These are my notes about the 2016 Chicago Colloquium on Digital Humanities & Computer Science held at the University of Illinois, Chicago.

Note: As always with such notes, they will have all sorts of unintended intertextual allusions due to my poor typing and the challenges of listening and typing a the same time.

Also, due to flights I missed the Friday afternoon workshop and most of the Sunday sessions.

Saturday, November 12, 2016

Karen Sieber: Visualizing the red Summer: facilitating digital research on the riots and lynchings of the Red Summer of 1919

Sieber showed us a neat site about the summer of 1919: http://visualizingtheredsummer.com/ . The archive itself has a nice interface where you see everything and then filder down.

Xi Rao and Jodi Houlihan: Southerly Wind: Exploring Chicago Chinese Immigrants’ Struggles in the early 1900’s

Southerly Wind is a game about Chicao's Chinatown in 1909. They used a murder case which was linked to Chinatown and led to tensions. They developed their prototype with Twine. They showed sketches which were quite compelling.

Paris Papamichos Chronakis: Bonds of Survival: Reconstructing the Social Networks of Holocaust Survivors

Chronakis started with a critique of how the death camps are studied. He wants to look at the social world - the social networks - of the camps. This is a way to recover the humanity of those in the camps. He wants to understand survival through the eyes of the survivors. To do this they gathered data about Thessalonican (?) Greeks in Auschwitz and created a network graph. This is not a quantitative project. They used testimony to identify social circles and investigated the nature of the circles. They looked at language/communication, common origin, and type of relation (family, neighbor ...).

He was fascinating on the politics of identity in the camp. The Greeks were isolated as they didn't speak Yiddish and weren't even considered Jews by their fellow campmates.

They have over 200 people in their databases. He showed visualizations that highlight the cosmopolitan nature of the community. What can the network visualization do? He believes it corrects the emphasis on the individual of other technologies (like videotaped testimony). The network viz offer a tool to move away from the perpetrator's gaze and look a the social relationships of the prisoners. They let us understand how identities were both forced on perpetrators, but also negotiated among prisoners. They let us understand what it meant to be a human in Auschwitz.

There were interesting questions for all three of the presenters. One question was about the role of narrative in both the design of games and in the

Another question for the first speaker was if she was going to interpret the archive on the site. Sieber responded that she wanted to create a site for students to develop their views. It was developed for teaching. I'm guessing she will write about it separately.

John Shanahan, Robin Burke, Ana Lucic and Megan Bernal: Reading Chicago Reading: Capture and Analysis of City-scale Literary Events

One Book One Chicago is a civic initiative that makes a natural experiment on reading. There is also data from libaries about circulation, demographics, text (HathiTrust), and social media. They made clear that they don't know who is signing out the book so they have to look at the demographics of the branches.

The goal of the project is to look at the sociology of reading. Ideally they could have a tool that could model how a book would do if added to a library. As OBOC is a repeating event it allows experimentation - they can test their models.

They then talked about running PCA on the branch demographics. They then clustered branches/books. It went by quickly, but it was clear that they had found some really interesting patterns. They also showed social media traffic around the OBOC event.

Christopher Forstall: “The Lannisters send their regards”: Intertextual Tools and Theory in the Age of Fandom

Forstall and his collaborator Walter Scheirer are interested in intertextuality. He started by asking us how many of us recognized the allusion to the Lannisters. This is important in Classics as there is a lot of allusions, references, quotation and so on without citation. Readers of the time would recognize the allusions. If we want to reconstitute the experience of reading at the time we need to model intertextuality.

They are doing "quantitative intertextuality" which is a dialogue between macroanalytical techniques and intertextual theory.

I can't help but wonder if we have a model for random allusion (as in unintentional.)

Then he talked about the Tessarae project at Buffalo where they used sequence alignment to try to find allusions. The problem is that not all two word passages are allusions - so how to tell what is an allusion. They created a model using student tagging and commentaries.

He then talked about Game of Thrones and how they tried their tool on the script and the social media.

Ilana Miller: What does a literary trend look like? Publishing books about Jews in communist Czechoslovakia

Miller is interested in how people were able to publish about the Jewish experience in the artificial publishing environment of communist Czechoslovakia. She is also interested in the issue of scale - what was popular rather than what i

She showed a graph of the distribution of relevant books. There is a big spike in 1964 - 1969. 1968 is important politically as it is seen as the end of the thaw. She showed another graph of the number of books published (rather than new books.) The number of book in the thaw tends to go up.

She noted the problem of how many of the most popular books have a Jewish character, but can't be said to be about Jewish experience.

1968 is a threshold, eve if the process starts in 1969 and Jewish books drop off from there are authors emigrate and bans/recalls were brought in. She had a dramatic graph showing the effect of the recall on domestic books. The 1990s, when things opened up again, look like the 1960s. The difference in the 1990s is that it is much more foreign books (especially the US).

This paper was a nice example of the value of small data.

