Interactive-Journalism-Chapter 3.pdf

3 Hacker Journalists, Programmer Journalists, and Data Journalists Brian Boyer was casually scanning Boing Boing, the tech blog, in May 2007. The previous day, the blog had featured its usual mix of oddball news with tech info—favorite podcasts, interviews with bloggers, Mac...

3 Hacker Journalists, Programmer Journalists, and Data Journalists Brian Boyer was casually scanning Boing Boing, the tech blog, in May 2007. The previous day, the blog had featured its usual mix of oddball news with tech info—favorite podcasts, interviews with bloggers, Mac up- dates, and an amusing feature on a robot that looked like a chicken, which had been taught to dance by Japanese scientists.1 But on May 24, Boyer found a different kind of post—an announcement that would change the trajectory of his career and his life. It read: turning coders into journalists (hint: add spellcheck, subtract skittles). Rich Gordon of Northwestern’s Medill School of Journalism says The Medill School of Journalism just won a grant that will allow Medill to offer master’s degree scholarships in journalism to computer programmers. It’s among $12 million in grants awarded via the Knight News Challenge.2 The general idea is to lure talented coders, immerse them in the practice of journalism, and then turn them loose to figure out interesting ways of putting journalism together. Boyer, who was designing medical records software for small doctors’ offices, said he enjoyed the “craft of making software,” but had lost interest in mak- ing software for businesses. He wanted to make something he was proud of: he wasn’t, so he was looking to do something other than what he had been doing in the business world. So when he came across the Boing Boing post, as he put it, “I Googled journalism and said, shit, that’s what I want to do. Literally, journalism is about enabling people to do better, to self-govern, and I said I wanted to do that... let’s give people information and build democracy from the grassroots.”3  Boyer became part of the inaugural Medill program, where he learned everything from how to cover crime to what libel laws were. And by 2011 he was running a full team at The Chicago Tribune. But his wasn’t your typical newsroom reporting team. Instead, Boyer’s group—The Chicago Tribune News Apps team—built tools for data analysis and visualization, ultimately creating software for new types of storytelling as well as for traditional text- based accounts. Programming was the starting point for their work and the beginning of their contribution to the editorial workflow. Boyer is one of a small but distinguished group of journalists who have come to the newsroom from the programming world. He is a hacker journal- ist—with a past in software and now in the newsroom. He is not alone: today, as journalism contends for relevancy in the digital environment, coding skills have become increasingly important. He is joined by other former profes- sional developers who now spend most of their time programming in the service of news. Some resist the label “journalist,” but to be successful they must understand the editorial workflow, demands, needs, and expectations, communicate according to journalistic norms, and, more important, expand the professional jurisdiction over work and knowledge for journalism. Accompanying Boyer, in newsrooms across the country—in greater num- bers in large newsrooms, but also in midsize and small newsrooms4—are his journalism-first counterparts, those who began working in journalism or in liberal arts disciplines and have taught themselves to code. These program- mer journalists think in terms of the story first, and they have a different way of thinking about interactive journalism. Adding to these ranks are data journalists, who work closely with data, often use computers to help with reporting, and may not always know how to code. These journalists descend from a legacy of computer-assisted reporting to work to manipulate data in the service of journalism—they are data specialists who may also create interactives. They, too, think about journalism differently from traditional journalists, often looking for numerical or categorical data rather than quali- tative evidence. These categories as defined here may not make much sense to interactive journalists themselves; indeed, they may find such distinctions to be arbitrary or overlapping. Those I’ve defined as hacker journalists might call themselves programmer journalists or data journalists, and these terms get further muddied outside of the United States in practice and in academic discussion. What, then, do we make of this labeling quandary? It is difficult for those within the field to remove themselves from day-to- day work to consider larger conceptual categories, even though they think they may be doing so. Nonetheless, the advantage of academia is that we can provide such a perspective to begin to draw outlines of the emerging 72 Chapter 3 field—academics have the luxury of having the distance to make distinctions and provide analysis. The empirical data here gathered from across fourteen different newsrooms around the world provide the evidence to draw these distinctions. Empirical data represents what people have actually said, and it offers a rationale directly from the words of the people who work inside these newsrooms. So these journalists essentially come to define themselves through how they talk about their backgrounds and perspectives, and I offer the analytical distinctions through the way they talk about who they are. The diversity of terms and descriptors used to define practitioners can be seen as varying across the work of different scholars. Katherine Fink and C. W. Anderson make the decision in their work not to create their own defi- nition of data journalism to begin their inquiry.5 Australian scholars Terry Flew, Christina Spurgeon, and Anna Daniel see the origins of interactive journalism as beginning with computer scientists who augment journalism.6 Sylvain Parasie and Eric Dagiral, who offer an ethnographic study of The Chicago Tribune, use the term “programmer journalists” for what I define as hacker journalists, but they conflate their study population with data-driven journalism.7 Cindy Royal, who spent a week embedded with The New York Times Interactive News Team, discovered that the team’s journalists had vari- ous backgrounds: some had worked in tech companies, others had worked in journalism and had journalism backgrounds, but all were united in the service of a journalistic endeavor.8 Others have also tried to break out interactive journalists into a specific typology. Wibke Weber and Hannes Rall define three groups of journalists working in interactive journalism: the programmer, the designer, and the statistician.9 Astrid Gynnild contends (curiously) that data journalism is only data journalism when it deals with open-source data.10 Mark Coddington, offering another attempt to define the field, distinguishes computer-assisted reporting from data journalism by noting that CAR is a “historical mode of quantitative journalism.”11 Even Adrian Holovaty, the first breakout interac- tive journalist, had difficulty with labels. He wrote a blog post arguing “who cares?” about whether his work was called data journalism or not, but then he used the phrase “journalism as programming” as a way to talk about the back end of his work. All of these definitions are so different that it’s important to take a stab at finding conceptual clarity using field-based evidence derived from an analysis of the patterns through which journalists talk about who they are and how they approach their work. But it is fundamental, now more than ever, to understand and to catego- rize interactive journalists. These terms are all quite popular, but used as buzzwords, there’s not much consistency about what they mean. Rather than Hacker, Programmer, and Data Journalists 73 leaving “hacker journalist,” “programmer journalist,” and “data journalist” at the level of buzzwords, only to be discarded in a few years, the words actually have some staying power in the discourse of journalism if they can come to mean something clear and definable. Similarly, these categorizations make it apparent that such terms do matter because they have specific histories and cultural assumptions built into their practices and identities. Categorization through empirical work helps us understand what is novel, old, and the same about these groups and as they compare to traditional journalism; in turn, such sorting helps us get a more complete grasp of the phenomena at hand. This chapter looks at the people behind the work and knowledge of this expansion of professional journalism. The reason to offer this perspective is that backgrounds and perspectives help us understand their approaches to work. A look at journalists’ ways of thinking opens an examination of how they will offer distinct claims to abstract knowledge that will inform and expand the larger journalism profession (further explored in chapter 5). This “people” chapter helps us see clearly that there are indeed new kinds of journalists—those who work with code and those who use computational work to approach data holistically—who signal by their presence just how journalism is expanding. Hither Come the Hacker Journalist Who is a hacker journalist? A PowerPoint presentation created by the Knight Foundation featured the idealized story of the hacker journalist. The hacker would fuse with the traditional journalist to become “Journalist 2.0.” On one side of the slide was the programmer/hacker, wearing headphones, a black t-shirt, jeans, and “chill clothes,” and was described as a “problem solver, pro- cess oriented” and a “builder.” On the other side of the slide was the journalist, wearing dorky clothes—a shirt and tie under a buttoned cardigan, holding a reporter’s notebook and wearing glasses. This traditional reporter had the words “Big Picture thinker, storyteller, wordsmith, contrarian, investigator” associated with him.12 After journalism school—or, perhaps, after experience with traditional journalism—the hacker journalist, or “Journalist 