 Welcome everybody. Mike Entany is here today to chat with us about press freedom and network journalism. Mike is a postdoc at Microsoft Research until late June, a fellow at the Bergen Center and he will be joining the University of Southern California in August as an assistant professor. So welcome Mike. Thanks. Thanks everyone. I'm going to stand if it's a little easier so I can actually see everybody when I'm talking. But yeah, thanks very much for your invitation to talk today. What I wanted to do is actually describe a project that I've been working on for a couple of years and it's around this idea of what press freedom might mean in an era of network journalism. So a quick way to think about this is there's a very different landscape of production of speech online especially. So one of the starting points is is there a meaning of press freedom that we can think about being present in all of the journalistic network practices and what my press freedom mean in this sense? Is there a difference between the speech clause and the press clause today? Is there a different identity for the press as distinct from people who are speaking online? So what I want to do is ask you to just hold on to those ideas a little bit as this idea of press freedom, this idea of network journalism and walk with me through this idea of what it might mean. So just to set the context a little bit and I'm sorry maybe a little hard to see with the lights but this is not surprising or new to anybody in this room but just this idea that these are stats from the Pew Center for State of the Media report. Basically online is just dominating the growth of news, right? We see generally a decline in newspapers, many other sources are flat line and online is the place where we see this growing. It's something I think we all sort of know intuitively but there's stats on it. The other thing I wanted to start with in terms of context is sort of there's a lot of talk or consternation right now about the meaning of social media and news. And at least what Pew found recently is that social media is not yet a huge driver of traffic to news sites. We see most people going directly to news websites or their apps finding stories or finding news sites through search or through websites that are particularly created by news organizations. So with at least through Facebook and Twitter is being a relatively small amount of traffic, yes. Can you tell us what this says? I'm sorry, yeah, sure. So basically we have about a little over a third of people going directly to news websites. It's just sort of showing where does traffic to news websites come from and we see a really small set of them coming through Twitter and Facebook recommendations. That's the only take away from that so don't worry about the details on that. No, it's a good question though. And then if we do dive into social media a little bit we actually see that friends and family especially on Facebook are a dominant source of referrals to news sites. So actually with journalists and news organizations playing a relatively small role right now on Facebook. So again the details of this is not super important but what I just want you to take away is this idea that we have this confluence of different actors. We have news organizations, social media companies, friends and family social relations, search companies that are driving traffic and there's sort of a big mess here, a big mix of trying to understand how news traffic flows. So that's one piece of context that I just want you to keep in mind. The other piece of context I want you to keep in mind is a story and again don't pay attention to the details too much. It's this dip in the graph that I think you can see that's the main point. There was a case that happened in 2008 where if you went to the Google news search engine and typed United Airlines, the first hit that you got back was a story about United Airlines filing for bankruptcy in 2008. Only problem was that that was a story from 2002, United Airlines filed for bankruptcy in 2002, not 2008. Google really quickly corrected the error and it actually has a really nice review of how where that error came about and it was basically to do with the search engine crawling the Sun Sentinel which posted the story from 2002, the Sun Sentinel posting a story in a way that was just difficult for the news crawler to parse up and parse it out. So Google had a really nice response and it was sort of this learning moment where we saw what was the interaction between the news crawler and the stories that were posted on our website but not before the United stock price took a rather large tumble and it gradually recovered but there was sort of this moment where again, confluence of actors, Sun Sentinel published the story a long time ago, posted on their website, Google News crawler crawling that site making some sophisticated but in this case, ill-informed guesses about what constituted timeliness and then some real world material consequences in terms of the United stock price. And the last piece of context that I want to provide, this is actually a set of cases that people in this room from the digital media law project worked on. There was a case where Enterline actually sued the Pocono Medical Center in Pennsylvania for a sexual harassment case. Turns out the Pocono record had written a story on that case and Enterline decided they wanted to actually gain access to the commenters who had written under the story. So not the actual story itself but people had written in the comments and because they thought there was valuable information and they wanted to know from the Pocono record who are these commenters, they wanted to contact them directly and this is a state level decision but the state basically found and said no, the