I’ve been playing with piktochart.com to make a promo poster for my upcoming full year, fourth year, seminar on the antiquities trade. See it in full here: https://magic.piktochart.com/embed/2078049-dark-poster
I’m working on a paper that maps the archaeological blogosphere. I thought this morning it might be good to take a quick detour into the Twitterverse.
‘archaeology’ on twitter
Here we have every twitter username, connected by referring to each other in a tweet. There’s a seriously strong spine of tweeting, but it doesn’t make for a unified graph. The folks keeping this clump all together, measured by betweeness centrality:
Let’s look at american archeology – as signified by the dropped ‘e’.
An awful lot more fragmented – less popular consciousness of archaeology-as-a-community?
Top by betweeness centrality – the ones who hold this together:
#Blogarch on twitter
And now, the archaeologists themselves, as indicated by #blogarch
We talk to ourselves – but with the nature of the hashtag, I suppose that’s to be expected?
Top by betweeness centrality
Top replied to
electricarchaeo (yay me!)
Put them altogether now…
And now, we put them altogether to get ‘archaeology’ on the twitterverse today:
Visually, it’s apparent that the #blogarch crew are the ones tying together the wider twitter worlds of archaeology & archeology, thought it’s still pretty fragmented. There’re 460 folks in this graph.
Top by betweeness centrality:
top hashtags (not useful, given the nature of the search, right? But anyway)
Top word pairs in those largest groups:
(once I saw ‘bars’, I stopped. Archaeological stereotypes, maybe).
Top mentioned in the entire graph
illumynous 9 bonesdonotlie 8
drspacejunk 8 drkillgrove 4
bonesdonotlie 8 capmsu 4
archeologynow 7 yagumboya 3
openaccessarch 7 drspacejunk 3
macbrunson 6 archeowebby 3
swbts 6 allarchaeology 3
archeowebby 6 openaccessarch 3
algenpfleger 5 cmount1 3
youtube 5 brennawalks 2
So what does this all mean? Answers on a postcard, please…
(My network files will be on figshare.com eventually).
I’m a fan of Terry Pratchett. I re-read his novels frequently because each time, I find something new to consider. I was recently reading Lords and Ladies, which is part of the witches’ cycle of stories set in Discworld. This passage resonated:
Cottages tend to attract similar kinds of witches. It’s natural. Every witch trains up one or two young witches in their life, and when in the course of mortal time the cottage becomes vacant it’s only sense for one of them to move in.
Magrat’s cottage traditionally housed thoughtful witches who noticed things and wrote things down. Which herbs were better than others for headaches, fragments of old stories, odds and ends like that.
[...]It was a cottage of questioning witches, research witches. Eye of what newt? What species of ravined salt-sea shark? It’s all very well a potion calling for Love-in-idleness, but which of the thirty-seven common plants called by that name in various parts of the continent was actually meant?
The reason that Granny Weatherwax was a better witch than Magrat was that she knew that in witchcraft it didn’t matter a damn which one it was, or even if it was a piece of grass.
The reason that Magrat was a better doctor than Granny was that she thought it did.
Take a look at any github page, and examine the readme page. Strikes me, there’s a lot of the witches about these code repositories. The parallel isn’t perfect, but I feel rather like poor Magrat. For instance (and taken at random*):
Install a Java Development Kit.
Install Git.git clone https://github.com/overview/overview-server.git
Which development kit? What version? How many flavours of PostgreSQL are there? What do I do with that? As I fumble towards dim understanding, I figure the folks who are building these things are more like Granny, and understand that any will do the trick, because they know what to expect and how to fix it if it goes wrong. Me, I need the right version the first time, because otherwise I’ll just make a hash of it – and I’ll have to teach it to someone! (Although I can git clone from git bash with the best of ‘em – now!)
I don’t have the tacit knowledge of experience built up yet. There’s just so much to learn! Like Magrat, I can write it all down, spell it all out, and in doing so, I’ll eventually become like Granny, where it just flows.
I look forward to that day. But for now, I’ll keep engaging in my research witchcraft, figuring out the bits and bobs that those far more clever than me have devised, and reporting back what I’ve found.
