We love technology when it works properly, but then again, when does it ever? The digital side of this project has some unique challenges that many traditional historians would not face. On my end of the project, the biggest roadblock is the presentation and usage of the testimonies. The documents already procured by Dr. Nieves are in JPEG format meaning that they are not editable in their current state. This does not present an issue when displaying the original document on the website, but for reading and studying, a transcript is preferable and much cleaner. There are only two potential solutions to remedy this problem with the first being the unbearable task of hand typing out each document’s contents to create a transcript. After using this method on three or four testimonies, I can tell you that it is both time consuming and frustrating. This leaves us with the second option which is to convert the JPEG files into an editable document such as a PDF where it can then be possible to pull the testimonies off the page into a word document. The best outcome would be for software to take the JPEG files and transfer them directly into text in one sweep. In the world of technology, this is usually never possible, and currently we are utilizing a multistep process.
For Truth, Lies, and Alibis I am using Adobe Scanner for my iPhone which allows me to use photography to capture JPEGs from the book and create multipage PDFs that I can then send to my computer via email. We tried using Adobe to recognize text from PDF’s in our initial experiments, but it was found to be extremely inaccurate and unreliable with even the slightest mistake of the alignment especially with longer texts. Currently, I am using SnagIt and its grab text feature that allows me to drag my cursor over the text that I am want to be pulled for recognition. Six testimonies from the book were chosen for Dr. Nieves’ talk this upcoming week on the project, and I used these excerpts to try out the software. It only took about twenty minutes to use Adobe Scanner to turn the pages into PDFs as the only thing that was really required was to snap a few photos and crop them. The remainder of the work using SnagIt took me around three to four hours to complete because I had to go through the motions of learning the quirks of the software.
For those wanting to use it, my recommendation is that you make sure that whatever file you are pulling text from is as clear and straight as can be to avoid the software understanding (and spitting out at you) the text as hieroglyphics. I found that the ideal size to grab is around a paragraph length and zooming in to help the software determine words. Despite its general effectiveness, I just want it to be clear that corrections and edits will need to be made, and you will have to sometimes hand type out some sections if the document is not readable. Overall, my current process is sufficient for my work with the novel because I am able to remake the photographs if necessary maximizing the effectiveness of the SnagIt tool. This unfortunately is not the case for the documents procured by Dr. Nieves as the JPEGs were created from photos of documents that are in South Africa. We are currently exploring other software options to tackle this portion of the project, but I will be plenty busy finishing up my work on the novel for now.
My work on the Soweto Project originally started with the reading of the book Truth, Lies, and Alibis by Fred Bridgland. My focus was to gather testimonies that had anything to do with Stompie Moeketsi and his death at Winnie Mandela’s residence on 585 Eagle Street in the township of Soweto, South Africa on the night of December 31, 1988. These testimonies were to later be added to the already large number of JPEG files (mostly court case documents and newspaper clippings) taken by Dr. Nieves while in South Africa. Something that is important to understand is that with a project with this sort of content and magnitude, a sense of direction is a necessity to the progression of work. We have all this information, but without the big picture in mind, the work we had gathered so far was just words and pictures on paper. Our ideal image of what the project was to achieve relied heavily on the mysterious idea of spatial history, an idea that has no clear definition. Keeping all this in mind, the reading and marking of testimonies proved straightforward as pages could be marked as you went along, and were easy to return to later.
After much debate on the subject, it finally clicked for me that what brings a space alive to most readers is the people within it, so what better way to capture spatial history than with the life stories of those most impacted by it. What this meant for my own work on the project was that we wanted to lock our focus on the life of Stompie Moeketsi. His death is important to that life, but it has only been told through the lens of how it affected Winnie Mandela and the African National Congress. It was a fact that Stompie had been killed that night in the courtyard of Mrs. Mandela’s residence, but that one moment did not define his entire life. Gathering this story will require more than just words from extant texts because those who knew him have to date said little about his life. The discussion of gathering new and original testimonies from those closest to him is exciting, but could also potentially cause problems for those involved. The lack of information about the Mandela property itself is also an issue for our spatial project and gaining access to it is going to be difficult. There are also several technological challenges that are quite daunting, but that is a beast better tackled on its own. One thing is certain to me, I want to tell this story, and I am willing to do whatever it takes as a historian to make sure that it is told.