In discussion I asked about modelling random allusions - what is the null hypothesis. That led to an interesting discussion about intentionality. Christopher rightly pointed out that the words an author chooses aren't all intentional - some might be a characteristic habit (authorship signature). Some patterns might have to do with language and so on.

Charles Cooney and Clovis Gladstone: Omnia aurea dicta — Examining reception of Lucretius in 18th-century England using ARTFL’s ECCO alignment database

Cooney and Gladstone talked about developing a system to follow the reuse of commonplaces in 18th century texts. See http://commonplacecultures.org/?. They used ECCO, a Latin corpus, EEBO-TCP and so on. This allowed them to look at reuse from antiquity to the late 18th century (excluding the middle ages.)

They also had to think about the navigational tool that made the reuses accessible. To test that their system was useful they looked at a test-case of Lucretius. They wanted to see what the after-effects of the "Epicurean Revival" when Lucretius is rediscovered. Cooney showed the search environment.

Then they talked about how Lucretius was used. There were a lot of uses of Lucretius to try to refute atheism.

Studies of reuse that used to be impossible now seem possible with large diachronic databases. One challenge is that even with

Joanna Gardner-Huggett: Visualizing the Impact of Feminist Art Collectives: ARC Gallery and Artemisia Gallery, Chicago (1980-1985)

Gardner-Huggett is looking at the impact of two feminist networks on the art world. There are archives of the two networks but little about the wider impact. Interestingly there are artists drawing networks. There are also projects using data visualization to study impact over space and time. Here conclusion is that the story is a mid-western one, not a national or international one.

Gardner-Huggett created a dataset of guest artists and used this to see where they were located and how they were connected. She talked about the dangers of "compensatory art history" - trying to add people to the canon which can hide how they were originally marginalized. With her visualizations she could see where people were studying and how schools could provide hints at personal relationships.

Why were there fewer guest artists from the South or Pacific North-West? Was it a lack of artists interested feminist art or a lack of networks? Gardner-Huggett concluded that while feminist ideology was important to the two collectives, but it wasn't necessarily important to the guest artists.

One node of graduate training turns out to be important, the School of the aRt Institute of Chicago. The School seemed to act as a network node and may have been teaching art in ways that resonated.

The paper was dense and nicely thought out. I liked how she was willing to recognize how her data

Paul Jaskot: The Problems and Possibilities for Visualizing Historical Journals with GIS: The Case of the German Construction Industry, 1914-24

Jaskot talked about the session is about visibility and invisibility. His topic is large - he is using geospatial tools to look at German construction activity. The problem is that the full

He investigates the scale of the problem through some datasets. Visualization is morphological. Mapping historical journals lets them see the context of the architecture/construction. History says there was little construction from the beginning of the war through 1924. Looking at the journals finds this was not true and there was German construction economy.

Art historians aren't use to using journals for more than reading. Jaskot is mining journals for the big picture. Constructing the database is part of the process. He talked about ambiguity and social history.

The project pointed out expected and unexpected construction. The act of building the database showed unexpected activity in town like Forbach that no one had noticed. It showed the attention of the architects at the time on a project in a small town. Jaskot talked of "dark matter."

Leipzig/Mockau is another example of an unexpected site of construction of housing estates in 1919 - 23. The previous narrative of utopian housing taking place after 1924 must be complicated.

Databases can give form to new forms of history between the chaos of history and reductive histories. Their hope is to be able to add sources to fulfill the promise of social art history. Eventually they want to look at forced labour and construction.

Smiljana Antonijievic Ubois: Supporting Humanists’ Digital Workflow

Ubois presented work done with Ellysa Stern Cahoy. They looked at workflows of scholars across disciplines. They ran a survey and interviews. Ubois has a book Amongst Digital Humanists.

She talked about how in the sciences digital practices are integrated in many more phases of reseach than in the humanities and digital humanities.

One of the biggest changes for scholars is being able to search and find evidence. Search usually starts with Google for discovery search. Then scholars use hard drives and cloud services to organize materials. For secondary sources scholars use citation managers. This gave scholars mobility - they could do their work anywhere. But there was a traumatic shift for some going from physical organization to digital.

When it comes to how scholars archive their stuff, humanists archive publications while scientists are likely to also archive data and slides.

She then talked about tools. They should be interoperable, connected, simple learn, intuitive, adjusted to disciplinary needs. They would like to have a single inteface where they can save citations and so on. Scholars are not rewarded for learning tools.

Most scholars don't feel they have any influence on tool development.

Phase II of their work tried to connect things - specifically creatging pluggable backend for Zotero. Scholarshere / Zotero for self archiving is a tool that lets scholars who use Zotero to be able to deposit their publications into university repositories. Very cool.

James Clawson: Who’s Afraid of Topic Modeling? Proposing a Collaborative Workflow (with Virginia Woolf)

Clawson started with a model for what we should be doing from importing data to a cycle of transforming the data, visualizing it, and modelling it. He talked about topic modelling as a model of a literary collection. He then talked about how difficult it is to get started with topic modeling which is why he has developed a workflow. His tool is Topic Kit. It is an R program that can be used without knowing R.