*Well, not totally at random. It comes from the Overview Project who have taken pity on me (and others!) and have worked very hard indeed to simplify setting up a development environment for their text analysis server, ‘Overview‘. Thank you Jonathan and Adam! I’m learning a lot from chatting with these guys, as they shepherd me through the process. I’ll be posting on that process soon, pointing out some of the tacit bits I found I had to uncover in order to make it work. Their platform, conceived for journalists should also migrate its way into history & archaeology as I think we’ll find it very useful!
I like 123D Catch, but there is the whiff of ‘black-box’ about it all. Sometimes, you’d just like to know what’s going on. There may also be times when, for various reasons, uploading data to a cloud service hosted in another country is just not the right solution for you.
There are many open source products though; right now I’m playing with VisualSFM. Download and install it; then download CMVS. Extract the zip. Within it you will find folders for various operating systems. Find yours, and copy the files within, to the VisualSFM folder.
Now you’re ready to go, as per the image below. Here’s a longer tutorial too.
You might however find it easier to use this bundle of all the bits and pieces you need, if you are familiar with python. Extract the zip, grab the folder that corresponds to your operating system, and move it to C:\ . Install python (I’m using Python 2.7). Then, open the command prompt (type ‘cmd’ in the ‘run’ box, Windows), navigate to the folder (On my machine, it’s now in c:\bundler, so I had to type:
C:\users\Shawn Graham> cd ..
C:\users> cd ..
C:\> cd bundler
…and the magic begins to happen. I got an error at first: ‘blah blah blah PIL missing blah blah’. PIL stands for Python Image LIbrary. Go here, grab the correct version, download it, and double-click to install. Then try again with the RunBundler.py command above.
So that’s running now on my machine; I’ll update here if it all goes wrong – or if indeed it all goes right!
I found this today: http://www.networkedcorpus.com for visualizing the results of MALLET topic models, by reference to exemplary passages. Its creators explicitly contrast this with index building by hand… more about all that later; here’s how to get it working for you.
See the documentation at https://github.com/jeffbinder/networkedcorpus
Install python 2.7, numpy, and scipy. Download the installers, and follow the instructions.
Then download the networked corpus zip file, and unzip it somewhere on your machine.
In the networked-corpus folder, there is a subfolder called ‘res’ and a script called gen-networked-corpus.py. Move these two items to your Mallet folder.
Generate a topic model as you would normally do, from the command line.
Then, type at the command prompt:
gen-networked-corpus.py --input-dir <the folder with your original texts in it> --output-dir <the name of a folder you'd like all the output to be in>
Navigate to that folder in your browser, open the index.html and voila.
Over the Thanksgiving weekend, the FutureFunder campaign in honour of my grandfather, to create an undergraduate research fellowship in digital history, achieved its funding goal.
I wanted to thank everyone who contributed. Whether that contribution was through donations, through sharing on social media, or through sending me emails making ‘hey, you should really talk to [person x]…’ connections, this could not have happened without the support and buy-in of the DH community, my colleagues, and the alumni of the History department. Kylie, Pia, and Ryan in University Advancement were also tremendous supporters, helping garner national media attention, making connections, and coming up with novel ideas about how to promote it further.
Thank you, all!
Now begins the *really* fun part! Over the coming weeks, I’ll be working with University Advancement, our department’s undergraduate committee, and digitally-inclined folks hither and yon to set up the formal parameters for awarding the fellowship. One of the conditions of the fellowship would be for the student to maintain an active research blog, where she or he would detail their work, their reflections, their explorations and experiments. It would become the locus for managing their digital online identity as a scholar. I think I will recommend that the student use Reclaim Hosting to do this, since RH’s whole raison d’etre and my own sense of what students need to be doing online mesh very well (and see this post on the future for RH!)
We already have great, digitally-inclined undergraduate students in the history department here. Rob is a HASTAC Scholar. Hollis organized a THATCamp. Oliver is getting into data mining. These are just the three easiest for me to link to. Others like Devin and Joe have done fantastic thinks using Voyant Tools. Matthew, Julia, and Zack have gone into networks in a big way. Allison has been developing expertise with Omeka.
