William Makepeace Thackeray's "Swift," [1] is one of the great literary hatchet jobs, a devastating critique of the life and work of Jonathan Swift. Yet, because it has influenced popular opinion and scholarly writing about Swift ever since, it has a crucial place in Swift studies. Therefore I was excited last summer to find a copy within a collection of Thackeray's works and to have the opportunity to add this find to my Gulliver's Travels World Wide Web site [2], a collection of texts and links useful to the study of Jonathan Swift's Travels into Several Remote Nations of the World.

Now in its fourth year, my site features an edited, annotated text of Swift's 1726 satire rendered in HTML format. It also includes a selection of original materials, a large directory of links to other resources, plus a growing collection of online editions of related texts. I have collected and reproduced more than a dozen other print documents on line, from letters to long essays, book extracts to complete pamphlets, along the way dealing with challenges in formatting them for the Web.

Thackeray's essay posed one significant, though common, formatting problem: The work is riddled with footnotes, many unusually long. Even in print it looked disjointed and was difficult to read in some places. [3] Designing an online format to accommodate the footnotes and do justice to the text was going to be a challenge.

The problem, simply put, is where to put the footnotes? I saw three basic choices: eliminate the notes, publishing the body of the work only; leave the footnotes intact, reproducing the original format, paging, footnotes and all; or reformat the document, relocating the notes in a manner more appropriate to the electronic medium.

Cutting the Gordian Knot

I had already solved a footnote formatting question once by removing them. The text was Samuel Johnson's "Swift," a work like Thackeray's in tone and importance, accompanied by "Notes, by Mrs. Alexander Napier." I paused only briefly over the conundrum of what to do with these footnotes before realizing that they were not worth the worry. Mrs. Alexander Napier's notes added nothing of value from the perspective of the Gulliver's Travels site. I wanted Johnson's words only and without further deliberation excised Mrs. Napier's. However, as a concession to the integrity of the original, I added

In the current case, however, the notes were written by Thackeray and therefore integral to the work. The essay was based on a series of lectures he had delivered. When he published them, Thackeray chose to expound upon and bolster his original points through notes. Removing them, in this case, was not a valid option.

A Paged Approach

One choice would be to render the footnotes as footnotes in the electronic edition, mimicking the original formatting of the print edition. That would mean pagination matching the pages of the original, with the footnotes represented exactly as they appear in print.

In HTML this can be accomplished in a couple of ways. One approach would be to create a separate file for each physical page, with the notes appearing at the bottom of the page just as in the print version (figure 1). Another possibility would be formatting the entire document in a single, continuous file, with pages distinguished by spacing, horizontal rulers, or table borders. Each page-worth of text, body and footnotes, could be formatted as it appears in print (figure 2).

Non-HTML formats offer even more features and capabilities in creating documents. The more robust standard formats, such as SGML and XML, and proprietary formats, such as Portable Document Format (PDF) and Flash., also can be used to mimic printed pages. Documents in PDF format reproduce very realistic looking pages.

Even with the tools and techniques readily at hand, I was not eager to reproduce the Thackeray essay as it appeared in print. Looking at the pages in front of me, I had to question whether footnotes were the best way of presenting this information. Even brief notes interrupt the flow of the work and intrude upon the reader. At points the Thackeray looked like two separate documents vying for space on the pages.

There are more general concerns. Should we go to extraordinary lengths to create artificial pages and their artifacts, such as footnotes, especially when there might be a chance to solve the problem in a new and better way? I have avoided creating electronic documents that look and behave like print as antithetical to the basic premises of electronic publishing . [4] What attracted me to the new medium was the opportunity to do something new and different. I did not want to be tripped up by footnotes.

The Page and the Scholar

At the same time, I have become more aware of the scholarly needs and expectations for the authority of published works. These often require a close correspondence between an original and its copies. Without a distinguished publisher's imprint or a scholarly institution behind my project, my electronic texts have to borrow their authority from their printed sources. Any liberties taken with the content and format risks breaking that delicate but all-important link.

Pagination, in particular, is an important reference point, often the only means of indicating location within a work. Poetry, drama, and selected other works (e.g., the Bible) use other reference systems, such as lines, scenes and verses, which can translate very well in electronic formats. However, most prose works are cited by page number, and the lack of pagination in electronic editions can be a deficit.

