Freedom on the Move and Ethical Challenges in the Digital History of Slavery

Building the Team: Centering Collaboration

Historians have used runaway advertisements before to demonstrate enslaved people’s persistent resistance, enslavers’ commitment to slavery, and the construction of race as a social category over time. Mostly, these historians built their own databases and research files by scouring newspaper collections from various archives and publishing collections of advertisements. Some of these collections have been digitized and some have not, and when we began FOTM, we had a vision for a comprehensive open-source archive that would be available and legible to researchers, genealogists, educators, students, and multiple publics alike. As Crystal Eddins notes, DH projects have the potential to engage with the long Black Radical Traditions of marronnage. Among the potential maroon characteristics that Eddins posits are the “creation of solidarity networks of people who share social position and liberatory goals.”1 From the start, we prioritized collaboration with programmers and the many constituencies we imagined would find the project useful. We designed the project to incorporate a crowdsourcing model to code and organize data points, asking users to help build the archive and find new ways to use it. As Mia Ridge has observed, at its best, crowdsourcing can connect “people, culture, history, and collections while providing the public with platforms for enjoyable, meaningful activity” (Ridge, 435).

But through team meetings, alpha and beta testing groups, and workshops with K–12 educators, we came to understand that the diverse publics we imagined using FOTM would not merely contribute time and labor to the project. We were not adding their response to our work in our revisions of the site. They would work alongside us as collaborators in building the archive and designing it, a welcome realization that nonetheless brought to the fore a set of knotty concerns (see Bailey; Thomas). Perhaps the most generative quandary has been the ethical issue of reifying the violence of reducing self-emancipated people to data to be “extracted” and “analyzed” from sources whose creators were intent on making them slaves when they so clearly stated with their feet that they were indeed free people. This is where computational humanities and analytics play an important role, but only alongside the digital humanities more broadly. Interrogating where, for example, linked open data might be useful, as in a project like Enslaved.org, alongside where and whether that approach does or does not repeat the commodification of enslaved lives is a key discussion point at FOTM. For scholars like Daryle Williams, defining core sets of metadata allows scholars to combine large datasets (of enslaved people’s biographical information) that may otherwise vary widely by empire, archive collection practices, and research preferences. For FOTM, the issues of collaboration and public engagement play as much a role as the metadata and formulas for searching the ads and has shaped our project’s inquiries.

The process of engaging and collaborating with scholars, students, and the general public in viewing, building, and designing the archive implicates them and compounds that ethical conundrum. The priorities and interests of the historians, programmers, and university grant and contract agents are not always in alignment, and efforts to foreground the experiences of the enslaved are inevitably compromised by orientations toward outcomes and resources. At FOTM, we are conscious of these and other dilemmas, many of which are endemic to the digital history of slavery and to Black digital humanities projects more broadly. We have attempted to build into the project itself ways to address these concerns. Even as we recognize that they are often ultimately irresolvable, we would argue that their potential damage and harm can be acknowledged, reduced, and made into productive, important, and integral components of FOTM and its possibilities. We believe our experience points toward an instructive model for computational humanities projects grappling with similarly challenging materials and variegated stakeholders.

Archives of Fugitivity: Fragments of Self-Emancipated Lives

For contemporary users, what is it like to transcribe data categories easily gleaned from each ad? How should the programmer deal with each data point? A dropdown menu? A blank to fill in? What do those choices mean at a computational level? What about at the interpretive level for users? What about users who encounter the long list of racial categories across multiple languages that appear in the advertisements available for transcription? What if they are deeply offended by the racist terms used for enslaved people in the past? The challenge most intrinsic to FOTM is also perhaps the most inextricable from it, which is that by amassing evidence of fugitivity, breaking down that evidence into discrete and detailed categories, and enabling it to be collated for examination, study, and research, we risk replicating some of the imperatives and hierarchies of slavery itself.

One cogent answer comes in the form of a found poem: “I/loved/her/but/that/white man/took/her/from/me.” From a photocopy of an advertisement found in the FOTM database, these words peek out from between redactions made with black permanent marker to conceal most of the words on the page. Concealed sections, chosen by an eighth-grade student, serve as a mimetic for archival silence. Those left visible attest to the way that even fraught sources, produced by enslavers, often reveal the pervasive resistance of the Black people they were so intent on enslaving. There, among words meant to transform a person back into chattel, to deny the self-emancipation evident in one person’s flight from bondage, a contemporary middle-school student left legible only the part that reveals one definition of freedom: love and intimacy. The collaborative mission of the project is explicitly to build with students like the poem’s author, not to anticipate their needs and teach them what they need to know. Time and again, the range of collaborative voices on the project have revealed things that did not catch the attention of professional historians.

