University Times

Will “publish or perish” ever become “post or perish?”

Where are the digital humanities headed? And will the academic establishment accept these projects — or even know how to assess the worth of databases and computer code?

Pitt faculty research in the sciences has taken up big data projects as a natural extension of existing scholarship, but humanities projects also have begun to move from the durable but not universally accessible medium of paper, presided over by lone scholars, to a plethora of online, worldwide, participatory media that go out-of-date seemingly every few years. How do faculty develop worthwhile, durable digital projects in the humanities? And how should the academy respond?

*

Itinera, says Drew Armstrong, faculty member in Pitt’s history of art and architecture department, “is a true scholarly project, based on the expertise of specialists” and an appropriate array of sources cited for anyone to examine.

But it is entirely online as an interactive map showing the global travels and interactions of artists, patrons, critics and others, starting with Armstrong’s area of scholarly exploration: the art world of 17th- to 19th-century Europe.

Clicking on a time, person or place will lead Itinera users to images of the source material, including letters, travel journals and art, to demonstrate how cultural concepts circulate with a British gentleman on his Grand Tour of Europe, for instance, or a Parisian painter training at the French Academy in Rome.

“For someone who doesn’t know this material, it is extremely hard to get a handle on,” Armstrong says. “There are certain kinds of information that can be represented in more effective ways in different media. I could send you to a lot of books, but you’ll never get a clear picture.” With Itinera, he says, “one has a much better vision of what is going on” in the art world at specific places and eras.

The idea behind the project “is to really think how the system lets us think about information that is not comparable to printed text,” he adds. It also is open-ended in several dimensions: It is expected to grow with the addition of new eras covered and other scholars contributing — perhaps in areas and languages outside Armstrong’s own knowledge. It also will have no official completion date.

“This is really highly experiential,” he says.

It is also experimental, even in 2014. Armstrong is tenured, so he need not worry whether his project will lead to academic promotion. But some current digital humanities projects at Pitt involve the visualization of texts through GIS (geographic information systems) data, or online representations of the worlds of people who inhabit particular texts, both nonfiction and fiction. How can more junior faculty devise or contribute to such digital projects and gain the proper evaluations and credit to move up the academic ladder?

Print journals long ago added online versions and more and more have been publishing only in digital editions, while the University Library System (ULS) has been housing digital repositories, maintaining online subject archives, hosting e-journals and digitalizing its archives and collection for many years.

But how can something like Itinera be evaluated properly? Armstrong is developing Itinera in partnership with his department’s Visual Media Workshop, run by Alison Langmead, who is the project’s technical and project director. Armstrong also has contracted with a Brooklyn firm, Whirl-i-Gig, to write the code. How can those project contributions be assessed critically?

“I can pursue this without putting my career in jeopardy,” Armstrong says of Itinera, but he acknowledges that “there’s no clear way to evaluate this kind of work.” Making obvious to its users that Itinera and other digital humanities scholarship has the backing of PhD researchers in an established academic department, and has been reviewed by more senior scholars, will put the stamp of more traditional scholarship on online projects, he believes.

Recognizing, too, that research presented only online will be supplementing, not replacing, traditional scholarly output may go a long way to easing its acceptance as more standard practice, he says.

“It all boils down to what resources are available,” Armstrong says, noting that, while Pitt has been supportive of his work, unlike a growing number of other academic institutions the University has no center for digital humanities. “There are a whole array of complexities [that come] from not having the institutional infrastructure, such as funding. What do you do when you run out of money? If I had more staff I could definitely come up with more projects.”

According to Aaron Brenner, head of the Digital Research Library in ULS, his department is trying to help faculty across Pitt manage large online research databases, which funders increasingly are requiring from researchers.

He points to the recently released Project Tycho (www.tycho.pitt.edu) database, which involved faculty from several Pitt schools, led by the Graduate School of Public Health, in digitizing and analyzing 125 years of weekly statistics for reportable diseases in the United States. “That’s an example of publicly available data sets in new kinds of research,” Brenner says.

He recognizes that assessing the scholarly merit of faculty work is a crucial question. “The current reward system may not take full account of the current scholarly activity,” he says. “I don’t know that there is a single answer to changing that, but it is a challenge.”

Elisa Beshero-Bondar, an English faculty member at Pitt-Greensburg, recently began work on Digital Mitford, an examination of the 19th-century writer Mary Russell Mitford, a prolific author of poetry, drama, fiction and letters and a contemporary of Lord Byron, but now little known.

The project will give Mitford’s works clickable annotations, including comparisons of earlier, heavily edited versions of her letters to their full manuscripts, and create connections among the people, places and events associated with her career and texts.

For many of today’s scholars of the 19th century, says Beshero-Bondar, the value in a new Mitford compendium would be the ability to mine it for Mitford’s references and contacts. That meant using open-source digital coding to make Mitford’s work searchable.