2.0,” would become social- ized into the newsroom, and in the middle of the slide he is shown wear- ing new clothes (a plaid shirt instead of the black t-shirt) and is associated with the words “translator,” “info distiller,” “impactor,” “data visualizer” (he’s holding a digital tablet now) and “pragmatist.” This hacker journalist, then, would bring his way of thinking to the needs of the newsroom.13 The slide noted that the hybrid figure was “loosely based on the life of Brian Boyer, 74 Chapter 3  “The Hacker Journalist,” Knight Foundation a programmer bored with creating e-solutions for insurance companies.”14 Hacker journalist, then, describes those who come from a programming background and are now part of the newsroom.15 Boyer says he, to the best of his knowledge, coined the term hacker jour- nalist.16 In fact, he owns the domain hackerjournalism.net. He wrote in 2008: “If you’re a hacker, you’re an especially good programmer. So, what are you if you’re a hacker journalist? Think about what photojournalists do—they tell stories with a camera.” The hacker journalist would tell stories through code. Hacker journalism is as much a type of journalism as it is a type of person who now works in the newsroom. And this hacker journalist comes from the outside, bringing in new skills, ideas, and perceptions of how to do journalistic work. The hacker journalist should never be conflated with anti-social hackers who have nefarious intentions—hackers who break into computers or steal passwords, who steal data from governments or news organizations, or who obtain credit card information illegally. Hacking is instead used here to mean a pro-social activity and outlook toward creating software. These program- mers “hack” out of the joy of building software for the common good. Hacking as a technical craft has a long history of innovation and creation with the goal of using code to improve society. This vision of hacking dates to the early computer culture of the 1960s at MIT—where email was devel- Hacker, Programmer, and Data Journalists 75 oped as an early “hack”—and it continues to imbue a spirit of community, openness, participation, and experimentation. Another way to think about hacking is as a solution to a problem, taking the easiest route to get something to work. Hacking values, one of which is “do no harm,” emerge from prin- ciples of the open-source software community.17 Newsroom programmers themselves and programming in newsrooms generally are tied to this hacker culture, which is inspired by an inherent desire to use code to do good and to share this work with a wider public. As hacker journalist Jonathan Stray of the Associated Press asked of tech experts who might become journalists, “Can you code, are you good at help- ing people learn about their world, and do you see how software as civic media might contribute to some sort of democratic or social good/making the world a better place?”18 Hacker journalists, then, come from programming and can better the world through applying code in the service of journalism. They are outsiders who leave behind industry, where they have sometimes been programming for a product they consider soulless and with low creative value. Now, they have the opportunity to program in the public interest. These programmers have been more than welcomed into the news industry, as their ability to construct computational solutions and innovations has been heralded and turned into aspirational myth—as we see from the Knight slide for the promise of the Journalist 2.0. Out of all the journalists who can program, though, journalists who were programmers first are in a distinct minority. Why They Come The reason programmers leave particular jobs is because they say they can’t find in commercial programming anything like what they might do in a news- room. Most of the hackers who have come to newsrooms have backgrounds in commercial IT, which they left because of the emphasis on client-side work and the limits on creative expression. The newsroom, with new challenges every day, offers hacker journalists a chance to build new projects to help illustrate social problems and issues in the public interest, which they argue would have otherwise been impossible in their old jobs. Consider what Ryan Mark, a hacker journalist who also came through Northwestern University’s program, had to say: “I think there are a lot of software development shops out there who are just kind of like factories and people put in their 9-to-5 and don’t really care too much about what they’re working on.... But in media and journalism it’s the total opposite. People really care about their work and really care about doing the right things.”19 76 Chapter 3  His colleague Joe Germuska noted, “I worked in interactive marketing for 10 years, mostly doing web development that wasn’t very public. Toward the end, I didn’t care about what I was building. The technology itself was not enough.”20 Software developers, according to Mark, Germuska, and Boyer, among others, seek to “scratch the itch” or find a solution through program- ming to what they saw as a problem—or just bothersome. Working in com- mercial software doesn’t offer that chance to explore or solve problems in the same way that coming up with creative solutions for news might. Other hackers across the world have also traded corporate jobs for the environment of public service. Tsan Yuan, a hacker journalist at the BBC, was employed in large Web site development for banks and insurance companies: “I came to a newsroom because I was interested in digital media, new media and the technical shape of news in a different way... being here has a huge impact.”21 In short, if you care about making a difference, leave behind the big bucks and work in journalism. Another British journalist, Alastair Dant of The Guardian, talked about his work in San Francisco at a series of “soul sucking JavaScript jobs,” which were a “lucrative dull enterprise.”22 While he managed to create a kids’ video game in his spare time, he wanted to come to The Guardian because of its commitment to good journalism at a time when the business model needed help. “The Guardian has a great liberal voice.... It’s doing high-quality jour- nalism.”23 For The Guardian to survive, it needed to innovate online, and Dant believed he could be part of this effort. At Al Jazeera English, Mohammed el-Haddad thought that he could be better integrated in world events with a job in the newsroom. Haddad was hired just after graduation. He said he came to Al Jazeera English at a time when the region was particularly active.24 He told MediaShift: “I was very fortunate to join Al Jazeera in 2011 just at the time of Egypt’s revolution. The energy was tremendous. I think we all had a sense of how big a moment in history that was going to be. For me, being fresh in the news business, it didn’t take long to see how the combination of traditional media and technology was able to impact hundreds of millions of people in the region.”25 Doing hacker journalism would be a way literally to change the way people were experiencing the Arab Spring and beyond. Hackers may also espouse a commitment to free speech; in fact, hackers have a reputation, at the most extreme, of freeing secret documents they believe belong to the public (like Sony’s internal emails, for good or bad). In this case, though, I did not come across any hackers who were interested in coming to journalism because of their commitment to free speech. Perhaps the commitment to free speech was implicit in their larger commitment to serving the public interest. Hacker, Programmer, and Data Journalists 77  Since journalism has the potential to harness new technical capacities, programmers have the ability to bring their unique skills together with jour- nalism’s distinct mission. The profession has created an opening for skilled individuals with knowledge of code to come in and change the way that technology is used in the editorial process; this has meant a place for hacker journalists inside the newsroom. It’s also clear that their belief in public interest melds well with the larger, overarching goal of journalism itself, suggesting that these programmers do not have to change their goals or aspirations when they come to the newsroom. Ways of Thinking This different background—the programmer who comes to journalism— presents an opportunity for thinking about journalism in a new way. The experience of building software and working in programming communities also means being connected to a particular set of approaches, problems, and solutions. The so-called “hacker” approach to software design perme- ates interactive journalism, and some of its ambassadors are these former programmers. Hacker journalists think about journalism as a site of play, exploration, and problem solving. AP journalist Jonathan Stray has explained what it means to approach journalism with this hacking perspective on his blog. In a post titled “Jour- nalism for Makers,” he discusses the important theme of hacking in creating and improving journalism. He points to the qualities inherent in people who come from programming to journalism and how they think about their work: [They are] geeks who like to understand very complex systems, and tinker with them. I want to borrow from the culture of “makers,” because maker culture plants a flag on this idea. It draws on the hacker tradition of technical mastery, the DIY aesthetic perfected by the punks, and the best disruptive tendencies of global counter-culture. It lives in online forums and nerdy meetups and on the dingy couches of hack spaces. This is the chaotic ecosystem that powers Silicon Valley.26 In Stray’s understanding, hacker journalism would bring together the best of hackers and the ethos of doing it yourself to try new things, to “tinker,” or make small adjustments and changes to see simply what happens, and share a different attitude that comes from an entirely different perspective. This “inquisitive passion for tinkering,” as Gabriella Coleman writes,27 is a cornerstone of hacking culture.28 Hacker journalists come from this hacking counterculture with the goal of being disruptive, making and doing through “technical mastery.” 