Done: "Finished" Projects in the Digital Humanities

Done: “Finished” Projects in the Digital
Humanities
Matthew Kirschenbaum
As the digital humanities continue to mature—theoretically,
institutionally, as a set of critical practices—there will be an
increasing desire to measure milestones, achievements,
completion, and closure. While one instinct might be to bristle
at such tendencies, and the traditional print-based scholarship
they seem to imply or infer, being pushed to formulate a more
fully realized set of responses can be a healthy exercise for a
rapidly expanding field. “Done,” as a shorthand convention
that has quickly become ubiquitous in the culture of professional
knowledge work, is emblematic in this regard, on the one hand
unequivocal and absolute, on the other hand useful only in the
context of low-level tasks that accumulate, often without finite
bounds or comprehensive structure, in the service of some
larger, less-defined endeavor. A server-side PHP upgrade may
be “done,” but what about the thematic research collection it
exists to support?
So how do we decide when we’re done? What does it mean to
finish something? How does the “open ended nature of the
medium” (a phrase we all pay lip service to) jibe with the reality
of funding, deadlines, and deliverables? What can we learn
from finished projects, both successful and unsuccessful? For
that matter, how do we define success and failure? Are “we”
the ones who ought to be defining it? If not, who? This panel
attempts to begin formulating responses to such questions, by
bringing together practitioners and project leaders from multiple
institutional settings and contexts. William A. Kretzschmar’s
paper traces the local history of one particular project, grafting
milestones and markers of what its gotten “done” to various
versions and releases, a common enough convention in software
and information and technology. His remarks raise issues not
only for how one assesses closure, but also digital preservation
and version control. David Sewell addresses the questions I’ve
posed from the perspective of a publisher. His work at the
University of Virginia Press’s Rotunda digital imprint, itself
unique in the field, offers a very different context, one where
finding some measure of “done” is a financial and professional
necessity. Finally, the team from the Orlando project offers the
perspectives of one of the largest and longest running
undertakings in the digital humanities, at the moment when the
project has gone to press. The co-authors discuss not only what
is done, but also what will stay done (and relevant) only by
virtue of the ongoing expenditure of effort and resources, in
the form of continual upgrades.
Large-Scale Humanities Computing Projects:
Snakes Chasing Tails, or Every End is a New
Beginning?
William A. Kretzschmar, Jr.
One of the motivating questions for this session is “What does
it mean to finish something?” As it happens, the word “finish”
can mean two things that have quite different implications for
large-scale humanities computing projects. On one hand,
according to the OED (sv) “finish” can mean ‘To bring to
completion; to make or perform completely; to complete.’ On
the other hand, the word can also mean ‘To perfect finally or
in detail; to put the final and completing touches to (a thing).’
In my own work of this kind, the American Linguistic Atlas
project (<http://www.lap.uga.edu>), we do neither of
these things. We cannot come to an end of the work because
we are witnesses and archivists of how Americans talk, and
they keep talking differently across time and space. Neither do
we think that our humanities-computing representation of our
research is capable of being finally perfected, of achieving some
perfect state, because the demands placed upon our research
keep changing. If we view the entirety of the Linguistic Atlas
Project as a “large-scale humanities computing project,” the
word “finish” is just not part of the deal. However, it is quite reasonable to ask, as our granting agencies must ask, “what do
you want money for this time?” or “did you accomplish what
we gave you money to do?” From this viewpoint, the Atlas
Project consists of a series of particular tasks or experiments,
each one of which is capable of being “finished” in both senses
of the word. In this paper, I would like to discuss the reality of
funding, deadlines, and deliverables, as they relate to the
sequence of tasks that make up the larger Atlas Project. In so
doing, I hope to show the special character of work done
deliberately as part of a sequence for a large-scale project, as
opposed to work proposed as a singular task. The
contextualization of the separate tasks leads to special cases of
what it means to “finish” the work in either sense. The point
of what follows is not the Atlas Project itself, but instead the
way that individual tasks respond to the technical and academic
situation at the time, and how our work and thinking over the
years must change so that we can avoid the charge of being the
snake that chases its own tail.