Pocono record did not have to reveal the identity of those commenters. The logic is a little bit narrow in terms of it said that information was available other places anyway but the end result was this sort of finding at this, again at the state level that the press can actually protect the anonymous commenters of people who are posting under its site. So we start to see this sort of broadening of what's considered maybe a reporter's privilege or shield law at a state level to include comments underneath the story. We find the exact opposite finding where the state of Illinois impaneled the grand jury that actually subpoenaed comments from the Alton Telegraph around a case where they basically wanted to know the identity of the IP addresses, the names, the dates, the times when the commenters had posted stories, posted comments underneath a particular story that the Telegraph had run but we actually found the exact opposite ruling in this case that said commenters are not considered sources, it's not considered under this umbrella of what we consider the press or journalism to be and there is no protection. So I offer these three starting points, this broad notion of what kind of stats we're dealing with in terms of social media and Facebook and Twitter and how news traffic is flowing and also this case of a search engine crawler interacting with a news story making some sophisticated guesses about what constitutes timeliness and then the court's decision about what constitutes a news story and what I want to argue is that these are all in a way stories of press freedom, these are all stories about moments when the press in a rather traditional form is interacting with some new actors and we have some confusion about what constitutes journalism, what doesn't constitute journalism and what I want to argue is that these are socio-technical moments. If we want to understand what press freedom means in each of these cases and some others I'm going to get to, we have to understand how code is generated, how it's understood, but also how the professional practices of the press are understanding their relationship to different technologies. The main point I want to leave you with is this idea that press freedom actually lives in what I call this network materiality, so it lives in this space where people are coming together and then a lot of our old assumptions about what constitutes journalism or press need to be updated in terms of this. But if I want to back up just a minute, I actually want to argue something slightly different to start with and as soon as you use words like press freedom or press autonomy, it kind of begs some questions about what do you mean by freedom, right? And I think in this country especially we've seen sort of this dominance of free market logics as constituting what free speech means. So this idea you send out information into the marketplace of ideas, somehow the marketplace sorts it out, truth emerges, it does this democratic work and people are able to make decisions about how to be citizens. That's sort of the general model. But what I want to argue actually, and there's this thread of both legal theory but also sort of philosophical theory around what freedom means, in a democracy autonomy actually means more than freedom from something, right? It means more than my right not to be interfered with by you. It means something actually positive. There's an affirmative notion such that for me to actually have autonomy, for me to realize myself, for me to do these things, discover truths, direct my own actions, understand consequences, understand what it means to participate in a public sphere, I actually need you all to do something as well. I need you to participate, I need you to actually say things that I can hear. So there's a sort of, this is the basis of a social contract that is the underpinning of what we mean by democratic freedom. So it means what I assert here and what I can argue is that freedom requires more than hearing just what markets or self-interest or your friends and I put in brackets here and our algorithms might tell you. There might be a different conception of what constitutes freedom that can underpin what we mean by press autonomy or press freedom. So that's, especially in a U.S. context, sometimes that's a bit of a radical departure from this idea that we're all individuals floating around completely formed without the need to actually interact and depend upon others, but we actually think that that's a different thing. So the second premise of this argument is to say that free press, it's actually part of this system of freedom of expression, right? If the press is not something that just happens to have been constructed through a bunch of different voices speaking, but maybe, and this is the open question of this project, maybe there's something that the press does that entails or engenders this freedom, this positive freedom that I was mentioning earlier. So this is the starting point, is to think about the press freedom as something that's about the press separating itself from others, about saying no, this is a moment where we are the press, we can think about the issue of the comments under the news stories as a moment where the press said we're going to carve this out as a reporter's privilege area, this is a press function. But we also have these moments where the press depends upon different institutions, the press depends upon search engines and Facebook and different institutions to drive traffic through it. So the press cannot be this completely separate institution. The press has to be in this constant negotiation of what it means to both be the press and not be the press. So this is the basis, or rather again sort of inflammatory quote for sometimes when I give this