Past experience working on the Gulliver's Travels Web site had illustrated this point. Many scholarly articles about the Travels cite the text using the pagination of the Davis edition (Herbert Davis, ed., The Prose Writings of Jonathan Swift [Oxford: Basil Blackwell, 1965-73]). Someone encountering a reference in the form (Davis, 56) could find the cited passage in my Travels only if it used Davis's paging. Without such correspondence, there is no way to locate the passage cited as (Davis, 56) anywhere outside of Davis. Thus, there is a clear advantage to using pagination in an online edition when it matches that of a standard print edition.

The edition containing the Thackeray essay was not from an important or standard edition, so there would be little advantage from that point of view. However, there can be other advantages to pagination. Around this time I encountered A Casebook on Gulliver Among the Houyhnhnms, an anthology, which provides the original paging for all the works reproduced there. A note indicates, "Bracketed figures throughout this book indicate the end of the page on which the selection appeared in the cited source." [5] Therefore, researchers are able not only to locate a passage in the original source, but can also cite from the copy as if they had the original. Would providing the original paging in the online version of Thackeray have similar advantages for researchers?

Lifeline

I found myself in a dilemma. My instincts and experience inclined me towards remaking Thackeray's essay in a new, more functional format, rather than recreating the printed page with all of its limitations. At the same time, I was aware that there might be advantages, in the scholarly world at least, to a closer correspondence between the electronic edition and the print original and disadvantages to dismissing the importance of that link.

In this moment of need I turned to the "18th Century Interdisciplinary Discussion (C18-L)" [6], an e-mail forum to which I have subscribed on-and-off for the past few years. The members are mostly academics, college and university faculty members, instructors, and postgraduates, with a shared focus on 18th century studies. The list generates twenty to forty messages a day during the academic year, taking on any and all issues with vigor. As its members were also a potential audience — researchers and teachers — for the results of my editorial work, the list seemed an ideal venue to test my concerns and ideas.

On August 22, 1999, I sent a message to C18-L outlining the issues described above, asking for their views. [7]

The main points I raised were "format and pagination" . . .

and footnotes. . .

The response was an eloquent near-silence. I received two responses; a longer note from a longtime correspondent, himself an editor of electronic texts, and a shorter but still-helpful note from another list member. Both agreed that reformatting the footnotes as endnotes made the best sense, commenting on different ways to manage jumping around in the text. The longer note also observed that the original paging was useful, providing examples of possible formats.

Otherwise, the 18th-century community was keeping its own counsel and I received none of the heated exchanges about research and scholarship in the digital world I had expected. Disappointing as this was, I quickly recovered, resolving that the decision was now mine to make.

Decanting the Contents

After due consideration, I decided to reformat Thackeray's "Swift" as an HTML document, ignoring pagination and rendering the footnotes as end notes. Despite my own interest and some external encouragement to indicate the paging of the original in the online version — it would not have been too difficult to include the bracketed page numbers in text — it was the layout of the print edition that tipped the scales in the other direction. The printed pages were so oddly broken up by the long footnotes that pagination seemed unimportant. Would I need to indicate the pagination of the footnotes as well? That would be even more awkward and disjointed.

Reformatting print footnotes as endnotes in an electronic edition is a reasonable compromise between all the needs and opportunities described above. It retains the information and takes advantage of some of the features of the new medium. It is also possible to create such a document using simple HTML, readily accessible to the widest selection of browsers.

The mechanics of creating the text were fairly simple. There was just one graphic, a portrait of Swift, which I scanned and inserted at the head of the document. The rest was text. In past projects, I scanned the whole work in one pass and then hand-edited the resulting file, extracting the footnotes and moving them to the end. I found this to be a tedious process. For the Thackeray essay I made two passes with the scanner, first the body of the text and then a second pass for the footnotes, making sure to include page numbers as reference points. This results in two separate files that can be edited side-by-side to create the final document.

There were just a few significant questions remaining: How should I link the text to the notes? How to present the links within the text? What action would carry the reader to the notes, or vice versa? And, third, how to present the notes in relation to the body of the text?