The Linguistic Atlas Project Web site has been notable over
many years for its twin goals of interactivity for research
(including the use of GIS) and making its data sets accessible
and available to the public. I first programmed a GIS system
on the Mac platform for our Linguistic Atlas data in 1990 (Kirk
and Kretzschmar 1991, 1992; Kretzschmar 1992). It was widely
used for teaching and research on American English in the early
1990s, and it immediately led to breakthroughs in how we were
able to think about language variation data (Kretzschmar 1994,
1996; Kretzschmar and Lee 1992). The immediate task was
“finished” in both senses, but the larger Atlas Project required
more developments. While the Mac system was widely used,
it was limited by restricted storage available to distribute data
sets (at that time, chiefly through diskettes).
As the next task in the sequence, we then ported the system to
the Web, which I first demonstrated at a conference in 1996
(Kretzschmar 1996b). We had been working on an interactive
ftp/gopher system as early as 1994, but when Web technology
became available we saw that it enabled perfectly what we had
been attempting from another direction. The Web allowed us
to make all of our textual data available, with many additional
GIS features for locating speakers and information. The Atlas
Web was a significant advance for both teaching and research
(Kretzschmar 1997, 2002a), in line with the goals for an
electronic atlas first set forth nearly a decade earlier
(Kretzschmar 1988). After a while, we wanted to do new things,
so we began work on a major revision of the Web site that
finally came on line in 2003, which added even more interactive
choices such as more flexible searches and tallies of the
speakers and language data set. However, we kept "The Old
Site" as a link on the new one, so that long-time users would
find what was familiar to them, and also for users who did not
want the greater complexity that came with greater flexibility
of use. We could not just move it, however, because “The Old
Site” had to run on a new platform and had to be compatible
with the extensive Python scripting that ran functions on the
New Site. The task of importing the Mac-based GIS system to
the Web was complete by 1996, but not finished until 2003
with the platform change and the final touches of the more
flexible site.
Still, the larger Atlas Project is nowhere near at an end. We are
now rethinking what the site should do, from a text-based
system to one that features audio and stored images along with
text and GIS. This change has become possible only in the last two years, as much greater network-attached storage has become available (measured in Tb, before long Pb). We are now the one of the largest clients at the U of Georgia institutional storage array (which we share with bio-informaticists, physicists, and
others usually considered to be power users) because of our
archival audio files, and we are just at the beginning for audio
and images. We now conceive of our new interviews as
conversational corpora, in which text transcriptions serve as
time-linked indices to audio files (Kretzschmar, Barry, and
Kong 2005; Kretzschmar, Anderson, Beal, Corrigan,
Opas-Hanninen, and Plichta 2006). While many users will want
to listen to our speakers, others will want to perform acoustical
analyses, now a strong trend in language variation research, as
we ourselves now perform them (Kretzschmar, Lanehart, Barry,
Osiapem, and Kim 2004; Kretzschmar, Kim, and Kong 2005).
Our next task is to integrate sound with text and to enable
acoustical research functions, while maintaining our interactive
GIS functions—a whole new set of tools and problems from
the previous task (Kretzschmar 2002b). New options for both
hardware and for text encoding make it possible to consider
the new site, just as the Web was a new option for the previous
task.
So, are we "finished"? Yes, with the GIS Mark 2 of "The Old
Site," but never finished either with keeping that site and its
successors available or with new approaches to the information
we keep as new technical possibilities and research demands
appear. We can complete particular tasks, and often we can
even “finish” particular tasks in the sense of polishing them for
improved use. Yet one experiment does not make the whole
line of research. One research proposal does not make the whole
research program. While we can succeed with each task, we
must always see tasks as part of the larger process that must
continue if future users are still going to be able to take
advantage of our resources. We would be the snake chasing its
tail if all we did was keep polishing a single task, but we do
well to make every end a new beginning as new possibilities
become available.
