Image

The Source Documents Behind the News (A librarian reads the news...)

Publications and Presentations

In my classroom...

Boats and Ferries and Bridges, Oh My!: Digital Humanities support models

This morning my coworker Sarah Calhoun and I presented again at the Oberlin Digital Scholarship conference, “Building a Distributed Collaborative Model for Digital Scholarship Support at Liberal Arts Institutions: A Mixed Metaphor Salad.” Our other co-presenter, Austin Mason, couldn’t make it to the session, but he gets the credit for finding and bringing to us the metaphor that drove the session. He’d attended a session by Liz Milewicz of Duke University and gotten permission to reuse her metaphor. And the metaphor in question? Various methods of crossing water from one shore to another. Here’s how we used it.

“Automobile crossing rope bridge. 1923.” Photograph made accessible by the Field Museum Library. http://www.flickr.com/photos/field_museum_library/4462494439/

“Automobile crossing rope bridge. 1923.” Photograph made accessible by the Field Museum Library. http://www.flickr.com/photos/field_museum_library/4462494439/

On occasion, supporting the digital humanities can feel an awful lot like trying to drive a car across a rope bridge. Either the infrastructure is inadequate for the project, or the project chose the wrong infrastructure to use to get across the river.

In an attempt to avoid this unfortunately pairing of project and infrastructure as much as possible, we propose thinking carefully about the infrastructure we put into place, and also about how to communicate clearly with researchers and with ourselves. We want researchers to find and use the right resources for the job, and we want to prevent siloed services that may duplicate effort or cause turf wars. And one way to think through these issues is to map out what’s happening on your campus.

Most campuses will have brave DIY-ers, who cross the river in daring ways using found objects (fallen logs) or special skill/access (base jumpers). These are important parts of creative research on our campuses, but they are not very repeatable or scaleable. They are not a great plan for a support model.

"Simple cable ferry, Gee's Bend, Alabama, 1939" by Marion Post Wolcott.

“Simple cable ferry, Gee’s Bend, Alabama, 1939” by Marion Post Wolcott. https://en.wikipedia.org/wiki/Cable_ferry#/media/File:Marion_Post_Wolcott_- _Old_cable_ferry_between_Camden_and_Gees_Bend,_Alabama.jpg

One of the low-barrier, low-overhead, repeatable methods of crossing might be a rope ferry. These work quite well either solo or with a ferryman, but they may not be terribly stable over time. Perhaps various semi-ephemeral things like free blogs or social media fit here, where researchers can get information up online, but it may not be stable over time if the researcher or the service move in new directions.

"Venezia - Ferry-boat Lido di Venezia"

“Venezia – Ferry-boat Lido di Venezia” https://commons.wikimedia.org/wiki/File:Venezia_-_Ferry-boat_Lido_di_Venezia.jpg

Or perhaps there are more modern ferries available. These are larger, more powerful, and require less effort on the part of the researcher. Maybe something like an institutional subscription to Omeka or WordPress fit here. They were for a lot of people interested in doing a lot of different kinds of things, and support can be pretty standard on campus.

"Brooklyn Bridge Manhattan"

“Brooklyn Bridge Manhattan” https://commons.wikimedia.org/wiki/File:Brooklyn_Bridge_Manhattan.jpg

But if you have a ferry running the same route over and over and over, and if you also have some money and staff for new construction, maybe it’s time to build a bridge. Some schools, for example, have a whole unit or a Center dedicated to supporting digital humanities. Some of these Centers are even iconic, like the Brooklyn Bridge. And I think that depending on the school, a massive, multi-lane bridge with tons of on-ramps and off-ramps might be a wonderful thing.

However, bridges really do require upkeep, they can become bottlenecks, they often have height and weight restrictions, and they may not serve all needs. Researchers are creative beings who may need to start in a different spot, end in a different spot, or get across in unusual ways. So at this point it becomes a matter of project portfolio management.

Vinopal, Jennifer, and Monica McCormick. 'Supporting Digital Scholarship in Research Libraries: Scalability and Sustainability.' Journal of Library Administration 53, no.1 (January 1, 2013): 27-42. doi:10.1080/01930826.2013.756689.

Vinopal, Jennifer, and Monica McCormick. ‘Supporting Digital Scholarship
in Research Libraries: Scalability and Sustainability.’ Journal of
Library Administration 53, no.1 (January 1, 2013): 27-42. doi:10.1080/01930826.2013.756689.

