Afterword: Dressing Old Words New
Michael Best
University of Victoria

Best, Michael. "Afterword: Dressing Old Words New." Early Modern Literary Studies 3.3 / Special Issue 2 (January, 1998): 7.1-27 <URL:

The contents of this paper:

Old and New

  1. The Internet enables access to an audience of remarkable diversity, and the opportunity to provide that audience with an equally diverse range of materials. Anne Lancashire's poll of the responses of Shakespeare scholars to an Internet Edition is revealing in two ways: the number of replies she received indicates that there is a real interest in the new way of accessing information; at the same time, however, the resources that scholars expect, or would like to see, are fundamentally the same as those they would like to find in a printed text, with the addition that the electronic space would permit greater comprehensiveness. It is clear from the respondents that they expect to use an electronic edition as an adjunct to a regular printed edition rather than a substitute for it.

  2. Any new venture in a field with as long a tradition of scholarship as Shakespeare studies will have to be seen as credible in terms of that tradition. Editors for the Internet Shakespeare Editions (ISE) will have to be at least as thorough in their editorial practices as the editors for the Arden Shakespeare and its peers; since the electronic editions of Shakespeare on the Internet thus far are inaccurate and outdated, and since the medium allows anyone with access to a server to put anything on line and call it an edition, the ISE will have to set up visibly rigorous scholarly standards if the editions are to receive wide scholarly acceptance. Thus one crucial component of editions in a new medium will be their fealty to the older medium and its standards.

  3. At the same time, an Internet edition will be a disappointment if all it offers is more of the same. Ray Siemens, Don Foster, and Ian Lancashire point to ways that an electronic edition can extend the text more than by simply providing what we expect from a print edition. In the discussion that follows I will examine both the old and the new: the process by which an Internet edition can preserve traditional editorial approaches, and the newer capacities that the electronic medium makes available to a variety of readers. Since this essay itself is appearing in the electronic form, I make a gesture towards the growing conventions of the medium by including headings of the kind that are found in most documents posted on the Internet.

    The Electronic Edition

  4. In the print medium, the word "edition" implies a fairly consistent set of expectations: the text will be responsive to the earlier history of publication of the work, and will provide some kind of scholarly framework, from the minimum of a preface explaining the provenance of the text, to the full scholarly apparatus of introduction and notes. The electronic edition at the moment, however, covers a far wider range of possibilities. At one end will be an erratically proofread, scanned text from a doubtful and outdated source, chosen because it is out of copyright; Don Foster comments that even such a leader in the field as Chadwyck-Healey seems willing to publish seriously flawed electronic texts.[1] At the other end of the range of electronic texts will be the kind of edition offered by the Cambridge University Press project on The Canterbury Tales, a CD ROM with deeply encoded text, full scholarly apparatus, and digitized images of the original manuscripts.[2]

  5. The ISE aims for something rather more modest than The Canterbury Tales project, partly because the medium of the Internet, unlike that of the CD ROM, has greater limitations both in the vexed area of copyright, and in the amount of data that a user is willing to wait for. The Internet is also a more rapidly evolving and less predictable medium than a CD ROM, where the editor/designer has more complete control over the software that is used to interpret and display the data.

    Stability in an Evolving Medium

  6. One of the major challenges of any Internet scholarly edition will be to establish consistency in a rapidly changing medium, and credibility in the scholarly world, where the experience of the uneven quality of those texts that are on line has justifiably made scholars cautious. The academic structure of the Internet Shakespeare Editions follows that of on-line academic journals like Renaissance Forum and EMLS, which in turn follow the practices of established print journals. The ISE has an Editorial Board, members of which are drawn from a range of interests in the profession; its mandate is to ensure that all materials posted on the site are of high quality, and have undergone adequate peer review.