The implications of the Atlas Project situation for other projects
are definite and important. First of all, each project needs to
consider whether it wants to be a single task, or whether it will
be one of those continuing efforts for which any particular site or tool is just a single stage in a larger, longer process. The
reality of funding, deadlines, and deliverables makes it easier
to propose and defend the single tasks, but in many cases we
will be more honest, and better off in the long term, if we
recognize that what we are doing is evolutionary rather than
singular. We may have to work harder to convince funding
agencies to support long-term work like the Atlas Project, but
we must do that nonetheless. At the same time, our experience
shows that we can indeed identify separate stages in our work
that have independent value, and are thus fundable under the
demands of deadlines and deliverables. Indeed, the demands
of funding agencies can help those of us with long-term projects
to organize our overall effort so that we keep it vital, interesting,
and in the forefront of what new technology and standards can
offer.
Bibliography
Kirk, John, and William A. Kretzschmar, Jr. ""The Analysis
and Interpretation of Dialect Databases by Interactive
Mapping"." ACH/ALLC Conference, Tempe, 1991. 1991.
Kirk, John, and William A. Kretzschmar, Jr. "Interactive
Linguistic Mapping of Dialect Features." Literary & Linguistic
Computing 7.3 (1992): 168-75.
Kretzschmar, William A. Jr. ". Linguistic Theory and Computer
Modeling of Linguistic Survey Data." ACH/ALLC, Paris, 1994.
Kretzschmar, William A. Jr. "Quantitative Areal Analysis of
Dialect Features." Language Variation and Change 8.1 (1996a):
13-39.
Kretzschmar, William A. Jr. "The LAMSAS Internet Site."
NWAVE, Las Vegas. 1996b.
Kretzschmar, William A. Jr. "Teaching American English
Online." Journal of English Linguistics 30.4 (2002a): 318-327.
Kretzschmar, William A. Jr. "TEI and Linguistic Interviews."
2002b. <http://www.tei c.org/>
Kretzschmar, William A. Jr., Jean Anderson, Joan Beal, Karen
Corrigan, Lisa-Lena Opas-Hänninen, and Bartek Plichta.
"Collaboration on Corpora for Regional and Social Analysis."
Journal of English Linguistics 34.3 (2006): 172-205.
Kretzschmar, William A. Jr., Betsy Barry, and Nicole Kong.
"Publication of Full Interviews from the Atlanta Survey
Project." ADS/LSA 2005, Oakland. 2005.
Kretzschmar, William A. Jr. "Interactive Computer Mapping
for the Linguistic Atlas of the Middle and South Atlantic States
(LAMSAS)." Old English and New: Essays in Language and
Linguistics in Honor of Frederic G. Cassidy. Ed. N. Doane, J.
Hall and R. Ringler. New York: Garland, 1992. 400-14.
Kretzschmar, William A. Jr. "Computer-Assisted Study of
American English Lexical Data." In From AElfric to the New
York Times: Studies in English Corpus Linguistics. Ed. Udo
Fries, Viviane Müller and Peter Schneider. Amsterdam: Rodopi,
1997. 239-47.
Kretzschmar, William A. Jr., MiRan Kim, and Nicole Kong.
"Vowel Formant Characteristics from the Atlanta Survey
Project." ADS/LSA 2005, Oakland. 2005.
Kretzschmar, William A. Jr., Sonja Lanehart, Betsy Barry,
Iyabo Osiapem, and MiRan Kim. "Atlanta in Black and White:
A New Random Sample of Urban Speech." NWAVE 2004, Ann
Arbor. 2004.
Kretzschmar, William A. Jr. "Computers and the American
Linguistic Atlas." Methods in Dialectology: Proceedings of the
Sixth International Conference on Methods in Dialectology.