When most users begin working with FOTM, they select whether they prefer to “contribute” to or “search” the database (Figure 11.1). Selecting “contribute” gives users the options to “extract data,” to “group advertisements,” or to “moderate” (Figure 11.2). The first option is the most involved. Choosing it takes users to a page of advertisements and asks them to pick one, transcribe it from a pdf image into text, answer a lengthy series of questions that disaggregate the text into several dozen metadata categories, submit their work into the system, and then, if they wish, begin anew on another advertisement. Users who select “group advertisements” are asked to examine a series of advertisements, determine whether any of them were placed repeatedly over time, and link them together as a group of duplicates. Users who select “moderate” are asked to proofread and check the accuracy of contributions already made by other users and thus to provide an element of editing and quality control for the database. Users who select at the outset to “search” the database, meanwhile, can look for keywords or pull together one or more categories of metadata from every advertisement entered by previous users.

Not every “runaway advertisement” ever placed is still extant, as not every American newspaper from before the Civil War is still extant, and so even at its most complete, the project will only ever be able to claim comprehensiveness of the surviving evidence base. Nevertheless, the number of advertisements that has survived is so tremendously large that we do not yet have a reliable way to make a precise estimate of it. Indeed, these advertisements are, by far, the most voluminous and systematic evidence base ever created for the phenomenon of fugitivity from slavery in the United States. Moreover, they are usually the only surviving written evidence documenting the lives and efforts of tens of thousands of individual enslaved people to achieve their own liberation.

Retrieving, documenting, and sharing that evidence can advance a deep and authoritative understanding of the phenomenon. It offers the opportunity to dignify those lives and efforts, and the harnessing of computational power to a digital archive of the evidence lets us see things at both fine-grained and holistic levels that might otherwise be impossible. Organizing and investigating the metadata extracted from the FOTM database can tell us, for example, about the ebbs and flows of flight from slavery by geographic location and time of year, about the linguistic skills and the presumed literacy of fugitives, and about the items they chose to bring with them when they attempted their escapes. We can see in the particular and the aggregate alike the ages of those who fled from slavery and how often they fled alone, with family members, or with one or more other enslaved people. We can even map the likely paths and directions they took while trying to elude their enslavers. Information about these and dozens of other aspects of fugitivity, ranging across thousands and thousands of instances of flight, would traditionally require immensely time-consuming and painstaking work, but they can be revealed in seconds by the FOTM engine. Moreover, and importantly, being able to parse the material from so many different angles and with the input of so many voices and collaborators helps counter, at least in some measure, the problematic gaze of the enslavers who created it.

An Ethic of Care: Participant Users and the Fraught Nature of Slavery’s Archive

Yet there is no denying that runaway advertisements still reflect the voices, priorities, and concerns of the enslavers and jailers who placed them, and that users inevitably “speak” in those voices when they transcribe them. There is no denying that asking users to enter information about the ages, physical characteristics, racial categories, skills, possessions, and rewards offered for fugitives can be a discomfiting request. Seeking out and entering such information so that it might be sorted and arranged, after all, arguably asks users to recreate the inspections and quantifications of Black people and their worth, which, as scholars such as Walter Johnson and Daina Ramey Berry have demonstrated, made their capital, reproductive, and labor values assessable under slavery.

The inescapable moral dualism of creating and analyzing the data in FOTM extends to users who perform searches as well. Currently, for example, a researcher who performs a keyword search of “lame” will retrieve 899 advertisements. Among them are accounts of Sampson, “an African” and “a Shoemaker by trade” who “walks lame in consequence of the loss of his toes by frost”; Albert, who was “lame from a defect in his hip”; and George Guy, who was “lame of one leg, and had on when he left an iron collar around his neck” (Charleston Courier). Such information can tell us about the conditions under which enslaved people worked, the punishments inflicted on them, the lengths to which they would go to escape their bondage, and the intersecting histories of slavery and disability. Retrieving the information also conjures up the violence of slavery and forced labor as a matter of course. As Jessica Marie Johnson has argued, the “devastations” of the archive of Atlantic world slavery easily “reproduce themselves in digital architecture” (58).

The problems here are unsettling, and they can seem especially so given that many if not most users of the project are likely uninitiated to the problems when they begin working on FOTM. Some users of FOTM are researchers affiliated with the project, the directors themselves, or other scholars. But the project website currently indicates that more than 11,000 contributors have transcribed advertisements, extracted data from them, or moderated content. The vast majority are high school and college students using it as part of a classroom assignment, members of the general public who have read or heard about the project and want to help, or genealogists who come to the site curious about what they might find and decide to do some transcription or metadata work during their visit. Some among these various kinds of users may understand that their activities on the FOTM site evoke some of what sustained the enslavement of the very people described on their screens. But it seems probable that the numbers who do or who have thought about the issue are relatively few.