Jennifer Vinopal (who was also this conference’s keynote speaker, though we’d planned to site her even before we knew that), and her colleague Monica McCormick wrote a fantastic piece on “Supporting Digital Scholarship  in Research Libraries: Scalability and Sustainability” in the Journal of Library Administration. In it they recommend having an infrastructure such that the majority of project can use standardized tools that are well-supported on campus. Then allow for creativity by building in support for projects that can mostly use those standard tools but with some standardized consultation services, or with a bit of custom tweaking. But reserve some capacity for a few truly custom projects that will require a lot of support (and try to make sure that these project can be “first of a kind” rather than truly one-off projects).  So perhaps you have a couple of docks for some ferries not far from your bridge, or a landing area for your base jumpers, but there’s communication and vetting involved in committing resources to these special projects.

And no matter your solution, close communication between support and coordination folks will prevent boating collisions, or the building of duplicate bridges. To mix metaphors quite wildly, at Carleton we think of our coordinating folks as a three-legged stool: Library, Humanities Center, and Academic Technology. Representatives from these areas try to make sure that things move forward in a coordinated fashion even when the actual support and work of digital scholarship happens in all kinds of places on campus. We’re calling it a “coordinated distributed model,” and it is still in its infancy. We are currently tackling the question of what infrastructure to build and for whom, which tools will be our standard and which will we cut loose, who will be involved and to what extent, and how will we make sure that people who need us will find us?

Exploring these questions, we handed out blank maps to participants and asked them to depict their campus’ current models, talking in small groups about how things work on their campuses, and comparing with others to find trends and themes.

Feel free to use or make your own. I just drew this sketch of a map on my iPad.

Feel free to use this map or make your own. I just drew this sketch of a map on my iPad. https://drive.google.com/a/carleton.edu/file/d/0B1hW_xoemoVvSklzYmJTNnFzWTA/view

One fascinating thing was that there were three participants from one college, and all three drew different maps. Others drew people drowning in the river, or people standing on one shore and gazing longingly at the distant shore. But one thing that became clear was that whatever infrastructure an institution adopts, it has to fit the local context. There’s no sense sinking a ton of money in a massive bridge if there’s no demand, or just because another school did it. And on the other hand there’s no sense leaving researchers to fend for themselves on campuses where lots of people have a similar need.

Then we handed out new blank maps and asked participants to think about an ideal infrastructure for their campuses over the next 5-ish years. There were so many interesting and useful responses. One person drew a crew of happy inner-tubers (beer implied), and someone at a different table drew one big inner tube. Most had more than one method of getting across. Others had thought about setting up villages or resorts on one side or the other to show the community that would be important, or close communication, or the bringing together of units that are currently separate. The creativity and thoughtfulness in the room was so inspiring!

At the end we asked participants what one thing they wanted to take back or change first at their institutions. My own answer was that I want to hand out blank maps to the people I work with on campus and see what we each think the current model looks like, and where we each hope it might go.

Data Management and the Humanities

Humanities Data: Document fragments from church archives in Cuba

Humanities Data: Document fragments from church archives in Cuba

This morning my colleagues Kristin Partlo and Sarah Calhoun presented with me on data management plans (DMPs). The session itself was mostly a conversation about some example DMPs from the collection of successful DMPs that was recently released by the National Endowment for the Humanities followed by conversations about how we might talk to faculty (using this DMP template as food for thought). But from this conversation a few themes crystalized for me.

Why would humanists care?

I think we often hear that “sharing the stuff of my research with the world” isn’t a huge motivator in the humanities. I’m not sure how broadly true that is, but it’s true enough with enough people that it bears thinking about. And I have heard skepticism, usually in the form of “nobody else will care about this stuff I’ve collected,” so how do we address that?

Sometimes I think maybe it’s a matter of vocabulary, so talking about creating “an archive” or “a digital collection” might capture imaginations where “manage/share your data” won’t. Even talking about how a bibliography is a described dataset can be useful because humanists are very familiar with practices of collecting, organizing, and sharing bibliographic information, usually at the ends of articles, books, and syllabi, but sometimes independently. Humanists have actually been doing data management since forever when you think about bibliographies. In this context, we care about versions (editions), standardized and encoded “data fields” (so that other researchers know if you’re referring to a title of a chapter, article, or book, for example), durable URLs whenever possible… Bibliographies are rich with descriptive and preservationist practices that can help inform management and sharing of files and information more broadly.

I think another aspect of why Humanists might care involves thinking about sharing with the sympathetic collaborator that is your future self. Your future self will forget the ins and outs of where you put or how you named your image files, or what the columns on your spreadsheet actually mean. Our computers are chalk full of the stuff of our research — PDFs, draft versions, images, audio, video. Knowing where all of those things are requires management of all of that data just so that you can find things again later.

Learn through analogy with the known

Kristin talks about how the monograph is largely self-describing. It has title, author, and publisher information in predictable spaces. There’s the table of contents, often an index and bibliography, and things like introductions and conclusions that describe the book for you.

Then there are style guides and formal or informal glossaries that people adopt, and these serve to help make your data (“data” writ large) understandable and consistent for other readers.

