[Humanist] 28.40 simulations in history; the model becomes the object

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Wed May 21 02:03:32 CEST 2014


                  Humanist Discussion Group, Vol. 28, No. 40.
            Department of Digital Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org

  [1]   From:    Scott Weingart <weingart.scott at gmail.com>                (263)
        Subject: Re:  28.36 when the model becomes the object of study

  [2]   From:    Dino Buzzetti <dino.buzzetti at gmail.com>                  (199)
        Subject: Re:  28.36 when the model becomes the object of study

  [3]   From:    "Sternfeld, Joshua" <jsternfeld at neh.gov>                   (9)
        Subject: RE:  28.39 pubs: gaming and history

  [4]   From:    Willard McCarty <willard.mccarty at mccarty.org.uk>           (9)
        Subject: digital simulations in history


--[1]------------------------------------------------------------------------
        Date: Mon, 19 May 2014 20:24:34 -0400
        From: Scott Weingart <weingart.scott at gmail.com>
        Subject: Re:  28.36 when the model becomes the object of study
        In-Reply-To: <20140519230918.D15FD65C2 at digitalhumanities.org>


Dear Willard,

There are plenty of situations where the map is used as a stand-in for the
territory. That said, I doubt even the most positivist social scientists
would argue "Everything that can be known about the object is in the model"
- a map as big as that territory would be unfathomably large in our present
historical & digital moment.

An early example that fits your definition of model might be Schelling's
segregation model from the early 1970s. Schelling, a sociologist, built a
few simple simulations of neighborhood dynamics. Simply, people of one
color would move to another location of n% of their neighbors were not the
same color as them. It turned out that this percent could get quite low and
racial segregation would still occur - for example, if each individual only
50% of their neighbors to be the same color as them, complete segregation
would still occur in fairly short order. Schelling used this to show, not
how neighborhoods actually operated, but that an intent of segregation was
not necessary to cause segregated communities - it could happen naturally
under fairly relaxed conditions.

Similar types of simulations have been picked up by sociologists,
archaeologists, philosophers, and even some historians (especially military
historians). For example, Patrick Grim, a philosopher, uses these types of
simulations to show how different communication network structures lead to
different styles of collaboration and information diffusion. Other models
are used to show whether it was strategies or sheer force or luck which led
to certain key military victories, or the extent to which societies relied
on a particular resource, and so forth. It is very much entwined with
counterfactual history, although departs from it in some key aspects.

"But then on the basis of this model you can make inferences otherwise
impossible. These inferences, let us say, check out, make sense, hold up.
They become what one knows." This is the case in each of the examples
above, but I doubt any of the people on the projects actually believe the
model is equivalent to everything they know about reality, or even that the
model itself is the "bottom" of the object of study. The model is simply
there to do (usually) one of two things: show whether a hypothesis is
sufficient to explain an observed affect (without necessarily saying it *was
so*), or to make a prediction/postdiction/interpolation/etc. to infer
something about human interactions that we lack the evidence to observe
directly.

Fittingly, these areas of study are also, usually, where the boundaries
between the humanities and the social science blur. As your friend from
Santa Fe suggests, both the observations and the model are important, and I
think people using models in these ways on both side of the aisle are
deeply aware they draw from each when embarking on these studies.

Best,
Scott Weingart

On Mon, May 19, 2014 at 7:09 PM, Humanist Discussion Group <
willard.mccarty at mccarty.org.uk> wrote:

>
>                   Humanist Discussion Group, Vol. 28, No. 36.
>             Department of Digital Humanities, King's College London
>                        www.digitalhumanities.org/humanist
>                 Submit to: humanist at lists.digitalhumanities.org
>