78 Chapter 3  Hacker journalists talked about how journalism was a problem that could be solved, and emphasized this rather than the story. Stijn Debrouwere, a hacker journalist working at The Guardian who came out of a commercial programming background, tried to explain what he saw as the kind of think- ing hacker journalists provided to the newsroom—a problem-solving skillset: This is an absolute cliché but I do believe it’s true: developers are trained to be lazy, and that’s a good thing. Journalists are used to manual work: looking through that same courts database every single day to see if there’s any new documents, poring through huge amounts of documents by hand, doing the same calculations over and over again, or they might simply give up stories because the challenge might seem insurmountable. You could never get a developer to do things like that: they’d write a small ap- plication so they never ever have to do the same task twice. And along the way you reap the benefits, because some of those applications can be used in very different contexts by other reporters, and other applications can be opened up and repurposed into news applications you can share with your audience. That’s a really valuable attitude and skillset, I think. I also think programmers are more used to thinking in terms of informa- tion rather than stories. That’s both a valuable asset and a shortcoming, because it means programmers spot valuable information we can share with readers, even when it looks boring or “not a story” to a reporter, but OTOH [on the other hand] it also means that programmers perhaps don’t have as great a feel for storytelling.29 Debrouwere points out two distinct ways of thinking about journalism. First, journalism is something to be hacked; second, journalism is information. Programmers bring to journalism a focus on the “hack”—looking for the quick solution to the big problem in order to find an easy way to do things; this solution, of course, comes from code. Journalism is a process that can be made easier through programming solutions. The hacker journalist offers the mindset that if the right code and the right methods are applied, then journalism will be efficient and answer the questions posed. Programmers also don’t think of journalism as stories but as a product with component parts—“information.” This suggests that to the program- mer, journalism may be able to be detached from a narrative and a longer context. Instead, journalism as “information” may be moved around, disag- gregated, and re-ordered. Each element a reporter uncovers for a story should be thought of like a line of code: information as building block rather than the end point of story. The information then becomes the scaffolding for other projects, which may or may not ever become stories. Hacker journalists approach their work with a perspective not found in traditional journalism. Rarely has the word “play” come up as journalists Hacker, Programmer, and Data Journalists 79 talk about their work. But hacker journalists take the idea from the spirit of programming that they can constantly test things out. They enjoy the process of crafting new ways of doing work. They like messing around, try- ing to get computers to approach journalism in new ways; they experiment knowing that they may fail. Ransome Mpini, a hacker journalist at the BBC, explained the attitude this way: “It’s pretty interesting. You have to have an interest in playing around with tech. There was a definition of a hacker that I liked, [which] I found [last night], that I thought was quite right—it went something like ‘a passionate skill and an innovative skill.’”30 Hackers are experimenting and innovating. The attitude is also one that ushers in a desire for doing work as playful and fun; they look at journalism as information and journalism as a problem that can be solved. These are not ways of thinking that have normally been found in newsrooms; they come instead from the programming culture that now feeds into newsrooms as code becomes more integral in daily news work, and coders more ingrained in the life of a newsroom. Hacker journalists also say that they will disrupt newsroom workflow. At a panel at West Virginia University, the head of The Seattle Times’ News Apps team, Lauren Rabaino, said that she found traditional workflow restrictive: “There is this newsroom structure of project management that is an assembly line.”31 She talked about how this assembly process was not nimble enough. Instead, she pointed out how “agile” software development was simply much better. Agile is one method for software development wherein production happens in rapid cycles with key players reporting in to assess progress. This agile development brings with it the opportunity to create prototypes rapidly, for good or for bad, and can work well for both long- and short-term projects. This prototyping again speaks to the temperament for experimentation that hackers bring with them to the newsroom. And there’s simply a cool factor that hackers bring to the newsroom. Hacker journalists are the people in black t-shirts. In 2010, after winning a Pulitzer with a major interactive project, Bill Keller at The New York Times pointed out how the newspaper suddenly had people with tattoos and pierc- ings working there.32 These hacker journalists may even have a look (ste- reotypically imagined as having bushy beards—a problem underscoring a more complicated gender issue), and there is an outsider, rebel mentality that nonetheless tries to build on a communal project, what you might find in open-source communities. Rabaino added, “We want to empower people. We want to be the cool kid in the newsroom.”33 In this respect, some leaders of these teams hope to infuse a new kind of energy and attitude. 80 Chapter 3  As a result, the attitude hackers bring to journalism is associated with a newfound approach to innovating in the newsroom. Journalists who do not come from a hacker background expressed what they saw this world of hacking as offering to news. Jon Keegan at The Wall Street Journal explained: “The idea of hacking and journalism comes a bit from tech hacking—it’s seeing how journalism can go in new ways and it’s not just solving a tech problem.”34 He suggests the term raises a key question: “Can you creatively solve the problem?”35 Other journalists in his newsroom also took inspiration from what they saw as the philosophy behind hacker journalism. One of Keegan’s deputies, Sarah Slobin, explained: “I learned that the hacker piece is that you can fig- ure it out yourself.... You have to figure out things you can’t do and if you can’t answer a question, you have to figure things out.... You also have a DIY mentality.”36 It inspired her, as a journalist and then programmer, as it “makes me do new things.”37 At the same time, The Wall Street Journal would never use the term “hacker journalism”; as a NewsCorp paper, it is especially sensitive to the word hacker because of the UK phone hacking scandal, where journalists hacked into potential sources’ voicemails. But the term still remained a source of inspiration at public radio station WNYC in lower Manhattan, about forty blocks away from The Journal’s headquarters. Head of the Data News Team John Keefe traced the idea of hacker journalism back to the idea of openness and community. He noted, “With hacker journalism, there is a sense among the crowd that they are do- ing good things. That is really cool. We are sharing and building on work.”38 In this way, the infusion of hackers into the newsroom represents this Journalism 2.0—the journalists are indeed offering a new way of approach- ing journalism problems. Not only is the platform different—through soft- ware—but also the process is different. Other journalists embrace this hacker mentality and draw inspiration from what hacker journalists have identified as some of the benefits they bring to the newsroom. Play, problem solving, experimentation, “making,” and “doing” are all terms that embody the action behind the code-based thinking these programmers bring. And as we can see from other journalists, this kind of approach is welcome. Are They Journalists? If the focus is on assembling code and writing programs, then this raises the question of whether this is journalism. Certainly, this is not traditional journalism if hacker journalists are writing scripts to generate interactives, Hacker, Programmer, and Data Journalists 81 from maps to charts to databases. Often, hacker journalists are working with more traditional journalists who will then do the work of finding stories, though hacker journalists can and do have Page One bylines. If the product is an application—and the hacker journalist isn’t offering analysis—are they just programmers in the newsroom? By self-description, many resist the label journalists. They don’t see them- selves writing the traditional story. Germuska says that he is not a “real” journalist, though he is proud to collaborate and get A-1 stories, which cer- tainly makes him feel like more a part of The Chicago Tribune news-creation experience. Mohammad el-Haddad of Al Jazeera English has only recently started calling himself a journalist—when I met him in 2012, he said, “I would not call myself a journalist.”39 But after two years of speaking about interactive journalism, with continued questions about his efforts in programming and journalism, his bio now reads “data journalist.” Haddad explained to me, “I have had to learn a lot about journalism.” Notably, as he said, the word hacker would simply not be used at Al Jazeera, either to describe what he does or to talk about the kind of journalism he works on—hacker comes too close to the kind of anti-social efforts Syrian computer terrorists continually engage in to undermine Al Jazeera English content.40 Most of those I spoke with suggested that while they cared about journal- ism deeply, they didn’t feel equipped to engage directly with the processes of daily journalistic work. Boyer told traditional NPR journalists in a meeting I observed that they were the ones in charge of the story, not him.41 At the same time, his group was collecting the data that would populate the search fields, and he had, on his team, an expert in getting these reports. Perhaps, then, he was working as a data collector and data presenter for interactive journalism, but not for traditional radio news. In this way, he was doing journalism work, even if his output wasn’t the work of traditional journalism. In the United Kingdom, hackers on news staffs deflected the label jour- nalist, arguing that they were much closer to developers that happened to enjoy working in a newsroom. A Guardian employee explained it this way: “I guess as a developer I see my contribution to the content production and the content of the story as an application of technical knowledge and web text and html to make more complete interactives to visualize something and make something interactive... to provide a different way to look at expert feedback—I suppose I am doing journalism but I am not a journalist by background.”42 In this respect, many of those with developer-type skills who come out of the technical world are not the traditional journalists we might expect from the newsroom. 