I now realize that I held an almost comically idealized vision of the people who created the metadata in the digital archives that were vital tools in my history education. As I combed through archives in search of photos, diaries, government documents, and newspaper clippings to help me make sense of my research for my senior thesis, my mind would occasionally drift to the erudite angel that made these assets available to students like me. This mythical archivist sat at a heavy wooden desk, wore a black turtleneck sweater, and gently pressed their gold-rimmed glasses up their nose as they dutifully typed archival metadata while jazz floated in the air around them – all for the benefit of over-caffeinated budding academics stooped over their laptops in the dark and who hissed like opossums when, invariably, vital files you swore to Beyoncé you downloaded are nowhere to be found.
When I leapt at the opportunity to be a part of Apartheid Heritage(s) and to work under Dr. Angel David Nieves, my initial confidence wobbled slightly when I dug into assignments to create and research metadata for such important and critical subject matters. I was a far-cry from the cerebral archivist of my imagination; rather than sitting in a richly appointed office, I sat cross-legged at my dining room table on Saturday evenings, listened to Amerigo Gazaway, and took breaks to wrestle with my dog while my partner caught me up on the latest UFC standings. Although Nieves clearly articulated that he wanted new perspectives and vouched for my abilities from the beginning, my doubts still lingered.
Quickly, I began to see that all the misgivings I had about myself were assets, and the idealized image of an archivist that I held in my mind began to crumble. As my work with Apartheid Heritage(s) progressed, I ran into ethical dilemmas that required collaborative solutions and new approaches to metadata that showed me that the future of the field did not lie with my imagined bespectacled graying archivist, but, dare I say, with people like me. Attempting to answer the moral dilemmas of metadata and digital archival work meant making decisions to intentionally challenge institutional thought when it is necessary and ethical to do so. Before I address some of the thornier issues I encountered, let me give a brief insight into the initial framework I worked within.
What is metadata?
Metadata is, on the surface, data about data. Its purpose is to describe data, such as a photo or postcard, in such a way that it is searchable and people can understand what it is, where it came from, and how or why it was created. On Instagram, for example, hashtags are a form of metadata. Imagine trying to find photos of adorable pitbulls without having the ability to search or follow #pibble. Without metadata, which in this example are in the form of hashtags, trying to find specific data becomes a daunting task. Unfortunately, metadata for digital archival purposes is much more complicated than the gram’s hashtag system.
There are specific rules and controlled language that partially dictate how and which words are used to describe the data in an archive. For Apartheid Heritage(s), Dublin Core (DC) provided the standards and prescribed language I would use to create metadata. The DC standards were chosen because, frankly, they were less complicated than other standards: they were straightforward, worked well with the Omeka platform that undergirds Apartheid Heritage(s), and they allowed a degree of flexibility that Nieves and I would use in attempts to deal with the moral dilemmas of metadata.
Again, Instagram proves to be a handy example because it uses certain standards and prescribed language in its application of metadata. Accounts are preceded by the at sign, @, and there can be no spaces in the name of the account: these are a form of metadata standardization and prescribed language. For example, dog behaviorist Cesar Milan’s account is @cesarsway, without any spaces or additional special characters, so it adheres to the standards and prescribed language of Instagram. By using standard language, Instagrams “metadata” is consistent and searchable.
Addressing ethical challenges
You may be thinking, “this is all well and good but where’s the problem? How can data about data pose moral or ethical challenges?” Well, I’m glad you asked. Metadata, in practice, can be very impersonal and dispassionate. This became evident when I looked back at the digital archives I haunted in school for some examples of how I, a brand new and somewhat self-taught metadatist, would approach this slice of academia. Generally the basic information for the item is listed, such as date, creator, location, etc, followed by a brief description without additional context to the item or the history surrounding it.
For Apartheid Heritage(s), we decided that it was not only inadequate to use metadata in such a sterile manner, but also unethical. How can item descriptions written absent of critical race theory or subject labels, which are prescribed and controlled language under DC standards, be chosen without addressing the white power structures that exist in archival language? We decided that, whenever possible, additional subjects and descriptive context would need to be provided in order to ethically archive the materials and begin the difficult business of decolonizing the archives.
