|[CHWP Titles]||[CaSTA 02-03]|
email@example.com |||| www.engl.uvic.ca/Faculty/MBHomePage/ |||| About the Author
CHWP A.29, publ. August 2005. © Editors of CHWP 2005. [Jointly published with TEXT Technology, forthcoming, McMaster University.]
[Abstract / Résumé]
KEYWORDS / MOTS-CLÉS: TAPoR, Shakespeare, hypertext edition, text analysis software, interface design; TAPoR (portail pour l'analyse de texte par ordinateur), Shakespeare, édition hypertexte, logiciels d'analyse de texte, conception d'interface.
|2. Searching and the dynamic text|
|2.2 Textual Analysis|
|Bibliography and Works Cited|
Master Ford is in disguise as Master Brook (a name he chose punningly for the stream a ford crosses). Brook/Ford has just been told by Falstaff that he will arrange an assignation for him with Mistress Ford. Thus Falstaff is unknowingly arranging for Ford to cuckold himself. When Falstaff leaves, Ford wonders whether he is actually experiencing this strange situation. In the Quarto of 1602, he asks himself:
For. Is this a dreame? Is it a vision?
(Merry Wives of Windsor , 3.5.122, Quarto text, 1602)
In the fuller version published in the First Folio of 1623, Ford mutters, sputters, or explodes, depending on the actor’s fancy:
Ford. Hum: ha? Is this a vision? Is this a dreame?
(Folio text, 1623)
The difference here is not especially significant in terms of meaning — there are many more teasing and profound variants in the play than this — but it is curious precisely because it seems so trivial: the addition of some huffing syllables, and the inversion of the order of vision and dream, dream and vision.
Variants of this kind are a wonderful opportunity for the electronic medium, where all kinds of possibilities could be explored: users could request the highlighting of passages where there are significant variants, invoke mouseovers of the variants, or — as I have argued in other forums — in some instances where modernization destroys ambiguities in the original, to view an animation that visibly performs the various possibilities inherent in the originals.
At a recent meeting of the Canadian Consortium for Computing in the Humanities/Consortium pour ordinateurs en sciences humaines, Ray Siemens pointed out the way that the advent of the Web has led researchers to focus on the potential of the hypertext edition, with the result that there has been a loss of interest in what he, taking a term used by Ian Lancashire, calls the "dynamic text" (Lancashire 1989; Siemens 1998, 1999). The dynamic text is one where links are created interactively by the reader of the text, rather than being created by an editor or author.
Now, while it is true that the first and primary aim of the ISE is to create hypertexts, in a form native to the medium of the Internet, it is our expectation that current technology will make it possible to envision a text that combines the fixed links of hypertext with dynamic links created by the reader. In this paper I adopt a more optimistic reading of dreams and visions than the irascible Master Ford: I will discuss progress on the design of the ISE’s hypertext editions, and will move on to discuss a vision of a kind of text that will be both hyper and dynamic.
As recently as a decade ago, hypertext was still much discussed at a theoretical level. George Landow (1991, 1992, 1994), J. David Bolter (1991), and others, wrote at length of the potential of the computer to link text and to create new structures for data. Pre-eminent in their claims was a belief that hypertext permitted a new way of organizing information that closely reflected current debates in critical theory. The unprecedented expansion of the Internet has rather overtaken this theoretical debate, as the concepts of linked data and lateral, networked units of meaning, have become commonplace—even banal—and as a new generation of students (and some scholars) reaches for a mouse as readily as for a pen. While some of the expectations of the theorists have been realized, especially in the way that those accustomed to the Internet move comfortably through a lateral rather than linear logical structure, the medium has developed in unanticipated directions. Two areas are of particular interest for any project that plans to make use of the new medium: the need to establish clarity in navigational conventions, and the increasing reliance of many users on search engines.
