21.276 CS & the humanities: what's interesting for both

From: Humanist Discussion Group (by way of Willard McCarty willard.mccarty_at_kcl.ac.uk>
Date: Tue, 2 Oct 2007 06:13:06 +0100

               Humanist Discussion Group, Vol. 21, No. 276.
       Centre for Computing in the Humanities, King's College London
                     Submit to: humanist_at_princeton.edu

         Date: Tue, 02 Oct 2007 06:09:35 +0100
         From: "Anna Bentkowska" <anna.bentkowska_at_kcl.ac.uk>
         Subject: RE: 21.273 CS & the humanities: what's interesting for both?

Dear Willard,

Do you intend to cover interactions of CS with art and art history?
There is a wealth of study cases and theoretical literature in this area.

You are absolutely right that in order to get a CS person involved in
the humanities research and 'get his/her blood to race' the problem
needs to be 'sufficiently scientific'. This is where - at least from
my perspective - the crunch lies. Rather than theoretical scientific
models, we (art historians) are after digital means (fairly simple
tools and applications) which help us to solve our problems. I have
recently surveyed hundreds of 3D visualisation projects (selection at
http://3dvisa.cch.kcl.ac.uk/projectlist.html). It was not unusual for
arts & humanities scholars to claim, rather arrogantly, that some of
the most technologically sophisticated solutions have failed to
answer basic questions of their discipline. I guess the brief of
your papers is to eulogise the benefits of collaborative research,
rather than support the argument of the 'two cultures'.

Practice-based research in art might therefore be of particular
interest to computer vision scientists, because it often involves new
algorithms (for example, Artificial Life art, starting with Craig
Reynolds's Boids in the 1980s through to most the recent research,
Paul Brown www.paul-brown.com et al.). This leads me to suggest
computational aesthetics as an interesting intersection of CS and the
humanities, where - rarely - both seem at equal footing.

A familiar area of CS in art history are studies into Renaissance
perspective, such as computation and visualisation of the third
dimension in pictorial spaces (vide Antonio Criminisi of Microsoft
and Martin Kemp, History of Art, University of Oxford). From the CS
perspective, the challenge of such studies is often that art is meant
to be ambiguous and free from rules. Even those artists (I mean, old
masters) who authored theoretical treatises on realistic
representation and geometrical perspective hardly ever applied what
they preached to their own practice. The scientists tend to call this
'inconsistent data which cannot be confirmed empirically'. George
Legrady, professor of Media Arts & Technology, University of
California at Santa Barbara, got it spot on: 'Whereas the artist is
by nature a generalist, as his or her work often depends on the
juxtaposition of sampled information from a wide spectrum of sources,
the scientific research model operates at the micro and specific
level. Research addresses a specific problem, and when a solution is
achieved, it requires the experiment to be repeatable. In contrast,
an artwork's value lies in its unique approach in addressing a
problem, and to reduce it to its integral components to understand
its inner workings may result in the dissolution of its essential
qualities in the process.' ('Perspectives on Collaborative Research
and Education in Media Arts', Leonardo, Jun 2006, Vol. 39, No. 3: 218).

My own interests include the use of Content Based Image Retrieval and
other advanced imaging techniques as an alternative to traditional
taxonomies and methodologies of art history. In particular, I saw in
these computational methods an advancement for traditional, and now
rather discredited iconological studies. CBIR for instance, has
offered methods of analysis and interpretation that are postulated by
the iconological theory, but difficult to achieve through
conventional means in art history. Articulating these issues in a way
that challenged CS people I was working with, was the most rewarding
process of getting to know my own stuff. I hope this worked both ways.

The least controversial and possibly the most successful are
applications of CS in documentation, conservation, restoration and
authentication of art (examples on request). A considerable number
of papers on these subjects were submitted by computer scientists to
the next year's IS&T/SPIE in San Jose, CA, Computer Image Analysis in
the Study of Art
While working on the programme of this symposium and reviewing
papers, I realised the (obvious?) dichotomy in the understanding of
theory and praxis by science and the humanities - which takes me back
to para 1. Two cultures after all?

Just a couple of possible threads which may help with churning the
ideas and context for your papers.

Best wishes, Anna

-----Original Message-----
From: Humanist Discussion Group [mailto:humanist_at_Princeton.EDU] On Behalf Of
Humanist Discussion Group (by way of Willard McCarty
Sent: 29 September 2007 06:42
To: humanist_at_Princeton.EDU

                 Humanist Discussion Group, Vol. 21, No. 273.
         Centre for Computing in the Humanities, King's College London
                       Submit to: humanist_at_princeton.edu

           Date: Fri, 28 Sep 2007 14:49:39 +0100
           From: Willard McCarty <willard.mccarty_at_kcl.ac.uk>
           Subject: CS & the humanities: what's interesting for both?

Dear colleagues,

As a followup to the question I just asked about interactions between
CS and the digital humanities, let me ask another. What are the
interesting problems & "grand challenges" arising in the digital
humanities that interest or should interest people in CS? Similarly,
excluding the funding concepts, what's in computer science research
that interests or should interest us in the humanities?

All comments welcome.

Many thanks.


Willard McCarty | Professor of Humanities Computing | Centre for
Computing in the Humanities | King's College London |
http://staff.cch.kcl.ac.uk/~wmccarty/. Et sic in infinitum (Fludd 1617, p.
Received on Tue Oct 02 2007 - 01:27:09 EDT

This archive was generated by hypermail 2.2.0 : Tue Oct 02 2007 - 01:27:09 EDT