Regardless of whether or not users are struck by the implications of their engagement with FOTM, the dilemmas of those implications are, in some ways, insoluble, and we do not attempt the fruitless task of resolving them. Rather, we confront them. We try to use them as an opportunity to encourage users and researchers alike to recognize how power is perpetuated through the historical archive, to see how engaging the archive through FOTM can be a small but not insignificant way to oppose that power, and to be purposeful and careful about how they use the data derived from it. As both Benjamin Schmidt and Crystal Nicole Eddins note in this volume, DH projects allow for a unique opportunity to engage sources and participants in this way.2 As Schmidt writes, one of the things made possible is having participants “change the assumptions of models” and work “with sources in the indefinitely reconfigurable ways that are now possible.” While FOTM, for example, presents itself as a database of fugitives from American slavery, visitors to the splash page of the project see it foregrounded as an endeavor in rediscovering the stories of self-liberating people. They can watch a short promo video that juxtaposes a voice-over reading of a runaway advertisement taken out by Thomas Jefferson with visuals centered particularly on the struggles and the flights of enslaved people from bondage. The FOTM website reminds users that even as the advertisements in the database were “created to control the movements of enslaved people,” they “ultimately preserved the details of individual lives” that enslavers could never completely control and that the voices of those who lived them are discoverable when advertisements are read critically and against the grain. And we remind users that working with the project is a responsibility. They drive the research and its outcomes.

Yet the act of transcription can result in users reinscribing the power relations that the documents evidence. Nearly any digital project that asks users to transcribe materials whose very existence reflects the selective and heavily mediated nature of the historical archive deals with this predicament. In the particular case of FOTM, the predicament means reinscribing the power relations and hierarchies of slavery. Users do not annotate each document. Instead, they extract metadata and make each document searchable by a list of terms set by the project directors. Users do not edit racist descriptors that advertisers used to describe the self-emancipated people who enslavers hoped to thwart and recapture. Users do not challenge each enslaver’s claim to the power to re-enslave the self-emancipated. FOTM thus simultaneously risks the realities of transcription never resonating with users and also alienating users who either come to consciousness about this duality while participating or who encounter the project and question how participation incriminates them in the discursive violence of the documents and physical violence they record.

Akin to the printers who featured these advertisements in their periodicals, users of the project complete the labor of making an enslaver’s or a jailer’s desires legible to a wider public. Trained scholars, researchers, and archivists are often aware of this reality: each ad represents an individual who fled enslavement, but the representation captured by each ad is the product of an enslaver’s views, desires, and hopes to make each self-emancipated person into a slave. Just because an ad includes a person’s height, an age, or a name does not mean that the information is accurate. Nor does an ad’s description represent how a self-emancipated person would have described themselves, the name their kin called them, or their true final destination. As Sharon Block and Daina Ramey Berry have shown, in both the eighteenth and nineteenth centuries, whites often estimated or approximated the physical characteristics of African descended people. Social markers like skin color, manner of dress, and even age were mediated categories that shifted by region, time period, and imperial jurisdiction. Using the ads as historical sources requires a sophisticated understanding of what, exactly, they do and do not leave a record of for those interested in consulting them.

But educators, students, genealogists, and laypeople unfamiliar with the extant archival materials of the eighteenth and nineteenth centuries may not be aware that each ad does not contain an objective description of its African American subject. Enslavers did, after all, construct each ad as an authoritative statement of their own power to make slaves of full human beings, to define the character and desires of Black people, and to entice readers to join them in enslaving people who had absconded. In fact, the privileged access enslavers had to the very medium of newspapers, and the right to dictate the information circulating within them, was a mechanism for monitoring and controlling the enslaved. The question for our team as designers of the crowdsourcing platform became how to impart this kind of knowledge and these sorts of understandings and lessons to users from diverse backgrounds with varying levels of experience with the archive of Atlantic slavery. We had to determine how to teach users to read, transcribe, code, and learn from each ad without falling into the trap that enslavers from hundreds of years ago so carefully laid for readers. The issue at hand here is not unlike the broader challenges faced by researchers who want to encourage the work and enthusiasm of “citizen scientists” with widely disparate backgrounds, knowledge, and training. We wanted to do what we could to have “users” of FOTM actually be “participants” or “builders” of FOTM. We wanted them not to be passive absorbers of the material they engaged with or simply processors of information; rather, we wanted them to be “citizen scholars” who would bring a critical eye to bear on their work and who would have a deeper appreciation for its ramifications. When building a dataset through crowdsourcing, those involved in the computational humanities must be attuned to how the process of data creation can end up relaying troublesome elements of the data to users. FOTM presents one possible model for addressing the issue.

Collaboration in Community: Participant Users and the Labor of Archive Building

We built the crowdsourcing platform of FOTM to foreground the collaborative nature of the project, and we ask users solely to transcribe ads as they appear. We chose to carefully embed moments of reflection right into the process of contributing to it. We remind users in each part of the transcription process that they are working with a form that we, the FOTM team, built. Furthermore, we explain that we built the platform to carefully code a set of historical documents that are themselves constructions of enslavers’ making. We worked with programmers to construct both the transcription platform and the forms for coding information that users harvest from the ads so that they serve multiple purposes: to capture information, to communicate the fraught nature of each ad as a historical source, and to train users to think like historians.