I’m excited to see what’ll happen next. Thank you, everyone, for supporting this Fellowship!
Here is a common plot in sci-fi literature and movies (based on a popular physics model): the world you know is but one in an endless range of parallel universes, where each one is slightly different. Who would ever have thought this would be a good starting point for archaeological discussions? Yet the meeting in Durham I recently attended showed that parallel universes might have more in common with archaeology than we think.
I was invited by Rune Rattenborg to join a workshop in Durham called ‘A Thousand Worlds: Network Models in Archaeology’. This concept of a thousand worlds can be interpreted in an archaeological research context in different ways. On the one hand, and most similar to the sci-fi parallel universes plot, you could think about the many different reconstructions of past realities that could all explain a single archaeological pattern. Literally thousands of hypotheses could be raised to explain a certain pattern, each of them suggesting different mechanisms driving human behaviour and ultimately its expression in the archaeological record. On the other hand, you could think about the many academic perspectives archaeologists find useful for understanding the past. Perspectives ranging from highly quantitative (you can place me in that camp) to very qualitative, from local to global, from scientific to philosophical, and from an explicitly present-day perspective to attempting to recreate past perspectives. Each one of these is a valid way of thinking about past human behaviour and behavioural change (or rather every configuration or combination of these perspectives).
Both of these interpretations motivated Rune to title his workshop ‘A Thousand Worlds’. He noticed that archaeologists interested in questions of past connectivity and those of us using network perspectives often address the challenges we are faced with in very different ways. The only common ground of most network perspectives seems to be that the relationships between entities are considered crucial to understanding the behaviour of these entities. For example, the romantic relationship between two individuals will affect the decision to stay in and watch a Hugh Grant romantic comedy or to go out for a beer with the guys. But Rune also noticed that each perspective allows for a wide variety of reconstructions of past realities. These two issues seem to confuse archaeologists who might be interested in using such a network perspective in their archaeological research. I totally agree with Rune’s motivation to create some order in this chaos. The main questions of this workshop therefore were: what different network perspectives are out there? What rules govern them? What do they allow us to do that we could not do before? And what are their limitations?
To some extent the meeting was successful in addressing these questions. A number of very different perspectives were discussed by selected proponents: I introduced an extremely formal network science approach, which was discussed rather more pragmatically by Anna Collar; Michelle de Gruchy highlighted some interesting challenges in a geographical context; another group of presenters (Kristoffer Damgaard, Eivind Heldaas Seland, Sofie Laurine Albris, Rune Rattenborg) used the concept of connectivity and explored how it could be reflected in archaeological and literary sources. Finally Ronan O’Donnell introduced the actor-network theory (ANT) perspective through a fascinating case study on a post-Medieval landscape in Northumberland, UK, from which the strong difference between the aims of the ANT and network science research perspectives became particularly clear.
Nevertheless, by the end of the meeting it became clear that we were not entirely successful in addressing the many questions we set out to answer. Eivind Heldaas Seland skilfully summarised each paper and formulated three key questions that require more attention: how can these different perspectives and approaches usefully work together? What is the added value of some of these compared to a more traditional description of our sources? How can we better use these perspectives in the future? The fact that we were unsuccessful at addressing these questions shows how complex and non-trivial they are (and we also ran out of discussion time). But for what it’s worth, I take this opportunity to share some of my thoughts on these questions, combined with some of the points I picked up from others during the discussion.
First of all, I believe the first question presents the false impression that the different network perspectives can and need to work together. I would argue that, many network perspectives do not need to and most of them do not work well together at all. This is because some of them (like ANT and network science) are designed to address very different questions. But even those approaches that have more in common, like the quantitative vs. qualitative use of network science, don’t necessarily need to be combined into an almighty network approach. There is no need for a great unifying theory or method in archaeology, not even for one that just focuses on questions of connectivity. Rather, I consider the different network perspectives as tools that function according to certain rules, and once these rules are known the tools have a potential to make small but crucial contributions to our knowledge of the past. I believe that if we are to ever achieve the full potential of these exciting new approaches for archaeology we will need to first critically explore them in isolation.
