Date: Tue, 4 Aug 1998 16:13:35 -0400
From: Brad Scott <bscott@routledge.com>
Subject: Re: 12.0157 TLS gleanings (Arden Shakespeare CD-ROM)
Willard
=20
As you have posted a summary of G.W. Pigman's review of the Arden=20
Shakespeare CD-ROM from the TLS to Humanist, I feel obliged to respond=
=20
to the criticisms he makes in it.
=20
By way of background, the CD-ROM was developed by Routledge (ie me),=
=20
but the entire Arden Shakespeare list was retained by the Thomson=20
Corporation when they sold Routledge a couple of years ago. In so=20
doing, these books and the CD-ROM were given to Thomas Nelson Ltd.
=20
Lest your readers think that it is all bad, in his review, Pigman=20
notes that even he thinks that the application has some strong points.=
=20
He cites the range of content, and the screen design, layout,=20
impressive array of links and (by and large) the ease of searching.=20
Not a bad start perhaps, though there are several important features=
=20
in the program which he does not mention at all.
=20
However, let's take the criticisms in turn:
=20
1. Obsolescence of the data. The CD-ROM includes all of the second=20
series of the Arden Shakespeare; it does not include any of the Third
series. This decision was taken since at the time the product was=20
under development, only three of the Third series had appeared, and=20
the new editions of all the plays were not scheduled to be published=
=20
until 2006. Given that, when we were designed the tool with our=20
consultant editor, Jonathan Bate, we decided to create a tool that=20
played to the strengths of the Second series, ie produce a program=20
that would support teachers and researchers explore the construction=
=20
of the 'edition'. By providing a range of core sources, such as the=20
Quartos, the First Folio, and the complete text from Bullough, it was=
=20
self-consciously designed to show to students and others the relation=
=20
between texts. In our minds, the play texts, notes, images and sources=
=20
were foregrounded and the Introductions (some of which represented a=
=20
different generation of critical thinking) were a less prominent part=
=20
of the resource. This criticism of the choice of Arden 2 was also made=
=20
by Jean Chothia in Computers and Texts=20
<http://info.ox.ac.uk/ctitext/publish/comtxt/ct15/chothia.html>, and a=
=20
response to it by Jonathan Bate and Nick Kind (the editor now=20
responsible for the project at Thomas Nelson)=20
<http://info.ox.ac.uk/ctitext/publish/comtxt/ct15/nelson.html>.
=20
2. Searching problems. I think there must be a rule that, no matter=20
how much time you spend talking to people about how they will want to=
=20
search electronic data, and no matter how much you try and accommodate=
=20
all the suggestions you get, you'll not please everyone. By taking the=
=20
route that we did -- ie to customise DynaText so that you could have=
=20
multiple synchronised frames within the window, thereby minimising the=
=20
difficulties of the user getting lost in hyperspace, unable to relate=
=20
the source of a link to its target -- we found that there were a=20
number of conceivable search scenarios that could not as a consequence=
=20
be supported (that across both glossaries simultaneously is a case in=
=20
point). Pigman also points out that you cannot search in both the=20
variant and commentary notes at the same time. True, this is again a=
=20
result of the decisions taken about screen layout, but it begs the=20
question - how many people would really want to? Certainly none of the=
=20
Shakespeare scholars we consulted. He also notes that one cannot=20
search the variants or commentary notes across more than one play at a=
=20
time. Again, this was something that (regretfully) resulted from the=
=20
initial choice of screen layouts; to accommodate such a change to the=
=20
way DynaText handles search results was in the end felt to not justify=
=20
the cost. Though we attempted to meet the functionality needs of as=20
many of our advisers as possible, in the end, some things just come=20
down to economics, and you have to make compromises.
