Computers, Writing, Rhetoric and Literature
Prev. | Next | Main
Considerations for Hypertext Scholarship
Content Considerations
One of the main strengths of hypertexts is their ability -- at least in theory -- to handle multiple variations in textual material. Thus, they are the perfect vehicle for creating archival, critical, and scholarly editions of textual works. Lavagnino [1] identifies four content-based challenges in the hypertextualizing of complex documents: selection, comparison, construction, and integration.
Selection includes choosing which texts and variants are going to be used. Hypertext offers the opportunity, which books do not, to decentralize the authority of individual texts. Therefore, will there be a central text, or will all versions be treated as having the same degree of authority and legitimacy? How will these choices be justified? Will facsimile versions be offered, or text-only versions? What differences do these choices make? Jerome McGann [2] argues for complete decentralization of textual variants in a hypertext site, so as to put as much choice as possible in the hands of the reader.
Comparison consists of providing the means for users to compare variants in texts
in both new and productive ways. Allowing the reader the choice of viewing any two or even
three versions of a text side-by-side is one of the potentialities of hypertext. Providing
links from one text to the next is another. Lavagnino imagines the development of tools
that will allow the user to make open-ended queries and generate comparative reports based
on variables the user specifies. For now, however, merely figuring out how to link and
display texts seems enough of a challenge.
Construction involves how a hypertext is put together. Is it in a form that a
reader can actually use? Construction is closely related to integration, which,
for Lavagnino, is the biggest challenge creators of scholarly hypertexts face. Noting
that scholarly print editions and print facsimile editions often get praised but not
used (the side-by-side facsimile edition of the folio and quarto of King Lear being a
notable example), he urges that hypertext designers find ways to show the value of
their work, so that hypertext editions and archives can be incorporated into the world
of literary criticism and textual scholarship. How can linked, compared textual variants
offer us insights into a work we could not otherwise achieve? We often think of the value of the work we do in electronic environments as self-evident, forgetting that there are many for whom our goals and methods are obscure.
Structural Considerations
The primary technical challenge in creating a complex hypertext is to provide maximum
readability while providing maximum informative value. Due to the nature of the medium
(and the current state of technology), these two goals are in conflict. Readability of
electronic text, for example, depends on how well it approximates the conventions
of printed text. In addition to being aesthetically displeasing, large numbers of links
within a text interrupt the reader's flow and (possibly worse) foreground the linked
words in the text. On the other hand, links are the heart and soul of hypertext. Links
allow a huge amount of data to be included in a document without bloating the surface
structure of the document.
For Jorn Barger, the advantage of hypertext is not what it allows you to include, but what it allows you to leave out.[3] In his view, a minimum of information should be included in each page for maximum scanability. Information the reader needs should be buried behind links. While this method works well for material written directly for the net, it causes problems when adopting printed material. Recall that his annotation of one paragraph of Finnegans Wake was over 70K. Nevertheless, Barger identifies many structural considerations in the creating of hypertexts.
For navigation, Barger stresses the necessity of providing a means of moving from
section to section without having to return to a higher level or to the main
page or to the table of contents each time. He also notes the importance of providing
full navigation information on each page, to assist users who might reach a particular
page on a site through a back door.
For decreasing waiting time, Barger suggests providing some means for the user
to distinguish between on- and off-site links, and to identify links that
lead to large JPEGs or GIFs. He also notes that care should be taken in establishing
separate pages for links. Those pieces of information that can be handled in-line
should be, to avoid unnecessary waiting for files to load.
DTD Proposal for Standardization of E-Text Versions of Basic Print Texts
There are numerous advantages to an electronic version of even a simple text.
While the readability of a text is decreased with current technology, the studiability of a
text is increased. Even with a basic text document, one can perform keyword searches,
instantly reference important passages, recall fuzzily remembered references, perform
concordance-like counts of recurring words and phrases, etc.
Because the process of typing out and marking up an electronic version of a printed text is
extremely time-consuming and tedious work, those versions that are on the Internet tend to
be widely-circulated and, once circulated, take on an air of definitiveness and authority.
This is especially true for texts disseminated by electronic-text centers and universities.
Once a text is distributed and linked to various sites, its accuracy is unlikely to be
questioned, especially in light of the practical advantages an electronic version offers for
research.
Unfortunately, many electronic versions of printed texts undermine their effectiveness
as research tools due to errors or casualness in typing and markup. The two electronic
versions of Ulysses now on the web do not match each other or any printed
edition. Though obviously based on different editions of the text, neither provide
any clues as to the edition they attempt to follow or why, the person(s) responsible for
the typing and
markup work, or the text's expected accuracy. These problems are confounded by further
dissemination, through which the origin of an e-text may be lost. One distinguished
scholar includes the Trent University version of Ulysses on his site with
a note that the novel was out of copyright and that "some kind soul put it on the
web."
For maximum value for reading and scholarly use, an e-text version of an established print text should include, at minimum, the following:
- Full title and author of the text
- The print edition the text is taken from. This is especially important in texts
that have undergone revisions. In many cases, e-versions replicate the earliest published print version, due to copyright laws, but not always.
- Notes on creation, including the creator of the e-text edition, the purpose
(scholarly, recreational, etc.), and the expected reliability of the text -- i.e.,
whether the text is an attempt at a definitive edition or merely a reference source.
- The address of the site where the text lives, within the text.
Optional considerations are:
- Page numbers that correspond to the print edition.
- Consideration given to how to treat print divisions in the electronic version and what size to make files.
File Size Considerations
For a text that already exists in a printed form, editors will usually have a choice between
(1) including the whole text in one file and (2) using "natural" breaks in the original, such as chapters, sections, books, or volumes, to create separate files.
The advantages of a single file are compactness and portability. Users only have to reference and/or download one file. Users can perform searches using their own text editors on the whole document, create their own concordances for key words and terms, find passages, etc.
The main disadvantages of a single file are size and unwieldiness. Many browsers will
simply refuse to load documents beyond a certain length. It is extremely frustrating
to wait several minutes for a document to load only to get half or three quarters of it.
Longer documents take more time to scroll through and to access key parts. The general
consensus is that anything over 30-50k could be problematic. Of course, this convention
is violated all the time. For some texts, it will be impossible not to violate these
constraints. The "Ithaca" chapter of Ulysses, for example, is about 160 pages.
Even a relatively short document may be 50-100k if it is heavily linked and contains a
few images.
The advantages of multiple files are compactness and economy. One can have a very economic
table of contents page with links to individual chapters (and then have those chapters link
to each other). This is especially useful if the document is going to be used more for
research than for general reading. If all the files are included under a search engine,
this economy is maintained. The disadvantages of multiple files are their lack of
portability and fragmentation. If the user wishes to scan or scroll through the entire
document, it is impossible. If the user wishes to consult the document off-line or
download it, it is difficult to use.
Prev. | Next | Main
Notes
[1] John Lavagnino, "Reading, Scholarship, and Hypertext Editions." (http://www.stg.brown.edu/resources/stg/monographs/rshe.html)
[2] Jerome McGann, "The Rationale of Hypertext." (http://jefferson.village.virginia.edu/public/jjm2f/rationale.html)
[3]Jorn Barger, "Hypertext and Navigability." (http://www.mcs.net/~jorn/html/net/navigable.html)