Secondly, the added value of these perspectives is more obvious than how they should be applied. Many in the audience seemed to agree that the concept of the network itself is a powerful tool to think with. It forces us to consider the potentially important role played by relationships between entities (however defined: humans, molecules, parallel universes), which might allow us to ask and answer new questions. For me the added value lies in the recognition that all archaeologists make assumptions about the nature of such relationships when they formulate hypotheses about past phenomena. It can be useful to think about these assumptions in terms of network concepts and, most importantly, there is a real need to be critically aware of their existence and formulate them clearly. Network science can help archaeologists to think about their assumptions of past relationships, to formally express them (in words and/or in numbers), and to evaluate their implications for past behavioural change and its reflection in the archaeological record.
Finally, the “better use” of such approaches and perspectives is not optional, it is necessary if they are ever to become useful within an archaeological research context. However, a critical use and application is not just a critical awareness of the rules that govern them. Rather, an equal if not larger effort should be afforded to the archaeological interpretation of network science results, or the differences in the interpretative process that a networks perspective implies. I believe none of the scholars that attended the Durham meeting would disagree with that. The studies they presented could be roughly divided into two groups: those that THINK through network and those that DO networks. I believe the former is more important than the latter, because there can be no doing without thinking. Although this sounds like an obvious statement it is worth emphasising it because the use of quantitative network analysis is too often treated like a “black box” approach, which it is not. Every network science study in archaeology, no matter how quantitative, aims to better understand (aspects of) past phenomena. When doing so, the scholar formulates a hypothesis, expresses their assumptions about past relationships and their roles, or at least clearly defines what they mean by the network concepts they use. Only after this phase of network thinking can a scholar proceed to network doing, which involves representing hypotheses/assumptions/the archaeological record as network data (points and lines, and what they mean). The ability to use advanced quantitative tools should not be an excuse for the post-hoc imposition of a theoretical framework that fits the results nicely; nor should the appeal of using fashionable network concepts lead to reluctance to formally express what is meant by them and to evaluate their implications for understanding past phenomena.
Even though none of the three key questions about the role of the networks perspective in archaeology can be conclusively answered at this time, I felt that its future is nevertheless bright. The diversity of possible approaches and perspectives is encouraging and will lead to critical research that promises to help archaeologists better evaluate what approach is useful for their studies of past connectivity, and what is not. Some of these approaches might require multi-disciplinary collaboration, especially the more scary and maths-heavy techniques in the network science toolbox. But archaeologists should never be tempted to outsource the network thinking part of the process. Critical knowledge of the archaeological literature and data leads to an awareness of the relevant research questions, and the same knowledge will lead to valuable interpretations of analytical results and research processes. There might be a thousand pasts out there, and there might be a thousand ways of reaching them, but this quest will always need to be undertaken by archaeologists.
Selected relevant publications:
Brandes, U., Robins, G., McCranie, A., & Wasserman, S. 2013. What is network science? Network Science 1(01): p.1–15.
Brughmans, T. 2013.Thinking through networks: A Review of Formal Network Methods in Archaeology. Journal of Archaeological Method and Theory.
Knappett, C. 2011. An archaeology of interaction. Network perspectives on material culture and society. Oxford – New York: Oxford University Press.
About the author:
Tom Brughmans is currently finishing a PhD in archaeology as a member of the Archaeological Computing Research Group at the University of Southampton and at the University of Leuven. His main research interest is to explore the potential of network science for the archaeological discipline. Tom blogs at archaeologicalnetworks.wordpress.com
I’m playing with NodeXL, looking at the ways ‘Canada’ (and Canadian history) are tagged on Flickr photos. Above is a pretty first pass. Here’s the zoomable pdf for your enjoyment.
In my HIST2809, Historian’s Craft this term, I’ve been asking for students to maintain a repository of their learning using Omeka.net. Every time we do an assignment or an exercise, that work is meant to go into their repository. The final exercise in the course is to build an exhibit of their learning progress. Here is the assignment prompt; I thought folks might be interested.
Omeka is not just for storing items. It is also for exhibiting them. Exhibits are built around the idea that you are telling a story with these items. You will have collected many different items over the course of this term.
Your exhibit should tell the story of your learning in HIST2809.