“I don’t think I could have started the Mitford project … while I was in the tenure track and getting used to my teaching load,” says Beshero-Bondar, who was awarded tenure in 2010. “When I was coming up for tenure, I had managed to produce a book and a couple of articles, but that was taking up all of my time. I would have been a little nervous with taking the lead on a project like this … It was uncertain how my colleagues would see it.

“On the other hand, I think this kind of work I’m doing leads to more traditional scholarly publications as well,” she says; her work on the Digital Mitford project already has led to scholarly conference presentation proposals and journal articles.

“Since I’m at Greensburg,” she says, “I might have a little more room to play with, oddly, than if this were at the Oakland campus, since Greensburg is a teaching campus. Since I’m tenured, I feel like I’m in a safe place to experiment here … and to use this project when I apply to be a full professor.”

This large project needs a team of researchers, Beshero-Bondar points out: “It involves working with Pitt resources but collaborating with editors around the country” as well as in England and Italy.

“That aspect is the other side of what makes it nontraditional for assessment as humanities scholarship. Usually, humanities scholars work alone. ‘How much work does this person put in?’” she expects academic assessors to ask. “I think we’re in uncharted territory at this point. It’s up to faculty committees: Is this the sort of thing we’re going to value or not?

Another reason the project was prohibitive for a junior faculty member: She needed time to learn a lot of coding methods, including XML (eXtensible Markup Language), which would allow her to index items inside the handwritten manuscripts reproduced online, and add notes for scholarly users. Next she wants to learn how to do her own versioning, the side-by-side text comparisons that pinpoint changes from one original or published version to the next.

“The people reviewing this need to be in a position to review the coding” when assessing the scholarly value of such projects, she says. “I don’t know if everybody would agree with me on that.”

And it is the researcher’s duty to reveal more of the digital design for critical inspection, she adds. She expects academic evaluators of such projects in the future to include scholars of both coding and the subject matter.

To learn the code, she sought the help of coding classes at the Women Writers’ Project, which began at Brown University and now is at Northeastern; a Pitt faculty collaborative on digital humanities resources, and the Computational Methods in the Humanities honors course taught by David Birnbaum, chair of the Department of Slavic Languages and Literatures.

Birnbaum’s digital humanities work began when he was a graduate student in the 1980s and could not find the right computer tools for the research he was attempting, so decided to write his own.

“A lot of people in humanities think about computers like they think about toasters,” Birnbaum says. “You put your bread in and it pops up,” but how the mechanism or electronics work remains a mystery. “And by God you’re not going to build one.”

In his honors course, he first helps the students define a research question for which computers will be most useful, then teaches them the coding they need to instruct the computers to serve them best.

“There are some things computers do better than humans and some things humans do better than computers, so it seems sensible to divide those duties,” he says. When a scholar wants to study the entire corpus of English literature for patterns, but cannot possibly read all the necessary works in a lifetime, that’s where computers come in. Once a computer is done digging through such a big data project, it still takes a human to assess and refine it.

For Birnbaum’s own projects on medieval Slavic literature, he creates digital editions of manuscripts so that they are searchable and accessible to all on his website. He also builds tools to explore these manuscripts. Other scholars no longer are dependent on Birnbaum’s indexing for their searches; now they can search down to the parts of words for the answers to their own scholarly questions. He also builds tools that allow him to move around in documents and graph visualizations of text patterns.

“I know there’s some concerns about the evaluation of digital projects,” says Birnbaum, who has tenure. But he believes it is easy for colleagues in his field to interact with his databases to assess how accessible the data is and how easy it is to work with. Can such projects be refereed successfully, as articles in journals are today?

“There’s no reason it can’t happen with digital projects,” he says.

University Senate President Michael Spring agrees.

“Using the computer as a device that’s capable of [handling] vast stores of data, we can do things today that took literally decades years ago — and not that many years ago,” he notes. “Therefore, digital humanities is here to stay, because things that were an [impossibility] we can now do.

“This is a matter really of time,” he adds. “As the younger faculty become the older faculty, these things will become accepted.”

His own research happens to focus on incunabula: books from the earliest era of mass production, 1453-1500. “The Greek term really means ‘a baby in swaddling clothes,’” he points out. So the word also is used for products of any industry in the early stages of development. “I tell my students that when we have new technologies — the horseless carriage, the cordless phone — we will name them by their predecessor because we don’t know they are going to turn into cell phones.”

“The transformation will be a little bit disruptive. But nobody will go back to monks” writing books by hand. “Paper has been a phenomenal medium throughout time. I love printed books. But I am slowly turning everything into digital form.” Even though he is very concerned about the obsolescence of digital formats, he says it takes less time to redigitize than to make the initial move from, for instance, photograph paper to JPGs.

“If you asked people 10 years ago when everything would go digital, they would have said 50 years, and it happened in a mere 10 years. We have begun the new age.”

Something that has been invaluable to UPG’s Beshero-Bondar is Digital Humanities Research at Pitt (DHRX), a website for Pitt faculty in every discipline and on every campus who want to share expertise about tools and collaborate on digital humanities projects. Among its organizers are Alison Langmead, faculty member in the School of Information Sciences as well as the history of art and architecture in the Dietrich School of Arts and Sciences, and Brian Beaton, another iSchool faculty member.

