The session will start in five minutes.
The session is starting shortly.
The session is starting shortly.
The secret history of the CIA’s interrogation program.
An almost unlimited role of dark operations.
Chapter and verse of what happened.
I will come back to that. I wanted to make sure that worked. All right. Everybody ready? Does somebody want to close the door? That’s a lot of noise. All right. Welcome, everybody. This is talking about every part of the pig. Let me explain that phrase a little bit. Anybody familiar with that phrase? Know what it means? Use everything you have. Use everything that’s available. So I’ve highlighted a couple key phrases in this session description here. Just because that’s what we’re going to end up talking about. There’s a lot of us in here, so I think we’ll do breakout groups. And we’re going to be talking about tools, practices, culture, in the context of open notebook reporting and structured journalism. Is everybody okay with the live transcription happening right in front of you? Is that super distracting to anybody?
Yeah, but it’s awesome.
(laughter)
Fair enough.
(laughter)
All right. Where is my mouse here? There we go. So… Some definitions to start with. We talk about open notebook reporting—it’s the idea of—it’s really about a goal. It’s the goal of trying to put as much of the raw material of your reporting out into the public. Whether you do this for transparency, you do it because it’s a way of creating more content, maybe you have other reasons. Whatever your reasons are, it’s the practice. And then structured journalism—some of you may have heard me talk about structured journalism before. Some of you may have heard me talk about structured journalism incessantly. It’s the idea of taking your reporting and organizing into a logical structure so you can reuse it later, rebuild it, things like Homicide Watch. Like… Give me some other projects. Now I’m a blank. Fire Tracker out in LA, if any of you are familiar with that.
Would any of the Vox stuff fit?
Yeah, I think of Vox that way. I actually think of Vox as having two structured journalism projects sort of at the same time. Their card stacks and their story streams. If you think about how… What’s their sports site called? SB Nation is organized around teams and players and sports. That’s a way of structuring content. So I think of it that way too. So this is… This session is not totally structured journalism, but it’s within that universe, and that’s where I’m coming from, when I’m talking about it. So these are points of focus. Tools, practices, and culture. And this is really going in order of what I think of as easiest to hardest. There are a lot of good tools out there. We make lots of tools. I make tools. Some of us in this room are just—we’re constantly making new tools and new ways to do journalism. That is in abundance. Practices are a little harder. Practices are about habits. And then culture is the hardest one. I forgot I wrote definitions for these. So practice is about habit. It’s what we do by default, in a lot of ways. And then culture is the bigger stuff. The more amorphous things. The—what do we want? What do we actually believe as organizations? As individual journalists? What are the things we’re striving towards, and what are the things we repel against? Or what are the things we—that make us uncomfortable? All those are culture. And so we’ll talk about that last. So I want to set this up with an example that I’m thinking of, but I’m going to have you all do your own examples. So this is a trailer for a film that Frontline did a month ago. And from one of our filmmakers, called Kirk Documentary group, that does a lot of our politics films. So I’m just going to play this so you can all get a sense of what a—when I talk about a big media investigation, this is a lot of what I have in my head, when I’m thinking of it.
I scribbled down after the name interrogator
The secret history of the CIA’s interrogation program
An almost unlimited role in terms of dark operations.
They called it torture.
Chapter and verse of what happened.
In other words, the CIA was lying.
The CIA said it was necessary.
We were at war. Bad things happen in wars.
Frontline investigates secrets, politics, and torture. Tuesday at 9 p.m. on most PBS stations.
So heavy stuff, right? So… I want you to put yourself in the head space of… You have some role on an investigative team. You are a reporter, you’re an editor, you’re a programmer, you’re a designer, you’re a web producer, whatever role you feel most comfortable in. And I want you to be thinking about a project that you’ve worked on, that is roughly this scale. Or something big. Something that is significant to you. And I want you to use that as sort of an example project, as we go through a couple exercises and a couple breakouts. And ideally, I’d like to have somebody in each group who is—knows the reporting of a project well enough to actually talk about the pieces of it. So let me just check the room real quick. How many of you are reporters? How many of you are editors? How many of you are programmers? How many designers? Wow, we have a really mixed group. Great. How many of you have done something recently in the past six months or a year that is, like, big, major investigation, throw all of your institutional weight behind it?
Our newsrooms or us personally?
