Blog Post

Spring Spotlight #3: Digital History’s Relationship to Human Rights Archives and Data Analysis

Addressing issues of access, preservation, and use of human rights related materials from sites around the world, the archivists, professors and practitioners featured in this Spotlight share their experiences and expertise, highlight new tools and resources, and engage the pressing conceptual, political, and practical challenges associated with human rights archives and data analysis.

Featured Practitioners and Projects:

  • Pamela Graham, Director of Area Studies and the Human Rights Center, Columbia University (contact: pmg16@columbia.edu
     

Can you describe your work for us briefly, provide a few examples of projects you have done, and give us a sense for what your day-to-day can look like?

 

Trudy Huskamp Peterson (Independent Archivist, Contractor, Consultant)

Day to day:  I am a contractor.  Each contract is different and so the daily work varies.  Here is my day on February 14, which is a fairly typical day. When I got up I skimmed the email that came in overnight and found that Swisspeace (an NGO in Switzerland) proposed a telephone conversation either that morning or the next about a training course planned for March in The Philippines.  I sent a quick message back that I could talk the next morning.  Then I went to the World Bank for a half day, where I am working on contract, primarily looking at access issues.  In the afternoon at my home office I completed a draft for a proposed course on globalization and archives for the East Asia Branch of the International Council on Archives (ICA) in Hong Kong and sent it off.  I answered an inquiry from a posting on an archives listserve as to whether bank account numbers should be protected by a privacy restriction (I said yes, if any one of the account holders is alive).  In the process of answering and giving the inquirer some background information, I went to the ICA website for the link to the Principles of Access to Archives and noticed that the Spanish and Portuguese translations were not there.  So I sent an inquiry to the ICA office as to why these weren’t up (answer: soon).  I got an email from the Society for the History of American Foreign Relations that a session I proposed in honor of Anna Kasten Nelson was accepted; I then sent emails to her sons informing them of this memorial event next June.  I answered an inquiry as to whether I would be willing to do a consulting visit in a country in Central America (yes).  Finally I turned to reviewing news media. I write a monthly newsletter on archives and human rights in which I report the previous month’s news, and to gather pertinent stories I read dozens of news feeds.  On this day I skimmed pretty thoroughly the New York Times and the Washington Post in paper and al-Monitor, Balkan Transitional Justice Daily, RAINbyte and others electronically.  

Projects: Three major projects that I have worked on in the last decade have been (1) the police archives in Guatemala, where I served as the archival consultant and trainer; (2) the Special Court for Sierra Leone, where I helped assess the state of the records, provided training to the staff, and both provided training and organized a preservation training course given by others for the Court and for the National Archives of Sierra Leone; (3) the Nuclear Claims Tribunal in the Marshall Islands where I am currently working.  Much of my first visit to the Tribunal was spent shifting boxes and crates and equipment in a storeroom to see what records were there; since then I have been appraising and describing the records and searching for money to begin a digitization project.  There is much more work to each of these than the simple sentences imply. 

 

Patrick Stawski, (Human Rights Archive, Duke University)

As Human Rights Archivist I am responsible for all aspects of the Human Rights Archive, a specialized collecting area within the David M. Rubenstein Rare Book & Manuscript Library at Duke University.  The Human Rights Archive supports research, scholarship and activism in human rights.  On a daily basis I work to build distinctive collections through purchases and gifts; develop public programs and outreach activities for the archive; and I work with students, faculty and researchers to facilitate their use of the archive’s holdings through reference support and classroom instruction.  I work closely with the Duke Human Rights Center and other departments and centers throughout the Duke community to help foment human rights education and activism.

Projects I am currently working on include the acquisition and appraisal of the Southern Poverty Law Center’s collection of extreme right-wing literature and print material; processing of the International Monitor Institute records, a collection of over 5,000 videos, databases, and related organizational records that document human rights abuses in the Balkans, Burma, Rwanda, Sierra Leone, and other areas; a video digitization and description project in conjunction with Duke’s Nicholas School of the environment and the United Nations Environmental Program (UNEP), focusing on the environment and human rights in post-conflict settings; a project called Rightsconnect to evaluate and propose a program for undergraduate education in human rights at Duke University; preparing for a public screening and discussion of the award-winning film “The First Year” which deals with the challenges faced by teachers in our public schools.

 

Ben Miller (Digging into Human Rights Violations)

My principal research is on how data media influence the collective memory of population level events.  Specifically, in terms of events, I’ve focused on genocides and mass violations of human rights such as occurred during slavery in the United States, under the Khmer Rouge in Cambodia, during the ethnic cleansings in Guatemala and Bosnia, under the Nazi regime during the Holocaust, and during the Rwandan Genocide.  Other types of events my work has taken on are the complex of disasters experienced in Tōhoku in 2011 and the more prosaic way in which people use bulletin boards to discursively map their cities. 