talk in the U.S., but Alexander Mikkeljohn was sort of one of the early theorists of this idea in a legal philosophical thought. And what he said is that in talking about the press as an institution of free speech, he said the point of ultimate interest here is not the words of the speakers, but the minds of the hearers. The press ensures not that everyone shall speak, but that everything worth saying shall be said. And this should raise some flags here because it should say, well, how do we know what it means for everything worth saying shall be said? What are the criteria by which that's going to be judged? What are the structural conditions that might entail that things that are being said or worth hearing are being said and that we're all having a likelihood of encountering them? So that's sort of the starting point is saying these are not easy questions, and I don't mean this in terms of government censorship or in terms of preventing people from speaking. I actually mean the opposite. I mean the likelihood that you're going to encounter ideas that are different from your own that you haven't sought out just to hear. Which brings me to this idea that press freedom is this problem of networked institutionalism. So I want to give a brief history of U.S. press freedom. It's always dangerous to do this on one slide, but what I want to argue actually here is that this problem is not new, right? This idea of the press having to separate itself from other people, the press having to carve out moments when it's unique, but also the press having to depend upon other people to perform its actions and actually achieve its goals, this is not really a new problem. And we see it again, so the Supreme Court actually, it's almost, it actually is kind of easy to summarize the Supreme Court's view on the press clause in one slide because the Supreme Court is actually really shyed away from making decisions based on press clause logic. We don't actually see a huge case law history here. We see a lot of invocations of speech clause and a lot of talk about what constitutes free speech, but the court in this country, the Supreme Court in this country at least, has tended to shy away from defining what the press is. And this makes sense in a way. We don't want to have sort of a two-tiered constitution where we say, if you are a member of the press and you have a press pass or you're employed by news organization, we actually, we kind of don't want to create a situation where those people have more freedoms than somebody else that's a dangerous moment when the constitution starts being fragmented in a way. So the court has tended to do it and we found that in its findings, so there's no federal shield law, for instance. There's no right of reporters' privilege to not be compelled to testify in front of a grand jury at a federal level. There are state levels, but not at a federal level. For a time, there was this right of reply and this notion of a fairness doctrine at a federal court. And we see that through cases like Red Lion or Miami Herald versus Turnello, but it was a slightly different finding. But we saw that actually sort of slowly erode, especially in the 80s through a lot of FCC decision-making, such that today we don't really have that same kind of logic. But for a time in sort of the 60s and 70s, this was this notion of fairness doctrine applied. But again, we saw the court sort of shying away from this idea of press freedom. Another set of histories really here to talk about press freedom is actually more in the sociology of the profession of the press itself. And what we see especially in this country since sort of the 1920s is this idea that journalism has really struggled with whether it's what James Carey calls an interpretive profession or whether it's called a scientiser objective profession. Is the press about going out and telling and reporting the truth in the objective news, or is it about making interpretations? And that's sort of an age-old debate in journalism, but we see the press constantly struggling with that. For instance, Daniel Hallen calls this the scientized view of journalism where journalists have this sort of physics envy where they want to go out into the world and report on experiments and claim a truth value of what they've done. And that's traditionally not really worked very well for a host of reasons about the social construction of knowledge. But we also see in the history of journalism this idea of rise of professional organizations, schools. We see schools of journalism starting to arise in sort of the 1920s and 30s, which are different from schools of speech communication and different from schools of English, where journalists are starting to rely upon sort of tricks of the trade, including using polling data for the first time in sort of the 1930s as a way of distancing itself from publics and saying, it's not me that's reporting this story, it's the poll that's reporting this story. This is not me as a journalist entering into and having a relationship with my sources, this is me relying upon statistical data. So again, we see this sort of this dance between journalists in the world as they're separating themselves. And one of my last sort of examples here is a neat one. There's a bunch of really great studies showing that even journalists in their use of language within stories do these very subtle sort of interpretations of the sources and of the events that they're reporting on. We see use of irony actually in a lot of journalist stories. So they will sort of report up. There's a great case of Lou Cannon reporting upon. Reagan went to South Africa in the 80s and said there was no such thing as apartheid. And Lou Cannon, this is the words of the president, you have to report on this in some way, but Lou