=20
3. Images v. text transcripts. Pigman suggests that, without=20
searchable transcripts of the Folio and Quarto texts, "serious textual=
=20
work is impossible". This position was not held by a large number of=
=20
other scholars we spoke with. In designing software applications for=
=20
the academic community, one will never manage to deliver something=20
that will suit everyone. We were trying to develop a tool that served=
=20
a purpose that was both different from and yet would support and be=20
supported by, those other electronic Shakespeare projects published by=
=20
Chadwyck-Healey and Georg Olms.=20
=20
4. Query language. We did not use TEI for the Arden. In part this was=
=20
because it did not support many of the functionality-driven features=
=20
that we needed to build into the data. In addition, I did not deal=20
extensively with SGML in the documentation (which, with a Help file of=
=20
over a megabyte, was perhaps a good idea); my plan was always to=20
provide that sort of user support on the web site, though since we=20
were not involved in the post-publication development of the project,=
=20
I was unable to initiate this. I would contend that the number of=20
users out there who are working with the Arden Shakespeare in a=20
sophisticated enough way to want to know about SGML queries is still=
=20
rather small. We designed an application that could be used=20
successfully by those that don't even know what SGML is (yes, there=20
are people out there like that, some readers of Humanist may be=20
surprised to hear). For SGML experts, we also intended to make the=20
complete SGML files available to electronic text centres, which would=
=20
support the kind of interrogation that Pigman desires. In addition,=20
Pigman cites the example of CUP's Wife of Bath as a model use of=20
DynaText. It is true it is a wonderful resource, but I remain to be=20
convinced that it is a sustainable model for the development of=20
electronic texts.
=20
5. Price. =A32500 is the same price as the Chadwyck-Healey Editions an=
d=20
Adaptations of Shakespeare, as well as being comparable with a large=
=20
number of other (non-grant-funded) electronic titles coming out of=20
commercial publishing houses. The simple fact of the matter is that=20
data capture and clean-up, and software development are expensive.
=20
I would also add that Pigman's review is unusually harsh in comparison=
=20
with the others. Andrew Murphy of the University of St Andrews has=20
written one which will appear in a future issue of Computers and the=
=20
Humanities in which, though he has some criticisms of the product, he=
=20
describes the program as "a package that is beautifully presented,=20
well conceived and very easy to use", and that the development team=20
"deserve considerable praise for mapping out the electronic terrain of=
=20
twenty-first century Shakespeare".
=20
Rather than read the reviews though, I would just urge people to have=
=20
a play with the application and see what they think. Demos can be=20
requested from Nelson at <http://www.ardenshakespeare.com/>.
=20
I believe the Arden Shakespeare CD-ROM is an innovative piece of=20
software which attempts to address many of the difficulties of=20
delivering complex textual material in an intuitive fashion. There are=
=20
bound to be many issues still to be resolved, many of which we at=20
Routledge have been working on in the development of our new=20
generation of electronic projects, which includes the Routledge=20
Encyclopedia of Philosophy and the Wellesley Index to Victorian=20
Periodicals. For these projects (also built using DynaText as their=20
core) we have simplified the screen layouts, and adopt (and extend)=20
TEI.
=20
As readers of Humanist are well aware, there are considerable=20
intellectual and resource issues in capturing and tagging data, as=20
well as in designing effective software. I remain convinced that it is=
=20
only through a constructive debate between _all_ parties engaged in=20
humanities computing (ie academics, librarians, archivists and=20
publishers) that we can develop and distribute resources that will be=
=20
valuable for both teaching and research.
=20
I look forward to continuing to participate in such debates at DRH98.
=20
Brad
=20
_____________________________________
Brad Scott
Electronic Development Manager
Routledge
11 New Fetter Lane
London EC4P 4EE
Tel: *44 (0)171 842 2134
Fax: *44 (0)171 842 2299
Email: <bscott@routledge.co.uk>
Web: <http://www.routledge.com/routledge/electronic/default.html>
_____________________________________
-------------------------------------------------------------------------
Humanist Discussion Group=20
Information at <http://www.kcl.ac.uk/humanities/cch/humanist/>
<http://www.princeton.edu/~mccarty/humanist/>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D