Dirk Gently’s Holistic Detective Agency or Network Science? Let’s get our networky jargon right!

We’ve all heard it before, people saying stuff like “everything’s connected to everything else”. Most often this phrase is used in the way Dirk Gently would use it in Douglas Adams’ novels. Dirk runs a ‘Holistic Detective Agency’ which means that what sets him apart from other more traditional detective agencies is that he solves crimes by figuring out how all people, things and events are related to each other. In practice, it means he has very few customers since this method can never be consistently applied within an acceptable time limit. In the end Dirk is always reliant on dumb luck to solve his cases, although he backs his decisions up with pseudo-scientific jargony nonsense.

Network scientists including archaeologists and historians rarely talk like Dirk Gently-style detectives (I believe anything between Sherlock Holmes and Poirot best describes the range of fictitious detective analogies for us applying networks to figure out the past, I’m on the Poirot side if you’re wondering). They are more specific in their description of how things are interrelated, thanks to their use of network science concepts: arcs, edges, actors, nodes, centrality, cliques, path length, etc. This sadly does not mean that these academics make themselves any more understandable to their audiences than Dirk Gently does, just because they happen to use jargon usefully and consistently.

When talking to network scientists you really need a dictionary at hand. I often catch myself assuming that many terms like, closeness centrality or ego-network, which have very specific formal definitions are widely known or at least intuitively understandable. It turns out this is definitely not the case: assuming jargon is widely known leads to bad communication, assuming it is intuitively understandable leads to no communication at best and bad science at worst. This issue is particularly problematic for us network scientists who try to contribute to archaeology or history, disciplines where network jargon is entirely unknown.

I am delighted that we are finally starting to overcome this issue. Archaeologists and historians: your networky dictionary has arrived! A while ago we published a special issue of Journal of Archaeological Method and Theory on archaeological network science. Because so many network science concepts were used over and over again in each paper of this issue, we decided to write a network science glossary. You can find this  network science glossary on the Tutorials and Resources page of this blog as well as in the original paper. The glossary was written with an audience of archaeologists and historians in mind. It provides unambiguous non-technical explanations of key concepts as well as a number of examples. All of us who worked on it really hope this will help archaeologists and historians in particular to start critically engaging with all the amazing new work that’s appearing, and to produce some of it themselves.

Check out the network science glossary on the Tutorials and Resources page of this blog.

It was first published as part of the introduction to our special issue of Journal of Archaeological Method and Theory on archaeological network science.

Please cite the glossary as follows:

Collar, A., Coward, F., Brughmans, T., & Mills, B. J. (2015). Networks in Archaeology: Phenomena, Abstraction, Representation. Journal of Archaeological Method and Theory, 22, 1–32. doi:10.1007/s10816-014-9235-6

Book: support networks for persecuted Jews in WWII

I find support and assistance networks extremely interesting! Mainly because they pose so many interesting missing data problems, and as an archaeologist I like a good data problem from time to time. These kinds of networks are very much based on trust, since once a person or connection is compromised it will have disastrous, often murderous, consequences for many in the network. This topic is explored for the case of persecuted Jews in National Socialism during World War II in Marten Düring’s work. He traced a number of different groups of people, how they got in touch with each other, and how they provided assistance to persecuted Jews. Marten told me in most cases the support networks grow slowly and are built on strong trusted relationships. Often new individuals will be introduced to the network through a common contact who has received assistance before and vouches for the individual. However, there are a few cases when individuals gambled and got in touch without a pre-existing well-trusted connection. Such decisions could be disastrous, sometimes leading to the entire network being rounded up by the Gestapo, questioned and sentenced (which is often why these support networks are documented and why Marten was able to reconstruct them). The ‘data problems’ I mentioned are a consequence of the sheer secretive nature of the support network: hiding the fact you offered support to persecuted Jews was a question of life or death. It is particularly hard to reconstruct support networks that were not caught by the Gestapo, and one can only assume that those that were caught are not entirely documented, that there are a lot of missing nodes and links. Marten Düring offers us an in-depth look at a few cases which are particularly well-known, thanks to his rummaging around in archives for years.

