task of editing, but these do not always address issues of subjectivity in the editorial function, how the editor contributes to shape the edited work through his/her deliberate choices between versions, forms, granularity, media and presentation. Rather, the impression one gets from reading many SEs and their statements is one of presumed intersubjectivity, repeatability, and cumulative force. This to be realised through the use of the critical apparatus, the stemma and the editor's account for techniques and methods applied, the level of textual granularity chosen, and the paths taken by the editor. All this is to enable the user-as-editor to follow such paths or to tread, as it were, different paths than the editor. With adequately and carefully applied methods and techniques, the scholarly editor supposedly draws the "correct" text of the edited work from one or several documents, affecting its text only in as much as she/he washes it clean from the dirt of corruption. An outcome of this is the idea that editorial practice and textual criticism are recreating original material, be it an abstract intentional authorial text or a particular document text such as the reception text or a manuscript text. An extreme but increasingly moot conviction claims it to be both possible and ideal to confine the editorial task to mere discovery and proliferation of the original, to being somewhat of a transparent medium in which the work can safely be transported to its readers. The editor then goes on to report his/her work and reproduce the work in a new document, the edition, which in turn can be used as working material for new scholarly endeavour. But we must keep in mind the simple fact that rather than recreating the departure documents themselves, scholarly editing engages in creating new, target documents, "similar" but all the same derivative to the departure material.

Arguments as to the degree of representational force of the SE work along an axis. At one pole, editing as textual transmission between documents is a relatively uncomplicated matter. The real challenge is then to generate methods and technologies for the transmission to be performed with little or no noise - i.e. transmissional noise can be annihilated. At the other pole, scholarly editing is an undertaking inevitably constrained by many medial factors, making transmissional noise inevitable. From the point of view of the philosophy of science, the axis is related to on the one hand idealism (where in its most extreme Platonist variant contents are disembodied, separable from the their physical document carriers and hence transportable in their entirety to other carriers) and on the other materialism (where the most extreme position would argue that texts are not only media typical but even exclusive to particular material media).

Given the promises of new media and web distribution to vastly enhance the spatial confines of the editorial material, or even to annihilate them altogether, a subscriber to the idealistic view might be tempted to plead for the makings of "total" digital archives, where every document witness and variant of every work of an author can be accessed in digital form in all manners of display and modes and for all kinds of purposes. The idea is also to enable a user to generate practically any type of edition she/he desires and thus partly or wholly fulfil the editorial task him/herself.¹⁷ Taken to an absurd extreme, it might turn into a notion that documents in their entirety can be reproduced, which is a kind of "mimetic fallacy".¹⁸

In an e-mail discussion list thread earlier this year on digitization and text encoding, Willard (2003) referred to two recurring fallacies in digitization debates and media theory as the 'complete encoding fallacy' and the 'mimetic fallacy'. I think both make way for simple replacement models.¹⁹ The complete encoding fallacy was defined by McCarty as "the idea that it is possible completely to encode a verbal artefact", the mimetic fallacy being "the idea that a digitized version will be able to replace its non-digital original". The two are closely linked. If it is possible to "completely" identify, formulate and unambiguously encode every aspect of an artefact, say a document, into that of another document, then by logic the target document ought to be in every aspect equivalent to the departure document. And if it is indeed equivalent, it follows that to a user it is of no importance if she gets her hands on the one or the other. And if that is of no importance, then there is little need for retaining both the departure and the equivalent target document in a collection. The target document can in other words replace the departure document because it is a perfect mimic of it, or at least perfect enough to get rid of the old one. Conversely, you can't make a case for mimetics if you do not believe it is possible to transfer all the potentially relevant aspects between media and between documents. To these fallacies, the idealistic disembodiment viewpoint is of course an intellectual necessity: you cannot legitimize mimetics if you do not subscribe to the possibility of completely separating document from information. In all fairness, these are just ephemeral names and were probably not intended by McCarty to be regarded as a definitively thought-through model of the state of affairs in current media theory, but rather as mere handles with which to manage the particular discussion thread at hand and to make some rhetorical points in that particular context. Nevertheless, I think 'mimetic fallacies' is useful as an explicit label when discussing problematic tendencies so far only hinted at implicitly in many discourses on the various processes involved in the production, distribution, consumption, and indeed translation of works, documents, and their texts.