Cannon as a journalist knows that there is apartheid there. How do you possibly do that? So he frames these whole stories in this ironic fashion saying, well, I guess Reagan thinks there's no apartheid. And it's thick. The language is thick with irony in these moments. And we also see journalists using even adverbial phrases, so obviously or suddenly or actually or these little moments of sort of using language to signal to the interested reader that, you know, wink, wink, I really have an idea of what's going on here. I'm not an objective reporter. So we see the press sort of doing that, lots of moments. And finally, the last thing we see is these sort of organizational routines in the press. So we see these ways of news organizations organizing a beat structure. So you always go to, you know, the city hall, the fire department, the police department. You always go to White House. You always go to Congress. This is where stories happen. And we see the news organization structuring itself in a way that it can reliably interact with these places. That's where news happens. So we don't see the press necessarily in the history sort of creating new beats all of a sudden. So there's a automobile section and a real estate section to most newspapers. There's often no labor section to a newspaper. You know, why? Why not? This is a moment when reporters do this. Another kind of neat finding in the history of journalism here is that we often find when you ask reporters, there's this ideal, you know, oh, you must write for the public and you must have this public service and this public interest. And a lot of them reliably say, are you kidding me? That's incredibly nerve wracking. I don't want to write for an undifferentiated public. I write for my friends and my family and my editors. So in sort of the press' negotiation of its relationship with the public, we see this incredible personalization of journalism, such that actually even when the idiom of insanity there you can see, even when news organization professionals talk about, what do you think of the people who write in on your stories? What do you think of the commenters or the letters to the editor? A lot of them report those people are crazy. I don't want anything to do with them. I don't ask me to engage with them. I want to go do my stuff and I actually don't want to be touched by this public. So I offer these sort of as a little bit of a landscape of saying that this idea of press freedom is constant negotiation of reporters and journalists separating themselves from public in a number of different ways, but also simultaneously relying upon the public, both for sort of cultural and monetary value. So what I want to say today, so if that's sort of the end of part two in a sense, today I think we're actually seeing this notion of press freedom being worked out in a bunch of new sites. And what I want to do is rather quickly go through some examples to give you a sense of where I'm seeing press freedom being worked out. These are again these locations, these tensions where this separation and dependency between the press and others is being worked out. And the first one is just quite honestly these sort of press economics where we see online revenue being a fraction of print revenue. So there's this idea of trying to create a revenue stream that can support quality journalism. This unstable relationship to aggregators trying to figure out either whether it's with Google News or Yahoo News or different kinds of aggregators like Facebook or Twitter, trying to figure out what it means to exist in relation to these aggregators. Philanthropic funding being increasingly a source. The LA Times just received a large grant to fund its reporting. And finally some calls for state sponsorship or taxing spectrum allocation. So there's this moment where the press is really I think quite creatively and quite robustly now trying to figure out how it recreates its economic footing. But we also see this idea of press freedom or again this distinction, this difference in separation in the form of user generated content. I mean CNNI report is an old sort of example in a well trodden one where we see this branding of content. Sometimes it's CNN content, sometimes it's iReport content, sometimes content flows from one side to another side but that's an editorial decision. Those are moments, little moments when the press says, you as public we're going to invite you into our branding or our moment. We see it also in terms of commenting systems, a wide I think variety in a diverse set of commenting systems. The Huffington Post has a badging system so you can become a more well respected comment around the Huffington Post. You can be again invited in through a gradation of systems where the more you perform and the better you are as a commenter you can be promoted through the Huffington Post's levels. New York Times has a similar verified commenter label that you can get so you can be moderated far less heavily if you sort of pass the New York Times's test to be a verified commenter. That test curiously relies upon confirming your identity through Facebook. So that's one of the things. So again we see this moment where your participation in the system of the New York Times and I understand why the Times is doing it. It's actually not a criticism at all. There's this ideal, this need from a Times perspective to verify identity but again here's this moment where we're relying upon a third party to do that so you have to ask about constituency. And finally just the last one here is this Washington Post, one of the examples of a social reader that have sort of tanked a lot in terms of usage patterns in the last sort of month or two. But again this moment where the Post says you want to read this story go through the