Ed. Alan Thomas. Clevedon: Multilingual Matters, 1988.
200-24.
It’s For Sale, So It Must Be Finished: Digital
Projects in the Scholarly Publishing World
David Sewell
Since late 2004, the University of Virginia Press has been
offering as part of its catalog a group of scholarly publications
that exist in online format only. Distributed under our Rotunda
imprint (<http://rotunda.upress.virginia.ed
u>), these publications are a mix of born-digital, digitized print,
and hybrid creations in digital humanities and social sciences.
Some of them began independently as self-published projects,
usually under the auspices of one or another digital center such
as the Maryland Institute for Technology in the Humanities or
the Electronic Text Center at the University of Virginia; others
were initiated by the Press. Whatever their origins, once
accepted as Rotunda publications they have all been subject to
conditions of production similar to those for our printed books,
namely contracts, peer review, marketing campaigns, and
agreed-upon scope and deadlines. Clearly the academic
marketplace offers an extrinsic definition of a finished digital
project: if it’s for sale it must be done.
From this point of view, there is nothing inherently different
between print and digital scholarship. Academic publishers
have long been in the business of imposing admittedly arbitrary
conventions, limitations, and deadlines on masses of scholarly
discourse in the interest of presenting them as discrete units
that can be offered in the marketplace as completed products
(not just for purchase, but also for authoritative review and
citation). And this always entails negotiation between the
author’s ideal creation (which may well be both theoretically
unbounded and technically unachievable) and the publisher’s
available time and resources. The result is a pragmatic
compromise, one that has worked reasonably well for both authors and readers. I would argue that the yoking of digital
humanities projects to theories of the open-ended text is in
many ways an unfortunate by-product of their initial emergence
during the ascendancy of postmodernist theory, with a
glorification of the unfinished that may harm their credibility.
When a publisher issues a scholarly article, book, or digital
publication, its status as “finished” represents a social contract
that the necessary stages of peer review, editing and design,
and quality assurance (proofreading, plus user testing for digital
work) have been performed.
Here is one possible typology of digital projects according to
their degree of completion at the time of publication:
1. Self-contained monograph-like objects. All content and
functionality that is ever intended to be part of the project
is present at first publication. Future updates are expected
only for corrections and migration as required to new
software environments
2. “Version 1.0” digital projects. All substantive content is
complete at first publication, but not all planned aspects of
the user interface. Future updates are expected to add
functionality that was not originally available.
3. Series-like projects. Content is added in discrete stages; it
may be quite large in extent, but is definable. All intended
functionality may be present in the first installment, or may
expand as in category #2. Print analogues would include
literary or documentary editions published in multiple
volumes over a number of years or decades, and ongoing
reference works where new and updated entries are added
periodically.
4. Truly open-ended projects. The nature of the content, subject
matter, and/or authorship is such that no particular state of
the project could ever be meaningfully said to represent
“completion.”
To date, Rotunda has published digital projects that fit into each
of the first three categories. Their work flows and timetables
have been more fluid and sometimes more tentative than for
the Press’s printed books, but not different in their basic nature.
It has always been possible to define what will constitute a
finished project, outline the steps necessary to arrive at it, and
establish reasonably solid deadlines for the process.
I will provide concrete examples for several Rotunda
publications of decisions we have made in conjunction with
authors about what could or could not be accomplished by a
deadline, the distinctions between Version 1.0 and Version 2.0
features, and the kinds of updates we have found ourselves
needing to make to finished publications. I will also discuss
why the truly open-ended projects of category 4 are more
problematic for traditional scholarly publishers; our experience
suggests that they are the most likely to require new
mechanisms and/or institutions for quality control and
sustainability.