To meet these goals, the team considered not only what we would ask users to do but how we would ask them to complete each task. Importantly, while these deliberations were driven initially by the FOTM team, they evolved such that the deliberations themselves became a collaboration between the team and project users. As users began engaging with the project, for example, we solicited feedback and listened to questions about the information that appears in advertisements, about why the project aimed to collect certain kinds of data, and about the uses to which such data might be put. We realized that we had failed to explain sufficiently both the problematic nature of the advertisements and the research and other purposes for which the data in them might be significant.

In response, we made a series of adjustments. We developed, for instance, a glossary of terms to guide participants unfamiliar with eighteenth- and nineteenth-century terms, particularly for race and appearance. The glossary allows us to reiterate that terms used in the documents are archaic, racist, and carry historically contingent meanings, and that cataloging them can further efforts to undermine their power even as it risks reifying their meanings. We built a tutorial to guide participants through the transcription process and to signal more clearly, as users learned the form and the process, what working with the ads truly means—namely, they would be handling a weapon that enslavers employed to hunt down and enslave Black people. The tutorial teaches users what the documents are as it teaches them how to transcribe and code them in the archive.

We designed a series of forms that lead users through the process of coding information in each document. We start with transcription as a way to reiterate that users are working with sources and not reading accurate life histories. We then ask for information on the advertiser(s), for details that the advertiser gives about the runaway event, information the ad contains about individual runaways, and finally, information about their enslaver. We begin with the source and then guide users through how each advertiser constructed the source. This signals, at each click to a new form, that the ad is a carefully constructed document and one specific action used by enslavers to enslave.

Careful choices of phrasing foreground the unstable nature of each document. We ask users, “What is the primary name given for this runaway?” rather than simply labeling the response box with the words “full name” or “runaway’s name.” This question reminds users that they are reading an ad placed by an enslaver or jailer, that the ad’s author and not the self-emancipated person supplied the name listed, and leaves open the question of what each runaway may have called themselves. We provide a question mark icon that, when hovered over or clicked, indicates that enslavers sometimes gave enslaved people names that they themselves did not use or prefer. We open the possibility that the ad offers one story, not the story of each runaway event. We ask users over and over again, in each form, what the advertisement constructed by enslavers and jailers says and not “what happened.”3

The form we collaborated on has a pedagogy that also ultimately involves the users in building a database with searchable categories. We do not just invite users to supply data or to transcribe uncritically. We challenge them, through the form we constructed, to consider what each ad can and cannot tell us. We also ask them, both informally and through periodic focus groups, to tell us how we can make the aims of the project clearer and more conscientious about its procedures and aims.

We hope and believe this mutually collaborative process has improved the project itself and made FOTM more responsive to and inclusive of its multiple audiences. We know it has made the directors of the project more acutely aware of the impact of every one of our choices. We understand and try to consider how each of our decisions has implications for FOTM, its meaning, and the arguments it makes and will potentially make. That the interface (Figure 11.3) presents an image of the original advertisement at all times for the user, for example, rather than disappearing once a transcription is made, serves as a subtle but important reminder of the historical tangibility of an enslaved person’s experience and of an enslaver’s aim to thwart their escape. Along the same lines, the logo we selected for FOTM includes both the male and female printer’s icons of fugitives, reveals the text of runaway advertisements through those icons as if they were transparencies, and arranges the title of the project so the words “freedom” and “move” stand out prominently. All were conscious choices that set fugitives themselves at the forefront and highlight their decisions and goals rather than those of their enslavers, while simultaneously alluding to the source material at the heart of the project and gesturing toward how that source material haunted the enslaved at every turn.

Considerations of how we ask users to engage with challenging content in the FOTM database, and what those engagements mean for thinking about the history of slavery and for the subjective experiences of the users themselves, are issues whose relevance is linked mostly to the front end of the project. User feedback and input are important for refining the project infrastructure, but users are not privy to most of the process of actually creating it. Nor are they exposed to plans and conversations that have been ongoing for years and continue to play out among project personnel about everything from grant proposals to building pathways through the user interface to which fonts to use in the website design. Yet those plans and conversations, those decisions and choices made behind the scenes, are the bedrock on which we ask users to construct a landscape. And fault lines lie in that bedrock because it rests on the expertise of individuals with wide-ranging skills and training and varying senses of the goals and significance of FOTM. Computational humanities projects of almost any significant size and scale will rely on extensive and cross-disciplinary personnel, which in turn demands an understanding of what every member of a team brings to a project and a purposeful consideration of how the various strengths of those members can work in tandem instead of at cross-purposes.