As anyone who uses the Internet regularly will have discovered, there is a bewildering variety of navigational signals on Internet sites. One fertile area for research in the development of fully hypertext editions of Shakespeare’s plays will be to test intuitive ways of signalling the differing kinds of links that are now possible. The range of interrelated materials available in a full scholarly electronic edition of a Shakespeare play is daunting, and the potential for losing one’s way in the maze of documents is considerable. For this reason, we have structured the editions carefully:
At the core of the text is the modern-spelling version — what is called in some e-texts the “finder” text. From it readers access either the original texts (and in a three-dimensional diagram these would be behind the modern text), or the commentary and scholarly apparatus. Readers access other materials through the commentary, so that they are always aware of their location in the finder text. This simplified diagram does not show the extensive cross-linking of the many sections of the site.
All texts are densely and intelligently encoded. The initial tagset is one developed specifically for the ISE; the text is then converted to standard XML. The advantage of the ISE tagset is that it is simple for non-experts to understand, and cheerfully allows for the important dual hierarchies of the transcriptions of the early texts as first published: the structures of the physical book (gathering, page, column) and the conceptual structures of the text added by later editors (act, scene, line). The modern text is relatively straightforward, and follows basic TEI guidelines closely. The early texts include detailed tagging to indicate the use of Early Modern type-forms: the long-s, and the many ligatures compositors used. We can now display the transcriptions of the Early Modern texts in an attractive recreation of the originals, though not all ligatures are yet enshrined in Unicode, so there are some limitations where the display cannot completely render the encoded text. Thanks to the generosity of the State Library of New South Wales and the British Library, we are able to link these texts to graphic facsimiles of Folios 1-4 and all important early quartos.
The Editorial Guidelines for the ISE require editors to distinguish between three levels of annotation, and one of the ways that we will be studying the hypertext edition will be to discover the most effective way of implementing this hierarchy of annotation.
Level 1: glossorial explanations or very short notes.
These might be invoked by a simple mouseover, with the word or phrase signalled by an unobtrusive underline: there are now a variety of attractive ways of creating pop-ups for mouseover actions.
Question: does one need the underline, or can we assume that users would mouse over words they are curious about to see if there is a note?
Level 2: longer annotations.
These annotations will be of the level of detail one finds in advanced scholarly editions like the Arden Shakespeare, where the editor explains a crux, illustrates a word usage by citing parallel passages from elsewhere in Shakespeare or other contemporary writers, comments on performance issues, or explores critical debates of various kinds. Notes at this level may contain links to further information: to any original texts as transcribed on the site, in order to make clear a textual matter; to full citations of the parallel passages quoted (often branching to other sites on the Web to do so); they might link to images of performance; or in some cases to critical articles (though good recent criticism is scarce on the Web for copyright reasons). Level 2 annotations could be signalled by an icon in the margin opposite the passage to be annotated, though this is easier in verse than prose, where lineation is fluid. In general, our aim is to keep the text itself as uncluttered as possible, avoiding multiple ways of underlining or modifying the font. When the annotation is invoked, a pop-up window might appear, or the text of the note could appear in a separate section of the page. Again, in the display of notes and commentary there is an opportunity to experiment with the interface and to test different strategies.
Level 3: extensive discussions of issues that require more detailed treatment.
These notes will deal with the kind of material that might appear as an appendix in a standard print edition, and would deal with major editorial issues. Level 3 notes will be linked from level 2 notes, and would open in a new window altogether.
In addition, we will develop navigational signals that will permit users of the site to branch to a given passage in early printed versions, either as transcribed, or as a graphic facsimile; in the more complex texts, there will be three different versions. Hamlet, for example, exists in two very different quartos as well as the Folio text, which is again substantially different: thus there are a total of six texts to link to, since each is available both as a transcription and as a scan of an original.
The improved effectiveness of major search engines on the Internet has radically changed the way that many users access sites. Last month, over 50% of visitors to the ISE came directly from a search engine; most came from Google.com (25%) but there were also many visitors from Yahoo (11%), Ask Jeeves, MSN and AOL networks, and the Google sites in Canada, the UK, Australia, Germany, New Zealand, Italy, France and Brazil. The total number of “hits” last month, by the way, was over two million; the number of requests for pages (a more useful figure) was just under half a million. This reliance on search engines can be problematic, since their algorithms emphasize the popularity of the sites they index rather than their quality. The principal usefulness of search engines, however, is that the links generated by them are interactively created by the user rather than pre-selected by the creator of a site. An additional aim of the ISE is to take advantage of the familiarity of users with search engines to create a dynamic text alongside, and integrated with, the hypertext edition. At present, I envision two separate areas where users will be able to explore and create their own pathways through the data collected for the editions: by interacting with a growing database of Shakespeare in performance, and by invoking tools that will allow sophisticated computer-aided analysis of the text.