Facebook app that is going to count it for us and I think many of us probably had that moment where you're like oh man really I just wanted to read that story. I didn't want to actually enter into this rather complex relationship with it but that's sort of again this moment where we see this press freedom being negotiated. And I think that's actually why you've seen this dramatic drop-off in the usage of that. You know another sort of reading precipitation is this idea of crowdsourced funding. I think a lot of people in this room probably know David Cohen, this spot.us, this idea of crowdsourcing funding for a story. So you put up a story say it's going to cost me 400 bucks to do this story. Who wants to add a little bit sort of kickstarter for news in a way so people add 20 bucks and then gradually it gets to this threshold. A student of mine at Stanford got one of her stories funded for the times the giant pool of garbage that people saw that was sort of a story that was popular in the times a few years ago. She got it funded through spot.us. And it's this again this moment of sort of you know relying upon or interacting with publics. A different version of that is AOL seed which is currently undergoing a little bit of revision so it's not exactly the story right now because they're re-jigging it. But it was basically the system that would go out and pay attention to search queries and say what are popular search queries that are happening right now and then AOL had a pool of stringers that it could rely upon and say hey it seems like we need some stories or some content around this particular topic because it seems to be really hot on a search engine. So again this is a journalism by demand almost that's generating stories based on search engine traffic. And I think we similarly see this in terms of professional practices. There's another source of pressure here on this idea of press freedom where journalists report actually saying they feel like they're under pressure to produce stories faster. They feel like they're under pressure to have a viewpoint within their stories more than they had felt before and to engage with their readers directly. So this is a moment where we're asking journalists again to not just go out and do you know Daniel Howland's 1920s, 30s, Scientized View of Journalism as going out and discovering news and reporting news in an objective way but we're actually asking them to be more like conveners of conversations. We're asking them to do more than we've asked them to do before and that's yeah. A trend for data directed journalism where people are actually relying on large data sets. Absolutely I'm going to get to that later but yeah no good excellent point. So again just another one here we see is this social media being used. I did a little study on social media policies and social media usage within newsrooms and we see this sort of real confusion right now about what journalists should do with their social media accounts whether they're journalists first or independent social media users second, what that overflow looks like who gets a story first, does their news organization get it first, did they get it first how do you source online if you can't verify the identity we see a sort of confusion happening there and similarly we see this explosion of new roles even in terms of news organizations where the old Ombudsman role is being sort of recreated and reproduced with a bunch of new roles whether it's social media editor engagement editor again this sort of this moment of struggle with news organizations I think trying to figure out what this means and second to last we see this again another site I think we see this negotiation of press freedom happening is in some of the intellectual property and the organizational systems that news organizations are starting to create so the Associated Press's AP3P system is one of the ones that's gotten sort of the most attention lately but essentially this idea of trying to create a little digital watermark within AP stories that you can then track stories as they flow across the web and have it be kind of like a content management system such that the AP can both track but also shut down uses of its content in different places so again we see the AP responding and saying we know our content's going to get out there we're going to try a sort of a technological solution that's going to encode some rather stringent views on IP that we have within our system we're going to create this watermark system this is a little document that actually the Neiman lab circulated from an internal AP confidential memo describing the system and the goals I didn't take it you can blame the Neiman folks for that or thank the Neiman folks for that either way the Guardian similarly again in this theme is doing a really cool thing with their open news list system so they're actually saying those old processes where stories would get assigned within news organizations in a rather opaque way where we didn't necessarily know why one reporter was working on something or not or how many stories they're working on the Guardian basically just publishes a Google doc that says who's working on what story where it is in the particular state of development so that you as reader anyone can go and sort of see this internal baking of news that's happening in the Guardian this again you know a lot of these are real double-edged swords right because again this is this moment where the Guardian's saying come on in public we're gonna show you you know how this asset gets made we're gonna invite you in and see the stories that we're producing but at the same time this can be an incredible starting point for influencing a particular journalist who might be working on a story if you are a source within it or