Take, for example, this photo of a mine dump near Johannesburg. Rather than simply identifying the subject of the photo at its surface (Gold mines and mining-South Africa; Hardrock mines and mining- South Africa; Twentieth century; etc) additional labels were added in an attempt to highlight the power structures that led to and were codified during apartheid – in addition to the community erasure that continues to happen. By including the subjects Industrial wastes and Hazardous waste sites, as well as a statement in the description as to the location of these dumps (always near Black townships), and the ongoing hazards caused by mine waste sites for Black communities, we have worked within the controlled language to speak to the complicated legacies of these assets.
Another example can be found in this press photo of a riot that broke out in downtown Johannesburg in 1941, seven years before the rise of the National Party and the formal beginning of apartheid. The caption on the back of the photo says that “thousands of police, soldiers and civilians participated in a serious riot” – but why did that riot break out? With a little research I learned that this riot was a result of the Ossewabrandwag, an Afrikaner pro-Nazi and anti-British paramilitary organization, harassing and beating white South Africans who supported or participated in warfare against Germany during World War II. By including this valuable context in the subject headings (Fascism-South Africa; Right-wing extremists) and in the description, an attempt is being made to challenge the white-washing and sanitization of archives and archival language. Further, I attempted to lead the user to understand that apartheid, like so many other modes of oppression, have a long and sinister buildup before they are formally recognized. The “pre-apartheid” period, to which this photo belongs, is not a discrete “from the apartheid era of history” but, rather, is essential in understanding the power structure under which apartheid would be built.
Hurray! Archives are being decolonized! The white-washing of history is being undone.
Eh…it’s complicated. There are many challenges that I continue to grapple with as I work on metadata for Apartheid Heritage(s). While providing contextual descriptions and thought provoking subject headings is an attempt to decolonize archives, I am still bound to their very controlled language. This is an issue for two reasons- one, I can only use subject headings approved by the Library of Congress and two, the lists are massive.
Navigating and applying the lists of approved subject headings require a degree of creativity because they continue to reflect white power structures and colonization. There are many examples that I have discovered and I am sure that as my understanding deepens and my research continues, many more will be revealed. For example, the only references to Nelson Mandela in 8,273 pages of small print are about a namesake park in England. Latin lovers is an approved Library of Congress subject heading but there is nuh-thing on Mandela beyond a park in England, a country that colonized South Africa and Black South Africans.
This illustrates the colonizing viewpoint of subject headings that complicates how I use the controlled language to appropriately describe assets in Apartheid Heritage(s). I struggled to describe, for example, the type of oppressive labor system that was established prior to formal apartheid, and certainly a hallmark of it during that period. Black South Africans mine workers were paid in company scrip, a type of coin currency that could only be used in company controlled stores. This promoted continued capitalist oppression and allowed white South Africans to remain largely in control of the markets that Black South Africans used.
Describing this type of labor in DC and Library of Congress subject heading language became difficult because it is a complicated system. Black South African mine workers were not slaves, so that particular subject heading did not apply, but they were oppressed and manipulated for the enrichment of white business interests, which certainly smacks of slavery. And they were paid – “kind of.” So how do I use controlled vocabulary to describe a “not slavery but kind of slavery where workers were paid but not really” labor situation?
Another issue I continue to deal with is navigating the lines between providing context and editorializing and ensuring that a certain level of historical emotional neutrality does not tip towards either accidentally normalizing egregious legacies or becoming biased itself. An example of this is the life and legacy of Winnie Madikizela-Mandela, the former wife of Nelson Mandela, and a seemingly canonized woman nicknamed “The Mother of the Nation.” While the challenges she has presented me warrant a lengthy post of its own, I will touch on why she presents an example of the ethical questions I am still attempting to answer.
