[Humanist] 27.163 readings on modelling

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Sat Jun 29 00:12:30 CEST 2013


                 Humanist Discussion Group, Vol. 27, No. 163.
            Department of Digital Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org



        Date: Fri, 28 Jun 2013 12:02:17 -0400
        From: Wendell Piez <wapiez at wendellpiez.com>
        Subject: Re:  27.147 readings on modelling
        In-Reply-To: <20130624205458.9796C2CDA at digitalhumanities.org>


Dear Willard, Michael and HUMANIST,

Joris makes some excellent points. The most important point he makes
is that modeling is central to DH and ubiquitous, but mostly implicit,
and almost always poorly understood. That's the first thing to say: we
are just starting to figure it out.

An analogy might be the development of a cuisine. In early days there
are practices, themes, recipes more or less codified, common
ingredients and techniques, lots of lore, some theory and principles,
and even a bit of chemistry. But there's nothing like a comprehensive
overview anywhere about what cooking is and how French cuisine is
similar to and different from Chinese -- to say nothing about what
grilling has in common with baking, vs how it is different -- how both
are "cooking" and yet they are very unlike, no one could or would
mistake one for the other, and practitioners of each are keen to
distinguish themselves from the other (even to the point that French
chefs may sometimes consider themselves to have more in common with
Chinese chefs than with French bakers). Cooking, of course, is way
ahead of digital media in all this.

To make it worse, in digital systems architecture and development
generally and in DH especially, models are built out of ... models.
It's models all the way down. (The data type primitives wired into the
languages in which applications are developed are models too.) Indeed,
one of the features of the layered architecture is that the designers
of models at one level can be experts in the particular capabilities
and affordances of the models out of  which their own models are
built, while paradoxically oblivious to their rationales and necessity
(or lack thereof) -- not infrequently, quite deliberately so, for all
that they (we) become religious about them. So one can design a
template for a family of Microsoft Word documents without knowing the
first thing about the application's data model or its underlying
object model (as opposed to its user interface and feature set), and
how they both enable and limit what can be done. Just so, TEI
practitioners may take XML for granted -- for TEI, XML is an
externality, and can (must) simply be assumed, which makes for
difficulties in communicating with those whose modeling technology of
choice is tabular data in spreadsheets, or relational databases, or
interlinked networks of HTML/RDFa-encoded texts. Etc.

For Michael, here are links to a couple of my own offerings in this area:

"Three Questions and One Experiment":
http://www.wendellpiez.com/resources/publications/ThreeQuestions-OneExperiment-slides.pdf.
An informal overview of the problems of data modeling for DH,
presented at the Brown University Workshop on Knowledge Organization
and Data Modeling in the Humanities (March 2011).

"How to Play XML: Markup Languages as Nomic Game". Paper at
http://balisage.net/Proceedings/vol3/html/Piez01/BalisageVol3-Piez01.html,
slides at http://www.wendellpiez.com/resources/publications/HowtoPlayXML-slides.pdf.
About markup technologies specifically, but broad in its implications.

Also very much worth a read, by Michael Sperberg-McQueen: "Runways,
product differentiation, snap-together joints, airplane glue, and
switches that really switch", at
http://conferences.idealliance.org/extreme/html/2004/Sperberg-McQueen02/EML2004Sperberg-McQueen02.html.

Best regards, Wendell

On Mon, Jun 24, 2013 at 4:54 PM, Humanist Discussion Group
<willard.mccarty at mccarty.org.uk> wrote:
>                  Humanist Discussion Group, Vol. 27, No. 147.
>             Department of Digital Humanities, King's College London
>                        www.digitalhumanities.org/humanist
>                 Submit to: humanist at lists.digitalhumanities.org
>
>
>
>         Date: Mon, 24 Jun 2013 09:53:12 +0200
>         From: Joris van Zundert <joris.van.zundert at huygens.knaw.nl>
>         Subject: Re:  27.141 history of markup? readings on modelling?
>         In-Reply-To: <20130620200510.B4EA53A3A at digitalhumanities.org>
>
>
> Dear Michael,
>
> If you were looking for more concrete digital and/or computational models
> for topics, events, text, style, and narrative etc., there's a lot to go
> round and Wikipedia (e.g. http://en.wikipedia.org/wiki/Topic_model) is good
> companion for starting a dive into literature I think.
>
> A bit counter to "quantitative/tractable models of research objects" you
> may also be interested in chapters 9 and 10 of Handbook of Knowledge
> Representation. More specifically chapter 10 shows the potential for
> "qualitative computational models", which to me suggests a closer match to
> humanities modeling:
>
> - Forbus, Kenneth D. (2008), "Qualitative Modeling". In F. van Harmelen et
> al. (eds.), Handbook of Knowledge Representation. (p.361)
>
> - Struss, Peter (2008) Model-based Problem Solving, "Qualitative Modeling".
> In F. van Harmelen et al. (eds.), Handbook of Knowledge Representation.
> (p.395)
>
> For some background/rational as to why qualitative modeling and
> computability is of essence I think Johanna Drucker's Humanities Approaches
> to Graphical Display in DHQ (
> http://www.digitalhumanities.org/dhq/vol/5/1/000091/000091.html) is still
> excellent.
>
> Modeling as part of the dynamic between humanities and computer science has
> been little studied to uncover its mechanisms and potentials. Althouh
> modeling arguably is applied in any DH project (and I'd argue in any decent
> research design actually), mostly DH modeling practices seem tacit and
> implicit.
>
> I think this is actually ground that remains to be covered, at least in
> (digital) humanities. That is: there are a lot of practices people apply
> for capturing, describing, and encoding topics, events, text etc.. These
> constitute concrete models, or model implementations, I guess. However
> there is not much explicit reflection on what constitutes modeling, or what
> a Model is for that matter. What do researchers assume modeling to be, what
> do they think it facilitates? Nor is there a lot of literature on multiple
> models for the same domain, and how models in such cases are confronted and
> compared to each other. There is even less, I suspect, on the analytic
> capabilities and underpinnings of various models applied in the humanities.
>
> And actually I'd be very interested in other people's views on this.
>
> Some (very) general pointers:
>
> - McCarty: “either a representation of something for purposes of study, or
> a design for realizing something new.”
> (Companion to Digital Humanities as you quote)
>
> - Minsky: “To an observer B, an object A* is a model of an object A to the
> extent that B can use A* to answer questions that interest him about A”
> (Minsky, Marvin L. (1995). Matter, Mind and Models.)
>
> - Root-Bernstein: "The point of modeling is to depict something real or
> imagined in actual or hypothetical terms in order to study its structure or
> function." (Robert Root-Bernstein and Michele Root-Bernstein (2003).
> Intuitive Tools for Innovative Thinking. In Larisa V. Shavinina (ed.),
> International Handbook on Innovation.)
>
> The last also has: "Modeling, as many practitioners have said, is like
> playing god, toying with reality in order to discover its unexpected
> properties."
>
> Best
> -- Joris

-- 
Wendell Piez | http://www.wendellpiez.com
XML | XSLT | electronic publishing
Eat Your Vegetables
_____oo_________o_o___ooooo____ooooooo_^





More information about the Humanist mailing list