One of the most effective ways of using the electronic medium in the presentation of a dramatic text is to include multimedia data. As I have suggested, the editor can create a hard link between specific passages and selected illustrations, extending the current practice in print editions of including a stage history of the play. But the medium invites a significant extension of this practice, with its capacity to archive large amounts of graphic, sound, and even video materials, and to store them in a relational database structure that will allow for multiple, user-generated, paths through the archive. Through an extensive performance database, the ISE will provide a research resource for scholars, students, actors, and all those interested in performance and performance criticism.
Initial work on the database can be seen on the ISE website. With the co-operation of Kenneth Rothwell, author of the definitive work on Shakespeare on film (Rothwell 1990), the ISE has developed a database of Shakespeare films, subsequently updated by Tanya Gough; we have devised guidelines for the acquisition of performance materials (ISE, “Guidelines” 2001) and have posted sample materials of the kind to be collected using a local production of Romeo and Juliet (Hardy 1998). In preparation for collection of data for the Performance Database, the ISE has established a close working relationship with the Shakespeare Theatre Association of America (STAA), an organization that boasts a membership of over a hundred companies and festivals that focus on the performance of Shakespeare’s plays. We are initially working with companies in Western Canada: Bard on the Beach (Vancouver), Shakespeare on the Saskatchewan (Saskatoon), and Free Will Players (Edmonton), but many other companies we have contacted through STAA have archives that they are willing to have digitised and added to the database.
The development of a flexible data model for the relational database that will hold the performance materials is under way; there will also be a need to develop Web interfaces both for accessing the data and for inputting it. The kinds of search screens that will be most comfortable and intuitive for users will have to be tested.
A second area where users would ideally be able to generate their own paths is in textual analysis. Here the ISE is especially fortunate, since the University of Victoria is one of the consortium of six universities in Canada to receive major funding dedicated to the establishment a Text Analysis Portal for Research (TAPoR). ISE texts have been donated to the portal to provide an example of intelligently tagged text for the various varieties of software to work on.
One way this might work would be for a user to select a word or phrase in the hypertext edition, then drag it to a text box within the workspace where there would be options that would allow him or her to view the word as used elsewhere in the play, or in the Shakespeare corpus as a whole (a concordancing function), or in related Renaissance texts — Shakespeare’s sources, the Bible, the Elizabethan Homilies (Lancashire 1994, 1997), for example — all in KWIC (Key Word In Context) view. Or the user might look up the word in such reference works as the Early Modern English Dictionaries Database (Lancashire 1999) or other freely available online reference works. More sophisticated explorations will also be possible, as words, phrases, and collocations can be traced through frequency distributions, through lemmatised versions of the text, and so on. A parallel project being undertaken by a co-investigator, Ray Siemens, is the development of a dynamic edition of the Devonshire MS, where all links will be created automatically by algorithms rather than as fixed links chosen by the editor. He will integrate into this edition the text analysis tools being developed at the University of Alberta by Stéfan Sinclair, HyperPo; through a similar collaboration it will be possible to make the same kinds of tools available for dynamic readings of ISE texts.
The big question is how to get the dynamic text to work with the hypertext. What I envision (or dream of) is a kind of workspace that works within a standard Web browser, and that brings the three components I have been discussing together: the hypertext, access to performance materials, and tools for textual analysis. A possible layout of the page would use up to four separate areas of the browser. On the extreme left, as a general evolution seems to be insisting, would be general navigational buttons standard throughout the ISE site; next — reading from left to right — would come the hypertext edition itself, with line numbers (as in a print edition), text underlined in grey for pop-up glossaries, icons to indicate further notes, and simple buttons at the top to link to the equivalent passages in the relevant Early Modern texts; there might be a further choice to view parallel texts in the same window. As I suggested at the beginning of this paper, an advanced view of the text might also allow the user to highlight words that are taken from a text other than the copy text, or to toggle between views that include visible variants. On the right of the screen would be the two work areas for querying the text or the performance database, one at the top of the page, the other at the bottom. In each case the options presented would be basic, and would be arrived at after some testing with users, with a button for more advanced searches (see Best 2003 for a link to a mock-up of this structure).