Either way. Something that you were like… This really better be big, because I have just put, like, a good chunk of my life and soul and blood into this. All right. Actually, show of hands one more time so I can see. Oh, wow. So… Most of you. How many of you are beat reporters? Think of yourselves as you do… Iterative… And I’m in that category too. By sort of default nature, culture, whatever. Even though I tend to work on these things more now. So that’s good too. A lot of my thinking about this comes from beat reporting and thinking about more iterative processes than, like, the throw everything at one project. So just so you all know where I’m coming from.
So Frontline does these. These are—every film takes about six months. And costs lots of money. And there’s a ton of reporting that goes into these. Like this much. So this is… It’s hard to see how much this is. This is the CIA timeline. That is the working document from Kirk Documentary group. I’m not going to show you anything that’s inside it. But to get a sense of what it looks like when they produce… This is what they work from, as they put their film together. This is Mike Wiser, sitting behind one of these. He’s one of the producers who works on these. The CIA one is about 700 pages thick. So… There’s a lot of stuff. And a lot of it never sees the light of day. Sometimes for good reasons. Sometimes just for other reasons. And when I was thinking about this session, my big question going into it is: How can we use more of what’s in that? How can we use more of those 700 pages as content? As something that our audience can consume and interact with, tell us we’re wrong about, tell us we’re right about? How can we use every part of the pig? So… That came out more pink than I expected.
I’m going to have you all work in groups. And this is gonna be something of a group brainstorm. And we’re going to go through tools, we’re going to go through practices, we’re going to go through culture. I want you all to break into groups and figure out an investigation—sort of an example project that you can work from. Have somebody in the group think about a project they’ve done recently that they can really talk about the process of it. They can talk about the reporting. Because I want you all to be thinking about, like, the stuff that you—that didn’t work. That didn’t fit, or that you didn’t know what to do—that you had, that was good, but you didn’t know what to do with it. So ideally I’d like somebody in each group who can talk about those and talk about the investigation in detail. If that makes you uncomfortable, to talk about stuff that your organization didn’t publish, either find a different example or… Find an example where you actually feel like you did use a lot of that stuff, and you made more use than you otherwise would have. Does that make sense to everybody? If that doesn’t make sense, that’s okay. Because… I’m used to talking about things that don’t make any sense.
So this is our first exercise. Assume that we’re going to talk about tools first. As you’re going through this, assume that no one will say no. Assume whatever cultural barriers there are, assume they weren’t there. Assume that any time anybody says—well, that’s hard—assume that would go away. These are assumptions you can’t make every day, but I want—for the sake of this exercise, I want all the nos to go away. But take a note of it. Take a note of where—in your organization, where you’ve worked, somebody would say—no, we can’t do that. We don’t have enough people for that. We don’t have enough money for that. Because we’ll come back to those. But for just this, think about the yeses and make everything a yes. So… We’re going to do color coded sticky notes. Are you all comfortable in the tables that you’re in as groups? Any group not have… Does every group have somebody who can talk about an investigation in detail? I realize none of you know each other, probably. So… We’re going to do sticky notes. Because I always like to brainstorm on sticky notes. Somebody from each table, come up and get just a wad of sticky notes. Just yellow for this group.
How is everything going? Good? Take, like, three more minutes to write down just your ideas, blast out any last, like, any of the things you’ve been circling around. Discrete tools. And then we’ll come back up and stick them on the board.
(tapping microphone)
Okay. Who’s ready to report back on tools? So for each group what I’d like you to do is have one person come up and very quickly give us, like, a one sentence synopsis of the story you’re talking about, and pick a tool that you talked about, and stick it up on the board. One or two. Anybody want to go first? I don’t know if the mic is working.
So I’m actually using an ongoing project, just because I thought we could get something out of it. So just to be quick and careful, are there any Canadians in here, besides… Oh, okay. Just because I work for a Canadian news organization… Cool. If you don’t tweet or talk about this project.
(going off the record)
We can go on the record for tools. So talk about how to tell stories with maps, using sort of… Timelines. Whether that’s literally over time, or age or whatever. So you can see sort of some of the trends, longer term. Because as much as we want to dig into the individuals, and we’ll allow for that, but also… I forgot one. But one of them was sort of in the dream world—a database that is really easy to update. Because we’re going to keep getting more information after this project launches. We want to it to be easy enough to encourage all these reporters to get into it and keep it growing. So sort of on that—continuing with story mapping. Building a narrative around it. So there’s the night lab tool, but what other tools we can use to build a narrative around it. Sound sites and other audio tools because we want to focus on the storytelling from the communities themselves and let them speak for themselves as opposed to always being that voice in between. So we want to come up with some cool audio tools.