82 Chapter 3  It’s still important to call hacker journalists “journalists” even if they might not consider themselves by this term. There’s considerable power in using the word “journalist”—and it is an important claim: these hacker journalists must be considered critical to the future of journalism and be recognized with this symbolic power. If we fail to consider hacker journalists as journalists, we are disregarding the value of what they add to the editorial process. Their product is software, certainly, but what they do is journalism in another form. The process of creating an interactive often requires hacker journalists to dig into data, gather information, and render it ready for news consumption. This collating and collecting is a critical element of journalism. Similarly, the integration into the editorial process signals the role these hacker journalists play as journalists: without them the final output—the piece of journalism— would not be possible. These hacker journalists have to understand journalism in order to do their work and must communicate with a news sensibility. To create inter- actives and work with other journalists requires news judgment, the abil- ity to understand the needs of the larger journalistic output, and therefore underscores the integration of hacker journalists into the newsroom. They would be unable to successfully do their work if they could not communicate across journalistic endeavors, as they do. And most of these hacker journal- ists readily accept, respect, and wish to pursue common journalistic aims, as Germuska, Haddad, and others suggest and as indicated by their willingness to learn about journalism and their hope of working to help further journal- ism’s larger goals. In fact, even before these hacker journalists come to the newsroom, they have an idealized sense of journalism’s public-service role. Their willingness to turn away from lucrative positions elsewhere out of a desire to contribute to the large project of journalism signals their acceptance and understanding of the aims of the profession. Thus, to move journalism forward and to think about how journalism must change, it is critical to recognize hacker journalists as fully integrated members of newsrooms. For journalists to make a claim to relevance, news organizations and the profession must claim what these hacker journalists do as journalism. This work is critical to establishing a unique entry point for journalism in the larger information ecology. Though they might not use the label “journalism,” their role in the news process, while different, nonetheless requires understanding the news process. So hacker journalists are journalists, and it is important to see them as such. Hacker journalism knowledge is critical for journalism to remain rele- vant—a defining pressure on the profession that helps create a subspecialty. Hacker, Programmer, and Data Journalists 83 Hacker journalists are needed to create something new in order to help journalism change to meet the needs of the digital conditions it faces. This need for abstract knowledge is critical to expanding what journalism can do in order to respond to the sociocultural, technological, and economic pressures it faces. New ways of thinking about and doing journalism are required, and hacker journalists help push the profession forward with their influence on the profession. Their counterparts, programmer journalists, or those who started first as journalists rather than programmers, help further expand journalism as a profession. Programmer Journalists: The Traditional Background Most journalists who are programmers do not come from the tech world. Instead, the majority of journalists who code have some sort of journalism background before coming to programming, either from journalism school or from working in a traditional newsroom with traditional duties. They have taught themselves to program and generally talk most about the importance of the story as central to their mission. This distinction, and the importance of identifying as a journalist first, is critical to their self-perception. Michelle Minkoff ’s tale of becoming a programmer journalist illustrates the flipside of the stories told by the hacker journalists, who began in a world immersed in code. Instead, Minkoff had virtually no idea that coding or programming even existed while she was an undergraduate English major at Brandeis. In an August 2014 tweet, she noted “2008! I was graduating Brandeis & blissfully unaware of code & future.”43 While a graduate jour- nalism student at Northwestern, Minkoff took a class that she describes as literally changing the future direction of her life. She recalled that her class, “Digital Frameworks for Reporting,” taught by New York Times programmer journalist Derek Willis, was so transformative that “life would never be the same again from that day” she entered the classroom.44 Her blog chronicles the steps she has taken to learn programming, from beginning efforts at learning coding languages, writing wish lists for programming journalism classes, and “thinking programmatically,”45 to her current and public efforts at going through far more complicated experimentation with “refactoring,” or improving code design, such as making it cleaner.46 She has gone on to have a storied career in just few years: she began as a “data developer/journo intern” on The Los Angeles Times’ data desk, went to PBS as a data producer, and now works as at the Associated Press as an interactive producer. Interestingly, her two most recent titles don’t have the word “journalist” or “reporter” in them. But Minkoff certainly sees herself 84 Chapter 3 every day in the process of producing journalism. To her, the titles she’s given—whether she is called a programmer journalist, a data journalist, or something else—mean little to her: “I fundamentally don’t care what it is called.... You do what your job is and you do what your job takes.”47 For her, though, unlike some of the hacker journalist types, the focus is distinctly in the service of the story, and she directly identifies as a journalist. “I am about the journalism.” She added, “No one sees the code, you see the analysis.”48 The ultimate goal is to “create a story through data and apps for people... to easily find things to zoom in to and drill down online.” In one of her posts, she described what it meant to her to be a programmer journalist—if she had to define it: “I don’t spend a percentage of my day in journalism and a percentage in programming—I spend most of my day in programming in order to practice journalism. I do other things, too: write documentation, pitch ideas, go to meetings, but always, always, always in the service of the journalism.”49 From the case of Minkoff, we see the progression of the programmer jour- nalist. These journalists often have no previous exposure to code—and often have backgrounds with no association with programming (they sometimes just start as liberal arts majors, like many journalists who didn’t go to journal- ism school). A number of these programmer journalists may learn some basics in school, but they continue also to teach themselves skills, experimenting with programming to make journalism better. Minkoff ’s focus on the journalism suggests a different orientation from the hacker journalists—for her, the prior- ity is always on the story, and her primary self-conception is as a journalist. Other journalists have similar backgrounds as Minkoff ’s, underscoring the difference between programmer journalists and hacker journalists. They, too, suggest that they have varying perspectives in approaching the work they do—with the major difference being the emphasis on thinking about journalism as it is traditionally idealized, though they also talk about adopting new attitudes and ways of thinking from programming into their work. The programmer journalist is often looking to use computational skills to work with data they have obtained either for themselves or from fellow journalists and then to help turn such material into stories. This adds to the messiness of defining journalists as programmer journalists, as data journalism also describes what many of these journalists spend their time doing. Backgrounds and Perspectives As programmer journalists talk about how they came to combine facility with code and journalism, we see a process of discovery that has developed outside Hacker, Programmer, and Data Journalists 85 the traditional technical route taken by hacker journalists who come later to journalism. Programmer journalists describe their backgrounds as either rooted in traditional journalism schools or in more humanities-based majors in college. They often tell a story of teaching themselves to code in their spare time. Many programmers are self-taught, too, and programming involves a considerable amount of new learning as languages change and improve, but these journalists are journalists before they ever become programmers. Sisi Wei, a journalist at ProPublica and formerly