  7. In one crucial way, however, the ISE site differs from even an electronic journal in its structure. An article once published, in print or in an electronic journal, will not be changed; it is a fixed artifact, and its readers know that in its published and final form it has been peer reviewed. On the ISE site, however, publication of texts is incremental: parts of editions can be posted as they are completed, and in order to make the most of the electronic medium of the Internet, the edition can be maintained in a way impossible in a fixed medium -- even a CD ROM. After initial publication, editors can add items to the bibliography; as more resources become available on the Internet, both within the ISE site and beyond, additional links can be added; as scholarship advances on other texts in the series, and in the burgeoning number of fine editions currently appearing in print or CD ROM, editors can add additional references and comments to keep the edition fully up to date.

    Maintaining the Published Text

  8. In theory, this kind of "maintained" text can be seen as a significant improvement on the fixed text we are accustomed to. The problem is that such a text may be seen as in some sense insecure; a normal text is finished, peer reviewed, and date-stamped, with all its potential incompleteness and imperfection. The Internet text would appear to have no such assurance for the reader. Part of the answer is to be found in the draft document prepared for the MLA Committee On Scholarly Editions, Guidelines for Electronic Scholarly Editions, and followed by such pioneers as the Renaissance Electronic Texts, and the electronic texts being developed at the University of Virginia, the University of Alberta, and elsewhere: each text includes as an integral part a summary of all revisions and emendations. The scholar is thus able to see in the one text its layers of revision, in much the same way as the title-page of a book which provides a history of its various editions -- though the statement of revision in the electronic text will in fact be far more detailed than will be found in a book.[3]

  9. The maintenance of a text can be made both straightforward and accountable if the kinds of change are categorized carefully. Some can be seen as automatic, or trivial; some will require further peer review before being posted on the site. An example of automatic revision would be the inclusion of internal cross-references and links to a new play as it comes on line; trivial changes would be of the order of correcting proof errors, or adding a recent book or article (print or electronic) to the annotated bibliography. An editor who wished to make substantive changes either to the text or to its supporting materials would be expected to go through the standard process of peer review before the changes were made.

  10. If the scholar accessing the text can thus be assured of its integrity, there remains a related problem for the scholar who is responsible for the maintenance of the text: that of gaining institutional credit for the work the maintenance requires. This is part of a larger question of the general credibility of scholarship in the electronic medium, and here again the MLA provides guidance. The approval processes of the ISE conform with those recommended in the MLA's Guidelines for Evaluating Computer-Related Work in the Modern Languages; in cases where maintenance of the text requires additional peer review, the materials submitted should be considered at least the equivalent of a refereed article; more minor adjustments are likely to be the result of the normal activity of a scholar who wishes to remain current with research in the field.

  11. An Internet edition has the potential for providing another, important way in which a scholar can be assured of stability in the edited text: like a CD ROM, the edition can include digitized versions of the original quarto and folio texts. The Internet Shakespeare Editions plan to provide both encoded transcriptions of the early texts, and graphic images of them. The only caveat here is the willingness of libraries to make their images available over the Internet; thus far the Furness Library at the University of Pennsylvania has led the way by putting high quality images of King Lear and Hamlet from the First Folio on the Internet.[4] The Folger Shakespeare Library has recently agreed to a procedure that will allow some of its images to be used on the ISE site. The advantage of encoded transcriptions is that they will be fully searchable in a number of ways: as well as a normal word search, the encoding of the texts will make it possible to search for specific items like speech prefixes, signatures, stage directions, and so on. The usefulness of this kind of text is demonstrated by the work of Don Foster in this collection of essays, and Ian Lancashire elsewhere.[5]

    Avoiding Obsolescence

  12. While the technology of print has evolved significantly in the last hundred years, the actual appearance of books, and the conventions with associated print editions, have changed very little. As I look at an Arden 3 edition beside an original Arden, I notice that the recent edition has a more modern typeface (and its paper is whiter beside the venerable pages of the earlier book) but the organization of the book, and the physical layout of the page have not significantly changed. The electronic medium, in contrast, is in a state of flux. Web browsers are updated almost every year, each iteration bringing more sophisticated programming tools to designers; the formats for digitized graphics, sound, and video are still evolving; the speed of networks is increasing; and the computers we use to view the data are still developing with a speed that is either exciting or irritating, depending on one's geek quotient and bank balance. The challenge in the design of a site that offers multimedia scholarship is to ensure that the data do not become out of date as the medium evolves.

