HASTAC 2015 Thurdsay notes

flickr, Chuck Holland

HASTAC 2015 was held at Michigan State University in East Lansing, Michigan. As always, I’ve tried to put my editorial comments in italics.

Connecting the Dots (opening plenary)
Scott Weingart, Carnegie Mellon

Used a visual model (a circle) to describe the extent of human knowledge. Noted that when doing a PhD, the idea is to nudge out a little bit from that circle and expand the scope. As he noted, while it may seem like a small contribution, it’s an “uplifting narrative.” Another way to think of it is that the knowledge is already known, just not to scholars, so what the scholar does is give it shape and form that adhere to the rules of research. Example: an anthropologist may ‘discover’ things about a given population, but the members of that community live those practices and traditions. I think of my own PhD research in this light. Unearthing, sorting, and creating a narrative are in their own ways about creating new knowledge.

Went further with his comments and noted that narrow specialization has created isolation in academia. That’s not a novel idea, but he was laying groundwork. At the end of this, he noted that his analysis was pretty biased toward a Westernized, masculine point of view, and promised to go in another direction

Got distracted a bit by his foray into German etymology. Believe he said that the root or origin of Wissen comes from the notion of trees and branches, but that seemed wrong to me. When hunting in Lexer and Grimm for more etymological background. Buch seems to be derived from the tree, but can’t find much evidence that Wissen is. Then again, I was doing this while trying to listen and follow, so my search was shallow.

Digital Humanities: Explorations in Ancient and Medieval Studies
Olga Scrivner, Indiana

Scrivner started by showing an animated manuscript as an indication of what we can do in the 21st century. Gave a quick review of the DH waves. Wave one was largely quantitative, while the second is more qualitative and curatorial, as she emphasized (based on others’ definitions). We are now, per Berry, in the third wave of computationality.

What can we do now with literary analysis? Visual analytics, for one, with graphs, maps, trees, etc. (Moretti). We can also use word clouds, frequency, etc. (Vuillemot, et al.). Social network analysis is another tool, graphing characters in works (e.g.- Rydberg-Cox with Greek tragedies). She described other tools/methods; in sum, there are many ways to apply a variety of tools to literary texts.

Goals for her work are:

• make medieval literature accessible to a large audience and make it interactive
• develop methods for multi-level annotation
• explore new visualization techniques

Their subject is medieval Occitan, which is apparently still a spoken language. It was the other administrative language–with Latin–in medieval France. In particular, the key text is the 13th century Romance of Flamenca (Roman de Flamenca). First they need to correct numerous OCR errors and annotate the text. They discovered during their work that there is also a scan of the manuscript that they would like to process and include. Use OxygenXML for the markup. They also pulled in an English translation. Use Annis (German developed) for multi-level annotation: word, lemma, part of speech, etc.

In response to a question, she noted that the parallel alignment between the Occitan and English texts allows researchers to create more information about “resource poor” languages. Another general question she got was why this would interest students, or what it gives us. In particular, how does a visualization enhance understanding beyond reading? Her answer, in part, was that it creates different ways to say or make points. Another response came from attendees, which is that one can quantify an impressionistic reading, which helps make a case or point.

Lightning talks: Networks in the Humanities

Networks in the Humanities: An Introduction
Jason Heppler, Stanford

Noted that “formal network analysis” is a product of the last 20 years, although the usage of networks in scholarship is of course older. Gave a quick primer on networks, showing simple and complex examples (e.g.- Roman transportation network). Asked how we can represent uncertainty in networks. Simple way: dashed lines to show tenuous relationships. Also, how do networks change over time, and how do we represent them as “fluid, evolving, changing relationship[s].” Called these “time-based” networks.

Can I Get a Witness?: Network Analysis of Nebraska Homesteaders
Rebecca Wingo, Macalester

Project started with an offer of research money. Who could say no? Goal was to identify fraud in duplicate homestead claims, although she ended up doing something else.

Ads in newspapers had to put homestead claims in the public eye, with four witnesses to the validity of the claim. Had to extract and standardize the names since the papers of the time had frequent misspellings. Mapped this as a network that showed how interconnected the claims often were.

Analysis allowed her to focus on specific individuals and trace their influence, often using other historical sources and qualitative analysis. Involved geomapping, etc. It tells a good and interesting story about how communities formed, and what connections brought them about.

Reconstructing Social Networks: The Challenges of Small Data
Brian Sarnacki, Nebraska

For a project around the Grand Rapids water scandal, he observed that he had incomplete records for groups. He learned from this work and arrived at three lessons for working with what he calls small data:

• Be honest – be clearer about methodology, and about the sources and their limitations. Why are things included, and how. Give up the “wow factor” for a more realistic visualization.
• Give context – “more can be more.” Put in other details, do mapping, add anecdotes. Do this so that the reader isn’t just given a visualization and “asked just to trust the creator.”
• Find inner peace – Exercise restraint. Leave room for the readers to come to their own conclusions and ideas. Build an argument, don’t present ‘facts.’

Mapping the International Dimensions of the Nicaraguan Revolution
Andrew Wilson

Used Palladio as an alternative to paper mapping and network analysis. Used it to track Sandino’s movements during exile. Showed that international borders were of little meaning; he went everywhere. Also applied it to research he conducted in the archives of Germany’s Green Party, documenting the connections between figures such as Gabi Gottwald to various Nicaraguan and international organizations.

Panel: The Embodied Digital Self

Images as Data: Cultural Analytics, Aby Warburg’s Mnemosyne, and the Relevance of a ‘Science with Names’
Stefka Hristova, Michigan Tech

Glad I just read an article about the Warburg Library and Aby Warburg in The New Yorker. Good background for this talk.

In a very loose nutshell, she recreated Panel 45 from Warburg’s work using colour images and then treated it as a data source. Analyzed the images based on their colour traits and drew connections between images based on their colour palette. Discovered that this brings very divergent images into contact (e.g.- Massacre of the Innocents and Annunciation to Zacharias by Ghirlandaio). Used ImageJ for this work.

Quantified Self: The Body as a Creative Contstraint
Meena Natarajan, UC Berkeley

Self-tracking is nothing new. We’ve done this with many data points for a long time. In some cases–diabetes–it can be a matter of life and death. For most people, however, it’s something people do that isn’t important nor something that they think critically about.

Her points were primarily about reductionism and false agency. Noted several types of quantified self archetypes, including the individual who uses it to track their own data to “speak back” to established medicine that has failed to diagnose their malady.

Find it nearly impossible to take notes during a read talk that lacks slides or other breaks in the narrative.

Using Bioinformatic Algorithms to Analyze the Politics of Form in Modernist Urdu Poetry
A. Sean Pue, Michigan State

He (linguist) worked with bioinformaticians to sequence Urdu poetry, to oversimplify things. This enables him to do metrical analysis that shows that various texts have been influenced by earlier texts and/or resemble them metrically.

He noted that there are connections between open science and the digital humanities.