of The Washington Post, told me that she had been a journalism student at Northwestern whose only introduction to programming was a quick intro to Flash programming in a journalism class, though Flash is now obsolete (yet it was helpful to her at the time). Wei took one Introduction to Programming for Non-Majors class to make sure she was learning the right things,50 but mostly, she spent time “learning programming through extra-curriculars and summers.” In her observation, it’s actually a “common problem in the room” that most journalists are programmers without CS degrees and are self-taught, mean- ing they won’t “confidently say they are developers,”51 though she thinks this confidence issue is changing. Emily Chow, a programmer journalist at The Washington Post, calls herself a “journalist-developer.” She tells a story similar story to Wei’s, also coming out of the Medill program, and she joined Wei for an independent study with a professional programmer journalist. “I taught myself to code,” she said. “We had a one-week Flash intro.... I thought I might be a writer or a photographer, but people coached me.”52 These stories are compounded over and over by some of the leading figures in the interactive news world, including Aron Pilhofer, head of The New York Times team, who had been a computer-assisted reporter before morphing into the self-taught program- mer journalist he is today. The New York Times’ Jeremy Bowers also shared his experience in an email: “I identify as a journalist → programmer rather than a programmer → ­journalist because I learned to write software ad-hoc after getting my job at a newspaper and was a political science/English major in college.” This ad-hoc programming knowledge still means, though, that Bowers can spend most of his day working in code, as I observed him doing when I visited NPR.53 Some journalists have had a lifelong interest in programming but focused on building a career in journalism. John Keefe of WNYC “programmed for money as a kid” but became a straight-up traditional news director for the public radio station. In his spare time, however, he began thinking how interactive journalism should tell a story and tell it “really well in a way that you can’t... easily in word or sound.” He explained that he got as “good as 86 Chapter 3 he could get” and relied on conferences and the help of other programmer journalists in New York to show him what he didn’t understand.54 In fact, he recounted for me how he learned some programming from a few fellow New York journalists in exchange for Chinese food. And Danny DeBelius of NPR is a programmer journalist who never imag- ined he’d end up working with code. He studied journalism at the University of Colorado. “I didn’t have a lot of digital skills,” he told me. “I think I took one electronic [journalism] class... maybe it was the intro to HTML.” He interned at the Boulder Daily Camera and began the “unglamorous role of getting content on to the digital CMS” for two years until the terrible grave- yard shift so burnt him out that he went to work in a guitar shop. DeBelius got hired at the Rocky Mountain News to design special projects, such as election coverage in 2004, and the rest of his career took off. DeBelius told me, “I think most would say I’d fall into the programmer journalist category as I do spend a significant amount of my day writing code.”55 These programmer journalists spend much of their time working with data. We will see this focus on data and programming in chapter 4, where we will look at the workflow of interactive journalism teams. Their work involves either rendering the data ready for particular presentation or potentially crunching the data for stories. The raw material for much of their software is data, though certainly not all of it. They must know how to work with data and how to make it into a software product that has an unstructured but clear narrative. Often, programmer journalists have the goal of creating stories through interactives wherein software is the output and data is the input. At times, they may be called data journalists, and many programmer journalists might call themselves data journalists because of their intense work with data. However, I want to suggest that data journalists may be data specialists and far less focused on data visualization and interactive presentation—though they may indeed do some of this. And many programmer journalists may also fit this description as data experts. One can be both a programmer jour- nalist and a data journalist; or one can be just a data journalist according to the separations suggested by the evidence compiled here (data journalists spend most, if not all, of their time specifically working with the actual ins and outs of data). Thinking about Stories Like Michelle Minkoff, programmer journalists identify first as journalists, though their titles might range from interactive news producer to data de- veloper to news application designer. Many programmer journalists say that Hacker, Programmer, and Data Journalists 87 they are always thinking about the story first. They have some theories on how they are different from hacker journalists who focus on development first rather than working with journalism. Nonetheless, what they bring to journalism is a distinct approach to thinking about how to create journalism. Scott Klein described the attitude he saw as pervasive from the journalists whose background was not in programming. “Software is outside of them and journalist is inside of them.... There is much less magic to software development. They work primarily as journalists.”56 Within the professional tradition of journalism, the programmer journalists affiliate themselves with a tradition of working within the goals of first telling a story related to public information and then thinking about how they tell that story in code. They are journalists first, not coders. Sisi Wei, who works for Klein, added that the perception of her team was focused on being journalist as well as developer. “[You] are not just a developer.... You are thinking about things editorially.” Still, she pointed out that the idea of being a journalist was “probably what people cared about most.” Jeremy Bowers, who was deeply invested in solving the technical problems of the NPR interactive projects when I visited the news organization, defined in an email what he sees as the different ways of thinking about approaching programming and journalism. There are some projects that programmers → journalists won’t enjoy building because they do not represent an interesting technical challenge. Many journalism problems are not particularly interesting technologically (though of course, some are). The result is that journalists → programmers are interested in solving these “easy” problems that have high journalism value and programmers → journalists might be more interested in solving more complicated technical problems that have a lesser journalism impact. He added: Many programmers → journalists I know become obsessed with not repeating themselves and thus want to build “platforms” or other broad tools. As a journalist → programmer, I have a higher tolerance for repetitive software.57 So for Bowers, the distinction is that hacker journalists may be more focused on the origination of new technology. They may be interested in solving new problems rather than thinking about how to work within what they have already created—and may be more focused on technical solutions rather than on the journalism output. This difference in focus might be a point of contention between hacker journalists and programmer journalists, yet it seems to be more simply a distinction rather than a point of disagreement. 88 Chapter 3  Notably, it is also clear that ideals from hacking have indeed permeated the newsroom. Programmer journalists do speak about what they do in terms of hacking culture; the major difference, however, is that programmer journal- ists make comparisons between the programming process and traditional journalism processes, looking for links rather than positing differences be- tween the two. These programmer journalists talk about rapid prototyping and experimentation, but they frame this kind of approach to programming in much the same way one might approach traditional journalism. Sisi Wei argued that journalism has much to do with the programming pro- cess: “I guess things in programming go hand in hand with journalism stuff; there’s almost an analog with journalism. Journalism is a lifetime of learning and new information, and with programming you have to have the willingness to try new things all the time because technology is changing so fast.... You are iterating and it’s very similar to writing drafts and cutting things out.”58 However, she also emphasized that since technology is rapidly changing, and as the tech world introduces new products and services all the time, a pro- grammer journalist simply has to be adaptable. As we have observed in the process of moving to online journalism in traditional newsrooms, journalists are certainly not always the most adaptable to new technological challenges. Minkoff similarly agreed that programmer journalism seemed to be a lot like breaking-news stories. She pointed out that programming interactives fit the model of story writing used at the AP, her current site of employment. We have tried the agile process—and I think that is a lot about what we do in a news meeting: you get multiple iterations of news and that’s especially the case at the AP. The first thing that happens is that you get a news alert and then you have a quick story—the 70-word story, and then you write through it, and sometimes that’s very intense, and that’s a lot like technology, like I’m going to show something but it doesn’t mean it’s the final product.59 These journalists see the importance of rapid iteration. The agile software development process focuses on quick development and on releasing beta versions before a final product. The idea that you could release a product for consumption that might not be completely error-free—at least from the programming side—seems different from the idea