Beneath “Mama Winnie’s” acclaim as an ardent anti-apartheid warrior, exaltation for being one of the first Black social workers in South Africa, and celebrity as a fashion icon who used traditional xhosa clothing as protest – beneath these lie a legacy of vindictiveness, atrocious human rights violations, torture, and murder. My challenge with creating metadata about Winnie Madikizela-Mandela is in knowing how far to push my descriptions so that I both ethically address the crimes she committed and resist pushing a bias, agenda, or viewpoint. This is a struggle and, at times, I have had to “take a break” from Winnie metadata for a while in order to let myself reset and step back, so to speak, from the emotions reading about her crimes cause to bubble up in me.
This difficulty balancing necessary critical evaluation from an unbiased distance is not limited to my work on Winnie Madikizela-Mandela – it is something I grapple with for nearly every asset in Apartheid Heritage(s). When I meet with the Apartheid Heritage(s) team these questions frequently arise (and so, so many more):
What is the context?
How can I challenge colonization?
Am I being biased?
Am I editorializing?
Am I accidentally normalizing?
How can I make this as searchable as possible?
The many moral and ethical dilemmas and questions that remain in not just my work, but in the field as a whole, have entire academic journals dedicated to discussing and addressing critical librarianship and information studies. Continued conversation, discussion, and growth have allowed me to grow in my understanding of this field and subject matter and, as a result, the metadata is evolving and growing with me.
Occasionally my mind returns to the esoteric archivist and metadatist of my imagination. Like my attempt to do my part to decolonize the archives, it is time for me to challenge the gatekeepers of knowledge. At the conclusion of a riveting lecture about finding Aztlán in meso-American codices, Dr. Davíd Carrasco paused to address the students in the hall. After addressing the many “firsts” that were occuring at highly esteemed universities across the United States – first Latina president, first Black chair, and so on – he told students that, instead of being in awe or feeling intimidated by these higher institutions, they needed to say “what took them so long? Why didn’t this happen ages ago? They have been missing out!”.
It is this spirit that carries me forward in my personal growth and that is evident in many academic fields. Rather than being intimidated that I did not match the same image as my pre-conditioned notion of what an archivist should be, it should be taken into consideration and appreciated for what it is. For history, archives, and other fields to advance there is a critical need for people who do not resemble the stalwarts of the past.
My growth as a historian, metadatist, archivist, librarian…whatever label is most appropriateis contingent upon continued growth, collaboration, and conversation. If you have any comments, suggestions, or perspectives that you want to share with me that will help me be a better and more informed human, please reach out to me at email@example.com.
San Diego State University Associate Professor of History and Digital Humanities, Angel David Nieves, and his Co-PI, Elaine Sullivan (UC-Santa Cruz) were awarded a $100,000 grant from the Andrew W. Mellon Foundation and the National Historical Publications and Records Commission towards the creation of a publishing cooperative for 3D scholarship and digital scholarly editions. The project, “Scholarship in 3D: A Proposal for a Digital Edition Publishing Cooperative,” intends to develop the necessary shared knowledge base and infrastructure for the successful publication of scholarly 3D digital editions, and to create new pathways to publication for scholars working with 3D content.
The planned cooperative will create prototypes to digitally publish and access historical collections for four projects currently in development by participating faculty. Institutional partners on the grant include USC, UCLA, UMass-Amherst, UT-Austin, UVA, Claremont Colleges, Hamilton College, Maynooth University (Ireland), the Alliance for
Networking Visual Culture, and the American Historical Association. Publishers on the grant include Stanford University Press, the Cotsen Institute of Archaeology Press at UCLA, and the University of Georgia Press.
The project is significant as it joins together two areas of scholarly inquiry in digital humanities that seek to identify and address the long term challenges of digital preservation of historical resources and to provide access to a range of content types, especially in the modeling of 3D reconstructions. The proposed Cooperative first meets at UCLA, Feb. 22-24, to begin their work. Lisa Snyder (UCLA) will host the meeting and act as our Project Director and site coordinator.
Prof. Nieves’s project, Apartheid Heritages: A Spatial History of South Africa’s Townships, brings together 3D modeling, immersive technologies, and digital ethnography in the pursuit of documenting human rights violations in apartheid-era South Africa. Nieves, who has been on leave at Yale University this past year as a Presidential Visiting Associate Professor, returns to the SDSU campus this fall as an active member of the Area of Excellence (AoE) in Digital Humanities and Global Diversity. The work on this year-long planning phase, if successful, anticipates extending other grant opportunities to graduate and undergraduate students working in digital humanities in the Department of History.