Toward Equitable Access and Collaboration

FOTM grew out of the research questions of historians, and the project was in many ways not especially original or novel in its conception. Historians have been gathering “runaway advertisements” for decades. Some compiled them simply as text that constituted “databases” before the internet existed, and they published them in book form that was then the most effective way to make the advertisements available to an audience mostly of scholars and genealogists. Technological advances led some in ensuing generations of historians to put sets of advertisements from particular eras, states, or regions online and to make them accessible and searchable in a fashion that presages FOTM. Many of those projects still exist and are ongoing, and they serve in some ways as valuable models for what we imagined FOTM could be. If what sets FOTM apart is its effort at comprehensiveness and its invitation for input from all sorts and any number of users, it most certainly draws on precursors with similar if usually more limited goals. It incorporates in some cases the data of those precursors, and it is part of a family of digital humanities projects focused on fugitivity (Windley).4 FOTM is also part of a movement of projects within the broad heading of slavery and data that are deeply invested in doing more than calculating figures from slavery’s archive and in fact demonstrate that the binaries and boundaries between digital humanities and computational humanities dissolve in the face of marginalized historical subjects.

Accomplishing comprehensiveness, allowing broad public access, and encouraging multiple audiences to contribute to the project, however, has also entailed and required participation from people with skill sets historians typically lack. Mostly working together with the Cornell Institute for Social and Economic Research (CISER), we have and continue to rely on the expertise of programmers and data specialists who understand how to translate information from the printed page into digital spaces, construct functional and logical data models and workflows, build an elegant programming interface, and manage and administer a database whose crowdsourced content changes and expands all the time. We draw on staff efforts in offices of sponsored programs at multiple universities to identify and pursue grant opportunities to fund the research and development that make FOTM possible. And we work closely with data services and grant managers who are vital bridges between historians and programmers. Crucial personnel who think about the long term and the big picture, they make sure that everyone involved in the project understands their roles at given points in time, that projected benchmarks are met, that plans for maintenance and sustainability often unnecessary in the traditional humanities but essential to the digital humanities are in place and carried out, and that progress is being made on grant applications and grants-in-effect alike. We consider every person in these capacities to be members of the FOTM “team” no less than the project directors, and we indicate as such on the project website to ensure that labor is equitably credited and acknowledged.

Somewhat remarkably, we have been fortunate that the relationship between the historians directing FOTM and the funders of the project has not been nearly as knotty as it sometimes can be. The grant agencies we have worked with appear to share our goals and understand why FOTM matters. They have asked only that the project carry out the plans detailed in grant proposals in a timely fashion, and to date, they have put almost no restrictions on our ability to reallocate portions of the project’s budgets as priorities and needs shift. Still, the varied personnel indispensable to FOTM bring varied ambitions and objectives to bear. Programmers and data specialists come at the project and the advertisements it comprises as an interesting “problem” to be solved and as a set of technological hurdles to clear. Staff in grant offices engage the project in terms of bringing funds to their respective universities, conducting those funds through appropriate channels, and ensuring that the bureaucratic particulars of applications and budgets are carried out effectively. Data services and grant managers, meanwhile, must superintend everything all at once and keep everyone on task. It is no criticism of any of them to observe that their respective responsibilities and talents mean that enslaved people at the heart of the project can fall out of focus. It is understandable that histories of slavery and fugitivity can drift from the center amid the very practical considerations of continuing to pay for FOTM and make it function well for users. Such practical considerations inevitably pull us as historians into such results-oriented calculations, and there is no avoiding those calculations altogether.

Take, for example, the lengthy discussion we had with all the FOTM personnel about how to gather metadata on the “race” of fugitives and how to ask users to help us do it. At the outset, most of the project programmers and other nonhistorians considered the matter an obvious one. Enslavers might describe fugitives in advertisements as, for example, “black,” “negro,” “mulatto,” or “yellow,” and collecting data was as simple as having users click a category or fill a text box with whatever description was provided. The historians on the project, however, observed that the physical descriptions of fugitives captured “color” as often as they did “race.” They noted that categories appearing in advertisements included terms such as “griffe,” “mustee,” or “octoroon” that are not only racist but are also sometimes archaic and almost certainly unfamiliar to users. They pointed out that descriptors did not necessarily capture anything objective or correlative to the appearance of fugitives other than what their enslavers saw and that asking users to pull the thread of racial descriptors from advertisements unproblematically was to reify the artificiality of race in any number of ways. In short, programmers saw the question of race mostly as a matter of ensuring that users could capture and indicate every category and descriptor, while historians saw the categories and descriptors themselves as impossibly unsteady and fictive.