He then walked us through some example. He had a cool way of drawing from a spreadsheet and he had cool visualizations comparing genres which I not sure how he did.

The code is available at Topic Kit on Github.

Anne Flannery and Tony Lauricella: Maps, Texts, and Possibilities: The Center for Ancient Middle Eastern Landscapes’ Geospatial Data and the Oriental Institute’s Digital Repository

Flannery and Lauricella started by talking about the issues around institutional workflow. The Oriental Institute was founded in 1919. They are building an integrated database for the Institute. The database is for the collections of the Institute. The collection isn't well catalogued and it is closed to the public.

They make heavy use of volunteers to enter data. Every week they then move new data online, after cleaning it up. They link data across departments.

Then Tony talked about the Center for Ancient Middle Eastern Landscapes. He showed data about sites in Afghanistan.

They have a remarkable amount of data and integrating it must be a challenge.

Geoffrey Rockwell: Information Wants to Be Free, Or Does It? Ethics in the Digital Humanities

I gave the keynote at the end of the day so, of course, I couldn't take note.

Sunday: November 13, 2016

Emily Forden: Mapping Patronage Networks in Ancient Pompeii

Forden talked about the 16th century drawings of Roman buildings have left us with the impression that the Romans lived in nice brick buildings with lots of light and windows. There is little evidence of such living. Rather the buildings were made of wood, crowded tenements, with no kitchens. Not able to have fires in the buildings, most people ate in eating places - often the same ones. Patronage was important for all sorts of reasons, but the rich leave in the summer so there is little wealthy patronage. Patrons would be visited in the morning and would throw a banquet in the evening. There would hierarchies of elite patrons, middle patrons and down to the clients at the bottom. Forden's thesis is that during the summer there would be little coinage circulating which would mean that low level clients had to shop with their patrons who would extend them credit.

Forden is using digitized maps of Pompeii. She highlighted bakeries and bars (where clients would have to get food.) She also looked at the mills that would feed the bakeries. She thinks that there must have been a lot of street vendors, especially around the amphitheatres. There seem to be both foot traffic spots and neighborhood spots.

To connect to patronage networks she looked at campaign posters. Communities would be marked with posters painted on walls. About a third are signed by women. Mapping the posters you can see different distributions which suggests different communities supported different candidates. Cluster analysis suggests a correlation with bakeries.

She talked about how many ancient historians push back on such statistical techniques.

Helen Davies: Resurrecting the map: medieval mappa mundi, multi-spectral imaging and deep mapping

Davies is working with medieval maps like the Vercelli Mappa Mundi using imaging. Medieval maps used to be ignored as they were seen as scribbles. She is using "deep mapping" from media studies. Deep mapping is the layering of cultural information onto maps - the idea is to apply the cultural and mythic information onto the space. The map becomes a organizing principle rather than the story. HyperCities and Google Maps are examples.

The problem is that maps often show our cultural biases which is as true now as back then. Thus we read the medieval maps for their cultural

The Lazarus Project went to Italy to recover the Vercelli map. She showed dramatic before and after images that show how you can recover information from the damaged sections. The Vercelli map is schematic which lets us draw inferences on the relative importance of different cities. If a city icon is twice the size of another that could mean that from the mappers perspective that first city is more important.

Digital maps are different than the physical object. One can zoom in on the digital and see beyond the visible light spectrum. On the other hand it is harder to convey scale, texture and materiality.

She showed a Photosynth view of the map which was fast and impressive (but Photosynth is going to be retired.)

Mark Moll: ‘From the shaken Kremlin to the walls of unmoving China’: Spatializing Russian Student Unrest

The focus on student unrest in Russia tends to be on Moscow and St. Petersberg. Instead Moll wants to talk about Tartu which is South of St Petersburg. Russian universities were often on the Empire's borderland to keep students out of the metropole. Moll is working on student matriculation records (Album Academicum). He has a database of where students came from before going to Tartu U (?), when they arrived, when they left, their religion and so on.

Moll has challenges with the different transcriptions of names across languages.

He showed a map of where the students came from (and what sort of school they came from.) The students mostly came from the baltic area, but there was still a spread. He showed a graph of in and out transfers. Around 1880 there a spike of incoming (after a Czar was assassinated.) Students may have started transferring to Tartu as there were troubles elsewhere in the Empire.

So what could be related to transfer spikes? There could be a contagion effect among students involved in unrest. There were changes in the statutes as Russian universities get new rules Russifying them. In 1893 the University was converted from German language to Russian, but that may not have changed the student culture. After 1893 it does seem that more students who only spoke Russian started coming in, but in 1888 there is a spike that seems do to unrest closing a number of other universities.

He concluded by talking about the challenges of multi-lingual DH and issues of cartographic distance vs spatial distance. Russification led to a decrease in radical activity (among Russians) while providing an opportunity for non-Russians to organize. Russian officials saw Tartu as an isolated city giving the activists room to create their own spatial network.

2016 Chicago Colloquium On Digital Humanities And Computer Science