  13. The best defence against obsolescence in the text itself is the use of a system of encoding text that is independent of computer platform or software. The native language of the Internet, HTML (Hypertext Markup Language) is one example: thus, for example, text is marked <I>italic</I> and it is left to the browser and the machine to figure out how to display italic text. The ISE is using a fuller tag set that defines more than the simple display tags of HTML; information concerning textual variants, special characters, and some functions of the language (stage direction, speech prefix, etc.) are tagged as well.[6] In order to display the text on a browser, the richer tag set of the ISE is translated into HTML; thus as the medium evolves, the basic text can remain unchanged, and the process of translation modified to take advantage of new features as they become available.

    Enjoying the New-fangled

  14. The computer has changed the way most of us work -- certainly those who read EMLS will be using computers for word processing and e-mail, and perhaps for some research activities. But the responses garnered by Anne Lancashire suggest clearly that scholars see the electronic text primarily as an opportunity to improve the speed and convenience of performing the tasks they are accustomed to, rather than as a means of doing something different.[7] Ian Lancashire points out that the Internet makes the text available to a far wider audience, one that may be more interested in simple explanation, or the nature of Shakespeare as a person rather than as a writer, and suggests that the ISE texts can provide this audience both with the materials it wants, and the opportunity to go beyond them. Ray Siemens and Don Foster point to the potential of new kinds of approaches to Shakespeare, involving techniques that could perhaps be accomplished with standard glossaries and concordances, but which would require the expenditure of so many hours that they would certainly not become a convenient way of reading the texts.

  15. Only time, and the provision of new textual tools, will tell whether the computer will actually change our experience of the Shakespeare texts. For a surprisingly long time even in the sciences, the computer was simply an adjunct to established methods of research. The recent explosion of the fields of the various branches of fractal mathematics ("chaos" theory), however, demonstrate that the computer has done more than speed up existing research; it has made possible a fundamentally different way of perceiving the universe around us. In an entertaining paradox, fractal mathematicians are claiming an astonishing universality ("Why earthquakes are like businesses are like magnets are like flocks of birds are like bacteria are like traffic jams are like ants..."[8]) while those in the Humanities are busy demolishing long-cherished concepts of universality in literature. The point here is not that Humanist scholars are likely to find that the computer restores the concept of universality in Shakespeare's texts, but that the physics of universality in critical systems could not have been explored -- or even imagined -- without the computer's number-crunching power. The experience in the sciences warns us that we may have no idea at all of the kinds of research or teaching that will be carried out with the use of electronic texts twenty years from now.

  16. While we wait for the word-crunchers to reveal new ways of reading Shakespeare, there are some important ways that an Internet edition can exploit the specific milieu of the Internet to its advantage. Internet users are very different from the readers of books. They like to chat, they don't like to read much on screen, and they don't like to feel lost in a maze of links.

    Informal Discussion

  17. The admirable urge of Internet browsers to chat and to contribute to a forum can be harnessed without compromising the principles of peer review and the integrity of the text: the key is to label the sections of the site clearly. The ISE uses the metaphor of a traditional library: the "Foyer" offers discussions of principles and the guidelines which the editions follow; the "Library" itself will contain only refereed materials; and there is an "Annex," a kind of electronic cafeteria, where those who visit the site can leave comments, and where un-refereed, but interesting, materials can be posted. At present the Annex provides twelve draft transcriptions of original Quarto and Folio texts, some articles on electronic Shakespeare, a digest of some earlier discussion about the design of the site, and a text-only "jumplist" of other sites of interest to Shakespeare scholars. As the editions are developed, the Annex will allow a fuller conversation about issues raised by individual editors. It is not intended to become a discussion group of the kind exemplified by FICINO, HUMANIST or SHAKSPER, but it will use creatively the desire of those who use the Internet to share their opinions.