My projects address the intersection of various technologies, frequently but not exclusively in the domain of human rights.  One project, Digging into Human Rights Violations (http://digging.gsu.edu), applies natural language processing techniques and a computational model of narrative to the understanding and analysis of human rights violations reports.  With funding from the NSF and Canada’s SSHRC, that project is developing a method for recognizing when an unnamed entity in one document in a corpus is identical to an entity in another document in the same corpus.  This problem, known in computational linguistics as cross-document coreference, is being taken on by a team of computer scientists and humanists from various fields including linguistics, literature, archival science, information science, and HCI.  Our approach is a phrase-based context modeling that extracts a set of trigrams of person-place-time.  These trigrams, reinforced by discursive context, are presented visually to a researcher to facilitate their exploration of a given event corpus and their attempts to find relevant documentation.

Another project less directly connected to human rights works explores possibilities for narrative storytelling in geographic information systems, and for data based histories of cities.  With internal funding from GSU, the team with which I’m working is developing a platform focused on Atlanta but adaptable to other cities that integrates historic maps (approximately 1,500), recorded sound, various data bases, and a curation mechanism so as to enable the construction of GIS-enabled stories that offer the precision of big data and the engagement of personal narrative.  One example is Hannah Palmer’s “I’m From Here” (http://www.atlmaps.com/imfromhere).

Each week is an attempt to balance keeping up with the developing research on digital archives and collective memory (such as W. Ernst’s Digital Memory and the Archive (UMN Press, 2012)), with the technical skills and models necessary to interpret these new archives (such as J. Chang’s LDA package for R (Feb. 15, 2013)), with the typical demands of being faculty (currently teaching an undergraduate theory course on simulations and interactive fiction and an independent study on the discourse of intervention relative to lawless vs. ungoverned spaces), with the administrative demands for each project (finding people and the funding to bring them on board), coordinating with established partners and reaching out to new ones, and writing to deadlines for various conferences and publications (most recently, EuroVis 2013).

 

Lu Xiao, (Digging into Human Rights Violations)

I’ve worked in various projects that covered different aspects of information science and information system research. The scope of my projects ranges from field studies through ethnographic and case study approaches, to actual software development, and evaluation of information systems through socio-technical approaches. Example projects include: understanding and supporting online deliberation with natural language processing tools; enhancing group decision making process with group visualization tools; and understanding researchers’ data analysis and data management practices in the study of primary data about human rights violations –part of the Digging Into Human Rights Violations data project (DIHRV). In the DIHRV project, my team has been reviewing research studies on human rights violations and interviewing researchers to understand the kind of data (format, source, etc.) and data analysis methods they used in their research.

 

Pamela Graham, (Director of Area Studies / Global Resources Division and the Center for Human Rights, Columbia University)

I direct a center and program within the Libraries that has the mission of collecting and making available human rights information.  This work represents half of my responsibilities as I also have other administrative roles within the library. My day-to-day human rights-related work can involve communication with donor organizations or individuals about archival collections that we hold, working with other librarians to identify and select relevant general collections related to human rights, and discussing a variety of ongoing projects. I have regular contact with students and faculty which is very helpful and important to understanding the full life-cycle of how information is ultimately used in research and teaching.  One of the exciting projects I have worked on is our Human Rights Web Archive, where we are capturing and preserving  a collection of human rights-related websites.  I’ve worked mostly on the content aspects of this project but also on advising on the design, usability, and research uses of the archive.

 

What are a few of the prospects, challenges, and risks for archiving born-digital records, especially as it pertains to human rights, truth and reconciliation, and the politics of memory? What are some of the risks of using “big data” in human rights?

 

Trudy Huskamp Peterson (Independent Archivist, Contractor, Consultant)

The prospect for technically preserving born-digital records is, in a word:  good.  Archivists around the globe now know that they have to save born-digital records, and many new tools are becoming available to help harvest and preserve these records. I think the concerns here are rather different from technical preservation.  Let me highlight three.

First, records creators have to be convinced that saving their electronic records is important.  Many of not most records creators are focused on the here and now, and records of five years ago are irrelevant to their daily work.  If they think about their past records, they probably still think of visible paper rather than invisible bytes.  And if they are interested in erasing their past, it is a lot easier to erase electronic records than to eliminate quantities of paper, as police departments from Paraguay to Lithuania have learned.

Second, and related to the first, archivists usually do not have enough clout to ensure that the electronic record is preserved.  The archival service within a hierarchy (in a corporation, in the government of country X, in an NGO) is usually buried in a general administrative unit or a tech department, neither of whom have an historical perspective and whose job is efficiency and cost-saving.  Then, too, some of us who choose to be archivists do not have the personality that makes us table-thumpers and attention-getters for our records programs, particularly in the face of determined opposition from someone at a higher hierarchical level—although I think archivists today are more assertive than we were a couple decades ago.