The first question is really how should I indicate the availability of a note? In the print edition, the presence of a note is signaled by an asterisk (*) or similar symbol, with a corresponding symbol at the head of the related note displayed at the bottom of the same page. In an electronic edition, there is something of a technical imperative to do away with such archaic symbols. The word or phrase — the 'source' — itself can be rendered as a hypertext link, highlighted in some form to indicate that it is live. The source text is linked directly to the corresponding note — the "target" — and clicking the mouse cursor on linked text makes a jump to the target note. [8] These links form a hard connection between the two points, without extraneous symbols.

The hard connection between hypertext links is good in theory but does not always work so well in reality. For instance, a hypertext jump does not indicate precisely which specific piece of text is being presented. Presumably, the reader is meant to see one of the notes now appearing on the screen, perhaps the top one, but how can one be sure? Also, without visible marks to indicate the relationship between points in the text and notes, a printed copy of such a document will be confusing, if not entirely useless.

Numbering entries is a standard method of organizing end notes in printed texts, and it is also an effective method of linking text and commentary in electronic texts. Providing sequential numbers in the text as indicators and linking these to the correspondingly numbered notes is a simple and direct solution.

The action linking the text to the notes is another design consideration. Using a hypertext link in HTML normally causes a jump to a new location, the target note replacing the source text on the screen. Though the browser's BACK button will usually return to the previous location, users do lose track of where they are and may not be able to find their way back to the main text.

There are a number of ways to keep the main text on the screen while displaying notes, allowing readers to hold their place in the body of the document while they refer to the commentary. Adding the argument target="_blank" 9 to the link tag will cause the target item to appear in a new browser window, leaving the original or source text in place. Another choice, HTML frames, present separate but usually related documents within individual panels of a single browser window. Some of the newer tools in the Web design toolkit, such as Javascript. and Flash., provide features that can be applied to this problem. For instance, "roll-overs" could make a related note magically appear on the screen whenever the cursor points to the corresponding location in the body of the text.

All of these schemes have the same fatal flaw: the different elements have to be stored in separate files. I initially tried the target="_blank" scheme for the Thackeray essay until I discovered that it would open a new window only if the target was in a different file. Thackeray's notes are so substantial, I wanted the two parts together in a single file, reducing the chance that the essay could become unlinked from the commentary. The single-file format also allows the text of the notes to be searched with the body of the essay and, again, makes for a more coherent document when printed. For the time being, I have decided that these needs override concerns about hypertext jumps, but I am still looking for a solution.

Conclusions

The current version of my electronic edition of Thackeray's "Swift" may be found at http://www.jaffebros.com/lee/gulliver/biography/thackeray.html. For all the deliberate thought that went into its design, it is a fairly simple HTML document. Or perhaps, because of all the thought that went into it, it is a simple document. With the exception of the one scanned portrait, the whole thing fits into one continuous text file. There is hardly a bell or whistle to be heard. Yet, even as I write this article, I cannot help but fiddle and tweak features, seeing what can be improved.

I need to remind myself that there is other work to do and the best use of my time is applying what I learned to the next project. There is still the text of Gulliver's Travels demanding my attention — a life's work — and a small but growing pile of new finds from the used book store waiting to be scanned, formatted and added to my Web site.

Lee David Jaffe is the creator, compiler, designer, editor, and sole proprietor of the Gulliver's Travels Web site. The site represents an intersection of several personal and professional interests, including electronic publishing, travel writing about Japan, and Irish literature. Jaffe is the author of two books — All About Internet Mail (1997) and Introducing the Internet (1996) — for Library Solutions Press [formerly http://www.library-solutions.com/] and of numerous articles and papers about libraries, computers, and networks. He is head of Library Computing, University of California, Santa Cruz, where he as worked since 1987. You may contact him by e-mail at [email protected].

Notes:

1. William Makepeace Thackeray, The English Humourists of The Eighteenth Century, vol. 10 of The Complete Works (Philadelphia : Lippincott & Co., 1876), 381-415.

2. Jonathan Swift, Gulliver's Travels, http://www.jaffebros.com/lee/gulliver/index.html

3. There are thirty-six notes for thirty-four pages of text. The total volume of the notes is almost equal to the main body of the text. In two places the notes take up all put two lines on a page and there are more than a half dozen cases where the notes take at least of half of a page. Pages 404 and 405 (see illustration) features only ten lines of text, with a single footnote taking up all but two lines on the first page and all but eight lines of the next.