Who’s next? Try to do…
Yeah. We talked about this in the context of a project where we looked at a whole bunch of reports about child deaths. So putting them online with document cloud was a big one. And then we also talked about this concept of normalizing formats, often through manual labor, so taking email files or whatever it is, putting it in a format that’s text that can be put online or searched by people.
Awesome. There’s that. Who’s next? Justin and then…
So a project that I worked on last fall at my last job… Was looking at alcohol on college campuses and around college campuses all over the country. And part of that was gathering data on every liquor licensee in the country. Or that was the goal. We didn’t actually get to all of the states. And we ended up not using a lot of that data, mostly because we had trouble matching it to data about the locations and size of college campuses. So we ended up scaling down the project quite a bit. But that meant we had a lot more data that we wanted to get out there, in case other people could use it. So ideally there would have been some tool or process to help distribute that, document it, and send out updates to it, as we understood more of what we were dealing with and received more of that data.
Let’s do one more, and then we’ll move on to the next round.
Oh, sorry. I’m sorry.
No, no.
Great. Briefly going off the record.
(going off the record)
(going back on the record)
So it’s a lot of audio transcripts, and some photos that were taken by reporters. But it’s a lot of audio. So what we talked about were—and reporting notes too. So it’s a large—we talked about a… Large, persistent, secure structured storage place for text, for typed notes that could be shared across a news room, something hopefully more structured and more share-friendly than Dropbox. Also more secure. Doing presentation and visualization of text and audio combined, which I think falls into soundsite-related stuff. Since it’s a whole bunch of interviews, natural language processing tools, and that ties into corpus analysis, which ties back into presentation and visualization. Thank you.
All right. So… Everybody else who didn’t come up… Come up real quickly and I just want you to stick your cards on here, and then we’ll have you present your story in the next round. So the next thing is… If I can get to it… Practices. So this is more about—assuming you have the tools. What would you want to be doing? What are your habits? What are your processes? And get a blue… Teal… Set of sticky notes when you come up. Does every table have some blue sticky notes? Or teal? All right. Go to it.
(breakout sessions)
So one more thing. As you all do this, as with the last time, assume that everyone will say yes, but keep in mind where people will say no. Where people will say—oh, that’s too hard. Oh, we don’t have people. Oh, we don’t do that. What are the roadblocks? Keep in mind impediments and then set them aside in your notes. And we’ll come back to them.
(breakout sessions)
(tapping microphone)
All right. Who’s ready to report back? I want to take the tables that didn’t get to report last time. One of y’all… Who wants to tell us what story you’re working on, as your example, and then give us some practices that you want to incorporate into your reporting process? You look like you’re ready.
We can go.
So one sentence description of the story.
Um… So ours is the undue force Baltimore Sun story. You guys may know about it.
Awesome story.
So a couple of things about the data—they got a lot of documents back. So a couple things that we suggested are things that we can do is—just have a general timeline of when you want things done. Do your research as a reporter—you should be doing your own research, but also do your research on what others have done, so you’re not repeating the same thing, but also you can either use that as supplemental information down the line if you need it… Also keep notes while you’re reporting, and share your internal knowledge with others. So reporters should share that knowledge with editors and editors should share that with producers, and with the developers, and so on and so forth. So communication. And annotating documents before you publish. Don’t wait until the last day to say—oh, we have to upload 500 pages to DocumentCloud. Can we annotate them and put them on the website today? Or, like, two hours ago.
You can totally do that, right?
Yeah. And that’s it.
Excellent. Who’s next? Who has not presented yet? Yeah, you guys were next up.
So at my last job, I worked—I was at the Washington Post and I worked on the Edward Snowden NSA documents, and one of the issues we encountered was there was a real lack of communication even within the team. It was kind of highly secretive, for obvious reasons, but it also hindered our ability to work efficiently. Especially on my end, where I was, like, working on the digital presentation of a lot of the documents. And I think we were talking, and some way of, like–some kind of tool that allowed a reporter to structure notes, interviews, and research, and, like, share them with editors and the other people who were working directly on the project—while they’re in the reporting process—would have been helpful to give us a bit more time, because we were always scrambling at the end. And redacting the documents before handing them off to us. In terms of something like… That a reporter should have been doing… Would have saved a little bit of a headache at the end, when something went online that should not have. So…
Thank you.