Third, it is not reasonable to expect every NGO and every local government to have an archives for electronic records.  While storage in the cloud is an option, it really does not serve the purpose of preserving a digital archives that can be made available to researchers, and it is too costly for small institutions.   What we have to have is a set of trusted digital repositories around the world that will manage the preservation of the master electronic records, while each local entity can control the access to them and provide the reference service using a duplicate reference copy. The Federal Archives of Switzerland, for instance, is providing preservation service for the scanned copies of the records of the Guatemala police archives, and the National Archives of Finland is providing preservation service for UMAM, a human rights NGO in Beirut.  Guidelines for such services are being developed; Audit and Certification of Trustworthy Digital Repositories http://public.ccsds.org/publications/archive/652x0m1.pdf is the easily accessible final draft of the recent standard from the International Standards Organization.

 

Patrick Stawski, (Human Rights Archive, Duke University)

The Human Rights Archive works primarily with the records and papers of human rights NGO’s and individual activists (as opposed to say state records or human rights scholars), and our users/patrons are scholars and researchers.  Therefore I can only speak to this issue from within this archival context (I imagine the response will vary depending on that context).

Our injest of digital records has included everything from boxes of diskettes to hard drives that contain entire servers of organizational records to crawls of human rights websites.  Content of these digital records also encompass all the record-types produced by NGO’s: board meeting minutes, memos, trip reports, presentations, databases, etc.

Duke Libraries began development of a digital records program years ago, so we now have a digital preservation environment, an Electronic Records Archivist who is in charge of digital preservation and provides the transfer and accessioning technology and protocols for injest of digital records.  Some aspects of our digital records program are still in the process of development, particularly description and access. 

This question seeks to identify challenges unique to the combination of digital records AND human rights work.  Digital records, of human rights or any other work, are a record format and have their own challenges regarding preservation, description, and access. Human rights records, whether analog or digital, are characterized by their provenance and pose different types of challenges: they may contain sensitive date which requires much thought and care when processing, describing and providing access; the content of the records may represent community memory and documentary heritage which complicates acquisition, appraisal, and access. 

As archivists are just now beginning to manage and provide access to the combination of these two, I don’t have too many real life examples to provide.  Certainly, human rights records in digital form will more easily lend themselves to processes such as redaction of sensitive data, secured/tiered access (assuming such a technological environment is present)and remote reference/access.  For example, using digital forensic technology we review large quantities of records for SSN’s, bank account numbers, and other types of personal data.  As we role out access beyond the traditional reading room environment we will begin to confront new challenges.

In respect to issues of the politics of memory, we might want to consider more broadly the information economy that encompasses archives, records, records users and creators, and the record making systems that determine what will and what will not be recorded. Even in the world of human rights practice such information economies are dependent on often unequal access to technology.  Not only records but the record systems which produce them can become points of contention and struggle, even within a human rights organization.  In the context of a transnational NGO, who determines what records will be made, who will and who won’t have access to them, and where they will be archived?  These politics strongly effect the shaping of the historical record and inevitably impact historical research.

 

Ben Miller (Digging into Human Rights Violations)

Big Data analysis is a complex, frequently black box set of methods that can introduce procedural errors resistant to diagnosis and correction.  Conversely, those errors are often already implicit within the data in a way that analysis reveals.  For example, consider Aaron Koblin’s “Flight Patterns” data visualization project <http://www.aaronkoblin.com/work/flightpatterns/>.  In his methodology, Koblin describes that visualizing the FAA’s flight data showed impossible routes – planes that arrived prior to departure or whose flight paths included fractured paths and hundred mile jumps – and that he manually redacted these physically impossible journeys from his data.  His test was simple and thorough – does the data violate the laws of physics – and not a viable test for error checking most human rights violations records.  For many reasons, all data sets include these and other kinds of errors; sometimes, because a particular data set is itself a concatenation of other data sets comprised of slightly differing schema; sometimes for technical reasons such as when a system invisibly truncates records to predetermined field lengths; and always, for human recording errors.  Relying on data simply because it is included in a rigorous schema, produced by a reputable organization, or automatically recorded by devices is something these formats invite us to do.  That, and the black box nature of most computational analysis present two of the most significant risks for working with collective digital memories. 