6. The SE as Rhetorical Tool

Let us return for a moment to Atkinson's article. His is, I would say, an incomplete recognition of the various aims and functions of textual criticism. It is also unfair to infer from Atkinson that scholarly editing and SEs always depend on textual criticism and therefore share its aims and functions. SEs are produced for a number of reasons by and for a number of professions and groups in society, using a variety of media, of bibliographical levels in the scope of the edited material selected, of granularity, and of editorial strategies, and theoretical programmes. In short, an SE is not only an iconic representational device, but a social and intercommunal instrument as well.

Furthermore, if we regard the SE as a genre, there are useful perspectives in genre theory to analyse bibliographical tools such as the SE. Among the many genre perspectives around in literary theory, linguistics, sociology and new rhetorics,²⁰ Carolyn Miller's much-quoted idea of a genre as "typified rhetorical actions based in recurrent situations" (1984, p. 159) strikes me as particularly fruitful. It emphasizes both a functional perspective of documents in use and the rhetorical dimensions of genres. Genres are more than the commonality of textual and visual patterns in documents, and Miller points to the socio-rhetorical situations that give rise to the documents. I think we can bring this perspective to analyse bibliographical tools such as the SE.

In a 2000 conference talk, Bethany Nowviskie made an interesting comment on the nature of SEs: " ( ...) a scholarly edition contains an editorial essay, which makes an argument about a text or set of texts, and is then followed by an arranged document that constitutes a frozen version of that argument. ( ...) the text of a scholarly edition is an embodied argument being made by the text's editor".²¹ This is as straightforward as it is an important observation, and it is in line with Miller's observation on the rhetoricity of genres. When the SE is looked upon as an embodied argument made by an authoritative instance of responsibility (what in Foucaldian terms might be labelled the "editorial function"), one can regard the SE as a text of its own, approaching the status of a bibliographical work in its own right. The edited work is, then, incorporated into the edition-as-work, or more precisely: the text of the edited work becomes a sign in the editor's text.

An editor might suppress or acknowledge her presence and influence in the SE, but the subjectivity is still there. The tension between acknowledged presence and presumed absence of the editor has a long history in textual criticism and scholarly editing. To intervene or not to intervene might in Greetham's use of words be described as a choice between an Alexandrian and a Pergamanian editorial ideal (1999, p. 50 f.). The former accepts and even presupposes intervention and corrections, laying the ground for eclectic editing, while to the latter interventions and corrections are theoretically awkward (and even come close to heresy), making way for the school of facsimile and best-text editing. The more explicit in an SE an editor's presence, the more the genre achieves authorial status. Conversely, the more an editor seemingly withdraws from the scene, the lesser its status as an authorial text. Going back to the idea of total digital archives based on diplomatic and facsimile editions: if in academic discourse scholars appear to want to "hide" their role as narrative writers,²² then such an archive promises - or threatens - to enable them to vanish altogether, inviting readers to step in and fill the creative, authoritative editorial function. As noted by Bjelland (2000), there might in such cases be problems when the archives in an attempt to achieve user-friendliness seeks to hide their markup, scripts and programming details, which actually disarms the user-as-editor.

Atkinson's description of textual criticism and scholarly editing activity has an idealistic flair about it, as if it were the sole objective of textual criticism and scholarly editing to recreate as accurately as possible one or several documents into a new document. As Henrikson (2002, p. 56) reminds us, if the purpose of an SE is little more than to carry the linguistic text of the work between generations of readers and media as accurately and objectively as possible, then there are certainly fast, cheap and reliable methods for accomplishing this by means of automation. But SEs, including the ones based on textual criticism, are more often than not producing new rhetorical documents, where both the editor herself and the readers of her/his edition are perfectly aware of the ways the edition text deviates from the edited document texts, and where this very reconfiguring, repositioning and recontextualizing of the edited work in fact is conceived of as a core value of the editorial work. An SE is rather an attempt at positioning the work in contemporary literary discourse.