By extending the traditional linear text by hypertext links designed by the editors, and further adding two kinds of interaction where links are forged automatically by the computer in response to queries generated by the researcher, a project of this kind will create a research resource that will stimulate fresh ways of looking at the texts. I like to think of the print texts we grew up on as the tips of icebergs of meaning, which the computer is now allowing us to explore in greater depth. An article by Jonathan Hope and Michael Whitmore (Hope 2004) shows how a tool developed for one textual environment—to assist in undergraduate composition classes (Collins and Kaufer 2001)—can produce unexpected and illuminating results when applied in a different context—Shakespeare’s plays. Here then my vision for the ISE: editions that integrate performance materials and tools for textual analysis into a sophisticated and scholarly hypertext edition. As the editions become more complete, they will become a powerful workspace for the user to re-experience the plays in ways not currently anticipated: the capacity of the computer to index, categorize, and analyze large quantities of data will provide the opportunity for serendipitous discovery as well as more conventional exploration.
Arden Shakespeare CD-ROM: Texts and Sources for Shakespeare Studies. (1997). Thomas Nelson.
ArdenOnline. Web site no longer available. The Arden Shakespeare site is found at <http://www.ardenshakespeare.com/>.
Best, Michael. (2002). “The Text of Performance and the Performance of Text in the Electronic Edition.” Computing in the Humanities 2002 36:3 269-282. [Uses Romeo and Juliet as an illustration of the value of electronic performance materials in criticism.]
-----. (2003) “Review of Christie Carson and Jacky Bratton, eds. The Cambridge King Lear CD-ROM: Text and Performance Archive.” EMLS 9.1 (May): 10.1-17 <http://purl.oclc.org/emls/09-1/bestrev.html>.
-----. (2003) A Shakespeare Suite. CD-ROM. Internet Shakespeare Editions, University of Victoria. <http://ise.uvic.ca/Annex/ShakespeareSuite/index.html>.
-----. (2003) Mock-up of page that includes tools for browsing performance and textual analysis tools. <http://ise.uvic.ca/Annex/Articles/mla/p6.html>.
-----. (forthcoming). “Standing in Rich Place: Electrifying the Multiple-Text Edition: Or, Every Text is Multiple.” Paper for the Shakespeare Association of America, Minnesota, 2002. Forthcoming in College Literature. Preprint: <http://web.uvic.ca/shakespeare/Annex/Articles/SAA2002/index.html>.
Bolter, Jay David. (1991). Writing Space: the Computer, Hypertext, and the History of Writing. Hillsdale, NJ: Lawrence Erlbaum Associates.
Bradley, John. (n.d.), TactWeb. August 2003. <http://tactweb.humanities.mcmaster.ca/tactweb/doc/tact.htm>.
Carson, Christie, and Jacky Bratton. (2000). The Cambridge King Lear CD-ROM: Text and Performance Archive. Cambridge, Cambridge University Press. Copyright is held by Royal Holloway, University of London.
Carson, Christie. (2003). “Designing Shakespeare.” 5 December 2004. <http://performingarts.ahds.ac.uk/designing-shakespeare>.
Collins, Jeff and Dave Kaufer (2001). DocuScope. Carnegie Mellon University. <http://www.betterwriting.net/projects/fed01/dsc_fed01.html>.
Dublin Core. (1995-2002). Metadata Initiative. <http://dublincore.org/>.
Faulhaber, Charles. (1997). “Guidelines for Electronic Scholarly Editions.” Modern Language Association of America, Committee on Scholarly Editing. 30 January 2002. <http://sunsite.berkeley.edu/MLA/guidelines.html>.
Finn, Patrick. “Reforming the Information Age: Formalism and Philology on the Net.” Mots Pluriels. October 2001. <http://www.arts.uwa.edu.au/MotsPluriels/MP1901pf.html>.
Google. Google: San Francisco, CA, 2001. <http://www.google.com>.