> --[1]------------------------------------------------------------------------
>         Date: Mon, 19 May 2014 06:33:15 +1000
>         From: Willard McCarty <willard.mccarty at mccarty.org.uk>
>         Subject: when the model becomes the sole object of study
>
>
> Responses to my question about studying models have, if I am
> understanding them aright, usefully illumine a blurry area I was not
> looking for but can certainly use. Yes, when one builds a model there
> are stretches of time when your focus is on that model. These temporal
> stretches lengthen. Conventional wisdom on modelling insists one must
> never forget you're dealing with a model, not with reality (conceived as
> something other than a model, unconstructed, "out there" to be studied,
> solid enough to bruise your toe if you kick it or, as Hacking says,
> real enough to be sprayed).
>
> But I was asking about any situations in which the constructed model
> takes over more or less completely from the modelled object -- no
> cycling back to the original to check things out, no comparing to see
> how well the modelling has approximated reality, since the model has
> become the studied reality. There are certainly situations in the social
> sciences in which details of that which is modelled cannot be observed
> directly. Note, however, that a friend of mine at the Santa Fe Institute,
> a mathematician who studies complex systems, says that these days
> "we use both" such simulations and analytic methods for which the
> object of study is in central vision.
>
> (Let us beware of semantic spread and so loss of power of meaning,
> when "model" becomes anything at all, a concept, an argument etc. For
> our purposes, esp for mine here, let's confine "model" to something
> made of software or some other lego-, Mechano- or tinkertoy-like
> components: something that runs and can be manipulated.)
>
> Let us say that hard work has won you total confidence in your model.
> Everything that can be known about the object is in the model. You're
> confident of that. Let's say you're right. But then on the basis of this
> model you can make inferences otherwise impossible. These
> inferences, let us say, check out, make sense, hold up. They become
> what one knows.
>
> Any examples of that happening in the humanities and interpretative
> social sciences? There are loads of examples in the natural sciences.
>
> I don't think this is an hypothesis about alternative realities,
> parallel worlds and the like. But it would seem very like counterfactual
> history, for example. It would seem to be about an assimilation of
> the computational that enlarges not contracts our intellectual life.
>
> More?
>
> Yours,
> WM
> --
> Willard McCarty (www.mccarty.org.uk/), Professor, Department of Digital
> Humanities, King's College London, and Research Group in Digital
> Humanities, University of Western Sydney

-- 
-scottbot.net  http://www.scottbot.net



--[2]------------------------------------------------------------------------
        Date: Tue, 20 May 2014 11:32:56 +0200
        From: Dino Buzzetti <dino.buzzetti at gmail.com>
        Subject: Re:  28.36 when the model becomes the object of study
        In-Reply-To: <20140519230918.D15FD65C2 at digitalhumanities.org>

I agree with Desmond's distinction--that in my opinion concerns 
models as formal representations. In other words, a formal 
representation containing variables vs constants.  In this 
sense, any technology we use is a model that affects the way 
we can process the information conveyed by the objects we 
represent and we treat as information carriers.

But there is a more general epistemological point of view, 
according to which we can only approach an object through 
a perceptual model.  So I would rephrase Willard's statement 
this way: "one must never forget you're ALWAYS dealing with 
a model, not with reality", contrary to the assumptions of naïve 
realism, so common among natural scientists. 

Yours,     -dino


--[3]------------------------------------------------------------------------
        Date: Tue, 20 May 2014 17:34:57 +0000
        From: "Sternfeld, Joshua" <jsternfeld at neh.gov>
        Subject: RE:  28.39 pubs: gaming and history
        In-Reply-To: <20140519231150.B810265E3 at digitalhumanities.org>

Willard,

Many thanks for alerting us to this article.  You may want to consider a related chapter that appears in the recent anthology edited by Brett D. Hirsch and published by Open Book Publishers, Digital Humanities Pedagogy: Practices, Principles and Politics:  Joshua Sternfeld. "Pedagogical Principles of Digital Historiography." pp. 265-290.  Accessed: http://www.openbookpublishers.com/reader/161.

While the article doesn't focus solely on video gaming, it does address in a broader context many of the same issues raised by Professor Trépanier, specifically the pedagogical intersection of historiography, new media theory, and information studies.  The piece includes a sample syllabus, and I can attest that I had students who selected an historically-themed video game for their final research project.

All the best,

Josh Sternfeld, Ph.D.



--[4]------------------------------------------------------------------------
        Date: Wed, 21 May 2014 09:54:13 +1000
        From: Willard McCarty <willard.mccarty at mccarty.org.uk>
        Subject: digital simulations in history
        In-Reply-To: <20140519231150.B810265E3 at digitalhumanities.org>

I'm happy to add to my list the following:

Champion, Erik. 2011. Playing with the Past. Human-Computer Interaction 
Series. London: Springer-Verlag.

Yours,
WM
-- 
Willard McCarty (www.mccarty.org.uk/), Professor, Department of Digital
Humanities, King's College London, and Research Group in Digital
Humanities, University of Western Sydney




More information about the Humanist mailing list