So while I agree with Atkinson on the iconical denominator, I would contend that textual criticism and scholarly editing are also hermeneutical and rhetorical activities, and the tools and documents they use and produce are equally interpretative and argumentative. In that way, and if we feel comfortable with Peircean distinctions, textual criticism and editing are also indexical activities, related to analytical bibliography, which places an SE along other axes as well.²³

7. The SE as Reproductive Tool

To sum up, the SE is a subjective, rhetorical device. It is moreover both a result of and a comment on contemporary values, discussions and interests. It is situated in time, in space, in culture and in particular media ecologies (of both departure and target media). To all bibliographical genres, using derivative target documents as representations of departure documents, these are factors imposing constraints on their iconic force. The situatedness limits the representational and moreover the remediating force of bibliographic tools, including the SE. There are no absolutes here. The SE obviously has representational and reproductive force, the very abundance and undisputable value of SEs throughout history testify to that truism. The interesting question is what factors are at work to limit or to enhance this force. Another important matter is what force and purpose the remediated material itself might have, that is, to what degree the SE is valuable as laboratory, as working material for new scholarly editorial endeavours. I am not talking about the value of SEs for historians, for literary critics, for studies in the history of ideas etc., but for the makings of new SEs based on textual criticism.

A claim has been put forward that digital archives can be used as the platform from which to construct new critical editions of high scholarly quality that differ in scope, intended audiences, bibliographical levels and underlying editorial strategies and theoretical programmes.²⁴ This is an interesting claim that has a nicely pragmatic ring to it, but I think we need to address the limitations of such presumed archives. One might also generalize the question and ask in how many cases earlier, print-based editing has been able to rely, partly or even solely, on the material contained in previous SEs, as a raw material basis for the production of new critical editions with little or no need for consulting the fontes, the original documents? I think the number is scarce, and I think there are several reasons for it. Obviously, the inclusivity, the simulating capacities, the modularity and the transportable flexibility of new media are considerably different than what is possible to achieve with printed codex editions, but are the principal problems as well considerably different?

Already the claim of the printed edition's reproductivity is questionable. It is based on the SE's supposedly scientific nature, in that it supplies reports of the editorial labour undertaken, a conscientious inventory of the extant material of the edited work, reproductive tools such as the critical apparatus or the stemma, arming the editor to undertake editorial research her/himself based on what the SE has to report. But to what extent do printed SEs really lend themselves to being such cumulative reports and reproductive laboratories? Are they at all being used that way? One can note the distrust that has been put forward in the reproductive force of, e.g. the critical apparatus, whose functionality might even be a chimaera.²⁵

Digital editing makes use of such print-born reproductive tools, but also fosters new ideas of how to accommodate reproductivity. Such ideas are normally founded on the inclusive, simulating, and hypertextual capabilities of new media, exploding the embryonic idea of synoptic and variorum editing in print media into full blown hypermedia display of several or all versions of works. Building a digital archive means bringing together and storing massive amounts of target documents. This is of course what any archive does, and we already have numerous prime examples of the beauty, force and hence the value of digital archives, many of which have been presented and discussed in earlier issues of Literary & Linguistic Computing. But the archives and their contained material will always be situated documents themselves, dependent on the kind of situational factors we have discussed above. If such archives are to be used as laboratories for generating new scholarly representational documents such as critical editions, i.e. turning the target documents into departure documents, one would have to stay alert as to the derivative status of the archived material in the first place. An SE based primarily (if not solely) on the derivative documents of such a digital archive will always to some extent depend on the inevitable choices made by the persons building the archive, on the historical, socio-cultural, cognitive, and media particulars and on the pragmatic purposes and theoretic values defining and framing the final derivative documents in the archive.

A transcriptional editing approach, for example, that equals the text of the primary documents and of the edited work to the linguistic, alphanumerical signs and the compositional structure of the text as interpreted by the editor(s), sets aside McGann's bibliographical codes, which in its turn might decrease the reproductive force of the resulting edition and its text for those researchers and students primarily interested in working further with precisely such bibliographical matters. A transcriptional approach aiming for faithfulness to the text of the studied documents, still faces a huge array of inevitable interpretive choices and has to make compromises and sacrifices of what to represent and what to leave behind (Robinson, 2002, p. 55). What epistemological approaches we bring to the editing process, what methods we use, at what bibliographical level we position the endeavour, necessarily determine the representational and laboratory strengths and weaknesses of the edition, making it more apt to some users and the editorial ideal they subscribe to, and less apt to others. A universal aim is doomed to failure because it is rooted in an assumption that both textual material and scholarly editing are context-free phenomena.