Scholars present the failure to forget, or to phrase it more technically, the lack of an ability to delete as a social function technology requires (recently,  in Viktor Mayer-Schönberger’s Delete: The Virtue of Forgetting in the Digital Age, Princeton UP, 2011), and most provocatively, in Matthew Kirschenbaum’s Mechanisms: New Media and the Forensic Imagination, MIT Press, 2007).  That aspect of digital memory – that no permanent and comprehensive deletion is possible – leads to the most dangerous aspect of digital record keeping; leaks are seemingly inevitable and permanent.  Therefore, all projects addressing personally identifiable information in a human rights domain have to take as their first mandate the confidential, secure storage of their data.  Given that big data analysis requires geographically and systemically dispersed teams running many cycles of iterative analysis, that mandate is a challenge.  Even proper anonymization  is a challenge (see L. Sweeney, “Foundations of Privacy Protection from a Computer Science Perspective,” Proceedings, Joint Statistical Meeting, AAAS, Indianapolis, IN. 2000). In the era of psychoanalytic criticism, it was sufficient to relabel a patient by first name and last initial, or possibly by a nickname: Anna O. and the Ratman being famous examples.  Given the scale and efficacy of techniques in data and text mining to determine ambiguous referents via coreferent research tactics, a set of practices excelled at by born-digital communities such as Reddit and 4chan, prior modes of anonymization are insufficiently secure.  The alternative is overzealous redaction in the manner of Guatanamo Bay interrogation transcript releases, something I wrote about here <http://60hzhumanism.wordpress.com/2010/11/06/on-redaction-fbi-4543-4551/>.  One current challenge facing those who work with digitized or born-digital primary source material with an end goal of generalizable research or historiography is to develop, in this new context, viable methods for the protection of identities while still documenting their stories with specificity.

 

Lu Xiao, (Digging into Human Rights Violations)

Different people may have different definitions of “big data”. With a lot more data available in human rights, it certainly has the potential of offering researchers a more comprehensive view of the situation than previously possible. This indicates more possibilities of analyzing the data, such as identifying more patterns about the human rights violation cases, allowing deeper understanding about specific cases, etc. Researchers have been using quantitative and qualitative methods to analyze human rights violation data. With the availability of big data, there might be a tendency to favor quantitative analysis (e.g., statistical modeling) because aggregation is possible, and because it can be very time consuming to conduct qualitative analysis. However, this method has its risks in human rights research. For example, quantitative analysis may miss rich contextual information, such as the behaviours and thoughts of the victims and perpetrators, which is important for understanding the whole issue and which is present in the big data. The presence of such risks makes it critical to develop software programs that facilitate the qualitative analysis of big data in human rights research.

 

Pamela Graham, (Director of Area Studies / Global Resources Division and the Center for Human Rights, Columbia University)

The challenges of preserving born-digital include the great volume of data that is and can be produced and the demands this places on libraries and archives.  How does one “curate” large quantities of data?  There are also issues related to access, some that deal with privacy issues, intelectual property issues, and some that are practical matters of how one can navigate and explore digital records. What does the concept of “arrangement” mean in the digital realm, and how should we create intellectual connections access between print records and digital?  Our scope of born-digital extends into the freely available web-based content created by human rights organizations.  The HRWA project has been a great learning experience for us.  There are curatorial issues of how to select sites for preservation given the vast quantity of material that is available, and how to make all of this data available and usable.  We’re also hoping to foster the development of new tools that can be applied against this curated collection, so researchers can visualize trends and patterns in the information presented on the web.  In the end, though, one still has to ground the use of big data in a solid understanding of the context and history of their research topic and not rely solely on big data as a method of explaining trends and events.  And we have to remember how much data and information is generated outside the reach of the digital, and how we should and can be preserving the full spectrum of documentation that is generated.

 

As students of the past many of us grapple with what anthropologist Michel-Rolph Trouillot astutely termed the “silences” of various archives with which we work, what are some of the challenges facing current and future historians with the corpuses of data and/or archives with which you work?

 

Trudy Huskamp Peterson (Independent Archivist, Contractor, Consultant)

Silences:  OF COURSE there are silences in archives.  Records are created in the course of doing something, either a transaction or an action.  A production manager responsible for getting enough Agent Orange manufactured to fulfill the requirements of the chemical company’s contract with the Department of Defense is not documenting how the product will be used:  he or she is making sure the product is manufactured as required and “on time under budget” and that what is recorded.  Colonial officials concerned about getting labor to work on, for example, water channels documented how many people came to the corvee and for how long; they weren’t concerned about the effect of such work on, say, pregnant women. 

Silences also occur because records were destroyed.  Some records are destroyed after archivists have appraised them and decided that they do not hold either information that is necessary as evidence of what the creating institution has done or unique information about persons, places, events or phenomena.  If that decision is wrong, the archivist is responsible.  Other silences occur when destruction happens without authorization, either because there is no archival authority or the archival authority is ignored.  And natural disasters and insects and rodents ravage records; I attempted to pick up documents in the storeroom at the Nuclear Claims Tribunal that fell apart from termite damage.

Data archives: I am not sure what you mean by “data archives.”  I think of “data archives” as institutions that preserve scientific data in structured formats that can be processed, often through the use of mathematical algorithms.  ICPSR is a model of this kind of archives.  Most of the records held in a data archives arrive as data sets, such as data from the annual survey of jails complied by the U.S. Bureau of Justice Statistics.

While it is true that data for classic data archives can be extracted from or extrapolated from archives (as Patrick Ball has notably done), I think plain text and audiovisual archives in electronic form will be the most frequent sources for historians in the future.  Here the challenge will be the volume of electronic records that is available and finding methods to use that documentary multitude.  The legal profession has faced this challenge in the process of discovery, and courts in the United States are now accepting disclosure based on documents identified through algorithms of eDiscovery’s “technology-assisted review and auto-categorization” (see, for example, the opinions in Zubulake).  In the future some historians will adopt such tools, but I also think that archivists will try to manage electronic records in ways and provide finding aids that will lessen the dependence on retrieval by mathematical models.