Although the website (www.dhrx.pitt.edu/), begun in fall 2012, is hosted by Pitt, the organization is not housed in any Pitt department, nor does it receive any Pitt funding. “We are just a group of people who share methodology and research interests,” says Langmead. “We’re a network of people doing work” — not merely using online archives to conduct research, but using digital and traditional materials to build online databases and more complex online presentations of research outcomes, akin to Itinera.

“It remains unclear how digital scholarship counts toward tenure,” adds Langmead, who, unlike Beaton, is not in the tenure stream. “Pitt is showing itself ever more amenable to this type of scholarship. But the waters are still muddy. We like to think of the research network as a new way forward.”

The voluntary collaborative group has been fueled, Beaton says, by a lot of new Pitt faculty coming from universities with central digital technology offices. “This has become a site for conversations about the future of academic publishing and the role of technology in teaching.

“It’s an opt-in support system,” he adds, designed also to help connect faculty with students who might assist in their projects. “To me, the DHRX is human infrastructure … to explore and make new kinds of academics work.”

Langmead, for instance, is teaming with Josh Ellenbogen, director of graduate studies and faculty member in the history of art and architecture, on a project called Decomposing Bodies. It aims to show how a pre-fingerprinting method of identifying criminals through body measurements, called Bertillonage, used the visual imagery it collected. Ultimately, as they have posted on DHRX, the project “seeks to create new means of understanding the implications and possibilities inherent in this 19th-century process of treating human beings as numbers and letters, and how this approach to the visible world might relate to the dawn of computing.”

“The challenge,” says Langmead, “is to allow these works to be accepted alongside other works … by showing their value.”

“People sincerely want to engage with them,” says Beaton, “but we don’t always know how.” Unlike the traditional book, with its thesis and proofs, what parts of a digital humanities project should be critiqued: Traditional factors, such as the sources? Newer factors, such as the visualization of the data?

Scholars have been trained to analyze writing, Langmead notes, but “we were not trained, for a very long time, how to analyze a database.” A critical theory of digital humanities analysis is in the works, and there are ongoing discussions in academic journals. “But the implementation on the ground is nascent,” she says. “We have to build the infrastructure we didn’t have.”

Adds Beaton: “There are huge questions about the sustainability of these types of projects,” due to technologies that get old or lose commercial backing, leaving materials created with these technologies hard or impossible to access. “That raises a lot of questions for people’s careers and portfolios.”

Although digital projects also have the potential to last longer than paper, even that longevity has its drawbacks: “You cannot walk away from most of these projects,” he notes. “There’s a large commitment of time.”

Of course, Langmead points out, digital projects also have an ability to change and grow with the input of new scholarship without having to re-emerge in a new edition, as would a book.

Beaton says he is starting to see informatics enter humanities disciplines with the creation of new certificates or dual programs: “It’s people articulating, ‘I need to know both my subject area and a set of tools when it comes to information technology.’

“And we’re seeing a lot of enthusiasm” from students, he adds.

“It’s a great time to be a digital humanist,” Langmead concludes. “But it is also a time of great flux.

“We aren’t the ones making tenure decisions. (But) we will be.”

Patrick Manning, the tenured Andrew W. Mellon Professor of World History and director of the World History Center, jokes that he is involved in “too many” digital humanities projects. An electronic scholarly publication he has just begun via ULS, the Journal of World-Historical Information (www.jwhi.pitt.edu), reviews not only books but databases, giving the imprimatur of senior scholars to these novel research products. He believes this practice of reviewing digital humanities scholarship and books on an equal footing should encourage junior faculty to venture further into digital scholarship.

Says Manning: “The purpose of that is to make it so that creation of good data sets can be evaluated and ranked in the scholarly world — a parallel way to the way books are … and publish them in a way they are successful.”

He also is leading the construction of the Collaborative for Historical Information and Analysis (www.chia.pitt.edu), a National Science Foundation-funded project to build a world-historical archive involving several universities, about which he has just published a book available on Kindle, “Big Data in History.”

In this spring term, for the first time, he also is teaching a small graduate course on digital history to prepare students to work on digital projects.

“It’s fun doing it,” he says of the digital humanities, “and I believe it is leading us in a valuable direction.”

That will include, he says, his next digital humanities endeavor focused on using diverse data sets to examine the peaks of economic inequality in America in the Depression and today. “Is that just an American phenomenon? If we look at other countries over the past century, would we find the same 80-year fluctuation” and other parallel fluctuations further in the past?

“We may find the patterns in social change that are long-term patterns and kind of unexpected. We’d think about the causation differently.”

The need of professors to publish isn’t going to end, he says: “Digital journals are gaining recognition in some fields [but] the big ol’ print journals are the ones you need to publish in” for the immediate future.

—Marty Levine