These are things that are familiar, so it’s easier to point to these things and remind ourselves that we’ve already seen the usefulness of self-describing units of scholarship and of somewhat standardized best practices. Now we just need to apply it to the digital stuff we’re working with more and more these days.

And in the past, libraries and archives were the main places that managed the sharing of shareable humanities data (primary and secondary sources), but the sharing involved researchers traveling to those collections (humanities “datasets”) to use them. Now individual researchers can create collections, or use collections without physically traveling from dataset to dataset. But this also means that researchers now have more responsibility to do some of the description and standardization that libraries, archives, and publishers formerly did a lot of. So yes, the work feels different, but it’s built on the same principles that humanists already value.

Formal vs informal data management

One of the interesting themes of our session and all the other data-related sessions I attended is that people talk a lot about the data management requirements of grant-funded projects. Meanwhile, I haven’t supported that work at all, but I have helped quite a few people manage their own individual or collaborative non-grant-funded projects. And I really think that data management becomes much more alive and broadly useful if I think about how the best practices identified and codified by grant funding agencies help us think about best practices for regular, every-day digital life, right down to the daily action of naming and putting a file into a folder on my computer. For me and for the people I’ve worked with so far, formal DMPs are alien things that don’t intersect with my life and research. But the spirit and practices reflected in those DMPs? THOSE I care deeply about. So perhaps shifting language from compliance to best practices, and shifting focus from grants to every day organizational practices, perhaps these things can help make data management less of an alien object that only scientists and social scientists ever touch.

News you can use: Obama signs bill modernizing terms referring to minorities in two sections of the U.S. Code

Seal_of_the_United_States_Congress.svgYou know you’re a librarian geek when you read the headlines telling you that Obama has signed a bill that will strike “oriental” and “negro” from two sections of the U.S. Code and think “What a good reminder that  full text searching through government documents is tricky.” And yes, that was me when I read those headlines on Friday.

And here’s the language from the bill itself.

(a) Office Of Minority Economic Impact.—Section 211(f)(1) of the Department of Energy Organization Act (42 U.S.C. 7141(f)(1)) is amended by striking “a Negro, Puerto Rican, American Indian, Eskimo, Oriental, or Aleut or is a Spanish speaking individual of Spanish descent” and inserting “Asian American, Native Hawaiian, a Pacific Islander, African American, Hispanic, Puerto Rican, Native American, or an Alaska Native”.

(b) Minority Business Enterprises.—Section 106(f)(2) of the Local Public Works Capital Development and Investment Act of 1976 (42 U.S.C. 6705(f)(2)) is amended by striking “Negroes, Spanish-speaking, Orientals, Indians, Eskimos, and Aleuts” and inserting “Asian American, Native Hawaiian, Pacific Islanders, African American, Hispanic, Native American, or Alaska Natives”.

First, I’m somewhat horrified to find the periods on the “wrong” side of final quotation marks — I guess Congress doesn’t go by standard U.S. style guides.

But more importantly, this means that if you’re full-text searching the U.S. Code, and you know to be careful to use outdated language when searching through historic documents, you should now OR in modern language if you’re searching through the Local Public Works Capital Development and Investment Act of 1976 and the Department of Energy Organization Act.

Here ends your communique from a librarian geek interested in search language.

A practical guide to the 2016 Georgia State eReserves Copyright Case for Librarians

The academic library world has been watching the Georgia State (GSU) eReserves copyright lawsuit with interest for half of my career, and now there’s a new ruling in the case. On the one hand, I recommend reading the ruling because there’s interesting language in Judge Evans’ Fair Use assessments for each item at issue, and every time I watch someone who knows a lot about copyright go through the four factor test I learn a little bit more about how I might apply it in my own work. On the other hand, it’s 220 pages long, so there’s that.

Oh, and another thing, I’m not a lawyer and don’t even play one on TV. I’m just a librarian who takes an interest in copyright but has no legal advice to give.

History of this case

In case you’ve forgotten how we got to this ruling, Cambridge University Press and a few other publishers sued GSU over book chapters that had been scanned and loaded into GSU’s eReserves system, 99 items in all, of which Judge Evans ruled on 74. The first ruling in that case was back in 2012 by Judge Orinda Evans (and I wrote a Practical Guide here). The publishers weren’t particularly thrilled with the outcome, so they appealed the case, and the appellate court handed down a ruling in 2014 that sent the case back to Judge Evans with instructions about how to amend her item-by-item Fair Use analyses (and I wrote a Practical Guide here). This Judge Evans has now done for the 48 items that the publishers thought were still at issue, resulting in the 2016 ruling that came out on March 31st. If you’re interested, here are all the details and filings involved in Cambridge University Press et al. v. Patton (this case’s official name).