A gap will probably widen between the finding aids that will be produced when archives digitize bodies of records that were originally in paper or audiovisual formats and the finding aids for born-digital records.  In the digitization process, the archivists have the opportunity to create “tags” for series and folders (and—in special instances—items) that will provide very direct entries for users.  With born digital records, archivists will probably continue to manage them using the tools that the creator used, plus some enhancements.  So the divide will likely be noticeable to users, at least for the near term.

 

Patrick Stawski, (Human Rights Archive, Duke University)

Above and beyond the silences already present in archives due to appraisal, description, and access policies?  Certainly one challenge will stem from the high costs and resources required to preserve digital records; only a small set of current archival institutions may be able to take on such expenses thereby possibly reducing the preserved and publicly accessible corpus of historical records.  Could this create large monopolies of historical records?

Again, here too we should consider the record creating practices of the human rights community and what those practices bring to light and what they obscure.  Such obfuscation will be transferred to the archive.  Another issue that comes up for me is how do we (creators and users) set boundaries on what will be available to history. For example, human rights work and by extension the records it generates are infused with both individual and group trauma.  Can historicizing this trauma reactive it?

If we don’t develop adequate discovery and access tools, even those records that are preserved will remain silent.  Researchers will have to evolve their methodologies and expectations as they consult born-digital records.

 

Ben Miller (Digging into Human Rights Violations)

A major hurdle for computational or statistical analysis of any archive is that the text be sensible to existing methods.  Rarely are the archives significant for historical research transcribed, only relatively recently has there been consistent coherent work to digitize them with sensible schema, and only in the last few years have there been well funded attempts to develop schema for digitizing structural and orthographic features of divergent texts (e.g., Laura Mandell’s Early Modern OCR project). The first challenge, therefore, is locating and securing access to a collection sensible to computational methods.  The second, as I indicated elsewhere, is that the methodologies are complex and transdisciplinary.  The third is retaining at the front of our minds that all a corpus can give evidence for is what’s contained within that corpus.  There is a disconnect between that corpus and the event to which it refers, and speaking to the event, instead of speaking to the corpus is always a strong temptation.  For example, Yale’s Cambodian Genocide Program’s Geographic Database <http://www.yale.edu/cgp/maplicity.html> contains the manually collected GPS coordinates of sites significant to the understanding of the scale and scope of the Khmer Rouge’s genocidal regime.  However, as was recently made clear after 13 years of research that there were approximately 42,500 ghettos and camps comprising the genocidal infrastructure of the Holocaust – a number far larger than was previously believed to be true -- even events that seem well documented are capable of demonstrating that our documentation and understanding is highly circumscribed (in The United States Holocaust Memorial Museum Encyclopedia of Camps and Ghettos, 1933–-1945, edited by Geoffery Megargee).

 

Pamela Graham, (Director of Area Studies / Global Resources Division and the Center for Human Rights, Columbia University)

Fascinating question.  We are trying to break some of those silences in several different ways. First, we are collecting and preserving materials that were not easy to locate or use in the past.  As we collect records of human rights organizations we bring the materials from warehouses and individual disk drives into a more open context that will support the use of these records for research and learning.  Records that have survived can be silent if they are not easily found or used.  Second, there are methodologically-reinforced or generated silences that exist and limit the use of records that have been preserved. We are purposely trying to engage and approach a wide variety of users who might not usually turn to archives or see these as sources to inform their work on contemporary and current affairs.  Historians are a natural beneficiary of our collections but we hope to encourage the use of these archives among advocacy groups, public policy scholars and political scientists. A third kind of silence can be created through the prevailing tools and discovery patterns for digital information. One could argue that human rights information on the public web is easily found and used.  But that is not always the case. Part of the rationale to curate and collect human rights web content is to offer the chance for researchers to work with this material in an environment that is different than the live web.  That is, with an interface and discovery tools that let you look at sites across time and one that is free of the ranking algorithms you will get with popular search engines.  Small organizations without a lot of web traffic may have a difficult time being “heard” or discovered on the live web.  For example, I recently searched for a report on violence in Pakistan authored by a human rights organization and in a Google search you would have to go through 5 pages of results to find it, unless you knew enough about the title, author etc. to do a more specific search.  This organization is drowned in an unfortunately large amount of information about violence and Pakistan.  Our web archive would change these discovery paths and offer other ways to locate specialized materials, and give a greater voice to less-well-known creators of information.

 

What has been your experience working with Historians and/or non-specialists how do you see the relationship between your own work and the emerging field of Digital History? 

