Ecologies of Data Part III: Circulating Significance in Civic Hacking

In the third contribution to "Ecologies of Data," Etienne Benson ruminates on his research efforts to track federal permits given to American trophy hunters of polar bears in Canada. In Benson's consideration of why, and moreover how, to build Data Refuge, he argues that the translation of environmental data into other formats and media is vital to make data meaningful to diverse communities. Catch up on "Ecologies of Data": Intro, Part I, Part II

Circulating Significance in Civic Hacking

By Etienne Benson

Like the proverbial shark, data dies when it stops moving. Or to use a more familiar metaphor: data refuges are only as good as the paths that lead into and out of them. One of the basic tasks of data rescue is thus to ensure that the data are preserved in a way that allows them to participate in a broader ecosystem of tools, practices, institutions, and projects. Often this requires the construction of tools of translation that allow data to circulate and to become meaningful in domains other than the one in which they were originally produced and preserved. Here I want to describe one such translation effort and some of the lessons I learned from it.

A few years ago I became interested in a peculiar form of government data—namely, federal notices about permit applications for the “taking” of endangered species and marine mammals in the United States. Since the early 1970s, such “takes,” which include killing, injuring, or harassing members of protected species and populations, have been prohibited by the U.S. Endangered Species Act (ESA) and Marine Mammal Protection Act (MMPA). However, because some forms of scientific research, conservation, and education involve prohibited activities—such as the attachment of radiotracking tags, the extraction of tissue samples, or the display of animals in zoos and aquariums—the laws include procedures for issuing permits after proposals have been reviewed by experts and the public. To inform the interested public of pending applications, notices are published in the Federal Register, a regulatory publication established during the New Deal that has been available online since the mid-1990s. The development of this permitting system was a contentious process shaped by the emerging environmental, animal rights, and indigenous rights movements and by public pressures for government transparency in the wake of the Pentagon Papers and the Watergate scandal.

Figure 1: A biologist labeling blood samples taken from a tranquilized polar bear. Source: U.S. Fish and Wildlife Service.

Figure 1: A biologist labeling blood samples taken from a tranquilized polar bear. Source: U.S. Fish and Wildlife Service.

As with much government-produced data, however, there is a great deal of translation to be done before they can be used in ways other than those for which they were originally intended.

Permit application notices are not the sort of “environmental data” that we usually think of when we hear the term—that is, they are not data that represent the state of the physical world around us or its effects on our health and well-being. Nonetheless, as a historian of science, I see these notices as a form of environmental data that illuminate the history of efforts to understand and conserve endangered species and those efforts’ ongoing legacies today. Such data offer a rich, if partial, view of research and conservation activities that require interaction with or proximity to individual members of protected species. As with much government-produced data, however, there is a great deal of translation to be done before they can be used in ways other than those for which they were originally intended. Often formulated in idiosyncratic bureaucratese and buried amid tens of thousands of other regulatory rules and notices, they are public without necessarily being publicly accessible.

Figure 2: Volumes of the print version of the Federal Register, which runs to tens of thousands of pages each year. Source: Office of the Federal Register. 

Figure 2: Volumes of the print version of the Federal Register, which runs to tens of thousands of pages each year. Source: Office of the Federal Register. 

In the mid-1990s, the MMPA’s process for issuing permits to “take” marine mammals was amended for a new and somewhat surprising purpose: to encourage American sport hunters to kill Canadian polar bears.

In the hope of making such data meaningful in a new context and for new purposes, and as a way of teaching myself something about the possibilities and limits of civic hacking, I developed a web app called TrophySource to visualize data extracted from a particularly interesting variety of permit notice. In the mid-1990s, the MMPA’s process for issuing permits to “take” marine mammals was amended for a new and somewhat surprising purpose: to encourage American sport hunters to kill Canadian polar bears. As enacted in 1972, the MMPA had prohibited the import of most marine mammal parts, including polar bear skins, skulls, and claws. In the meantime, Canadian polar bear populations had been thriving, Inuit communities had developed lucrative guide operations serving non-indigenous trophy hunters, and American hunters had grown increasingly frustrated at being locked out of the Canadian hunt. In 1994, responding to these opportunities and pressures, the U.S. Congress amended the MMPA to allow sport hunters to be issued permits to import trophies from Canada, with the understanding that the money they spent in Canada would go partly toward maintaining its world-class polar bear research and conservation system.  

Figure 3: Notices of permit application for polar bear trophy imports. Source: Federal Register, Vol. 73, No. 92, Monday, May 12, 2008, p. 27005

 

From 1997 to 2008, more than 900 permits were issued to American hunters from every state except Hawaii to import the trophies of polar bears harvested from approved Canadian populations. Because it was framed as a conservation measure and implemented under the procedures established to allow “taking” under both the ESA and MMPA, the permitting process nominally involved the same public notice in the Federal Register, the same expert review, and the same 30-day public comment period as a scientist’s study or a zoo’s captive breeding program. An accident of the way the trophy import program was shoehorned into an existing legislative framework, the “public” nature of this procedure was largely pro forma. As far as I know, there were no public hearings held about individual permit applications, and few if any applications were rejected during the eleven years that the program was in operation.

 

Nonetheless, this legislative quirk meant that significant data about permit applicants—including their names, towns, and the source populations from which they had acquired their trophies—had become public information. Collecting and visualizing it, I thought, might tell me something about the kinds of people who hunted polar bears while serving to call attention to the complex and contradictory ways in which trophy hunting was linked to conservation programs.