One of the main things that the appellate court asked Judge Evans to do was a much more thorough analysis of the “Market Effect” (Factor Four), delving into year-by-year sales and permissions licensing fees to see what the effect of unpaid nontransformative copying really might have been on those two markets. This plus the appellate court’s insistence that there should be no “mechanistic” Fair Use analysis means that the new ruling isn’t very generalizable to on-the-ground decision-making in our libraries and classrooms. Still every new ruling helps to clarify the boundaries of Fair Use, so this one is still worth paying attention to.

In this case, of the 48 items under consideration, 4 were ruled to be copyright infringement, and all of these 4 had previously been ruled infringement in the 2012 version of the ruling. Interestingly, of the 48 items, only 2 decisions changed as a result of the appeals process:

  1. One item had previously been rejected because the publishers hadn’t sufficiently proved that they owned the copyrights in the first place. This one was now ruled to be Fair Use, so the publishers must have provided that proof at some point.
  2. One item had previously been ruled infringement, and this time it was ruled to be Fair Use.

If you’re a real nerd, here’s a spreadsheet I made to track the item-by-item analysis and compare their outcomes between 2012 and 2016.

Highlights of what librarians should know

  1. This ruling makes it clear that there is no transformative nature to eReserves. Judge Evans says, “the excerpts are nontransformative because they are mirror-image copies of a part of the book” (page 5). I’ve heard some people wonder if readings like this might be transformative for various reasons including presenting only excerpts in juxtaposing items in ways the original publishers didn’t intend. This ruling makes it clear that rearranging how students experience the readings is not a transformative use.
  2. Because the copying was nontransformative, Factor Four (the “Market Effect” factor) was given the majority of the weight in the Fair Use analyses for each item.
  3. GSU’s status as a non-profit educational institution, and all of this copying done in the service of teaching, Factor One (the “Nature of the Use” factor) always favors Fair Use. One useful thing to note is that the professors in the courses were questioned about exactly how each of the readings figured into their courses, and this information informed both Factor One and also Factor Three, the “Amount and Substantiality” factor. If the amount copied was “narrowly tailored” to the professor’s pedagogical goals (see page 22 for example), that helped tip things toward Fair Use.
  4. For each item, two markets were considered: the market for sales of the full book, and the market of permissions fees for the book chapter(s) when available. The publishers worked very hard to make it look like if there are licenses available for the chapters, then faculty and librarians should pay those fees almost without thinking. Nancy Sims (J.D., M.L.I.S.) and Kevin Smith (J.D.) strongly advise the opposite. Smith’s contention is that there will only be a market for those licenses if we pay for the licenses all the time. If we don’t pay for the licenses unless we really have to, then there’s no market to effect. I think what he has to say on the matter warrants a nice long quote:

    “We should resort to paying for licenses only very rarely, and when there is no other alternative.  The simple fact is that the nature of the analysis that the Court of Appeals pushed Judge Evans into is such that licensing income for the publishers narrows the scope for fair use by libraries.  To my mind, this means that whenever we are faced with an e-reserves request that may not fall easily into fair use, we should look at ways to improve the fair use situation before we decide to license the excerpt.  Can we link to an already licensed version?  Can we shorten the excerpt?  Buying a separate license should be a last resort.  Doing extensive business with the Copyright Clearance Center, including purchase of their blanket campus license, is not, in my opinion, a way to buy reassurance and security; instead, it increases the risk that our space for fair use will shrink over time.” (emphasis mine)

  5. At the appellate court’s demand, Judge Evans was directed to pay careful attention to Factor 2 (the “Nature of the Work”) factor. This is the factor where if the item is creative, that makes copying less fair, and if it’s not creative, that makes copying more fair. (At the extreme, lists of phone numbers in the phone book are not creative enough to be copyrighted.) Some people have painted this with a very broad brush and said that fiction is creative and non-fiction is not creative. That’s obviously not nuanced enough. It was really interesting to watch Judge Evans evaluate each excerpt looking at writing style and whether the author provided opinion or evaluation. One item’s analysis on this factor ended with, “Author opinion, subjective description and evaluative expression dominate. Factor two disfavors fair use” (pages 36-37). Only 2 of the 48 items in this case came through with a “favors fair use” on Factor Two. The other 46 were either neutral or disfavored fair use.
  6. Given all of this, but #3 and #5 in particular, it was reinforced for me how important it is that the faculty member do the fair use analysis before putting things on eReserve. There’s no way that library staff would be able to carefully weigh the four factors without reading the works and knowing exactly how they will fit into the syllabus.
  7. And finally, I really appreciated Kevin Smith’s point that since very little of the Market Effect analysis is possible without court orders, making Fair Use decisions in “Good Faith” becomes all the more important. That way employees of non-profit educational institutions or libraries are not subject to statutory damages. (And I would add to his point that keeping records of these good faith decisions is very important.)

Further Reading (by people who know more than I do):