Ultimately, we settled on something like a compromise in which users do indicate racial descriptors that appear in the advertisements while also having their attention drawn to the constructed and evolving natures of race and racism. Users have access to a pulldown menu of descriptors and can include new ones in a text box, but above that box they are asked to tell us “what racial category is assigned” to a fugitive in an advertisement rather than “what race is” the person. Hovering over an informational link next to the question, meanwhile, informs users that categories of “color” and “race” were inventions enslavers deployed to keep the enslaved in bondage, and that although many of the categories are offensive, collecting them allows us to “map changes in how racism shifted its terms across space and time.”5

Ours is not a perfect solution. But it satisfied, at least provisionally, the aims of the programmers and the historians alike, both of whom learned from the discussion a great deal about the mindsets, needs, and concerns of the other. Moreover, this quandary, and others like it, has led all of us to grapple more deeply with the significance and import of FOTM. Eddins notes that one characteristic of marronnage that can be present in DH projects is the “disruption of capital accumulation processes that extract resources from Black spaces.”6 The historians have come to think harder about the compromises required when turning the experiences of enslaved people into data for the sake of better understanding those experiences. The nonhistorians have developed new sensitivities and appreciation for a painful yet liberatory past and the complexities of translating that past into the present. Many of them have said that being asked to think in these new ways invigorates a sense of mission unlike projects they have worked on before, even as most of our programmers have worked with or for universities and university faculty for years. While the capital here is in the form of the power to shape the archive, this social capital typically is not distributed to participants. FOTM is stronger for users at the front end thanks to the dialogue they may never see at the back end. Many projects in the computational humanities could surely benefit from having the kinds of dialogues described here, which both produce greater clarity about a project’s stakes and a more effective presentation for users and audiences.7

Publics and Partnerships

As we now aim to provide greater access to a greater range of publics through a museum kiosk that can be duplicated in multiple institutions, K–12 lesson plans and resources for educators, and expanded access for laypeople, we continue to grapple with the totality of what we are asking of users. Engaging users as collaborators and partners in the project of archiving this collection of sources means recognizing that they provide more to the project than volunteer labor. Especially as the project becomes a part of how classroom teachers deliver content about the period of Atlantic slavery, we are asking users to take on an interpretive mindset even as they transcribe advertisements. We are asking them to learn to think like historians, particularly about what those advertisements can and cannot tell us about the enslaved people and acts of resistance about which the ads purport to present accurate information.

We view this as an opportunity for a frank reckoning with the many realities of preserving, studying, and teaching America’s past and the important dilemmas of which educators, students, and laypeople should be more, not less, aware. The history of slavery, enslavers, and the enslaved is not a history that wraps neatly into enjoyable activities and lessons. The “history is fun” or “people in the past, they’re just like us” framing, which accompanies some attempts to engage people in learning about the past, does not fit. Typical “living history” models for other topics are very fraught with multiple histories coming to bear for public historians engaged in interpreting sites of Black history. From the National Museum of African American History and Culture in Washington, D.C., to sites such as the Whitney Plantation in southern Louisiana, there are certainly many examples of public history locations that grapple with slavery and its legacies seriously. Too often, however, we still see mock slave auctions, auction posters as creative projects, and “history role-play” lessons that are perhaps the best of the worst of trying to “funnify” or “gamify” a history that should be engaged with care. Field trips to the farms and plantations where enslavers treated people of African descent as property should have a decidedly different tone for students than a field trip to an amusement park. Instead of providing the many publics we are trying to reach with the tools to study and learn the history of American slavery by couching the history as “fun,” we ask participants to view FOTM as their own but also as everyone’s collective effort to amplify enslaved people’s resistance and the violent reality of American slavery.

Emphasizing partnership and collectivity is particularly important because it is easy to slip into the language of commerce when speaking about crowdsourced DH projects. Proposals for funding are rife with language like “stakeholders,” “sense of ownership,” “investment,” and “returns.” In our own conception of the project, we moved away from asking users to “take ownership,” a particularly unsuitable term given that each advertisement in the collection is a part of enslavers’ attempts to turn a person into property. Instead, we want users to join us in collaboration, to share the archive they helped to build, and to walk away with a new language for why the project is important. The ads, after all, draw each user’s attention to both the active project of enslavement and Black people’s pervasive culture of resistance.

Our system is evolving and imperfect. Our translation of historians’ understanding of historical categories sometimes falls short. For example, the historians, programmers, and other team members engaged in a lengthy conversation about how FOTM would code for the sex and gender of historical actors. But the result still does not quite capture the nuance of gender and gender expression for which we had hoped. Now, our form asks, “Based on the ad, do you think the enslaver was male or female?” and then the form provides the categories “male,” “female,” “other,” and “not provided.” We signal that we are asking the user to read the ad and supply information from the ad, rather than provide accurate information about the biological sex of a particular person. Again, we encourage users to question the way that enslavers read Black people’s bodies.

We acknowledge that the ads are not always complete or clear pictures of the individuals they concern and that, in fact, biological sex is not binary. But the way that we have worded our questions and presented users with possibilities still reinscribes binary understandings of sex as either “male” or “female.” “Other” and “unknown” are exceptions and implicitly less valid than the first two choices. We do not account for the many gender identities possible for self-emancipated people and instead ground our question in biological sex. While perhaps asking about biological sex is a less anachronistic way to discuss how the historical people in question may or may not have thought of themselves, our question closes rather than opens possibilities for users.