This is true for digital imaging and the choices of parameters (such as colour and resolution) that need to be made in the process of selection, interpretation, copying, formatting and reproduction of the images. This is also the case for the seeming simplicity of transcription, which is refuted already by several contributors in this issue, as well as by others. Transcription involves inescapable choices of particular textual features and fixed levels of granularity at the expense of others. And encoding by markup brings additional questions of hermeneutics and interpretation that add further to the subjectivity of the editing endeavour.²⁶ A user entering a digitization archive faces material that is encoded and thereby, arguably, already interpreted. Hypermedia archives further hypertextualize some intra-, inter- and extra-textual relations and navigation routes, and leave others dormant. What possible bibliographic work can be done with the archive material is thus already to some degree predefined, which of course will be awkward to any archive hoping to function as reproductive laboratory fulfilling scientific ideals.²⁷ A user is in the best of archives free to manipulate, recombine, and rearrange some of the material, but this freedom is not without limits. A user-as-editor that within the confines of for instance analytical bibliography wishes to ascertain chronological relations between two or more primary documents of which there are target representations in the archive, and whose focus of interest is more oriented to the form than to the logical structure of the primary documents,²⁸ probably needs access to different arrays of significant components than a user-as-editor oriented to the history of ideas trying to frame the primary documents or the work they contain in their socio-historical contexts.²⁹

I think the legitimacy of the reproductive assumption can be discussed, at least as far as we're talking about the makings of new critical editions based on the target documents in previous critical editions and archives. This is not to say that such digital (or printed, for that matter) archives cannot have editorial reproductive force at all, but rather that such force will always be delimited by the inescapable fact that the archival documents are derivatives, and that every textual decision behind them "inevitably reflects particular approaches to literature, and that the resulting text may be inappropriate for certain purposes" (Tanselle, 1995, p. 14).

The reproductive force of an archive will depend on many crucial factors, such as if we are dealing with the editing of works whose originals, archetypes, etc. are long lost and that therefore have come down to us only in the n^th generation, each generation being a derivative translation of previous ones, possibly (but not necessarily) accumulating errors, deviations, and other effects of such historical "whispering down the lane".³⁰ In other words, the distance between the contained documents and the originals of which they are derivatives, comes into play. With each generation of media translation, the distance is in principle increased between originals or archetypes and their derivatives, both historically and textually. The force of a laboratory for critical edition generation, then, that is based on derivative documents positioned several remedial generations from the primary documents, is affected by this circumstance, and it will inevitably define what kinds of new editions one can hope to generate using such derivatives. The users of such archives will, to paraphrase Tanselle (1995, p. 14), tie themselves to the historical moment in which the archive document containing the text was produced.

The reproductive force further presumes that the ambiguity inherent in the transcription, coding and encoding of the material can be disambiguated and decoded (Burnard, 2001, p. 35). Another factor is the theoretical distance between the departure project the material was taken from, and the target editorial project in which the material is to be reused, that is whether they differ in theoretical aims and programmes, in intended uses and audiences, or whether they are similar in these matters, the latter case arguably enhancing the reproductive force.

At the end of the day, we are facing questions of scientific theory: to what degree does the SE as genre and scholarly tool lend itself to the kind of sequential cumulativity of collective disciplinary knowledge that is ideal in the discourse of the sciences? Do we regard the SE as a more or less pure iconic tool, emphasizing the versatility of the textual material in editions, or as a more or less situated text, emphasizing its bonds to the particulars of time, culture, media and individual editorial or lectorial tastes and biases? Is the SE an autonomous or a constrained bibliographic tool? What answers we provide for these questions affect what reproductive force we ascribe to the SEs. One does not necessarily have to choose sides here, but rather discuss what we lose and gain with each perspective. Questioning the reproductive potency of digital archives is not necessarily the same as dismissing the considerable value of constructing large digital archives. I simply want to argue for caution in libraries and archives engaging in digitization projects: they need to stay on a pragmatic path and not be tempted by any siren songs of universal reproductivity. To quote Walsh (2002) in a recent debate on the versatility and reproductive force of XML: "It may not be possible to achieve one input - all outputs, but surely one input - many outputs is an entirely practical goal."