4. "By late in the decade, millions of Americans had abandoned their computers and turned to the immensely popular new VirtuLib 2000, a $14,000 device that enables the user to experience, with uncanny realism, the sensation of reading a book." Dave Barry, "The '90s: Dave Barry looks back at the decade," San Jose Mercury News (C:1,4) January 2, 1994. Pardon my resorting to low humor, but Barry's comments capture for me the essence of what is ridiculous about most "electronic book" proposals.

5. Milton P. Foster, ed., A Casebook on Gulliver Among the Houyhnhnms (New York : Crowell, 1966 [1961]), 4.

6. 18th Century Interdisciplinary Discussion <[email protected]>; subscribe at <[email protected]>; Web site http://cac.psu.edu/~bcj/c18-l.htm

7. The full text of my message to C18-L is available below.

8. In HTML the source or "reference" is tagged with an href anchor <a href="#target">source</a> and the target is marked by a name anchor <a name="target">target site</a>. In this example the # in href="#target" signals the browser to look for an anchor assigned the name "target."

9. In the link <a href="target.html" target="_blank">link source</a> the setting target="_blank" will cause the document target.html to open in a new window.

Links from this article:

Gulliver's Travels, http://www.jaffebros.com/lee/gulliver/index.html

Samuel Johnson's Swift, http://www.jaffebros.com/lee/gulliver/biography/johnslife.html

Figure 1

Figure 2

Figure 3

Appendix 1

Date: 22 August 199

To: 18th Century Interdisciplinary Discussion <[email protected]>

From: Lee Jaffe <[email protected]>

Subject: format for online text and notes

Cc:

Bcc:

X-Attachments:

I recently obtained a copy of Thackeray's "The English Humorists of the Eighteenth Century"

(The Works of William Makepeace Thackeray, in Twelve Volumes, London : Smith, Elder & Co.;

Philadelphia : Lippincott & Co., 1876., volume X.) which includes his (in)famous essay on Swift. I plan

to render the Swift essay in an online format because Thackeray's views on Swift exemplify the

Victorian judgement of the satirist's misanthropy and continues to form the basis for so

much current writing about Swift's life and work (notably including the recent Glendinning

biography).

The process of scanning this essay has raised several decisions and challenges about how best to present it

online. I thought that this group might provide a representative point of view from the scholarly side of

the table as to what would be most useful.

* Basic format and pagination: I've tended to approach online rendering with a focus on the content, rather

than format (more interested in the wine than the bottle). I'm not rendering manuscripts or work that has a

special visual focus. Also, the item isn't unique. There are many editions of this text with different

paginations. Therefore, I've planned to take no account of the book pagination in the online version.

Similarly, I have taken no notice of line breaks, even closing up words hyphenated in the print copy. Is

this right thinking?

While it would be possible to render the pages as printed — at one extreme using .pdf format to take virtual

snapshots of each printed page — I can't see the value. Any scholar would be nuts to use my edition as a

source for citing back to the print edition, but should for academic integrity's sake, either cite the online

edition as-is, or go back to a printed version if pagination is critical. Am I reading these issues correctly?

Am I stepping over a line which would render the online edition useless?

* Footnotes: The work — originally from a series of lectures — is heavily peppered with

footnotes, and I'm trying to determe, 1) whether to include the notes and 2), if so, how to

present them.

Including the notes is no minor task. There may be as much text in the notes as in the body of the essay.

I think that the text stands alone - though changed - without the notes, so they are not essential to the sense

of the essay. If the notes are an editor's hand, I'd prefer to exclude it. If they are Thackeray's,

I'm ambivalent.

If I include the notes (my current inclination), how do I present them in a way that acknowledges their role

within the printed edition and still works within an online environment? I could go back to presenting the

entire text in the original paging with the footnotes in context. Frankly, I don't think the notes worked all

that well in the original. There is one note, for instance, which takes up all but two lines on one page and

all but eight lines of the facing page. It is a typographer's nightmare. In an .html document, presenting

such notes as they appear in the print version would destroy the flow of the essay even more than it does

now. My inclination is to reorganize the notes as end notes, with links from the * to the relevant item.

This is the only way I can see to present this information effectively. I'd like to hear what

other think.

Thanks for your consideration,

— Lee Jaffe