If an advertiser misgenders a self-emancipated person after capturing them and insists that they are male, for example, or if that advertiser uses only male pronouns and insists on a name typically given to men in the period, a user may just go with the advertiser’s word. After all, we ask, “Based on the ad . . .” We also close off the diverse gender expression possible for fugitives from slavery that printers might have signaled with their use of printing plates known as stereotypes. Stereotypes came in sets and provided printers with stock images that they could use to draw readers’ eyes and to signal what content each section of a periodical contained. More than once in the archive, a newspaper uses a female stereotype absconding with a bindle to accompany an ad that describes a male enslaved person. Was it a printer’s mistake? A missing stereotype from a set? Or did this combination signal something else to eighteenth- and nineteenth-century readers? We may not know exactly, but FOTM is currently not asking for this metadata from users. This issue requires revision and careful consideration. Our team is committed to both.

Crowdsourcing an archive of ads that were originally placed by enslavers for the purpose of re-enslaving those who took to their feet and defied enslavement remains a challenge and a messy process. One way that our team has committed to an ethic of care in this work that combines pedagogy, service to various communities, and knowledge production is to start with the assumption that revision to both our plans and our project can and should be ongoing. While we can embed efforts at harm reduction and acknowledgment, we cannot as a team of historians, programmers, and project managers completely avoid implicating ourselves and project users in the violence that our sources evidence. This is part of the work: fully acknowledging the harrowing reality and legacies of American slavery involves, on some level, the acknowledgment that no one can escape its echoes. But as Catherine D’Ignazio and Lauren Klein observe in Data Feminism, “counting and measuring do not always have to be tools of oppression. We can also use them to hold power accountable, to reclaim overlooked histories, and to build collectivity and solidarity” (123). While the shadows of the historical subjects who appear in these extant advertisements may have been cast by slavery, those subjects, numbering in the hundreds of thousands, have also left us evidence of their persistent quest to define themselves through resistance and defiance.

Willis Holmes advertised in the New Orleans Daily Picayune at least thirty-one times over the course of more than two months in 1844 in his efforts to recapture Nancy. His advertisement remained the same but for the fact that near the end of March, he increased the reward he offered from $10 to $25. He seems to have finally stopped advertising about two weeks later. Whether or not that meant his elevated reward led to the results he wanted or that he abandoned any hope of taking Nancy back into slavery is unknown, and we do not ultimately know Nancy’s fate. Like the broader archive of slavery itself, the archive we have created in FOTM leaves as many questions as it provides answers. As we continue to build it, alongside our team of partners and our growing set of users, we hope it shows an ethical and responsible model for other computational humanities projects that deal with difficult materials and that it remains part of larger conversations about them. If we have learned a great deal so far, we know we still have a great deal to learn.

Notes

1. 1. See chapter 7 in this volume.

2. 2. For Schmidt and Eddins, respectively, see chapters 7 and 9 in this volume.

3. 3. To date, we have not gathered data on how often users turn to the glossary, the question marks embedded in the interface, or other explanatory materials as they work on the project. Such data would be potentially illuminating in any number of ways and could help further improve and refine FOTM as a resource. But when users come to the project, they are asked only to provide an email address to create an account, and they provide no other identifying information. This is in part because we know that many users of the project are middle and high school students, and ethical (and sometimes legal) issues might arise were we proactively to collect data about how particular users engage with FOTM.

4. 4. Lathan A. Windley published what are probably the best-known series of printed volumes of runaway advertisements, a four-volume set of eighteenth-century ads entitled Runaway Slave Advertisements. Other print volumes include Graham Hodges and Alan Brown’s “Pretends to Be Free,” and Billy Smith and Richard Wojtowicz’s Blacks Who Stole Themselves. Windley’s collections, supplemented with other digitized materials, appear online at “The Geography of Slavery in Virginia,” based at the University of Virginia (http://www2.vcdh.virginia.edu/gos/). Other online advertisement collections include “North Carolina Runaway Slave Advertisements, 1750–1865,” based at the University of North Carolina-Greensboro (http://libcdm1.uncg.edu/cdm/landingpage/collection/RAS), and the “Texas Runaway Slave Project” at Stephen F. Austin State University (https://digital.sfasu.edu/digital/collection/RSP). Though not focused on fugitivity per se, other important digital projects focusing on the liberation of the enslaved and its aftermath include “Visualizing Emancipation,” based at the University of Richmond (https://dsl.richmond.edu/emancipation/), and “Last Seen: Finding Family after Slavery,” at Villanova University (https://informationwanted.org/).

5. 5. FOTM does not currently link to data and materials on the historical construct of race and racial categories outside the project itself, but as the database becomes larger, more extensive ancillary and related materials and data on this and other questions may get added to the project website, freedomonthemove.org.