Yeah.
All right. Who else has not presented yet? Yeah.
I’ll just talk while I’m walking to the board. But we’re kind of piggybacking on two different stories, but pulling out the best of what works for the workshop. But in this case—the Chronicle did a long-term project on gentrification in the Mission. So there are a couple of practices that we used, and also referred to other media companies that might have used similar things. But long-term analysis is a key kind of thing. Partnership and collaboration can extend your news report. So you have, like, limited resources in your newsroom, but through partnerships, you can really continue yours news report long-term. So that was kind of cool. And immersion—so, like, there’s an example in LA where they set up shop in the neighborhood for, like, two weeks, and interviewed townspeople and stuff. So that’s, like, a really good example of a practice, I guess, to extend the news report.
Cool. Does anybody else not… You all.
So I work for an alt weekly in Vermont. Called 7 days. And we just recently did a story about… So our state runs hard alcohol sales. So we were able to get all of the data. Like, the quantities of every single—you would be shocked how many stores sell a ton of Fireball and 5 o’clock vodka. So we got all of this data. All of this state salary data. All sorts of things. And we had meetings—several months in advance—and then a month in advance and kind of gradually over the course of the project. And then a reporter a couple weeks before the story ran jumped in. And started doing the reporting. So we had this cool map. We had a bunch of charts about alcohol sales. And then he had this whole story about whether it makes sense for the state to be running this. But the charts and the data didn’t talk to the story at all. And a lot of people really liked the map. Because they wanted to know what their hometown liquor store was selling. But it didn’t add to the story. They were cool. In kind of… Separately. But they… So I think as the digital editor… Being in more constant communication with our reporter, who’s not all that used to working with data stuff, he doesn’t do a whole lot of, like, Excel crunching or whatever—being in more constant communication about kind of what he was looking at, and what we could do to kind of add to that, and sitting down with him. Because he was the one who knew how the story was being crafted. So larger meetings weren’t necessarily effective, because he wasn’t quite—he wasn’t at the point where he was even crafting the story when we were meeting. So constant communication. Maybe via email. And finding some comparable stories in advance, I think. And sharing our expectations and hopes for the project overall might have helped to steer the project.
Okay. Thank you.
Is there anybody else who hasn’t—any groups who haven’t presented yet? I can’t tell what you’re pointing at. Is this is the last one? Good.
So I used to work for an investigative team for the Palm Beach Post, for my last job. So we did this project on police shootings. The reporter started working on it a year or two ago. Long before the Ferguson shooting happened. But the problem was—and I think it’s kind of a common theme—the reporter was going at it alone. The team was not involved. And he had all sorts of written notes, which we—the tool we really wanted to collect was convert all these reports into some sort of structured data that was query-able and can be easily understood and shared across the team. So some things that we came up with was—to have sort of a quality checkpoint, milestones to make sure our data assumptions were right and have a more directed public records request than get all that we can find. Have better organization. And have some sort of structured notes system to share across the teams. That was about it. Oh yeah, and of course involve all the people required up front.
I’m sensing a theme here. Okay. So everybody has somewhat presented? You all haven’t? Okay. Do yours real quick and then we’ll get to what’s next. Where we’re going.
So we went a bit meta. And one of the things we wanted was a sort of metadata editor who sat between the people doing the writing and the people doing all the research. To collect all the reporting that was being done, and sort of catalog it in a standardized and structured way. So when everyone else is going through and using the same research, maybe producing a new story off of older research, that all the metaconsiderations around whether or not—say, someone decided not to use this source, because they thought it was iffy—that the second reporter goes—oh, maybe I should take a different look at this, or find another source. And stuff like that.