 

Trudy Huskamp Peterson (Independent Archivist, Contractor, Consultant)

Work with historians:  What is my experience working with historians?  If you mean serving as a reference archivist and working with historians as researchers, the answer is I have worked with historians who treated me as a valued colleague and with those who seemed to consider me a “fetch and carry” minion.  But of course I work with historians in many other ways, too:  I have been the president of the Agricultural History Society, served on the Council of the American Historical Association and on the executive board of the Society for History in the Federal Government, am currently on the Historical Advisory Board of the Department of State and so on.  In those organizations I have always been treated as a full professional.  Finally, I always use history and historians as I prepare to work in a country that I don’t know.  If I am asked to consult in Country X, I get in touch with a historian I know who specializes in X and ask what I should read.  Then I try to read 3-5 books of solid history before I go.  I am not trying to turn myself into a scholar of the country or region, but I am attempting to get context for the current situation in the country.  And I know of no better way to do that than by reading history.

Digital history: As to Digital History, that is the work of the user of archives.  My job as an archivist is to ensure that records and personal papers are selected and retained that will provide evidence of the institution or individual that created the records.  That responsibility includes all media used to create records.  How the user chooses to work with the media is, of course, that person’s professional option.

 

Patrick Stawski, (Human Rights Archive, Duke University)

Our collections are consulted by a wide range of researchers from many discipline and backgrounds: theologians, anthropologists, public policy, political scientists, documentarians, faculty, students, and independent researchers.  For the time being I think the best thing I can do is listen to our researchers to understand their research needs and concerns regarding digital records.  Duke Libraries has dedicated Digital Scholarship librarian who works with library staff and library patrons on digital scholarship projects.  I am currently collaborating with her and a group of Duke faculty and students on a human rights and the environment project.  I’ll report back on what I learn as the project progresses!

 

Ben Miller (Digging into Human Rights Violations)

The field most in demand for collaboration with digital humanists, and I imagine the same is true for digital historians, is computer science.  Developing research projects that provide tangible, discipline-valued outcomes for both humanists and computer scientists hasn’t been the challenge – what has been is finding collaborators with available time or with relevant backgrounds.  That holds true at all levels of practice from undergraduate research assistants to senior faculty colleagues. 

One of the most valuable pieces of advice I received was for a failed grant proposal.  A reviewer suggested that for that project, an attempt to use topic modeling to automatically identify the appearance of specific types of rights violation in witness statements, we would benefit from consulting an archivist.  That consultation led to an increased sensitivity to touches, the lifecycle of records, and how to document both for born-digital and digitized documents.  I find discipline specific practices as relate to the understanding of records, text, and authorship to have enriched my principally literary and computational understanding of those systems.

 

Lu Xiao, (Digging into Human Rights Violations)

My research and training experiences have been multidisciplinary, and I’ve had the opportunity to work with people from non-technical backgrounds, e.g., Education, Psychology, and Library Science. In this DIHRV project, my team has interviewed researchers in Law, Political Science, Economics, and History. And we expect to reach out to more researchers in different non-technical disciplines. Because my team has been focusing on understanding the users’ needs for the software programs, we did not encounter situations in which technical/non-technical background matter a lot. When we seek feedback/expectations about the software program from the researchers, we focus on issues at the high level, in other words, concerns that researchers may have for a qualitative analysis program that filters the information. I know little about human rights research, and interacting with them has been a great learning experience for me about the subject.

 

Pamela Graham, (Director of Area Studies / Global Resources Division and the Center for Human Rights, Columbia University)

Historians have been among the main users of our archives and have been great supporters of the work that we are doing.  I am also working actively with our Human Rights MA program here at Columbia, which attracts persons who often have one foot in the academic world and one foot in the world of advocacy and policy. I am hoping to build more interest in how our documentary record can inform their non-academically oriented work.  Digital History, or digital humanities more broadly, is actively being cultivated and explored here.  I think there are a lot of opportunities going forward – for example, so much of human rights advocacy and networking lives fully in the digital realm. Our most recent grant from the Mellon Foundation to support our Web Archive is focusing on the development of tools and techniques for analyzing digital content.  I think it will be very important for us to connect with those who use technology and tools for analyzing digital records.  But I can repeat what I said earlier – those tools and techniques are not a replacement for understanding the context and history of topics and events being studied and that context is vital to interpreting the outcomes of digital analysis.

 

Where would you say the future of archival work around human rights is headed and how might we best prepare for work in this field, either directly or through collaboration?

 

Trudy Huskamp Peterson (Independent Archivist, Contractor, Consultant)

Society for the History of American Foreign Relations, John Lewis Gaddis has an article on writing the biography of George Kennan.  In it he quotes Kennan’s essay “The Experience of Writing History”:  “the spectacular mechanical and scientific creations of modern man tend to conceal from him the nature of his own humanity and to encourage him in all sorts of Promethean ambitions and illusions.  It is precisely this person who, as he gets carried along on the dizzy pace of technological changes, needs most to be reminded of the nature of the species he belongs to, of the limitations that rest upon him, of the essential elements, both tragic and helpful, of his own condition.”  That caution applies—really applies--to archival work in the area of human rights.