I believe this study will prove invaluable for better understanding support networks and the missing data problems they pose. I see particular similarities with networks of the trade in licit antiquities, organized crime and really any type of so-called ‘dark network’. This work offers a reminder of how the study of the past can help us tackle challenges in the present.

Marten’s work was recently published by De Gruyter as a book, check it out here and find the abstract below.

Also keep an eye out for Marten’s chapter in the forthcoming ‘The Connected Past’ edited volume to be published by Oxford University Press early in 2016 🙂

Why did people help Jews hide from the Nazis? This study examines interactions between helpers and aid recipients using the methods of social network research. Based on six Berlin case studies, the author looks at the social determinants for willingness to help, trust formation, network effects, and the daily practice of providing help from the perspectives of helpers and aid recipients.

Three more days to send in your Digital Classicist Berlin abstract

Those trained in the ways of Classics and fearless of the computer beast: submit your abstracts now to the Digital Classicist Berlin seminar series. The deadline for the call for this year’s edition is in three days, at midnight 17 July. I had the pleasure of presenting my work at this venue once (see the video on my bibliography page). Aside from giving the talk in the lovely fancy hall of the German Archaeological Institute which makes you feel like a little Schliemann, it does attract a good audience of experts who are not afraid to ask you the difficult questions (which is exactly what you want by the way). There is plenty of time for discussion after your talk, and I definitely gained some unique insights and new contacts through the experience. Definitely consider applying.

Deadline 17 July

Digital Classicist Berlin website

We are pleased to announce the Call for Papers for the fourth series of the Digital Classicist Seminar Berlin [1]. This initiative, inspired by and connected to London’s Digital Classicist Work in Progress Seminar [2], is organised in association with the German Archaeological Institute and the Excellence Cluster TOPOI. It will run during the winter term of the academic year 2015/16.

We invite submissions on any kind of research which employs digital methods, resources or technologies in an innovative way in order to enable a better or new understanding of the ancient world. We encourage contributions not only from Classics but also from the entire field of “Altertumswissenschaften”, to include the ancient world at large, such as Egypt and the Near East.

Themes may include digital editions, natural language processing, image processing and visualisation, linked data and the semantic web, open access, spatial and network analysis, serious gaming and any other digital or quantitative methods. We welcome seminar proposals addressing the application of these methods to individual projects, and particularly contributions which show how the digital component can facilitate the crossing of disciplinary boundaries and answering new research questions. Seminar content should be of interest both to classicists, ancient historians or archaeologists, as well as to information scientists and digital humanists, with an academic research agenda relevant to at least one of these fields.

Anonymised abstracts [3] of **300-500 words max.** (bibliographic references excluded) should be uploaded by **midnight (CET) on 17 July 2015** using the special submission form [4]. Although we do accept abstracts written in English as well as in German, the presentations are expected to be delivered in English. When submitting the same proposal for consideration to multiple venues, please do let us know via the submission form. The acceptance rate for the first three seminar series was of 41% (2012/13), 31% (2013/14), and 40% (2014/15).

Seminars will run **fortnightly on Tuesday evenings (17:15-19:00)** from October 2015 until February 2016 and will be hosted by the Excellence Cluster TOPOI and the German Archaeological Institute, both located in Berlin-Dahlem. The full programme, including the venue of each seminar, will be finalised and announced in September. As with the previous series, the video recordings of the presentations will be published online and we endeavour to provide accommodation for the speakers and contribute towards their travel expenses.

[1] http://de.digitalclassicist.org/berlin/
[2] http://www.digitalclassicist.org/wip/
[3] The anonymised abstract should have all author names, institutions and references to the authors work removed. This may lead to some references having to be replaced by “Reference to authors’ work”. The abstract title and author names with affiliations are entered into the submission system in separate fields.
[4] http://de.digitalclassicist.org/berlin/submit