Humanist Discussion Group, Vol. 32, No. 257.
            Department of Digital Humanities, King's College London
                   Hosted by King's Digital Lab
                       www.dhhumanist.org
                Submit to: humanist@dhhumanist.org




        Date: 2018-12-07 08:55:59+00:00
        From: Max Kemman 
        Subject: Re: [Humanist] 32.252: influence of digital humanities

Dear André, dear humanists,

Thanks for this interesting exchange. My research concerns the more
classical question of how DH influences historians, but in that pursuit I
have interviewed information scientists, software developers, and
computational linguists as well. So while my research did not concern their
practices sufficiently that I can provide definite conclusions to your
questions, I do have some findings that provide suggestions for further
investigation or discussion.

A significant influence is that computational practices are grounded in
applied scenarios, rather than in theoretical/mathematical improvements.
Consider much NLP research for example, which tends to employ perfect
English ASCII data to test whether some algorithm works. In contrast, in
DH, the data is much more messy and fuzzy, with incomplete OCR, incomplete
fields, and different spellings or languages on the corpus or even document
level. The challenge is that 'real-world' data is much more complicated
than readily available data, requiring data cleaning and modelling.

When results do come out, the question is transformed from "how much more
efficient or effective is this" to "is this meaningful". For example, one
computational linguist working on semantic drift complained that the most
popular method is to just cut up a corpus in periods of 5 or 10 years, and
run software for each period. Instead, he collaborated with historians to
find meaningful moments when words might have changed meaning.

These may not be so much theoretical or methodological changes, except the
increased complexity of data modelling, but are significant enough that
some of my interviewees said they would always collaborate with 'domain
experts' from now on, while one even feared that he would have difficulty
getting a position in a purely computer science group since his research
had deviated so much already.

One problem for a more thorough review of this might be the diversity of
disciplines that are involved in DH. While the humanities is already a
broad and ambiguous grouping, it is possible to provide meaningful results
with respect to history or English departments as two dominant
sub-communities. For the computational domains, there are commercial
software developers, research software engineers, computational linguists,
information scientists, and computer scientists (with all its
subdisciplines: AI, information retrieval, database engineering, etc),
which all have different incentives for participating in DH, and experience
different influences.

With kind regards,
Max Kemman

PhD candidate Centre for Contemporary and Digital History
University of Luxembourg
PhD: Digital History as Trading Zones
www.maxkemman.nl




_______________________________________________
Unsubscribe at: http://dhhumanist.org/Restricted
List posts to: humanist@dhhumanist.org
List info and archives at at: http://dhhumanist.org
Listmember interface at: http://dhhumanist.org/Restricted/
Subscribe at: http://dhhumanist.org/membership_form.php

Editor: Willard McCarty (King's College London, U.K.; Western Sydney University, Australia)
Software designer: Malgosia Askanas (Mind-Crafts)

This site is maintained under a service level agreement by King's Digital Lab.