you want to be a source within it you now have an incredible avenue in potentially to figuring that out so we're asking journalists now to you know be open and be public and be different kinds of ethical figures as they're producing their stories and finally we see this sort of notion of what I call this distributed human machine intelligence where these are these moments especially and this is I think sort of the most exciting place and it gets to that gentleman's question is these are places where data and human decision making are interacting to produce journalism and to produce news but it's it's very very unclear what press freedom means in this context so authority from aggregation Google News being a standard example Technerati being one of the older ones but then things like trending articles this was last night for me apparently on Facebook on my phone Mark Zuckerberg's wedding is a trending article for me so these are algorithmic determinations of stories that I should be paying attention to that are that are done through an algorithmic process on the flip side we see this algorithmic making and consuming of news so some folks at Northwestern a narrative science group that that grew out of there did a a new project called stats monkey where they actually said you know little league game games don't get reported on really but we've got all these stats out there we've got all these and baseball is this incredibly sort of formulaic and structured game that we can generate news stories pretty easily so they actually are generating stories that look an awful lot like sports stories based on stats from little league games and again you sort of think wow that's that's great because little league games are you know traditionally under reported you're not necessarily going to have a bunch reporters going out there and doing it but we have this moment of relying upon data to generate these stories and you have to ask about what is that interstitial the intermediary layer look like of how the algorithm is actually constructing narrative in the way you know think back to the my examples I gave of you know journalists subtle use of irony or adverbial phrases to sort of do this very subtle framing and subtle distinction where does that subtlety exist in an algorithmic construction of news it's not to say it's not there but it's to say what kind of news are we creating in those systems and then finally as last month's Thompson news Thompson Reuters news scope where you can actually make investment decisions based on Thompson Reuters analysis of news stories of particular companies so you can sort of hand over your portfolio and say please go analyze the news make buy sell decisions for me based on the news feed that Thompson Reuters is is algorithmically processing so again it's this in almost a sense there's no need for a reader in that case right there's no need for actually somebody to consume a story that's been produced we can have algorithms talking directly to algorithms this is a University of Tokyo again I'm going further further farther field here University of Tokyo had this a journalistic robot this is a robot that was trained to actually go into a room build a visual image of that room and then go in the next day or periodically afterwards and recognize what's new in the room and actually generate a news story based on what it encountered as new it's a very narrow version of what news means in terms of sort of difference in timeliness but again it's this notion of algorithmically implanting a notion of news value into a story and then drone journalism and there's some really nice folks at University of Nebraska they're doing some work on this to say this is sort of a big data generation problem for news organizations to say no instead of going out and trying to get into a bunch of news feeds and news sites maybe we as news organizations or news professionals can actually generate our own data but it raises this question of what we do with all that data so what I want to argue is that all of those sites all of those places now I went through them kind of quickly but what I wanted to do is give you a wide sort of view these are things that I'm calling news where in a sense and the reason I sort of introduce this term news where is to say I think we need some sort of structured approach understanding or to viewing what press freedom might be in all of these socio-technical spaces so it's not just looking at code it's not just looking at what journalists are producing now but it's looking at this middle ground this confluence of journalists and technologies working together and I import here from for people who study science technology and society studies this notion of infrastructure and infrastructure sounds like sort of a boring word but I think what's really great about this as a theoretical framework is infrastructure is all this stuff that's kind of invisible in the world it's stuff that we don't see it's stuff that we use and encounter every day but we don't think about where this infrastructure comes from we often don't notice it until it breaks down one person's infrastructure is another person's application if you if you talk to somebody who works for a railroad that railroad line is not just you know invisible infrastructure that you you ride upon as you're on a train but that is the end goal that's that person's job to maintain that and to do it and what I want to argue here is in news where the reason I think infrastructure is a good analogy or a good metaphor to use is that we can enter into this conversation at many points along the way we can enter into it from you know the drone journalists we can ask how that data is gathered but we can also enter into it from a news story process and say you know where was that notion