[Humanist] 26.298 social networking in Beowulf, the Iliad & the Ulster Cycle

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Thu Sep 13 08:24:24 CEST 2012

                 Humanist Discussion Group, Vol. 26, No. 298.
            Department of Digital Humanities, King's College London
                Submit to: humanist at lists.digitalhumanities.org

        Date: Wed, 12 Sep 2012 14:07:04 +0200
        From: Tom Brughmans <tom.brughmans at yahoo.com>
        Subject: Re: [Humanist] 26.297 social networking in Beowulf,the Iliad & the Ulster Cycle
        In-Reply-To: <20120912052308.526F9290CDC at woodward.joyent.us>

Dear Willard,

That article sparked a lot of discussion on popular media channels (see
Guardian and Scientific American posts below) as well as some
multi-disciplinary discussion lists I am part of. This fact alone makes it
already an interesting paper. Below I summarise some of the comments I made
about this paper on other lists, apologies for cross-posting!

I think it is creative to trace networks in ancient sources, and I think
that might tell you something about how the story is constructed and the
role of characters within the narrative. Maybe. But the characteristics the
authors suggest are indicative of real social networks have been identified
in other networks as well, like the airport network or the WWW. They are
just some properties that are common for many real-world networks. What that
structure means is a completely different thing. Maybe the Iliad was just
created in a way that 'makes sense' and its original audiences could
associate with (and this is an argument for the structure and narrative of a
text, more than whether it's fact or fiction). That still does not mean it
is based on a real social network and I don't think it works as an
indication for how real an ancient text is.

As far as power law policing goes: I don't think any of their degree
distributions show a power law at all. There does not seem to be enough data
and the tails of the distributions just don't seem to fit very well. Just
saying they are right skewed distributions indicative of some kind of
hierarchy in the network is as far as you can take it (and good enough for
their aims it sounds to me).

The authors provide no references to the archaeological sources that support
the historicity of the events in the Iliad. But proving archaeologically
that there was a war in Troy and claiming that a specific social network
structure connected its protagonists are two very different things. I would
like to see the archaeological data for that...

I think if the authors would have pitched this article as an exploratory
exercise it would have been fine, rather than claiming results about the
fact/fiction nature of these texts.

The paper was also featured in the Guardian and it is worth going through
the comments. Aside from all the usual rants by insane people, one of the
authors, Ralph Kenna, actually commented on the Guardian article about the
paper. Kenna does seem to defend the historical (and even archaeological)
significance of their findings, rather than pitching it as an exploratory
exercise. Kenna also announces this message will be spread to a Humanities
audience in a different paper soon.


And here is the full quote of Kenna's comment:


Hello John,

Ralph Kenna here, one of the paper's authors.

I want to respond to a couple of the points that you make in your article
and give a little bit more insight into what our paper has found.

The first thing to say is that this research was not funded by the taxpayer.
The work was funded by the Leverhulme Trust, whom we acknowledge in the
paper, so can you please correct the remark about it being funded by
readers' tax contributions?

Secondly, your remark about 'tell-tale signs of being fictional' seems to
miss the entire point of the research. Our research did not focus on events
or individuals in either contemporary texts like Harry Potter or the ancient
texts we focused on. Moreover, the fictitious texts were only used as a
control group. The main purpose of the research was to compare the three
ancient texts from a social networks point of view, using standard tools of
statistical physics.

What we've found, which has been of wide interest to the general public and
warmly welcomed by archaeologists, is that the texts Iliad and Beowulf
contain something like real social networks, as does (and this is contrary
to mainstream academic opinion) large parts of the T'ain. This is of serious
historical significance and I do wonder why such flippant comments are made
about a serious effort to evaluate our ancient history.

The main goal of our research was to evaluate the artificiality of the
T'ain, a culturally significant Irish text. Mainstream academic opinion
claims it is wholly artificial. We have found, however, that the majority of
links between characters in the text does reflect real-life social networks
and this opens up a new area of investigation for Celticists and other
scholars. Indeed, a future less technical paper will be prepared for a
humanities audience.

Physics - and in particular sociophysics - uses network theory to help us
understand society in new and exciting ways. Dismissing this kind of
research as somehow frivolous and unnecessary is akin to a celebration of
ignorance. Our research reflects new and exciting perspectives, through
which we are able to evaluate our current and historical culture.