Awesome. I love the idea of a metadata editor. All right. Come to the end. The really hard part. So a friend of mine likes to say that culture eats strategy for breakfast. Culture is—strategy is what we mean to do. Culture is what we do without thinking about it. It is deeper than habit. It is, you know, what’s in your gut. It’s the stories we tell in our groups. So… For this last exercise, I want you to think about all of the impediments for tools and practices that you’ve kind of set aside. And I want you to think about… As you’re starting this story that you’ve been talking about, whatever story it is, how you would frame the mission of that story internally and externally to overcome some of those obstacles. Your goal is to use every part of the pig. Your goal is to use as much of your reporting as is reasonable or as you can. And I want you to write one or more mission statements. You can write it in bullet points. You can write it in (inaudible) language. How many of you have written a mission statement for a story or a reporting project before? It’s a weird thing, but it’s… Any of you read Peter Clarke’s Writing Tools? He talks about this a lot and I borrow from it a lot. It’s the idea of… Before you start, set aside… This is what I want this story to accomplish. This is what—or what I want to do with this. So I mentioned one of the projects I mentioned— Homicide Watch. That was a site that I helped run for a long time. So we had a big grand mission for the site. Right? Mark every death. Remember every victim. Follow every case. We also had this other mission that we talked about a lot. It was part of our open notebook approach. And my wife, Laura, who was the main reporter on this in DC, would say—everything that I would have on my desk, in my notebook—should find a way onto the site. And that’s really what informed our approach. In terms of what we—how we built the site. And some of the tools we made available. So we had a courts calendar because we kept a courts calendar. We had to know when we showed up in court and made our calendar public so other people could see when we were likely to have stories coming. And the (inaudible) scheduled their stories. We kept a documents library so that people could read those and we could go back and find—I think this was in a charging document a while ago. Let’s go back. It’s already searchable because it’s in Document Cloud. Thanks, DocumentCloud people. And we kept a structured database. It allowed us to make more of our content public. So anything on my desk or in my notebook should be on the site. That’s not going to work for everybody. It’s not going to work for the Snowden files, obviously. Probably not going to work for secrets, politics, and torture. It might not work for police shootings or some of the stuff in Baltimore. But there is a standard that you can find. And what I want you to try and do is think about what you would tell the reporter, the editors, the designers, the programmers, the web producers, at the beginning of the project—like, let’s make this our goal. And try to articulate that in something that will fit on a sticky note. Or many sticky notes.
Just as a response to what you were just saying about—that approach wouldn’t fit for all these projects. You can always make that your goal after the launch. Especially if it’s something like—a couple of people have done these police-involved shooting databases. So until you launch it, it’s not public. But you prepare it in a way that it can become public and be a sustained resource—it’s a great—
Yeah. And that’s a really good point. And that’s actually a good way to think about it. Are you putting all of this online at once? Are you trickling it out before you launch or after you launch? Is your main thing going to be in the middle of your releasing things? Think about time. Think about what you release. Think about how you release it. Is it processed? Is it like… Are you releasing… Are you writing stories based on all your data, all your notes that you have? Or are you just dumping things on the web and putting things on Github, like the (inaudible) project does. Or 538 does? So think about that. So we don’t have much time left. So I want to let you get to it. Write your mission statements. Be bold. And if you have… If you need help, let me know.
(breakout sessions)
And if you want to put your practices up on the board, do that too.
(breakout sessions)
All right. Who’s done? Who’s still working?
So… I don’t want to keep you too long, because we’re actually five minutes over time. But it sounds like there’s great conversations happening. So if you want to keep going with this, please do. We have 25 minutes until the next session. So… There might be beer outside, though, so I don’t want to keep you from that. Priorities are important. If anybody, real quick, wants to share part of their mission statement, just stand up and do it where you are. And if you are feeling really bold, you can maybe tweet something. I did say to keep it short for a reason.
I’ll do one. Okay. So we came up with three bullet points, two of which are closely related. First one is to assume that anything you do may be used in the future. If you’re a reporter, you should not assume that once you’re done with the story all this reporting is going to be thrown in the trash. And if you’re a developer, you should write documentations that tell you how to redo what you already have figured out once and are likely to forget. And then the other thing was to establish some sort of work flow for everybody at the outset of the project. Which enables, like, some sort of searchability and structure throughout. So if you’re a developer, you can tell…
Forget the so if, if you don’t mind.
No so if.
Appreciate it. Because we’re out of time. Go through real quick—mission statement.
Help reporters finish the sentences the data started.
What else?
Just wanted to—know the victim, know the perpetrator, know the cause.
Okay. Who else?
Publish as much as you can. Do it with care.
Who else?
Treat every part of reporting as if it was going to be published.
I like it. Who else?
Edit well, honor the story, only use content that’s germane, but be relevant to all your audiences.
Everything published and push the story forward.
Cool. Do you have something? Anybody else? All right. Thank you, all.
(applause)
Go forth and have beer!