Manually downloading the relevant notices would have been straightforward if time-consuming, but I wanted to see if a more automated approach was feasible. Luckily, a web interface and easy-to-use API had been developed as part of the Federal Register 2.0 project, which was launched in 2010 as part of the U.S. National Archives’ Open Government Plan. This API vastly simplified the process of accessing the hundreds of polar bear permit applications scattered throughout tens of thousands of pages of regulatory notices; it is especially useful for tracking regulatory developments as they occur, although in the case of polar bear trophy permits all the relevant notices had already been published. Using this API, I wrote a small program in Python to download all the polar bear permit notifications and to parse out the hunters’ names and towns of residence, the dates of application and permit issuance, and the polar bear populations from which the trophies came. On the basis of that data, and with the help of open-source toolkits such as the Leaflet mapping library, I wrote a short program in Javascript that made it possible to visualize and explore all 900-plus applications on an interactive map of North America. 

Figure 4: A screenshot of the TrophySource web app. Source: Author

Figure 4: A screenshot of the TrophySource web app. Source: Author

While the web site that resulted accomplished the technical goals I had set out for myself, the project had some serious limitations as an exercise in civic hacking. For one thing, by the time I began working on it, the import program was no longer active, having been rendered moot by the listing of the polar bear as “threatened” under the ESA in 2008 due to the threat of anthropogenic climate change. Even if the program had still been in operation, however, it is not clear what a geographical visualization of the data could have been used for, beyond simply informing the public about the number and general location of American polar bear hunters. Deeper connections to ongoing issues of advocacy and public awareness were missing. Moreover, from a scholarly standpoint, while there was a certain interest in seeing where polar bear hunters lived, the only clear patterns that emerged were ones that would have been easy to confirm without such a tool. It did not surprise me, for example, to find that virtually all polar bear trophy hunters have typically male names. I am quite sure that there are more interesting patterns to be found, but the data available from the Federal Register alone did make them visible. Among other hypotheses, my hunch that Texas would pan out to be the preferred habitat of the American polar bear hunter did not pan out.

Why do some projects in civic hacking fail while others thrive? I have found some conceptual resources for answering this question in the work of Bruno Latour, particularly a 1995 article initially published in the journal Common Knowledge under the impish title “The ‘Pedofil’ of Boa Vista.” Later republished as “Circulating Reference,” the article traced the transformation of soil samples into scientific facts by a team of Brazilian and French scientists. The “pedofil” in question was the perverse nickname given by the soil scientists (a.k.a. pedologists) to an instrument known as a Topofil, which allowed them to precisely measure distances in the field. For Latour, the significance of the “pedofil” was that it was one of a series of devices and practices that bridged the gap between the particularities of fieldwork and the disembodied facts that would later circulate through the scientific community and beyond. They were what ultimately brought words and things into relation—not through a sudden leap over an ontological abyss, Latour argued, but through the circulation of reference along a chain of careful and partial transformations. If any one of the links in the chain failed—if, for instance, the logbook that told scientists where and when a soil sample had been collected was lost—the data would lose its value, becoming once again what it had been at the beginning: just dirt.

Just as chains of reference give scientists confidence that the numbers and images they are manipulating in the laboratory will tell them something about the world, these devices and practices give members of the public ways of making data meaningful in their own lives.

Something similar happens or, just as often, fails to happen when a dataset is “made public,” or when a new app based on open government data is developed. Behind the dataset lie the chains of reference described by Latour, which assure users that it accurately represents something in the world—for example, that the notice of a permit issued to a polar bear trophy hunter published in the Federal Register ultimately tells us something about the death of a bear. To believe this, we need to have faith in the techniques and institutions that link bears, guides, hunters, customs agents, environmental NGOs, federal regulators, scientific advisors, and many other actors of diverse kinds. But simply ensuring the integrity of that chain is not enough to make the data accessible and meaningful to a broader public. “In front” of the dataset, so to speak, lie not chains of reference, but chains of significance, comprised of series of mediating devices, practices, and institutions that make it possible for people to care about the data in particular ways and for particular purposes. This does not happen by itself; for the dataset to become a matter of public concern, a new assemblage of mediating devices and practices must be constructed. Just as chains of reference give scientists confidence that the numbers and images they are manipulating in the laboratory will tell them something about the world, these devices and practices give members of the public ways of making data meaningful in their own lives. 

The success of such projects depends on recognizing that a piece of code, no matter how sophisticated, is at best only a single link in a longer and more complex chain of significance, and that the real test of success is not whether the data is represented accurately or even beautifully, but whether it becomes capable of circulating in ways that bridge the gap between data and matters of public concern.

Ultimately, this is the main lesson I learned from my foray into civic hacking: that the success of such projects depends on recognizing that a piece of code, no matter how sophisticated, is at best only a single link in a longer and more complex chain of significance, and that the real test of success is not whether the data is represented accurately or even beautifully, but whether it becomes capable of circulating in ways that bridge the gap between data and matters of public concern. This happens, as Latour’s work on facts suggests, not through a single cynicism-defying leap, but rather through the patient construction of tools and practices that make data progressively more meaningful to more people as one moves along the chain. Data that are both true and meaningful require building and maintaining the chains that extend in both directions—those going back toward the world they are intended to represent, and those going forward toward the people who might use them to make a better world.

---

Etienne Benson is an Assistant Professor in the Department of History and Sociology of Science interested in the history of ecology, environmentalism, and human-animal relationships. He has been a postdoctoral fellow at the Harvard University Center for the Environment, a research scholar at the Max Planck Institute for the History of Science in Berlin, and a visiting scholar at the Rachel Carson Center in Munich. He teaches courses on environmental history and the history of science and technology. Check out his work at http://etiennebenson.com.