Your exhibit should:
- be built around at least five items, including your assignment 1 & 2 originals
- incorporate (either copy or link to) 2 items from SOMEBODY ELSE’S Omeka repository, providing citation to the original item location (see the forum at the top of cuLearn for the URLs to other people’s repositories)
- link outwards to at least three other sites or sources (eg an item in a library catalogue, public zotero page, Wordle, Voyant Tools corpus, existing online exhibit, photo gallery… etc).
The point of this exercise is:
- To learn how to make an exhibit in Omeka, which is an industry standard in cultural heritage circles.
- To see how the assumptions built into the platform constrain or enable various kinds of storytelling. ALL digital resources have assumptions about how the world should work built into them, from Google to JSTOR to the Digital Public Library of America. Working inside Omeka.net gives you a glimpse of how these things work from the creator, rather than consumer, side.
- To learn how to analyze digital sources as we would any other source
WHAT YOU WILL SUBMIT:
1. A 500 – 1000 word reflection that analyzes your exhibit under PAPER headings, with the URL to your exhibit, with a final section discussing your process in building the exhibit.
We will be grading this document, not the exhibit itself.
So: have a title page with your name on it, your exhibit title, and the direct URL to that exhibit.
Then, for the reflection/analysis, discuss your exhibit AS IF you were considering it as a primary resource, explicitly using the PAPER headings.
You will tell us about
- your purpose (obviously, you want to tell us about the evolution of your learning, but you might have other goals, too, that are expressed through careful use of colour, or … ),
- your argument (the way you arrange things, force particular paths through the material, or…),
- presuppositions (your worldview as it pertains to the role/value of digital work, perhaps; you might feel that this is a waste of time, or you might love playing and learning with digital tools; or you might be ok with digital but see them as mere tools whereas someone else might think of them more like paint & clay, as things to create with: how does that effect what you’ve done or reflect within it?),
- epistemology (what has been chosen? what has been left out? why? to what end?)
- and of course, related…
(The ‘R’ part might be the hardest: read, cite, and consider your exhibition in the light of this article http://dare.uva.nl/document/215092 Jose van Dijck, Search Engines and the production of academic knowledge. International Journal of Cultural Studies,13(6):574–592.)
- INCLUDE a final section that tells us about the problems/potentials you experienced in building this assignment. In what ways does Omeka lean towards particular kinds of stories or paths through material? Does this matter?
A rubric will be provided. The balance of points will be towards your reflection/analysis, rather than the aesthetics of your exhibit.
-> You could have a bare-bones, ugly, exhibit: that would be perfectly ok. We’re not grading on aesthetics. But aesthetics do make a difference for the visitor to your site- an analysis of a bare-bones, ugly exhibit would need to reflect on what that design choice does, for the visitor, in terms of your purpose, argument…
I certainly want you to be thinking especially carefully about the ways ‘argument’, ‘epistemology’, and ‘related’ are reflected in your exhibit.
I’m continually fascinated by ways digital media can expand who gets to be a historian, who gets to be an archaeologist. Crowdsourcing expands our readership, too.Open peer review projects allow the potential readership for a volume to have a dialogue with the authors while the project unrolls.
My Futurefunder campaign adds a new facet to this. I’m trying to crowdfund direct tax-deductible donations to a fund that would support undergraduate students as they work on various digital history and humanities projects around the department. The Dean of the Faculty of Arts will match funds if we reach the halfway mark ($2500); the fund is currently only about $800 shy of that point!
I needed to do this. I kept finding that I was pulling funds from various nooks and crannies to send students to THATCamps, to help them get to DHSI, to set up laboratories for exploring data mining, to publish and work with me on projects. I found I was spending weeks a year writing research grants that, when boiled down to their essence, were all about finding funds to train students. This, it seems to me, is a very appropriate idea to take directly to the public, rather than the Tri-council agencies. I was very excited to be interviewed by the Globe and Mail about the project (the story appeared this past Saturday), and the fund has really picked up steam. I would be happy to chat with folks who are interested in this campaign (this experiment!). I would be extremely happy to chat with folks about the amazing work the undergraduates around here do, in digital history.
One last push folks, one last push!