  18. The cost of connect time, and the inconvenience of reading from a screen, mean that most Internetters are "hit and run" readers: they want to find something, download it, and move on or disconnect. Thus the ISE site should be easy to navigate, and provide a variety of ways of searching for specific -- or vague -- information. Basing the data structure of the site on a powerful database software from Oracle[9] will allow for a variety of kinds of searches (act.scene.line of the Riverside/Bevington/Hinman editions; occurrences of "love" and "death" within five lines; feminist approaches to Cymbeline); the software also provides the capacity for "fuzzy" searches; the fuzzy search has the potential to restore to the rigid binary world of the computer some of the serendipity we all relish as we browse the stacks of a library looking for something in particular, only to find something delightfully different and satisfying.

  19. Ian Lancashire points out that the actual audience of Internet editions range from the marginally curious general browser to the serious scholar using the power of machine-readable text for new kinds of research. The materials that can be made available range from the modern, annotated text itself, to graphic reproductions of original texts and digitized performance materials in graphic and video formats. It is one of the particular talents of the computer that it can provide an almost limitless capacity for choice, so that theoretically a site could respond to the needs of widely different users. The concomitant danger, however, is that the range of alternatives will be confusing; the wide appeal of Shakespeare makes it desirable that the ISE should provide choice for its viewers, but there will inevitably have to be careful management of navigational aids through the editions, and some degree of limitation in the range of choice available if the site is to succeed.

  20. Careful design can exploit fully the power of the computer and machine-readable text, without the process becoming daunting. Default display can be designed to satisfy the most requested formats: not every user will want to see the text of Othello with variant readings from Quarto and Folio highlighted in different colours, as in a sample page on the ISE site,[10] but the choice should be available for a user who clicks on a link to access a variety of different types of display. The encoding for the ISE texts includes the provision for the editor to include three levels of annotation: basic explanations (think Bevington or Signet), full scholarly discussion (think Arden, Oxford, or New Cambridge), and hypertextually linked further exploration of cruxes, rather like the appendixes to some of the Arden editions.

  21. As the Internet expands, and offers increasing numbers of quality sites, the ISE texts can be linked to resources outside the site itself. The challenges here are twofold: to monitor the external sites to ensure that they are stable, and to make some kind of evaluation of their quality so that the user will not be invited to link to a site that is unreliable in its scholarship -- or will be warned of its limitations in advance. Increasingly, the first problem can be monitored automatically, with "robot" programs that check all links to see if they are valid. The issue of quality is less easy to resolve; one solution might be for the editor of a play to provide a kind of Michelin rating system, indicated by colour or icon.

  22. The potentially bewildering mass of supporting materials (sources, criticism, performance records, collation, glossary, bibliography, transcripts of original texts) can be made manageable by providing a relatively uncluttered modern text as a starting point, with all links to further material to be made from there. The increasingly familiar convention of icons to signal additional resources can leave the text even of the notes straightforward; thus links to the Collation, Shakespeare's sources, and to performance materials can be signaled unobtrusively.

    Performance and Multimedia Materials

  23. The development and maintenance of an archive of performances of Shakespeare is another area where the ISE will be exploring the new-fangled. A full discussion of the process for acquiring these materials, and for their assessment is beyond this paper; it is an area, however, that promises to distinguish the ISE texts substantially from both print and CD ROM editions of the plays. Though it will take some time for the archive to be of sufficient inclusiveness to be useful, it will in due course provide raw material for critical and cultural studies of changing trends in production, and has the potential for changing the kinds of assignments we can ask of our students.

  24. The main obstacle for the ISE to overcome in this area is the Byzantine complexity of copyright law on the Internet.[11] Until the question of "fair use" has been settled, the ISE will not be posting video clips of movies; rather the site will cooperate with major festivals to develop the performance database and to update it annually. The database will include various kinds of graphic materials: director's production notes, prompt book excerpts, costume and set designs, photographs of the production, principles of casting, type of space used, video segments from the play (where available), and information about sound effects and music.[12]

    Conclusion: Getting There from Here

  25. Perhaps the greatest initial challenge to the editors and those using the Internet editions will be to exploit the capacity of hypertext to its fullest. Paul Werstine shows how a general editorial issue can be elegantly illustrated by directly linking the discussion to the examples it refers to. As scholars, we are accustomed to the discipline of linear argument; as writers of hypertext we have the opportunity to explore the multilinearity we theorize about but seldom practice except perhaps in a graduate seminar or the bar at a conference.

