Hi there,
Responding to Dieter's post below: first I'd like to thank you for your XDOM libraries, which I'm beginning to use for some of my Delphi projects. Great work!
I agree that scholars currently have little to gain from developing tools, from the point of view of their academic careers. However, there's another group of individuals like me, who are working in the academic context as programmers and in similar roles, who do this kind of work for a living. John Bradley would be another good example. Our work initially got going in the realm of language teaching support (Computer-Assisted Language Learning), giving birth to quite widely-used tools such as our Hot Potatoes programs; recently, we've morphed from a Language Centre into a Humanities Computing and Media Centre, and our work is focused more and more on HC, digital documents and encoding. Our experience has been that when we release tools which gain approval and acceptance, the university is generally pleased and appreciative; they've also helped us spin off some of our work commercially, to everyone's financial benefit. While individual departments and tenure committees may not (yet) give much weight to this kind of work, other parts of the university administration are more supportive.
The emergence of centres such as ours, which have relatively stable workforces (as opposed to the ad-hoc temporary hires associated with grant-supported projects) means that tools tend to be rewritten and updated steadily, which gives them more credibility (Hot Potatoes is now at version 6, for example, and has been out since 1997). This model of tool development, where a centre with long-term staff creates tools for the use of several projects, and maintains them over time, is much more likely to be successful than the case where an academic working on a temporarily-funded project hires in programmers to write something for a specific purpose, releases it, then moves on to the next piece of research, leaving the code to languish like an abandoned vehicle in a field (which is what "open-sourcing" a project often turns out to mean).
Cheers, Martin
At 09:37 AM 29/06/2005, you wrote:
Date: Fri, 24 Jun 2005 13:14:24 +0200 From: Dieter K?hler d.k@philo.de Subject: [dm-l] Tools for humanities computing (WAS: Are markup languages obsolete?) To: dm-l@uleth.ca Message-ID: 5.2.1.1.0.20050624114827.025b2050@pop3.philo.de Content-Type: text/plain; charset="iso-8859-1"; format=flowed
I often think that the free "tools" aspect had been underestimated in humanities computing. Of course, there exist a couple of good examples like the TEI XSLT style sheets or some open source archival software etc. Nevertheless, the situation is not satisfactory and I wonder what are its causes and by which means it could possibly be improved. Briefly summarized the main causes, I can think of, are the following:
- Most software development in the humanities takes place in an ad hoc
fashion: People have specific problems and develop specific solutions.
- There is a lack of institutional support for developing tools for others.
Tools are only by-products.
- If there is institutional support for developing tools for others, these
tools need to be sold in order to re-finance the work.
- It is not advisable for a scholar trying to build an academic career on
developing tools for humanities computing.
- There exists no academic infrastructure *focused* on developing tools for
the humanities, ie. a specific society, journal and annual conference.
Since one of my main research interests is considered with the development of tools for humanities computing, I would be very interested in the opinions of others on the above list of causes. Perhaps together we could find ways to improve the situation.
Dieter Köhler
Institute of Philosophy and Centre for Multimedia Studies University of Karlsruhe Germany
______________________________________ Martin Holmes University of Victoria Humanities Computing and Media Centre mholmes@uvic.ca martin@mholmes.com mholmes@halfbakedsoftware.com http://www.mholmes.com http://web.uvic.ca/hcmc/ http://www.halfbakedsoftware.com