Computer-Mediated Discourse, Reception Theory, and Versioning

This paper will address how computer-mediated discourse provides new opportunities and challenges in two areas of literary criticism, Reception Theory and Versioning. Although extremely different critical modes, they can be viewed as belonging to opposite ends of the space-time continuum, with Versioning taking advantage of the computer's ability to enhance our understanding of literature through space, and Reception Theory through time.

Versioning is a relatively new development in the area of textual criticism. Since the end of the Second World War, the basic theory under which most textual critics operated was to provide readers with a text that most closely mirrored authorial intention. This philosophy of editing produced texts which, by and large, never existed in the author's lifetime. They were eclectic texts: the editor, armed with his intimate knowledge of the author and the text, assumed the role of author-surrogate to create a text which mirrored final authorial intention. To do this the editor swept away corruption which had entered the text through the publication process by well-meaning editors, compositors, wives, heirs, etc. He also swept away any ambiguity left by the author herself. Thus, in the case of narrative, choosing a bit here from the copy text, a bit there from the first English edition, a sentence there from the second American edition, and a few lines from the original manuscript, that elusive but canonical authorial intention could be restored. In the case of poetry, editors were forced to choose one published version of a text over another, and substantively ignored authorial ambiguity, such as Emily Dickinson's, who left in many of her poems alternative readings of certain words.

By the mid-1980s, the monolithic approach to textual editing began to lose favour with a new generation of textual critics, such as Jerome McGann and Peter Shillingsburg, who, responding to new critical discourses, including Reception Theory, viewed the text as a product, not of corruption, but of social interaction between several of any number of agents: author, editor, publisher, compositor, scribe, translator. It was also recognised that authorial intention was often a fluid state; particularly in the case of poetry it was possible to have several "definitive" versions of any one work which represented the wishes of the poet at a particular point of time.

One reason, I would argue, that newer theories of textual criticism took so long to be developed was that until the advent of the HTML, the World Wide Web, and the spatial freedom of the Internet, textual critics had no suitable medium to display a fluid concept of authorship. Any attempts to demonstrate anything but a monolithic text which represented final authorial intention was doomed to failure. As early as 1968 William H. Gillman, et al. undertook what was to be a definitive edition of Emerson's Journals and Notebooks in six volumes. Lewis Mumford writing in the New York Review of Books put paid to this editorial method with his review article entitled "Emerson Behind Barbed Wire":

The cost of this scholarly donation is painfully dear, even if one puts aside the price in dollars of this heavy make-weight of unreadable print. For the editors have chosen to satisfy their standard of exactitude in transcription by a process of ruthless typographic mutilation.

In 1984 Hans Walter Gabler's Synoptic edition of Ulysses encountered the same resistance from both the editing community and the Joyceans. These early efforts at representing the fluidity of authorship were doomed to failure because of the two-dimensionality of the printed text. 'Reading' such texts as Gillman's and Gabler's became impossible for all the arrows, dashes, crosses, single underlining, double underlining, footnotes, endnotes and asterisks. I would thus argue that the theoretical stance to present anything but a monolithic text representing someone's final intention (which was more likely than not the editor's) was a product as much of the medium as of concurrent theoretical modes, eg New Criticism.

Hypertext has the spatial richness to overcome the limitations of the book's two-dimensionality to present, not only works in progress, but the richness and ambiguity of authorial intention. No longer do editors have to choose between the three Marianne Moore poems entitled "Poetry", but all three can be accommodated within the hypertextual archive. Furthermore, all of Moore's revisions can also be displayed. Depending on the skill, expertise and needs of the user, a single monolithic "Poetry" can be displayed (possibly for a secondary school class), two versions, or even all three versions could be viewed (for a Freshman poetry course), or all three versions and several of the manuscript drafts (for a graduate-level course on research methods). By utilising a markup language such as SGML, all these texts could be encoded to create an "Ur" version of the poem in which lines across versions can be displayed and compared.

No longer does space and the cost of publishing images in traditional formats have to guide edited editions. Facsimile versions which were only produced for the most canonical of authors now can be produced and "published" at extremely low cost. Furthermore, languages such as SGML or XML facilitate the linking of image and text files, annotation, notes, links to other relevant texts, and so on. As with my previous example, all or some of this apparatus can be turned on depending on the needs of the audience.

The needs of the audience brings me to my second theoretical mode than can find richer expression in digitisation, Reception Theory. Reception Theory seeks to provide present-day readers with a snapshot of a text's history across time, and how previous generations of reader-response have gone into shaping our conception of the text. It also seeks to make available to present-day audiences the historical, psychological, social and/or semantic codes of the text as it was received at some point in the past. If, meaning of a work is created in the interaction between text and reader, hypermedia has the potential to create a three-legged stool in which present-day readers overhear the dialogue created between past readers and the text.

As with my previous example, no longer does the cost associated with the production of printed material have to be the prime consideration in constructing a reception theory text. As it is now economically feasible to produce facsimile archives for un-canonical authors, reception theory archives can be constructed across time providing the reader access to objects that would have been unthinkable only a generation ago. For example, take the construction of a reception theory archive of W. B. Yeats's poetry. No longer does the author of the text have to content herself with providing one or two black and white reproductions of the first Cuala Press editions of Yeats's early poetry to demonstrate the semantic codes embedded in those early editions. In a digital edition, it is possible to include full colour shots of these texts, in addition to the Macmillan editions, which, as a standard trade publication, are stripped of those codes. Furthermore, the early reception of these poems was no doubt influenced by other objects of the Arts and Crafts movement in Ireland and England. These objects, prints, paintings, even wallpaper, could be digitised to create a lexia of non-textual meaning. As with my previous example on Versioning, it is possible to conceive a digital archive that could serve many audiences, with features turned on or off as necessary.

Hypertext can provide a vehicle for accessing the ways in which readers realised the aesthetic interpretation of the text if those readers left a record of that experience; that trail can be textual: critical (reviews, articles, critical texts), personal (letters, diaries), creative (the re-writing of a creative work). It can also involve other media, a painting or ballet based on a poem, myth or folktale. These acts of interpretation can be presented to present day users in much the same format as we are used to in a two-dimensional article, i.e., a block of text with hyperlinks (rather than footnotes) to relevant primary sources.

On the other hand, we have not yet realised a form of critical discourse which does not mirror the two-dimensional spaces we have used for the last 500 years to express ideas. Hypertext criticism in future will, no doubt, embody a new form for critical discourse which is shaped by the new medium. Criticism which takes advantage of the three-dimensional space of the computer is so new that the various paradigms of editorial/authorial intervention have not been fully realised, no less understood by both the creators and users. A case in point is that we still do not have appropriate language to describe these new objects of computer-mediated critical discourse: terms like "article", "book", "collection of essays", "review" etc. have meaning appropriate to the printed word, but possibly not to the digitised one.

And indeed there are costs associated with this new medium which will govern, to a large extent, the scope and content of archives: copyright costs, the production of machine-readable versions of texts, the cost of digitising images, hardware, software, etc. One's ideal archive is governed by a balancing of costs, and tradeoffs, such as deeper encoding vs encoding of a greater number of texts, will play a significant part in the creation of resources.

In addition, the searching, retrieval and display of objects in a digital archive will, by default, reflect the bias of both the editor/author and the system designer. Yet, unlike critical theory presented in a two-dimensional space, much of the bias will be invisible to the user, as it will be buried, for example, in SGML or XML encoding. Thus there will be new challenges in learning to "read" this new model of literary discourse, which will, in turn, no doubt, foster new theoretical modes.