of putting out a fast story. In the case of the AP, that fast story can’t have any errors—the quick first draft matches well with the idea of the quick first draft of an interactive, but there is more room for experimentation within building software. These programmer journalists much more clearly identify with journalism than hacker journalists. They are journalists first and then programmers. What they say about how they approach interactive journalism puts journal- Hacker, Programmer, and Data Journalists 89 ism at the center of what they do—the story comes first. On the other hand, some of their identity and ways of thinking about themselves and how they do what they do are distinct from previous kinds of journalism—a focus on experimentation and a work product based in code. For the first time ever, journalists equate the importance of knowing code with doing journalism. Programmer journalists work in both worlds in order to further the journalistic process. This provides evidence of the expansion of the profession as journalism adopts individuals with new skills. Similarly, these journalists embrace ways of thinking that are not found in traditional journalism. The focus on trying new things that require distinct adapta- tion has not been shown to be particularly prevalent among the majority of journalists, though they are trying. The ability to put something imperfect out for audience consumption, such as experimental beta versions, signals a departure from journalistic norms. These are commonalities, though, with their counterparts, the hacker journalists, signaling the consistency of both of these specialists within the larger occupational subgroup’s development. One might think that there would be some conflict between hacker journalists and programmer journalists in their slightly different approaches to thinking about journalism work, and while workflow does vary depending on who manages a team, as we will see in chapter 4, programmer journalists and hacker journalists work side by side, well-integrated into newsroom teams. Data Journalists: Uncertainty in a Definition To many people in journalism and academia, data journalism is an all-encom- passing term for the type of work described here. But it’s not entirely accurate. Not all interactive journalists are data journalists, and not all data journalists are interactive journalists. Data journalism can include numbers, statistics, names, categories, documents, and other kinds of information that can be coded and arranged in ways that computers can help analyze in the service of journalism. Howard’s definition of data journalism (reiterated once more) brings some clarity: “the gathering, cleaning, organizing, analyzing, visual- izing, and publishing data to support the creation of acts of journalism.”60 But data journalists should be broken out as a specific category, one that stands on its own. Based on the empirical data, data journalists are distinct; they are primarily working with data in the service of stories, actively trying to tell stories with data, and spend most of their time working specifically with data. Certainly, some of these data journalists may be programmer journal- ists or hacker journalists, who may also at times work with data, but data journalists employ data as their primary focus in their work—and many of 90 Chapter 3 these journalists do not see themselves primarily as working with code. In fact, some data journalists may not do any coding at all. Rather than being associated with programming, the term data journalism has its historical origins in the American term for earlier ways of working with data in the newsroom: “computer-assisted reporting” (CAR) and its practitioners, computer-assisted reporters. Data journalism is, in fact, an evolution of CAR. Colloquially put, working with data does not an inter- active journalist make; instead, for a data journalist to be included under the umbrella of interactive journalism, he or she has to be working on in- teractives. These terms matter, but let’s be explicit: data journalists can be interactive journalists, but they aren’t interactive journalists if they aren’t making interactives—this should make sense. If they aren’t making interac- tives, they’re probably a lot closer to CAR journalists, discussed below. And it doesn’t quite matter, in fact, whether data journalists can actually code—just as long as they’re creating interactives. This seems complicated, but when you read the empirical discussion and how journalists talk about themselves, it makes sense. Thus, it’s important to look at how data journalists talk about what they do and how they understand their work. Even though not all data journalism is interactive journalism, much of it can be included under this emerging subspecialty. As Alexander Howard points out, everyone in newsrooms uses computers: all journalism is computer-assisted now. CAR might be said to represent a point in time when data was used simply in the service of the story, whereas now, the data can be a story on its own. Data journalism seems to be a bet- ter term to describe and account for the rise in data and also for the more comprehensive ways of using computation and visualization to look beyond just anecdotes to overarching systemic analysis. Data journalism brings the entirety of the data set to the public, at least as much as possible, whereas CAR journalists would likely use internal databases sharing just key details for their analysis. As Howard and others point out, perhaps the public’s first introduction to data journalism in interactive form may be coverage of WikiLeaks. The Guard- ian, The New York Times and Der Spiegel all put out searchable, interactive databases and maps, allowing users to browse through the massive amounts of war logs and, later, diplomatic correspondence. Nate Silver, statistician wunderkind, relied on his data journalism skills to accurately predict the 2008 and 2012 presidential elections. In 2012, he drew 20 percent of The New York Times’ online traffic.61 Other prime examples, as Terry Flew and his colleagues suggest, include the disclosures in The Guardian of the expenses of parlia- ment ministers.62 Insight into nearly everything, from the 2010 earthquake Hacker, Programmer, and Data Journalists 91 in Haiti to data about campaign contributions, can now be found at many newspapers and some broadcast and cable sites across the world. Today, however, what some call data journalism differs from CAR report- ing in seeing the end product not as a story but instead as a “productive artifact” of “information filtering.”63 As Powers explains, the computer pro- grammer and the Web developer were long singled out as non-journalists and thought of as unrelated to editorial work; they were considered “different tribes.”64 In interactive journalism, the programmer and the journalist are brought together as one in the context of interactive journalism, moving beyond just using computers to enable journalism but using computers to actually create alternative forms of news. In data journalism, computers aid journalism, but the work may not be directed toward a visible, tangible, us- able output for news consumers. Data journalism is a contested term. Is everyone in the newsroom doing data journalism if they work with data in some way? Is there a level of sophistica- tion? In what way is their work a legacy of computational work? Journalists I spoke with tried to articulate the differences between past practices and the way CAR still works in the newsroom—as both CAR and data journalism involve using a computer to analyze data in the service of a story. Aron Pilhofer, head of Interactive News at The New York Times, explained in 2011:65 A lot of times, people doing CAR have skills very similar to folks that work for me [in the interactive news team].... X has exactly my background as a congres- sional reporter and did a lot of database analysis, what we would call CAR back in the day, and is now a coder. The difference is that interactive news is primarily focused on telling stories via web applications and making th[ose] public facing, using data to construct a narrative, or creat[ing] tools. He tried to articulate whom he saw as a data journalist within definitional constructs: [Who are] data journalists versus programmer journalists? In the last three years, a lot of terms have been coined. Data journalism is a loose term that back in the day meant someone who does CAR. Now the boundaries are much broader. What unites us is that we are building Web apps. CAR was fundamentally in pursuit of doing [written] stories. The Web app doesn’t have an obvious lead. I don’t think it matters what terms you use. You don’t have to be a developer to be a data journalist under this broad definition. Pilhofer brings out a number of critical points in this interview. First, he distinguishes interactive journalism and data journalism from the traditional 92 Chapter 3 forms of CAR. Interactive journalism, and today’s data journalism, is focused on the Web application. While data journalism certainly has a legacy in CAR, CAR was about a traditional story; data journalists can create tools or Web apps for public-facing data. Another key observation here is his distinction that data journalists do not have to be developers. This suggests that while data journalists are in- vested in a world of code, they do not actually have to be creating code. So programmer journalists may know code and may write in code, and some programmer journalists may consider themselves also working with data or even call themselves data journalists; data journalists do not necessarily have to work with code to be considered under the broader umbrella of “what unites us”—which is working to the Web application. Thus, in this rendition of the term, data journalism can indeed be different from traditional data reporting because it has this online element. Data journalist Matt Stiles at NPR explained his perspective on CAR versus his current work: “It is different because data is working on a spreadsheet versus working on JavaScript or Python for a project.... It’s not that dif- ferent from CAR but there is an