8. Coda

I started the article by positioning the SE as a bibliographical tool, a valuable and privileged one. Such tools are governed by values, epistemologies and interests that need to be identified and formulated. I think these kinds of discussions, invigorated by the advent of new media and distribution technologies, make us better equipped at identifying the nature, strengths, and weaknesses of the tools, what forces but also what limitations they have, e.g. as iconic representations and socio-historical genres, changing with time, space, social context, and media. Recognizing what interests and worldviews are at stake in, for instance, the making of digital archives might reduce the risk of us expecting the wrong things from them. It might assist in avoiding the traps of mimetic fallacies and replacement models in library management when funding, engaging in, and conducting digitization projects. It might make us cautious as what we can reasonably expect from such projects.

We need to see SEs and other bibliographical tools not as neutral prolongers of the life of the works and documents but as filtering media affecting them and our way of perceiving them. This might make it easier to understand what the tools can and cannot do, where they come from, what intellectual, cultural, symbolic heritage they bring with them, and where they might be going.

Notes

1. The term new media, taken from Manovich 2001, is preferred to other labels such as electronic or digital media.

3. In digital editing projects, technology and methods might to some extent set the scholarly editorial agenda (van der Weel, 2001, para. 14).

4. Among the overwhelming amount of reconsiderations at conferences, in yearbooks, e-mail discussion lists, and scholarly journals from at least the last decade, see several previous issues of Literary & Linguistic Computing (e.g. 15.1, "Making Texts for the New Century"), TEXT, Editio, Studies in Bibliography or the recently (2002) launched Variants: the Journal of the European Society for Textual Scholarship.

5. See Tanselle (1995) or Vanhoutte (2003) for overviews of variety in types. SE typologies are discussed more at length by Meyer (1992). The genre variants can be classified according to varying principles, such as the media used in the editing process (e.g. image-based or text-based), the media used when distributing the edition (e.g. printed versus electronic; codex versus CD, web or hypermedia), the use and intended audience of the SE (e.g. commented student editions or archival facsimile editions), editorial theory and strategy (e.g. intentionalistic, genetic, text sociological or new philological), the bibliographically conceptual scope of the edition in relation to the documents studied (e.g. facsimile, transcription, diplomatic, synoptic, eclectic or variorum).

6. This complexity is a result of versioning factors due to the fact that the document clusters are spread out in time, space, and across media.

8. Actually, they are "similar activities" in two ways: they are similar to each other, but they also both engage in iconic representation, which in its Peircean sense, as discussed below, is understood as a representation, whose object it is to be as "similar" as possible to an original.

9. In this particular context, I am using the word to allude to a Peircean iconic sign, which as we might recall, refers to the denoted object solely by its own characteristics. An iconic sign is linked to its object by virtue of similarity. A portrait of someone, for example, is an icon.

10. " ... if the relationship of enumerative bibliography to its Object does indeed involve the reproduction of portions of that Object, and if that of descriptive bibliography is the same only with greater precision and in greater detail, then there must be a final level of such a representational relationship at which the goal is to reproduce the Object with maximum precision and in every detail. The name we give to the discipline practiced at that level is, of course, textual criticism. That is to say, then, that the activity of textual criticism, by virtue of its most elementary function -- the reproduction with the greatest possible accuracy of sign systems which either once actually existed or which were intended to exist -- can and must be described as the concentration of the essential representational activity of enumerative and descriptive bibliography onto a single, total document." (Atkinson, 1980, p. 68).

11. A term suggested by both Grigar (2002) and Hayles (2003). There are many labels in use for this process, e.g. transition, transcendence, or transmission. The labels are neither haphazard nor equivalent, because as metaphors they convey something of the underlying understanding of the process. The translation metaphor is useful in its emphasis on the derivative status of the results of the process, whereas the other terms mentioned suggest that what is being transferred goes through more or less intact (or they are indifferent to whether there is a change or not). 'Remediation' (Bolter and Grusin, 1999) has roughly the same connotation, but I think the translation trope is better at making us grasp the degree to which the process creates something new. Cf. also Eco (2003) for a fresh semiotic view on the issues of sameness and translation.