6. 6. See chapter 7 in this volume.

7. 7. See chapter 7 in this volume.

Bibliography

1. Bailey, Moya Z. “All the Digital Humanists Are White, All the Nerds Are Men, but Some of Us Are Brave Journal of Digital Humanities.” Journal of Digital Humanities 1, no. 1 (Winter 2011). http://journalofdigitalhumanities.org/1-1/all-the-digital-humanists-are-white-all-the-nerds-are-men-but-some-of-us-are-brave-by-moya-z-bailey/.
2. Berry, Daina Ramey. The Price for Their Pound of Flesh: The Value of the Enslaved, from Womb to Grave, in the Building of a Nation. Boston: Beacon Press, 2017.
3. Block, Sharon. Colonial Complexions: Race and Bodies in Eighteenth-Century America. Philadelphia: University of Pennsylvania Press, 2018.
4. Charleston Courier. May 20, 1822. Available at Freedom on the Move, freedomonthemove.org. Accessed July 27, 2020.
5. D’Ignazio, Catherine, and Lauren F. Klein. Data Feminism. Durham, N.C.: Duke University Press, 2020.
6. Earhart, Amy E. “‘Can Information Be Unfettered? Race and the New Digital Humanities Canon.” In Debates in the Digital Humanities, edited by Matthew K. Gold, chap. 18. Minneapolis: University of Minnesota Press, 2012. https://dhdebates.gc.cuny.edu/read/untitled-88c11800-9446-469b-a3be-3fdb36bfbd1e/section/cf0af04d-73e3-4738-98d9-74c1ae3534e5.
7. Earhart, Amy E., and Toniesha L. Taylor. “Pedagogies of Race: Digital Humanities in the Age of Ferguson.” Debates in the Digital Humanities 2016, edited by Matthew K. Gold and Lauren F. Klein, 251–64. Minneapolis: University of Minnesota Press, 2016. https://doi.org/10.5749/j.ctt1cn6thb.24.
8. Hodges, Graham R., and Alan E. Brown. “Pretends to Be Free”: Runaway Slave Advertisements from Colonial and Revolutionary New York and New Jersey. New York: Garland, 1994.
9. Johnson, Jessica Marie. “Markup Bodies: Black [Life] Studies and Slavery [Death] Studies at the Digital Crossroads.” Social Text 36, no. 4 (137) (December 2018): 57–79.
10. Johnson, Walter. Soul by Soul: Life Inside the Antebellum Slave Market. Cambridge, Mass.: Harvard University Press, 1999.
11. Lothian, Alexis, and Amanda Phillips. “Can Digital Humanities Mean Transformative Critique?” Journal of E-Media Studies 3, no. 1 (2013). https://doi.org/10.1349/PS1.1938-6060.A.425.
12. McPherson, Tara. “Why Are the Digital Humanities So White? Or Thinking the Histories of Race and Computation.” In Debates in the Digital Humanities, edited by Matthew K. Gold, chap. 9. Minneapolis: University of Minnesota Press, 2012. https://dhdebates.gc.cuny.edu/read/untitled-88c11800-9446-469b-a3be-3fdb36bfbd1e/section/20df8acd-9ab9-4f35-8a5d-e91aa5f4a0ea.
13. New Orleans Daily Picayune. June 13, 1845, and September 15, 1846. Available at Freedom on the Move, freedomonthemove.org. Accessed July 27, 2020.
14. Noble, Safiya Umoja. “Toward a Critical Black Digital Humanities.” In Debates in the Digital Humanities 2019, edited by Matthew K. Gold and Lauren F. Klein, 27–35. Minneapolis: University of Minnesota Press, 2019. https://doi.org/10.5749/j.ctvg251hk.5.
15. Ridge, Mia. “From Tagging to Theorizing: Deepening Engagement with Cultural Heritage through Crowdsourcing.” Curator: The Museum Journal 56, no. 4 (2013): 435–50.
16. Risam, Roopika. “Beyond the Margins: Intersectionality and the Digital Humanities.” Digital Humanities Quarterly 009, no. 2 (September 2015).
17. Smith, Billy G., and Richard Wojtowicz. Blacks Who Stole Themselves: Advertisements for Runaways in the Pennsylvania Gazette, 1728–1790. Philadelphia: University of Pennsylvania Press, 1989.
18. Thomas, William. “What We Think We Will Build and What We Build in Digital Humanities.” Journal of Digital Humanities 1, no. 1 (Winter 2011).
19. Williams, Daryle. “Digital Approaches to the History of the Atlantic Slave Trade.” Oxford Research Encyclopedia of African History, November 20, 2018. https://doi.org/10.1093/acrefore/9780190277734.013.121.
20. Windley, Lathan A. Runaway Slave Advertisements: A Documentary History from the 1730s to 1790 Virginia. Westport, Ct.: Greenwood, 1983.