Orlando Done! The Tension Between Projection
and Completion in Digital Scholarly Research
Susan Brown
Patricia Clements
Isobel Grundy
Orlando: Women Writers in the British Isles from the
Beginnings to the Present was published on the world wide
web, by subscription, by Cambridge University Press in June
2006 (Brown, Clements and Grundy). Its freshly researched,
scholarly, literary-historical prose is encoded in an XML
application designed for this purpose by the Orlando research
team and repeatedly revised and tweaked during the course of
research and writing. The searching functionality and the
production architecture were created at a later stage but largely
by the same team.
The project’s cofounders were new to digital humanities
research, so our notions of scholarly process and completion
related to conventional print publications. As Claire Warwick
has noted, the idea of what is “complete” or “publication-ready”
in academic culture has emerged from a complex set of human
factors relating to such matters as the attribution of credit by
institutions and funding structures, as well as the conception
of what is required intellectually for a product to be done
(Warwick 368). Such factors undoubtedly entered into how
Orlando got done. Although our projected milestones for the
project broke out a number of steps, our research plan and our
proposal to our funding agency had projected a single moment
of completion when the planned electronic history would be
ready alongside several related print volumes of scholarship.
Once underway, we found getting done more challenging than
anticipated, as the enormity of what we had undertaken on the
technical side became evident. This is clearly a risk of
methodologically experimental research of any kind, and
particularly relevant to digital humanities work. One thing that
seems crucial in the design of digital humanities projects is to
design projects modularly, with a number of discrete and in
some way publishable deliverables. Orlando struggled for
funding in later stages as a result, we believe, of research design
that end-loaded deliverables, combined with a funding
environment that was not attuned to digital humanities work.
It soon became clear that it would be necessary to stage the
completion of the project. We uncoupled the electronic from
the print publication in terms of timing, so that the former stands
alone initially, even though we will integrate the print volume
material, now being written, with the existing published
textbase. Although over the course of Orlando’s development
we felt some pressure from both prospective users and from the digital humanities community to publish at an interim stage
a small portion of the electronic materials, we held back for
two major reasons. First, we had a strong sense that there were
intellectual demands for a critical mass of materials with a
certain degree of coverage. We wouldn’t be “done”, for
instance, without having completed the materials on Virginia
Woolf and, because much feminist work has resisted the
establishment of a small canon of female writers at the expense
of others, Woolf needed to be situated in relation to her less
prominent contemporaries. Staging releases of material is
increasingly common for digital electronic projects, but we
would contend that a certain critical mass remains necessary
to establish scholarly confidence in quality of a resource,
although what that critical mass is will differ from project to
project. Secondly, because our customized tagset required us
to build a fairly complex XML delivery system, we felt we
needed a relatively “complete” interface, with some usability
testing, that could demonstrate to users some strengths of the
markup into which the project invested so much time and
intellectual labour. For these strengths of the system to be
apparent, we again needed a critical mass of materials for
hyperlinking and search results. Research conceptualization
and publication options are thus both crucial determinants of
what “done” will mean for a particular project.
Orlando’s decision to publish with an online press, in one sense
made it easy to know when the textbase was done. Published
is traditionally done (though even then, glitches in getting the
textbase up and running on the publisher’s server meant that
we actually celebrated the publication some days before it
happened). But published electronic projects don’t get put on
a shelf in a library. Being unconstrained by the materiality of
print reinforces the arbitrariness of deciding that something is
done in the sense of complete, defined in the Oxford English
Dictionary as “Having all its parts or members; comprising the
full number or amount; embracing all the requisite items, details,
topics, etc.; entire, full.” In this sense, Orlando remains undone,
and our contract with our publisher recognises this fact by
stipulating for updates.
Regular updating of content (more entries, expansion of existing
entries, more contextual material) has already, therefore, begun
with an update adding new material in January 2007. This is
an obvious publicity tool. Enhancement of functionality (quite
apart from action rendered necessary by, for instance, the advent
of a new version of Internet Explorer), though it may be less
productive of instant, visible returns than enhancement of
content for most users, is nevertheless going to be vital for
achieving the status of a first-stop general reference work, and
for retaining its currency in the evolving electronic environment.