If anyone reading this comment would like to know more about the research,
please do read the paper http://iopscience.iop.org/0295-5075/99/2/28002 .



Maria Konnikova wrote something for Scientific American recently that uses
this paper as an example of how humanists are increasingly using scientific
methods, and why that is bad:

Carron and Kenna are not humanities scholars and (as far as I know) no
humanities scholars really contributed to their research. They are
physicists publishing in a Physics journal. So arguably this is the worst
paper to choose when you want to make the point that a scientific method for
humanities research is not a good thing. Aside from that, as we heard on
this discussion list and on others, their work just does not stand up to
close scrutiny. Their network analysis and their "humanist" interpretation
of the results are very questionable. So I have the impression that
Konnikova recently read the paper which triggered her to write out some
ideas she already had about scientific methodologies in the humanities. She
does not seem to be a specialist in network science or classics, but she
does raise some key concerns we also seem to have on these lists. But still,
this is a really bad choice of examples to support her point.

Aside from the Carron and Kenna article, I think Konnikova wants to make an
argument that sounds a bit too much like defeatist relativism to me. I don't
agree that all scientific practice is necessarily reductionist in its
purpose, although it might be usefully reductionist in its approach. Not
qualifying and contextualising scientific results is bad science by all
standards. But then again, there are reasons why "traditional" humanities
research is performed the way it is. Probably because it is useful in
providing that qualification and contextualisation. Her arguments why
history is a good example of this sounds more like a type of sampling issue
and can be applied to all social sciences if not many natural sciences. I
think it is worrying to label the trend towards a "scientific" humanities a
bad one, because clearly the best insights are gained when approaches are
mixed (which Konnikova seems to acknowledge at the end of the article).

What do you think?

I am curious if Carron and Kenna (2012) will spark even more discussion. If
anything, their article has been a very welcome contribution to science
already, because it sparked heated debate across disciplines!

Best wishes,


PhD studentArchaeological Computing Research Group
University of Southampton

On 12 Sep 2012, at 07:23, Humanist Discussion Group wrote:

>                 Humanist Discussion Group, Vol. 26, No. 297.
>            Department of Digital Humanities, King's College London
>                       www.digitalhumanities.org/humanist
>                Submit to: humanist at lists.digitalhumanities.org
>        Date: Wed, 12 Sep 2012 06:20:37 +0100
>        From: Willard McCarty <willard.mccarty at mccarty.org.uk>
>        Subject: social network analysis
> A colleague and old friend has alerted me to an article in the NY Times
> describing the work of two scientists who have applied social-network
> analysis methods to ancient stories - Beowulf, Iliad, and the Ulster
> Cycle (Tain Bo Cualinge - Cattle Raid of Cooley). See
> http://www.nytimes.com/2012/09/09/opinion/sunday/the-social-networks-of-myths.html
> for more. The original paper, attached, is Pádrig Mac Carron and Ralph
> Kenna, "Universal Properties of mythological networks", Europhysics
> Letters [EPL] 99 (2012): 28002
> (http://iopscience.iop.org/0295-5075/99/2/28002/pdf/0295-5075_99_2_28002.pdf).
> See below for the attachment link.
> The aspect that seems interesting to me is the notion of analyzing the
> social links between characters, a kind of structural analysis of basic
> relationships - who is connected to whom - as a general measure of the
> possible historical nature of the story. The authors compare the epics
> to Facebook or whatever example of actual social networks.
> Yours,
> WM
> -----
> *** Attachments:
>    http://www.digitalhumanities.org/humanist/Attachments/1347427243_2012-09-12_willard.mccarty@mccarty.org.uk_2192.2.pdf
> --
> Willard McCarty, FRAI / Professor of Humanities Computing & Director of
> the Doctoral Programme, Department of Digital Humanities, King's College
> London; Professor, School of Computing, Engineering and Mathematics,
> University of Western Sydney; Editor, Interdisciplinary Science Reviews
> (www.isr-journal.org); Editor, Humanist
> (www.digitalhumanities.org/humanist/); www.mccarty.org.uk/

More information about the Humanist mailing list