Part of the answer depends on what you mean by human rights archives (I am assuming you are asking about the records, not about archival institutions).  First, there are the archives of human rights organizations themselves, from Amnesty International to the Mack Foundation in Guatemala.  These need to be protected and preserved, and I think there is a reasonable chance that most of these institutions will take care of their records.  Second, there are the official bodies that are supposed to have something to do with human rights:  human rights commissions, United Nations High Commissioner for Human Rights, government lawyers in a civil rights unit, ombudsmen and so on.  These archives also have a reasonable chance to be protected.  Then we come to other parts of government, including those that may be tempted to commit human rights abuses; to the corporations that also have a business responsibility to protect human rights (see the 2011 report to the UN High Commissioner for Human Rights “Guiding Principles on Business and Human Rights”); to faith-based organizations (think of the sexual abuse charges against the Catholic Church); NGOs (the Boy Scouts of America also with sexual abuse charges).  Is it truly possible to have the foresight to save all these records?  When cases of human rights abuses are uncovered and documents are found that substantiate the claims of victims, it is a remarkable demonstration of the power of archives.  But we cannot expect to be so wise as to identify and save all such records or powerful enough to prevent their destruction in some cases.  We—both archivists and human rights activists--have to try to preserve these records, but as Kennan says, “Promethean ambitions” face real human limits. 

The International Council on Archives’ Human Rights Working Group is preparing a document that will state the archivists’ responsibilities for records that have human rights implications.  That will give the profession a basic statement of best practice, which will at least be a lever to use in discussions with records creators and will put a floor under archival practice in this area.

Preparation:  Education is key.  I think archivists should have a graduate degree in history and have experience doing research in primary sources.  I think that is essential in order to make good appraisal judgments, to describe records well, and to provide quality reference service.  I worry that the archival profession is too focused on techniques and not enough on content.  I worry that not enough historians choose to become archivists, bringing the special skills of research and analysis to bear on the work.

And I worry that the archivists and historians are drawing apart.  The Society of American Archivists, the American Historical Association, and the Organization of American Historians used to have a joint committee that met a couple times a year to talk about mutual issues.  This died two decades ago, and despite my urging to both historians and archivists it has not been restarted.  This divide is going to become even more a problem as historians start to use online materials more or less exclusively and, as the recent study from Ithaka S+R said, spend “the majority of their time in the archives informally digitizing materials for later review and analysis.” (“Supporting the Changing Research Practices of Historians,” http://www.sr.ithaka.org/research-publications/supporting-changing-research-practices-historians) We need to talk to each other, to work together, and support the crucial preservation of and access to archives.  Only by working together can we assure the future of the past. 

 

Patrick Stawski, (Human Rights Archive, Duke University)

I do see a need for more collaboration between the record-creators and archivists so that preservation and access concerns are addressed early on in the record life-cycle.  For example working with human rights litigators to understand and make best us eof Case Management Systems.  There is already collaboration around the creation of digital repositories and digital access (Hathi Trust, California Digital Library, etc.) and we can expect more collaborations in the future.

 

Ben Miller (Digging into Human Rights Violations)

Although the level of expertise necessary to produce projects like With Criminal Intent, a project arguably about the development of the application of justice in modern England, is only possible in a large team, learning how to properly phrase the research questions in a way comprehensible to the entire team takes a transdisciplinary background.  Preparation in historiography with a sensitivity to language is a given, but given how much of contemporary data based human rights research is predicated on data visualization and corpus linguistics, fields themselves predicated on statistics, that would seem to be the most relevant starting point for training.  The other technical area upon which human rights work with large text corpora relies is basic text analytics: not the development of new modes of analysis or algorithms, but simply being able to process a corpus through, for example, a topic modeling library in Python or R and getting a feel for the steps involved and the outcome of statistical interpretation and presentation of humanistic data.  Absent that, humanists have to overly rely on technical collaborators in a way that asks them to provide services, rather than produce research.

 

Pamela Graham, (Director of Area Studies / Global Resources Division and the Center for Human Rights, Columbia University)

I think we will need many kinds of specialists at the table. I am not trained as an archivist and I have a more programmatic and curatorial role in our environment.  I work closely with archivists and appreciate the enduring elements of their training that will transcend the shift into the digital:  the ways they approach problem-solving, and how to approach the arrangement, description, and access to records.  We need to have good partnerships with technologists and with the user communities to understand how information is being created and how researchers in different contexts wish to use the information. This is very important since the ability to create and disseminate information is growing exponentially.  We definitely will need continuing and deeper collaboration among archives that work in the field of human rights.  Both between university-based archives and between academic archives and other kinds of human rights organizations. I don’t think we can afford unnecessary duplication of  investments and effort in what may be very complex and expensive digital infrastructures to support archiving and research uses of archives.  