This continuing updating of both content and functionality
means that the core team must be held together and the project
must continue to find funding. As part of its strategy of
sustainability, and with the purpose of giving Orlando a
continuing institutional identity, the textbase has been licensed
to the University of Alberta. This gives a much broader
constituency an interest in the project’s success, and ensures
that the team of this moment forms a continuity with the future
as well as the past. It would be easy to cast all of this
proliferating responsibility somewhat negatively, as though
continuing work on an already-published project were a burden,
but the fact is that all of the previous research, writing and
encoding that has been done towards publication will work
together with user feedback to enable new and newly informed
research. Both expansion of content (even into new areas) and
enhancement of functionality represent continuing research
potential, even while they ensure that “done” can – need? --
never be anything but a figure of speech.
Apart from the issues of sustainability, the degree of Orlando’s
long-term success still hangs in the balance. Publication is,
potentially, the beginning of the road to success, not its end.
We would define success not merely as the selling of
subscriptions (though obviously Orlando needs to do well in
the marketplace if a substantial portion of its funding needs are
to be met from royalties), but also as the establishing and
maintaining of a pattern of heavy use of the textbase by its
subscribers: indeed, a pattern of reliance on it as the first stop
for either specialist or general information. We would further
define it as having users exploit the broad range of possibilities
represented by the encoding.
This abstract, written four months after publication, is in every
way provisional, but it is clear from early (and so far highly
encouraging) user feedback in the form of emails and a single
brief review that for most of our users the electronic side of the
textbase is ancillary to the content. They comment on content,
apparently assuming that the new resource should stand or fall
on content alone. Nevertheless, those who built the textbase
understand that this feedback must be to a large extent
influenced by the users’ responses, not directly articulated and
perhaps not even consciously formulated, to Orlando’s
functionality.
At this early stage most user messages come from the
literary-specialist part of an audience which will hopefully also
take in computing specialists and general readers. Most
messages comment on the coverage of individual authors (or
very particular groups of authors), on whom the commenting
user tends to be an expert, so that feedback is skewed both
towards the literary and towards the individual author entries,
which are in electronic terms the least sophisticated part of the
text. Several user messages have praised the “links”. This likely
means the hyperlinking of words tagged as names, organization
names, titles, or places. Such linking is an obvious and
elementary feature of myriad electronic texts. It seems therefore
that users who said they appreciated the links probably had in mind the way that the encoding sorts and organises the contexts
in which linked words are used, and the very particular design
of the links screen to convey this information. Messages which
praise the searching, or the navigation, are even harder to
decode.
We look forward to more comment on functionality and
production, and further reports from general-interest users who
have begun, for instance, by looking at coverage of gardening,
as well as to a breakdown of statistics about kinds of use. In
addition, we want to undertake standard usability testing. Log
analysis would also be invaluable if we could overcome some
technical hurdles. This sense in which work on Orlando remains
“undone” should be common to most digital humanities
projects: if we are to evolve useful tools and resources, both
our own and those to come, we need carefully to assess how
people approach and use them.
This project directs its research towards two practically
inexhaustible fields (women’s literary history and the capacity
of computing, and specifically of extensive XML markup, to
serve the needs of this humanities topic). Neither of those fields
becomes closed to further investigation by the fact that the
project is “done” in the sense that it has reached the public. For
this project, so closely focused on its major deliverable, the
new, post-publication phase has simply opened a third area of
enquiry: that of the relations between Orlando and its users.
Bibliography
Brown, Susan, Patricia Clements, and Isobel Grundy. Orlando:
Women’s Writing in the British Isles from the Beginnings to
the Present. Cambridge: Cambridge University Press Online,
2006. <http://orlando.cambridge.org>
Warwick, Claire. "Print Scholarship and Digital Resources." A
Companion to Digital Humanities. Ed. Susan Schreibman, Ray
Siemens and John Unsworth. Oxford: Blackwell Publishing
Ltd, 2004. 366-82.