12. To philosopher Nelson Goodman (1969), this "sameness" in fact constituted the criterion of works of art and their texts (which he referred to as notation). As long as the notation is "correct", the work is intact. When notation is altered, you end up with a new work. I will refrain in this particular article from a thorough analysis of Goodman's rather mechanical approach to works, texts and documents, but merely point out that it has little or no room for versions, variants and (near-)equivalents that enable bibliographic clustering of "almost-like" objects into works and work families. To a textual scholar, Goodman's definition also comes close to reductionism: it reduces the work in all its complexity to one of its components, the linguistic notation.

13. Obviously, I am referring here to sequential translation, i.e. across time. Granted, technologies have for centuries made parallel media translation possible, with multiple instantiation and proliferation of works. In such cases, each production process results in a multitude of differing documents that constitute the work and that for technical reasons might travel across media. This is however a primarily technical translation as opposed to the intentional sequential translation that has as its primary task to extend the work beyond the time span of the departure document.

14. In particular if by text is implied not only the linguistic text expressed in linear sequences of alphanumeric characters along with punctuation, but also the accidental textual particulars (expressed in typography and other visual markers) McGann (for some reason) chooses to label bibliographic codes (McGann, 2001). If the former aspect is normally subject to authorial intention, the latter is more the result of collaborative acts, including typographers, printers, and editors (cf. Bjelland, 2000). The bibliographical codes are probably subject to media translational changes to a higher degree than are the linguistic characters we normally define as the pure text.

15. Cf. Robinson (1996). David Levy (2000, p. 26) notes, as have many others, the affects that translation has on textual content of a document, but relates this knowledge to the contextual, one might be tempted to say social constrains of a document's properties and essence: "Differences will always be introduced in copying; the trick is to regulate the process sufficiently so that the resulting differences are of little or no consequence and that the properties of greatest consequence are shared. Determinations of which properties matter are made in the context of purpose and use." (Levy, 2000, p. 26; my italics)

16. See Tanselle (1974) for a stringent discussion on the claims of scientific properties versus bibliography and textual criticism.

17. An early passage by Price-Wilkin (1994, section 4.1) is indicative: "With proper markup, an edition can be viewed in as many ways as the reader desires. It can be a variorum, a study edition, a critical edition, or historical evidence. The form the edition takes is defined by the user's needs or preferences."

18. If the editorial role was transferable to the reader, then the latter would have to have access to all the original documents that the editor would have had in making the archive. This means the original documents themselves have to be included in their entirety in the digital archive, which is impossible. You cannot possibly computerize "everything" about an author or even a work or even a document (And what makes up "everything"? Cf. Greetham, 1996, p. 44).

20. By, e.g. Bakhtin, Todorov, Halliday, Swales, or Bazerman. For an introduction to genre studies, see Freedman & Medway (1994).

21. Nowviskie (2000) (my italics). There is a parallel in Bjelland (2000, p. 8): " ... the edition itself makes certain statements as to the nature of Shakespeare as an "author" and his "canon". To be sure, these statements are made implicitly, not explicitly as in the more politicized prefatory material. Yet as Michel Foucault would have recognized, these implicit statements play a powerful role in preserving the boundaries of a discipline as they also work to conserve a particular tradition."

22. Bazerman (1988, p. 14) makes the following observation on the scientific article as written genre: " ... to write science is commonly thought not to write at all, just simply to record the natural facts. Even widely published scientists, responsible for the production of many texts over many years, often do not see themselves as accomplished writers, nor do they recognize any self-conscious control of their texts."

24. e.g. Ore (1999, p. 143), "Det elektroniske grunnarkivet bør - så langt som mulig - være grundig nok og komplett nok til andre typer utgaver, men også til nye tekstkritiske utgaver basert på andre prinsipper" ( = "The electronic main archive ought to - as far as possible - be thorough and exhaustive enough to serve other types of editions, but also new critical editions based on other principles" [my translation]). Ore's reservation "as far as possible" is however important and goes to the heart of my argument.