online component.”66 Stiles distinguishes current data journalism from past CAR: old spreadsheets would be Excel, whereas today’s data journalists would use more sophisticated programming languages online. Head of the News Apps team at ProPublica, Scott Klein, a programmer journalist who works with data, went further in his attempt to delineate the evolution of data journalism: CAR in the 70s or 80s meant working with computers but not in an interactive way. You would look at a data set with a class prediction [for example] and go and find and write a story about a few of your examples. The newer path is to expose entire data sets to people in a clear and honest way—[asking] how do we show it to people, with analysis and stats and clean data.67 Today, as Klein went on to say, manipulating data sets is in the service of the whole body of data, “instead of finding two or three examples and working hard over weeks to find anecdotes” [as you would with CAR].68 CAR, then, was about showing just a few data points, and data journalism means mak- ing all data points available to the public, or making it possible, at least, to examine all data points in a systematic way. The efforts must be focused on “how do we show it to people” or rendering the data sets in a way that people can easily manipulate. The output is a news application, not a static story. In addition, old-school CAR was not interactive. Hacker, Programmer, and Data Journalists 93  Some journalists focus explicitly on what’s different about the journal- ism—and speak less about the interactive elements, suggesting that data journalism does not have to have anything to do with an online component. Instead, data journalism is an intellectual approach. Journalist Derek Willis of The New York Times gives this definition: “Basically, if a journalist works with tools that assist in interviewing, analyzing or conveying data for stories of any kind in any medium, then that person could be called a data journalist. I think it can be divided into roughly two camps, with some overlap: journal- ists who work with data for analysis, and journalists who work with data for presentation.”69 The first descriptor seems to describe much more closely the traditional CAR journalists, while the idea of data for presentation seems to more accurately describe the way I categorize data journalists here. Willis does not even speak about the idea of visualizing data in his definition. Some journalists have attempted to work together to standardize some form of a definition. At the 2011 MozFest (a media and computing festi- val hosted by the Mozilla Web company), journalists from Europe and the United States came together to create the Data Journalism Handbook. Paul Bradshaw of Birmingham City University wrote the introduction, which attempted to tackle the fluid idea of data journalism. In it, he began with the rejection of the idea that the definition of data journalism was simply journalism done with data. He noted that twenty years ago, journalists simply thought of data as a collection of numbers “mostly gathered on a spread- sheet.”70 But at the time, he notes, “that was pretty much the only sort of data that journalists dealt with.” The difference today, Bradshaw observes, are the “new possibilities that open up when you combine the traditional ‘nose for news’ and ability to tell a compelling story, with the sheer scale and range of digital information now available.”71 He adds to this definition that it involves programming or software to automate gathering data, combining information, or find- ing connections between “hundreds of thousands of documents.”72 Data journalism, he notes, is “often told with interactives—and the data may be a source or a tool to tell stories.”73 Every example in the book he offers is an interactive. But data journalism is not always about interactives, according to this handbook. Despite disagreement around the definitions, the guide is significant in its own right because it signals an attempt to standardize and codify the subspecialty by creating a sense of accepted practices with the larger goals of furthering this kind of journalism. Notably, in Europe some journalists tend to use the term “data journalism” to describe interactive journalism as a whole, not just activities associated 94 Chapter 3 with processing data. But depending on the organization, one could be called a developer or a data journalist and still be working on interactives either way. According to Pilhofer, an American, “Data journalism is just what they call interactive journalism over there.”74 Yet this is not always the case. In some newsrooms, such as Zeit Online, data journalist Sascha Venohr explained to me that “developers” did the programming, whereas he did nothing with code and instead “worked on interactives with data,” a distinction that seemed to be more and more con- fusing as I asked him for clarification.75 At The Guardian, data journalists kept the DataBlog and worked with publicly available data, writing posts. The “developers” rendered the complicated interactives for major projects and stories such as the Olympics, yet this kind of work still required data— but these individuals were more skilled with code. The BBC mirrored The Guardian’s distinctions, with “developers,” “designers,” and “data journalists” on its News Specials team. This suggests the terms are just as confusing in Europe as they are in the United States. Whether all programmer journalists are data journalists is a matter of degree and depends on self-perception and perceived amount of working with data. Almost all U.S. journalists I spoke with who were programmer journalists said that they might call themselves data journalists if the term were loosely applied. Many of these programmer journalists are working in what Fink and Anderson characterize as key components of data journalism: data procurement, graphic design, and statistical analysis.76 But many are quick to point to the “the real data journalist,” thus identifying the journalist procuring the data or doing statistical analysis as distinct from programmer journalists or hacker journalists. Data journalists are interactive journalists when they distinguish what they do by having this additional online component. Findings here confirm that data journalism is difficult to define and comes out of a larger tradition of CAR. When there is focus on Web apps, software, and/or interactivity, data journalists are connected most directly with the larger subfield of interactive journalism. Not a Coder Pilhofer helps us to understand that not all data journalists are programmer journalists, as not all of them code. Some data journalists can’t code at all and instead focus their time on cleaning, sorting, and interpreting data. Others use less sophisticated tools to build interactives. They remain integral to the Hacker, Programmer, and Data Journalists 95 construction of the subfield, though. Their work still relies on additional code layered on top of an existing Web site. The code they use is often already programmed within existing software. They could not create interactives without working in this world of code. Simon Rogers, a journalist at The Guardian and head of DataBlog, de- scribed himself as a data journalist. But he doesn’t code and instead relies on existing libraries and templates to create visualizations for his work. His goal was to make data more easily available, searchable, and interactive for users. He told me: “After 9/11, I began collecting a lot of data sets, and I asked if I could have an open platform data blog. I was hoping to see if we could give the collected information of the paper via a data blog to give it a longer life.”77 However, he emphasized that he was not a programmer or a coder in any way.78 But he was using the free tools that made it possible to create the kind of interactives—primarily the free Google Fusion program, which offers a variety of options for creating interactives where the code is already complete, from charts to maps, and relies extensively on spreadsheets. Other noncoding data journalists may work with tools as simple as Microsoft Excel or Access, or tools that enable interactives like Tableau, Carto DB or Data Wrapper. Some journalists do learn the basics of code to help input data, scrape data from Web sites, adjust interactives, or become more proficient and begin on the road to programmer journalism. These journalists don’t work with sophisticated code, but they see what they do as critically integrated into a traditional reporting project. Brian Boyer called his staffer Matt Stiles the “real data journalist” on the NPR team. Stiles explained why: “I have a reporter sensibility. I am the guy with a phone on my desk. I am the one with the FOIAs. I am the reporter getting the data.”79 He noted: I wouldn’t call myself a coder. I can write some basic code in JavaScript, Python, SQL and R, but I’m not a programmer by training or skill, like the other members of the NPR team.... I do make interactive graphics, but I can’t be trusted to build a complex Web application. I’d need help with that. I’m also slower than a lot of “coder/journalists” at even basic code like CSS or HTML. The difference between me and (some of) them is that I got ten years of traditional news reporting experience before I started hacking a bit.80 Stiles brought up a number of key points in his self-assessment. He sees himself as a data journalist, and he does define interactive graphics as part of his daily work. But he doesn’t do any sophisticated coding, though he 96 Chapter 3 knows at least some code. He sees himself as having and using traditional journalism skills to do his work. Other journalists agreed but noted some differences in their expertise. Mona Chalabi, a data journalist who works on The Guardian’s Reality Check blog and renders her work in Google Fusion for visualizations, explained what she does this way: “As a data journalist I suppose for me the way that I explain it to other people is that I work with numbers—that’s what concerns me.”81 She said that she visualizes the data to help make it easier for people to engage with the data, though