54

3 comments

Thanks for this insightful commentary Joseph! 

Memory of the World is another UNESCO project that invites submissions of material similar to crowdsourcing. The recently issued Vancouver Declaration on Digitization and Preservation highlights some of these critical issues of accessibility and involvement and a group of Tunisian archivists have also recently issued recommendations on the right to access information. 

I was involved in a discussion recently about a group of radical archivists who are trying everything they can to get in and digitize police and government records that they suspect will decompose before they are declassified. The stakes of these struggles over access to information could not be higher, especially in the reactionary age of "wikileaks." You're right to call our attention to the role of activists, survivors, people most directly involved, the people who are daily fighting to know what happened to a loved one and over the future of their community, as well as the broader public. I know Trudy has worked along side countless activists, and Patrick Ball, of the Human Rights Data Analysis Group, who was originally going to be featured in this Spotlight had to prepare last week for an upcoming trial in Guatemala. I think in many cases the distinction between archivist and activist may not hold up under closer inspection...

For people who write about the past, this question of engaging a "catalogue of atrocity" as you put it also made me think of the political stakes of the narrative tropes we inevitably deploy, made famous by Hayden White and so brilliantly applied to C.L.R. James' Black Jacobins by David Scott in Conscripts of Modernity. Saidiya Hartman's Lose Your Mother and "Venus in Two Acts" (Small Axe, 2008) provide two other inspiring examples of finding a way to move beyond irony and unintended consequences as the dominant, if often unacknowledged, modes of writing history.

I wonder too how historians can better engage the public at a time when Bill O'Reilly's Killing Lincoln is one of the best selling history books on Amazon? 

65

Human rights archives, like almost everything else, have exploded online over the past decade. This unprecedented effusion of information is a boon for researchers and activists alike. It also creates new theoretical and methodological challenges unknown to previous generations. Perhaps appropriately, this roundtable exhibits a careful mix of enthusiasm and skepticism.

A common thread across all of these responses is the need to step back from “big data” and pay closer attention to context and contingency. Although many of their projects rely on extremely large data sets, the respondents were nearly unanimous on this point. Lu Xiao warns of the temptation “to favor quantitative analysis (e.g., statistical modeling) because aggregation is possible, and because it can be very time consuming to conduct qualitative analysis.” Trudy Peterson calls for greater humility in the face our “Promethean ambitions.” Patrick Stawski suggests closer attention to the political economy of “record making systems.” Pamela Graham asks researchers to remember all of the material "generated outside the reach of the digital." And Ben Miller seeks to combine “the engagement of personal narrative” with “the precision of big data.” The latter is a worthy standard, I think, and all too rare in digital history projects, which tend to vacillate between these two poles.

Another theme that emerges from these replies is the need for effective outreach. The archival projects on display here are tremendously diverse, ranging from East Asia to Europe to Africa, from the Holocaust to the Khmer Rouge to modern NGOs. Yet, as Graham argues, even the most well-preserved records "can be silent if they are not easily found or used." One way of managing this wealth of material, as Stawski points out, is to improve collaboration between archivists and record-creators. Certainly advances in file sharing technologies, such as cloud storage, will make this task easier in the future.

Both Peterson and Miller call for closer dialog between archivists and historians. This is not a luxury, they argue, but a baseline necessity. Peterson would mandate that archivists earn a graduate degree in history. Miller requires what he calls “a transdisciplinary background,” with a special focus on topic modeling and computer science. Yet the amount of time required to master a single discipline, never mind three or four disparate fields, can be enormous. So an even greater degree of communication between specialists is desperately needed.

One question I would pose to the respondents is how they envision the relationship between human rights archives and human rights. Or to put it another way – how do human rights activists factor into this discussion?

The issues raised here led me to reflect on my own experience helping to launch and grow the Culture of Peace News Network (CPNN). Initiated by UNESCO over a decade ago and continued with the support of the UN General Assembly, CPNN offers a platform for reports and discussion on the global progress of human rights. Public participation, transparency, and interactivity are core values of the network. Long before the rise of Wikipedia, Facebook, or Google, in the heady early days of PHP and MySQL, CPNN was open for user submissions, peer review, and debate. It was arguably among the very first websites to crowdsource its content, many years before that term was invented. Although somewhat dated now, it continues to generate new data. Submissions cover local and regional events and experiences and are especially strong for Africa and Latin America. As of 2012, the site attracts around 100,000 unique visits every year. It is a living, breathing archive. Recent digital history projects, such as the Hurricane Digital Memory Bank, History Harvest, and Legacies of British Slave-ownership, also place special emphasis on public participation and interactivity. This model invites everyone to contribute as an archivist, a curator, or an historian.

How will human rights archives of the future engage both human rights activists and the broader public?

54

I've said it before and I'll say it again - historians need better publicists.

I'm interested to learn more about Memory of the World. And thanks, by the way, to you and all the participants for taking the time and the effort to make this happen.

49