25. "The apparatus eventually becomes a mere cemetery of variants" (Vanhoutte, 1999, p. 202).

27. Cf. Rockwell (2003, p. 215): "The tools [in large electronic text projects] are deployed not for general use but to make available the research of a specific project in the ways imagined by that project. However, original research consists of asking new and unanticipated questions ... "

28. Bjelland suggests that apart from the somewhat crude and quite general <fw> (forme work), there are few tools in TEI to manage the particulars of interest to analytical and historical bibliographers. "Although this difficulty might be overcome as more analytical bibliographers become involved in the creation of hypermedia environments," (a development that might currently appear as a trifle utopian) "how they will reconcile the inherent epistemological tension between TEI Guidelines designed to describe the document's logical structure and the bibliographical emphasis on its format (i.e., the form of the document) is by no means clear. Since all elements of a given document must be embedded with the structure-oriented DTD ( ... ), at present there is no way to make the page itself the bibliographical equivalent of a DTD. In other words, within the language of the TEI Guidelines the "forme work" will always be of secondary importance ... " (Bjelland, 2000, p. 25)

29. " ... none of these archives, no matter how sophisticated they may be, will be able to present every "bibliographical code" of significance in such a way that its nature and importance are readily apparent to the user of the archive." (Bjelland, 2000, p. 24; her italics)

30. I'm thinking primarily of the massively versionalized classical works with long lost originals and even archetypes. If we're discussing 18^th or 19^th century works, where we do have extant authorial manuscripts, first print editions, proofs etc, and where the bulk of documents and versions is not overwhelming, the perspective and the possibilities are altered, but not, I suggest, to the degree that allows for the automatic generating of tailored editions serving quite different and, more to the point, rival editorial strategies and theories.

References

Andersen, Jack (2002). "The Materiality of Works: The Bibliographic Record as Text." Cataloguing and classification Quarterly 33.3/4: 39-65.

Atkinson, Ross (1980). "An Application of Semiotics to the Definition of Bibliography." Studies in Bibliography 33: 54-73.

Bazerman, Charles (1988). Shaping Written Knowledge : the Genre and Activity of the Experimental Article in Science. Madison WI: University of Wisconsin Press. http://wac.colostate.edu/books/bazerman_shaping/

Bijker, Wiebe (1995). Of Bicycles, Bakelite, and Bulbs: Toward a Theory of Sociotechnical Change. Cambridge, MA: MIT Press.

Bjelland, Karen (2000). "The Editor as Theologian, Historian, and Archaeologist : Shifting Paradigms within Editorial Theory and their Sociocultural Ramifications." Analytical & Enumerative Bibliography 1: 1-43.

Bolter, Jay David & Richard Grusin (1999). Remediation : Understanding New Media. Cambridge, Mass.: MIT Press.

Burnard, Lou (2001). "On the Hermeneutic Implications of Text Encoding." In: D. Fiormonte & J. Usher (eds.), New Media and the Humanities: Research and Applications. Oxford: HCU. 31-38.

Eco, Umberto (2003). Dire la stessa cosa: esperienze di traduzione. Milano: Bompiani.

Freedman, Aviva & Peter Medway (eds.) (1994). Genre and the New Rhetoric. London: Taylor & Francis.

Goodman, Nelson (1969). Languages of Art : an Approach to a Theory of Symbols. Indianapolis: Bobbs-Merrill.

Greetham, David (1996). "Textual Forensics." PMLA : Publications of the Modern Language Association of America 111: 32-65.

Grigar, Dene (2002). "MOOtextuality". Text Technology : The Journal of Computer Text Processing 11.1: 163-179.

Haas, Christina (1999). "On the Relationship between Old and New Technologies." Computers and Composition 16: 209-228.

Hayles, N. Katherine (2003). "Translating Media: Why We Should Rethink Textuality." Yale Journal of Criticism 16.2: 263-290.

Henrikson, Paula (2002). "Kampen om litteraturhistorien : romantikerna som filologer." In: L. Burman & B. Ståhle Sjönell (eds.), Text och tradition. Om textedering och kanonbildning [Text and Tradition. On Text Editing and the Creation of a Literary Canon]. (Nordiskt Nätverk för Editionsfilologer. Skrifter. 4). Stockholm: Svenska Vitterhetssamfundet. 42-58.