Hardy, Linda. (1998) Production of Romeo and Juliet, Phoenix Theatre, University of Victoria. <http://web.uvic.ca/shakespeare/Theater/Rom/UVic1998/>.
Hope, Jonathan, and Michael Whitmore. (2004). “The Very Large Textual Object: A Prosthetic Reading of Shakespeare.” Early Modern Literary Studies 9.3 / Special Issue 12 (January, 2004): 6.1-36 <http://purl.oclc.org/emls/09-3/hopewhit.htm>.
Images Canada. August 2003. <http://www.imagescanada.ca/index-e.html>.
Internet Movie Database. (2003) August 2003. <http://www.imdb.com/>.
ISE (Internet Shakespeare Editions). (1996-2003). “Foyer.” <http://web.uvic.ca/shakespeare/Foyer/>
-----. (1996-2003). “Library.” <http://web.uvic.ca/shakespeare/>
-----. (1996-2003). “Shakespeare’s Texts in Original Spelling.” (1996-2003). <http://web.uvic.ca/shakespeare/Annex/DraftTxt/index.html>.
-----. (2001). “Guidelines for the Acquisition and Copyright of Performance Materials.”
-----. (2004). “Shakespeare on Screen.” Eds. Kenneth Rothwell et al. <http://web.uvic.ca/shakespeare/Library/SonF/index.html>.
Lancashire, Ian, ed. (1994, 1997). Elizabethan Homilies 1623. RET Editions 1.2. <http://www.library.utoronto.ca/utel/ret/homilies/elizhom.html>.
-----, ed (1999). Early Modern English Dictionaries Database (EMEDD). Toronto: University of Toronto, 1999. <http://www.chass.utoronto.ca/english/emed/emedd.html >.
-----. (1989) “The Dynamic Text: ALLC/ICCH Conference,” Literary and Linguistic Computing 4.1: 43-50.
Landow, George P. (1992). Hypertext: The Convergence of Contemporary Critical Theory and Technology. Baltimore: Johns Hopkins University Press.
-----. (1994). Hyper / Text / Theory. Johns Hopkins UP: Baltimore, MA.
-----, and Paul Delany, eds. (1991) Hypermedia and Literary Studies. Cambridge, MA: MIT Press.
Levenson, Jill L. (2001). “Romeo and Juliet Prompt-books.” Centre for Reformation and Renaissance Studies, University of Toronto. <http://www.library.utoronto.ca/crrs/Rommain.html>.
MLA. (1997). Modern Language Association of America, Committee on Scholarly Editions “Guidelines for Electronic Scholarly Editions.” <http://sunsite.berkeley.edu/MLA/guidelines.html >.
Rothwell, Kenneth. (1990). Shakespeare on Screen: An International Filmography and Videography. New York: Neal Schuman.
-----. (2003). Shakespeare on Screen: An International Filmography and Videography. Internet Shakespeare Editions, University of Victoria. <http://web.uvic.ca/shakespeare/Library/SonF/>.
Siemens, Raymond. (1998) “Disparate Structures, Electronic and Otherwise: Conceptions of Textual Organisation in the Electronic Medium, with Reference to Editions of Shakespeare and the Internet.” The Internet Shakespeare: Opportunities in a New Medium. Early Modern Literary Studies Ed. Michael Best. 3.3 / Special Issue 2 <http://purl.oclc.org/emls/03-3/siemshak.html>.
-----. (1999) “Shakespearean Apparatus? Explicit Textual Structures and the Implicit Navigation of Accumulated Knowledge.” Text: An Interdisciplinary Annual of Textual Studies 14. Ann Arbor: U Michigan P, 2002. 209-240. Electronic pre-print published in Surfaces 8: 106.1-34 <http://www.pum.umontreal.ca/revues/surfaces/vol8/siemens.pdf>.
Sinclair, Stéfan. (2003). HyperPo. Computer program for textual analysis. <http://tapor.ualberta.ca/HyperPo/>.
STAA [Shakespeare Theatre Association of America]. (2002). Home Page. Internet Shakespeare Editions, University of Victoria. <http://www.uvic.ca/shakespeare/STAA/>.
Text Encoding Initiative. (2001). <http://www.tei-c.org/>.