  26. A further challenge, as Ray Siemens suggests, will be for scholars to become "dynamic" readers of Shakespeare and the associated texts. This will involve becoming accustomed to the ways of viewing the text that the computer makes not only possible but increasingly intuitive as the Internet interface becomes more flexible, and as we become familiar with the new tools for the analysis of text that are being developed. Fifteen short years ago, word processing was an arcane activity requiring the mastery of a complex "command line" syntax which generated frequent intimidating error messages from the computer, and equally curious "dot" commands that instructed the output device on the very basic formatting that was available. Now we are accustomed to working on a screen that directly imitates the page we plan to print to, and the commands have been reduced to the simplicity of a mouse click on an icon or menu choice. In the same way, text analysis programs are beginning to move beyond the stage of requiring a resident guru to implement and design the commands needed. As tools of this kind become available for use over the Internet, they will be included as a part of the ISE site, at which time the kinds of research currently being pursued by Ian Lancashire and Don Foster will become more widely used and integrated into other kinds of research.

  27. An electronic edition has two potentially conflicting aims: to remain faithful to the well-established traditions of the discipline, and to forge new tools and ways of conceiving the text it preserves. If this sounds like a paradox, it is a familiar one. A logo for the ISE might be the familiar emblem of the dolphin and anchor, "festina lente," as we make haste slowly towards the full realization of an Internet Shakespeare.[13]

    Anchor and dolphin

    [Return to top]


[1] On the subject of accuracy in preparing electronic texts, the MLA Guidelines comment:

One very reliable method of manual transcription for printed materials is to input the same text twice, by two different people, who do not necessarily have to know the language involved, then use a collation or file compare program to find the differences. (Faulhaber, section I.C.4)

The experience of Foster, however, suggests that this method is more fallible than might be expected, since the Chadwyck-Healey texts are prepared by this technique.

The technique of double-keying employed by Chadwyck-Healey comes under scrutiny of a different -- though related -- kind from Paul Werstine, who echoes a concern voiced in discussions on HUMANIST. In a passage from his original paper, he commented:

If we are to link any great number of digitized texts of English dramatic manuscripts to digitized early printed texts, we will have to depend on the Chadwyck-Healey archive, which has been purchased by my university and many others, and in using Chadwyck-Healey, I would suggest, we sink into an ethical and textual morass. The Chadwyck-Healey texts, I've been told, came into being under despicable socio-economic conditions, the data entry, in the euphemisms of Harris-speak and Bank CEO jargon, being outsourced to offshore companies for double-keying.

Unpacked, this language means that Asian labour was paid a negligible wage under unspeakable working conditions so that we can use the results in our computers. But the method of double-keying does not produce texts with the reliability of either the scholarly edition or the facsimile. What's more, for dramatic manuscripts, Chadwyck-Healey used already mediated Malone Society reprints, some of which have themselves been shown to be textually quite unreliable, largely because the Society's first secretary refused to leave London to verify transcriptions of manuscripts housed elsewhere. It is hard to see how anything we might idealize as truth could be reared on the foundations Chadwyck-Healey offers.


[2] "The Canterbury Tales Project from Cambridge University Press will make available in an electronic form, over a ten-year period, full transcriptions of the text of every manuscript and pre-1500 printed edition of Chaucer's Canterbury Tales, together with digitized images of every page of every manuscript and early edition, collations of all these texts, and analyses of the textual tradition based on the transcriptions." Cited from http://www.cup.org/Chaucer/ctpbl urb.html. [Back]

[3] A good example of this kind of accountability is the practice of many "shareware" software programs that provide a detailed history of enhancements and bug fixes. ISE editions will maintain a list of this kind as a separate page, from which users can access the latest materials. Since CD ROM versions of selected texts will be issued by the University of Toronto Press, this page will become a reference point for those using the CD in keeping it up to date. [Back]

