Whose facts matter?

A Cautionary Tale



Mainstream journalism practitioners pride themselves on the pursuit of verifiable facts.

Today, core journalism norms are being challenged, starting with the pursuit of objectivity.

Meanwhile, computational journalists are developing sophisticated new newsgathering methods. One goal of these efforts is to improve efficiency; another is to raise audience's views of news credibility and engagement.
Others advocate for "solutions" journalism or "peace" journalism - but critics say these approaches verge on advocacy.
A coalition of tech and nonprofit funders recently launched the $14 million News Integrity Initiative to promote news literacy.

One News Integrity Initiative partner, the Democracy Fund, recently announced grants to 20 projects across the country aimed at combating fake news and restoring public trust in news media. Some of these projects recruit community members as volunteer fact-checkers. Others develop software to detect hoaxes and fake news bots.

But what - and who - is missing from these conversations?


We have an opportunity to develop shared understandings of "truth" and "facts" by developing more inclusive methods of news design, reporting and engagement. This goes beyond the 50-year-old imperative to diversify news staffs and sources, although that's important. It goes beyond the laudable advice to journalists to engage people of differing perspectives.

An inclusive journalism epistemology requires new conversations about how our methods, sources, business models and deployment of technologies might reinforce bias and perpetuate invidious distinctions along our "fault lines" of race, class, gender and geography. In this 2013 video, Dori Maynard, the late director of the Maynard Institute, recalls the missed opportunities at the dawn of the Internet age:

Here we had a chance, back then, when we really could have built in structures that ensured that we were as inclusive as possible. Instead, the people driving the conversation told us, 'Don't worry your pretty little head about that. This is all going to be taken care of.' What we ended up doing was replicating the structure of the traditional media...
[On] the issue of race in... technology, ... you heard, 'Oh, no. The Internet is going to solve all this. Race and gender will no longer matter.'"(Maynard)
In his book, Whose Global Village, Ramesh Srinivasan warns against, "the assumption that the technology "naturally" shapes interactions between members of a community and those outside it that are positive for all involved."

Srinivasan also insists."We cannot simply trust our gateways to the digital world as if they are democratically designed platforms." Instead, we should "consider the intentions behind technology initiatives, and the way they are framed."

Indeed, tech companies using artificial intelligence to generate news and organize content have found their programs can magnify and reinforce bias.


What intentions and values guide the push toward computational journalism, solutions journalism and news integrity? What are our unexamined assumptions about the nature of journalistic truth, the value of hierarchical databases and algorithms, and the relationships between news providers and their publics?


This essay will consider these questions in the context of evolving notions of objective journalism, and in light of emerging scholarship on collaborative technology development. Part one will focus on an exchange of letters between journalist Ray Stannard Baker and sociologist, historian and journalist W.E.B. Du Bois over Baker's coverage of race. Both men cared about facts. Baker's journalism was arguably solutions-oriented, and he sought feedback on his drafts from Du Bois.

Du Bois took issue with aspects of Baker's reporting that he felt reflected fears and stereotypes. He was only partially successful in persuading Stannard Baker to make revisions. In additon, there were important sources both men appear to have ignored - or at least, failed to acknowledge.

Part two will examine the Baker-Du Bois exchange in light of critical scholarship on the concept of objectivity in journalism and the social sciences as it emerged from the mid-19th to the early 20th centuries. In particular, we will consider the work of David Mindich, Khalil Muhammad, and Natalie Byfield.

These scholars have identified ways in which the "just the facts" creed of early 20th-century journalism was limited by ingrained biases of race, class and gender. Mindich identified how race, class and gender bias shaped the editorial judgment of 1890s-era newspaper editors. Byfield and Muhammad have argued that these biases would be reflected in "objective" news reporting practices throughout the 20th century, even as newsrooms sought to diversify their staffs.

Part three will explore the requirements for inclusive and culturally competent computational journalism. Here, we will look at a case study involving the use of artificial intelligence in homicide reporting at the Los Angeles Times. The Times' Homicide Watch project had worthy goals: to report on each victim's life and death respectfully, and to explore the larger social contexts of the city's crime problem. As we will see, the results were mixed.

Part four will consider how emerging best practices for collaborating with communities to create culturally responsive technologies might inform the creation of news products. Here, I will draw up the work of Ramesh Srinivasan as well as the research experiences of myself and others in fostering interdisciplinary computing collaborations designed to meet the information needs of specific communities.

Share your thoughts

Go to Part One