Here is a first look at Apartheid Heritages‘ new bookmark design, featuring the site information, a project summary, our institutional partners, and a preview 3D model of one of the archived locations. (Graphic design by Greg Lord, @gplord // Featured model by Jack Hay, Hamilton College Class of 2019.)
I came to this project from a technical background. I am a 3D designer and a software engineer but my work on this project has led me from the safe confines of Computer Aided Design into the deep and troubling history of “native housing” on a large scale during apartheid. It is the design of these houses that gives the project clarity as these rudimentary and hastily wrought plans constantly remind one of their legacy.
My experience in architectural design software when I joined this project had been centered around upscale homes with the luxuries of sustainable design and other costly construction, but these were no shingle-style summer homes on the ocean. In Soweto the floors were often packed earth and the furnishings simple and industrial. Open space was provided for in neighborhood plans to accommodate the military for crowd control.
My work on this project began with the process of modeling these simple township houses using ARCHICAD. ARCHICAD is an industry-grade architectural software that I had experienced while working for an architect close to home. I have also had experience with a number of CAD tools in the past. After speaking with Professor Nieves about the project I decided that the ARCHICAD toolset and out-of-the-box photo-rendering options would be a good infrastructure to model with.
As I completed each model, I would add more and more detail; bringing complexity and specificity to the industrial building materials. I used complex profiles to model components in a 2D view before extruding them into the 3rd plane. I also took advantage of the rendering engine to generate images that added realism to the models for presentation.
The destination and use of these models is still being defined. My goals for work over the summer include creating a larger 3D model to present the individual models in a realistic setting and configuration, an ARCHICAD-centered video tutorial series, and an online 3D library. Meanwhile, my daytime job this summer consists of work in cloud infrastructure and software development at EBSCO Information Services where I hope to bring some level of experience to the process of building a database from the ground up.
If you are interested in the 3D modeling side of the project, I urge you to watch my DHi intern’s presentation (~30 mins) which covers the breadth and depth of 3D work on the project. I also hope to release a video series that covers the tools and techniques that I use which will be available on the DHi website. This will cover the technical elements of ARCHICAD and the process of drafting from planning to completion. I hope to document my experiences with the larger visualization and the library as well.
This post arrives early in the evolution of this project. Drafting a 3D model in powerful architecture software is the easy part.
Throughout 2014, Gregory Lord (Hamilton/DHi Lead Designer and Software Engineer) and Kevin Xiao ’15 (Hamilton/DHi Student Intern) began to design and implement the early version of the Soweto Virtual Environment. This work doubled as both an active DHi development project, as part of the Dangerous Embodiments grant (funded by the NEH, partnered with the University of Arkansas, and later University of Mass., Amherst), and as a student training and development project, training Xiao in the use of 3D modeling and game design tools.
Xiao participated in the project as a 3D modeler and scene designer for the Soweto virtual environment, learning the fundamentals of 3D modeling over the course of his participation in the project, studying under Lord.
To develop this model, researcher and project director Angel David Nieves, Ph.D., provided the development team with blueprints for a variety of houses that comprise the represented section of the Soweto township, along with the Orlando Methodist Church.Using this data, Xiao and Lord designed 3D models that used these blueprints, making sure to capture both the layout and dimensions of each building.From there, the team used a combination of both historical and modern photographs and images to texture the houses, bringing them closer to a photorealistic appearance.
Finally, the team imported each of these models into the Unity engine, where they combined the models with tools that allowed them to create an accurate street map from Open Street Maps data.By putting all of these together, and reconciling the building locations with maps of the region, the team recreated an accurate representation of a particular block of the Soweto township, serving as proof of concept for what will later be the full township.
We’re proud to offer a first glimpse at the upcoming tool for the Soweto Historical GIS Project — a 3D mapping and historical recreation platform built on the versatile Unity 3D engine:
The Social Justice History Platform is a software platform designed to represent geographic and spatial data within an enhanced interface designed to contextualize locations and objects alongside the primary source documents that provide their historical narrative, and a range of related multimedia objects (including video, audio, images, and text).