not all of her work is always interactive or always visualized. Each year at the NICAR conference, data journalists can learn more code— the conference program now has a serious programming component in- tended to help data journalists become more proficient with using code to both analyze and present stories. For instance, the 2014 schedule included a workshop that promised: Give us four hours and your laptop, and we’ll send you into NICAR with a fully- functioning data-crunching machine and the knowledge to use it. One of the biggest hurdles to learning programming is the often-bewildering process of setting up your computer. Promised setups included the following programming potential add-ons: Participants will walk away with a fully functioning dev machine (on their personal laptop) that includes: VirtualBox, Ubuntu/Xubuntu, csvkit, Python, Git, Django, SQLite, MySQL, PostgreSQL, PostGIS, PANDAS, Ilene, virtualenv/virtualenvwrap- per, QuantumGIS, Node.js, NPM, Ruby, Rails, RVM, Bower, Grunt, Fabric, Yeoman, CIR news app template.82 Other sessions included: NewsCamp, getting started with data viz, a mapping mini boot camp that promised: “ArcGIS, QGIS, PostGIS, TileMill, GDAL, GeoDjango and probably a few more acronyms for good measure”; A PyCar mini boot camp, which advertised: “[We] teach journalists basic program- ming concepts using the Python language.”83 Thus, we can see that there are no bright lines between programmer jour- nalists and data journalists, though programmer journalists do spend time fluent in code and journalism. Data journalists identify as working closely with the analysis of data, and see themselves as concerned with the presen- tation of this data for the public. And when they express what they do, we see how they associate data journalism within the umbrella of interactive journalism. Not all data journalism is interactive journalism, and not all data Hacker, Programmer, and Data Journalists 97 journalists are interactive journalists. However, this close association with programmer journalists as well as the story output and goals of interactive journalism signals the importance of including this subgroup in the larger understanding of interactive journalism. Data journalists have the closest legacy to traditional journalism. Com- puter-assisted reporters have been part of journalism since the 1960s. Their work with data has been included in many award-winning stories, and con- current with the rise of graphics, in graphical representations as well. Data journalists work with categorical, numeric, and document-based data across all vectors of social life, rather than on relying on qualitative accounts for journalism. The emergence of this subgroup of reporters signals an expan- sion of the profession as journalists have needed new skills to account for more sophisticated data. Increased computing power has given these data journalists the tools to not only process more data but also to work in an interactive environment—without knowing more code. Rounding Up Definitions and People The goal of this chapter was to help elucidate the types of people who do the work of interactive journalism. To understand the formation of the subspe- cialty of interactive journalism, it’s important to understand the backgrounds and self-perceptions of the actors involved in carving out this new element of journalism. We find three key groups: hacker journalists, programmer jour- nalists, and data journalists. These categories work more as Venn diagrams than they do necessarily independent categories. In some cases, backgrounds are similar; in other cases, ways of approaching journalism are similar. Hacker journalists come to newsrooms from programming backgrounds. They often work primarily in a coding capacity, though some do share A-1 bylines and work with data analysis. They may not identify as journalists but are called journalists here because they must know how to communicate in a newsroom workflow, build products for acts of journalism, and are integrated into editorial workflow. Programmer journalists are as fluent in code as they are in journalism, but they started with no background in coding and identify first as journal- ists. They speak primarily about their focus on the story, often work with data (though their manipulation of data may not be as sophisticated as that of data journalists, who spend more time with data), and their output is primarily interactives that serve stories on their own and complement traditional news stories. 98 Chapter 3  Data journalists may sometimes be grouped under the programmer jour- nalism category but also may be journalists who are not fluent in code. These are the kin of the old-school computer-assisted reporters, who relied on computers to produce a few key insights for stories. Now, data journalists work to reveal far more data and often use computer programs in the ser- vice of interactives. Their focus on an online component ties them into the larger subfield of interactive journalism, though not all data journalists will be interactive journalists. And it should be noted that at some large news organizations, like The New York Times, the graphics desk works on inter- actives, too, and some of these visual journalists know incredible amounts about data and code (including “Snow Fall”), and some visual journalists argue that they specialize in data visualization but cannot program, even if they do a little, like Alberto Cairo, interviewed in this book. It is helpful to have conceptual categories because they map how journalists talk about themselves. Self-perception gives us insight into how journalists ultimately see their jurisdiction over work. These divisions help us understand similarities and differences in ways of thinking, which ultimately helps us determine how these backgrounds contribute to the development of abstract knowledge that expands the profession. There are particular aspects, such as the various ways journalists came to programming, that do suggest some natural divisions of how they describe their own backgrounds and approaches. Similarly, these distinctions help show how the subspecialty is negotiating definitions as it develops internal cohesion. When it comes to working to- gether, differences among these smaller subsets of interactive journalism complement and enhance the functionality of the product and process. Some journalists reading these descriptions are likely to resist these lines I have drawn, however faint or permeable some of the divisions may seem to be. Journalists may likely want to claim that they have all of the qualities that I see as actually distinct among these groups. What hacker journalist doesn’t want to say that they think about the story? Programmer journalists and data journalists of course want to claim that they are deeply inspired by program- ming perspectives, and indeed they are. But there are differences, and these differences, while perhaps of degree, do reveal themselves in the data. It matters to include hacker journalists in the discussion of “Who is a journalist?” because hacker journalists need to be recognized as included in the larger development of the subfield. Their role is integral to creating the work product of interactives. It’s important to recognize programmer journalists because it underscores how journalists are now learning different skills and approaches beyond what any journalists ever before have consid- Hacker, Programmer, and Data Journalists 99 ered journalistic work. And the discussion of data journalism reveals how dealing with data in the newsroom has evolved and how this is integrated into the larger subfield. The rise of the subspecialty of interactive journalism becomes clearer as we define these different actors within this new form of journalism. We see different ways that external pressures may influence each group. Recall that journalism is challenged not only by economic pressures but also by advances across the development of a more sophisticated and engaging Web, which underscores the importance of making the journalism profession expand and adapt. There is little contest that emerges between professions. Programming claims no jurisdiction on this public-service enterprise to provide individuals with knowledge about their communities. Rather, journalists take skills and thinking from programming and apply this without any sort of evidence of internal professional tension. This signals a departure from traditional no- tions of the development of professions, where there is latent contest between different ways of thinking. It may well be that interactive journalists are ac- cepted into the newsroom with little resistance because they are viewed as so critical to change and the future of journalism. Newsrooms appear open to these changes, even if they suggest at times a somewhat different approach to news work. And internally, journalists within traditional journalism are not approaching these coding projects and data interactives asking to do them or hoping to take over these roles; rather, the backgrounds distinguish these individuals as having skills and perspectives no one else can claim. Similarly, an acceptance of the dominant norms of professional journalism across these new kinds of journalists may help integration—their work in the service of a story, a public-service orientation and aspiration to work, and the belief in traditional journalistic practices, even if the output is an interactive. We see how journalism is expanding its professional domain as journalists take up code. These journalists help the profession make a claim to relevance. Never before have journalists—editorial members of the newsroom—taken up code in these larger numbers and in this significant a way to the extent that it is literally changing the form of journalistic work. This development makes hacking relevant within the newsroom and makes journalism more adaptable to the demands of the digital environment. But it is important to see what, exactly, these journalists do inside newsrooms to create this kind of new journalism. To this end, we look at how journalists work and how they explain what they do. 100 chapter 3