[4] See the University of Pennsylvania's Center for Electronic Text and Image, http://www.library.upenn.edu/etext/. Michael T. Ryan writes that the library "certainly will be doing more facsimiles from the FF" in the future. (Private e-mail of 1 December 1997.) [Back]

[5] Ray Siemens' article cites several studies by Lancashire and Foster that exemplify the kind of research that the electronic text makes possible; Foster's contribution to this collection is a further example. [Back]

[6] In establishing its tag set, the ISE is again following the recommendation of the MLA Guidelines in using SGML (Standard Generalized Markup Language) in a form compatible with the Text Encoding Initiative (TEI). See Faulhaber (ed., section I.A.B). The current development of eXtensible Markup Language (XML) from SGML may well provide more powerful means of viewing the texts and materials on the site. See also Rifkin. [Back]

[7] Though Anne Lancashire's survey was explicitly aimed at scholars, it is almost certainly true that our students will be equally conservative, despite the fact that in many cases they will be more computer-comfortable, if not more computer-literate, than current Shakespeare scholars. In two courses I teach partly on-line, I ask my students to become familiar with the research resources for Shakespeare studies that are currently available on the Internet; as a part of the assignment I ask them what materials they did not find. The list, though less formal than a survey, is revealing: they looked for study aids (plot summaries, character sketches, and so on), and for criticism of the plays. Once again, they seek the kinds of literary materials they are familiar with. It is worth noting that students give much higher priority to supporting critical materials than the scholars in Lancashire's survey. [Back]

[8] "One Law to Rule Them All," New Scientist 156:2107 (8 November, 1997): 30-35 (cited from the cover page summary of the article). [Back]

[9] "ConText provides many powerful text retrieval features. In addition to exact word searches, ConText handles: stemming, to match plurals, past tenses, and other alternate forms of words; fuzzy-match and sounds-like, to match misspelled words and other 'close' words; and synonyms. ConText also provides advanced linguistic capabilities that can actually analyze the content of a document, producing 'themes' and 'gists'. Themes are words or short phrases that represent the main concepts in a document, while a gist is a collection of the most important paragraphs in a document." ("Leveraging the Oracle ConText Option In Your Web Applications," Oracle Developer's Programme Newsletter, December/January 1997.) For general information on Oracle software, see their site at www.oracle.com. [Back]

[10] The Othello text can be viewed at http://web.uvic.ca/shakespeare/Annex/DraftTxt/Oth/. [Back]

[11] At the time of writing this article, a bill under consideration by the US Congress would bring some reasonable order to the field (see the announcement by the National Initiative for a Networked Cultural Heritage [NINCH], September 17, 1997, reported in HUMANIST Vol. 11, No. 408, November 17, 1997). A useful gateway to discussions about copyright is "Copyright and Fair Use," Stanford University, http://fairuse.stanford.edu/. NINCH maintains an excellent page at http://www-ninch.cni.org/. [Back]

[12] A draft of the guidelines for the acquisition of performance materials is posted at http://web.uvic.ca/shakespeare/Foyer/PerfGuide.html. [Back]

Crab and Butterfly [13] The emblem and its passage of commentary can be viewed on the fine site at Memorial University: http://www.mun.ca/alciato/. Even more appropriate might be the illustration of the motto that appears in Geoffrey Whitney's Choice of Emblemes (Leiden, 1586; see http://www.mun.ca/alciato/whit/w121.html), where the same paradox is figured by a crab and a butterfly, creatures of different elements. I am not, of course, suggesting that traditional scholarship is crabby, electronic scholarship flighty. Both illustrations are provided to the site by courtesy of Glasgow University Library. Quoted with permission. [Back]

Works Cited

Responses to this piece intended for the Readers' Forum may be sent to the Editor at L.M.Hopkins@shu.ac.uk.

1999-, R.G. Siemens and Lisa Hopkins (Editor, EMLS).

(LB, RGS, 19 January 1998)