International Federation of Library Associations (IFLA) (1998). Functional Requirements for Bibliographic Records : Final Report / IFLA Study Group on the Functional Requirements for Bibliographic Records / [International Federation of Library Associations and Institutions. IFLA Universal Bibliographic Control and International MARC Programme, Deutsche Bibliothek, Frankfurt am Main]. München : Saur. (UBCIM publications ; N.S., Vol. 19). http://www.ifla.org/VII/s13/frbr/frbr.pdf

Levy, David M. (2000). "Where's Waldo? : Reflections on Copies and Authenticity in a Digital Environment." In: Authenticity in a Digital Environment. Washington , D.C.: Council on Library & Information Resources. 24-31. http://www.clir.org/pubs/reports/pub92/pub92.pdf

McCarty, Willard (2003). "Data modelling for the history of the book?" Humanist Discussion Group 16(26.02.2003).509. London: King's College, Centre for Computing in the Humanities. http://www.kcl.ac.uk/humanities/cch/humanist

Meyer, Heinrich (1992). Edition und Ausgabentypologie : eine Untersuchung der editionswissenschaftlichen Literatur des 20. Jahrhunderts. Bern: Lang.

Miller, Carolyn R. (1984). "Genre as Social Action." Quarterly Journal of Speech 70:151-167.

Ore, Espen (1999). "Elektronisk publisering: forskjellige utgaveformer og forholdet til grunntekst(er) og endelig(e) tekst(er)." In: L. Burman & B. Ståhle Sjönell (eds.), Vid texternas vägskäl : textkritiska uppsatser. (Nordiskt Nätverk för Editionsfilologer. Skrifter : 1). Stockholm: Svenska Vitterhetssamfundet. 138-144.

Price-Wilkin, John (1994). "Using the World-Wide Web to Deliver Complex Electronic Documents : Implications for Libraries." The Public-Access Computer Systems Review 5.3: 5-21.

Robinson, Peter (1996). "Is There a Text in These Variants?" In: R. Finneran (ed.), The Literary Text in the Digital Age. Ann Arbor: Univ. of Michigan Press. 99-115

Robinson, Peter (2002). "What Is a Critical Digital Edition?" Variants: the Journal of the European Society for Textual Scholarship 1:43-62.

Rockwell, Geoffrey (2003). "What Is Text Analysis, Really?" Literary & Linguistic Computing 18.2: 209-219.

Smiraglia, Richard P. (2001). The Nature of A "Work": Implications for the Organization of Knowledge. Lanham, MD: Scarecrow Press.

Svedjedal, Johan (2000). The Literary Web : Literature and Publishing in the Age of Digital Production. A Study in the Sociology of Literature. Stockholm: The Royal Library.

Tanselle, G. Thomas (1974). "Bibliography and Science." Studies in bibliography 27: 55-91.

Tanselle, G. Thomas (1995). "The Varieties of Scholarly Editing." In: D. C. Greetham (ed.), Scholarly Editing : a Guide to Research. New York: MLA. 9-32.

Tillett, Barbara (1991). "A Taxonomy of Bibliographic Relationships." Library Resources & Technical Services 35.2: 150-158.

Van der Weel, Adriaan (2001). "Review : Stijn Streuvels, De teleurgang van den Waterhoek. Edited by Marcel De Smedt and Edward Vanhoutte." TEXT 14. http://www.textual.org/text/reviews/vanderwe.htm

Vanhoutte, Edward (1999). "Where is the editor? : Resistance in the Creation of an Electronic Critical Edition." Human IT 3.1: 197-214. http://www.hb.se/bhs/ith/1-99/ev.htm

Vanhoutte, Edward (2003). "Display or Argument : Markup and Visualization for Electronic Scholarly Editions." In: Th. Burch, J. Fournier, K. Gärtner & A. Rapp (eds.), Standards und Methoden der Volltextdigitalisierung. Beiträge des Internationalen Kolloquiums an der Universität Trier, 8./9. oktober 2001. Stuttgart: Franz Steiner. 71-96.

Version mounted on the web in May 2004
URL: http://www.adm.hb.se/personal/mad/llc.htm