of irony or journalistic interpretation in the algorithmic construction of news so this is largely what I already said but conceptually these are these sort of shared and embedded but largely invisible kinds of logics that I think are structuring these relationships between presses and publics and when I you know make this move then to say well operate like what is it that you're actually studying then like this is sort of a an abstract notion of what news where might be but this is the moment then to look at sort of the algorithms the interfaces the practices capital flows how advertising revenue flows and ask at each step along the way where this idea of a right to hear or whether freedom to hear exists in this confluence of technology so I think I'm just going to skip because I already said that so I want to shift now last the last bit of this is a study that I did of one particular piece of news where so if I've given you this sort of overview of what I mean by press freedom and I've asked you to sort of buy on to my sort of you know quasi socialist view of us depending upon each other for freedom I'm Canadian by the way I should admit that in the moment so I've asked you to sort of buy into that that notion of that philosophy of freedom and I've given you a little history of where press freedom has been worked out and how this isn't really that new a problem and then giving you a little bit of a run through of these places that we see it being worked out in network technology today what I want to do now is just dive a little deeper into one particular kind of news where and it's one that's a little bit understudied these days but that's it's a notion of an application programming interface or an API and this is something that's kind of new that news organizations are doing as they're starting to create application programming interfaces or API's into their systems so for those in the room of those in the crowd you know haven't heard of what an API is or don't really know what I mean by an application programming interface I think there's actually for me anyway sort of a relatively straightforward way of thinking about this so I want you to pause for a second and think about your experiences at the post office because what I want to argue is that the post office is a kind of API right I go into the post office and I if I put a stamp on an envelope if I address it the right way if I put you know the two in the center and the from and either on the back or in the top left corner and I know that I do it and I plan ahead maybe a couple days although given post office the US post office interactions lately I hope this doesn't become a historical example for the post office but I know how to behave in relation to the post office right I know how to behave in order to get something done I can get my letter across the country or across the world as long as I know how to follow the rules of the post office what I don't need to know is how the post office works right I don't need to delve into the details of how that letter is actually read whether the stamp is read to be the correct amount of postage or not I don't need to do that the post office is this way that we've all learned to interact with and this is this notion for again for the program in the world the idea of application programming interface is sort of a table cloth it conceals complexity it's something that says I don't want to invite you into the guts in the details of my system but I'm going to help you figure out how to relate to a system so this is what news organizations are starting to do and you should think about this in relation to some of the stuff I said earlier about how the press distances itself from publics but also relies upon publics at the same time the idea that freedom is something that sort of requires dependency but it also requires separation this is the press struggling with an identity in a moment this is an identity crisis the API is one place to see it so what the API is in a news organization sense is it's kind of a software tool kit so you can go to well this is actually what it is it's a system of rules and associations it's the software tool kit that lets you actually use news organization data in a way that you wouldn't have used that before for programmers mostly although there's some inroads to letting non-programmers or people without technical abilities use it but these are sort of the rules and associations by which news organizations let you into their content so I did a study of three different news organizations APIs this is the Guardian, the NPR and the New York Times and these are sort of three leading USA today also has one Mastry has a whole set of APIs that it actually uses but these are sort of the three leading API news organizations that had APIs and what you do with these APIs is you basically as a programmer you say you have a bunch of data within your news organization I want to build an app with that data I want to rely upon it I want to repackage it I want to visualize it in some way I want to actually make something new with that data we have a few different examples here this is one built with the Guardian the Guardian has a whole bunch of data on where nuclear warheads are pointed in the world and actually you can so put in your city and country and see the number of nuclear warheads that are pointed at you if you want to do that that's kind of a thing NPR had a really nice one, Reverbiage has a really nice one where it actually takes NPR headlines and pays attention to the metadata and then replots it on a globe of the world and you can actually see sort of more of a geospatial sense of where stories are coming from and what stories are about so again it's a different way of taking a news feed that NPR is providing and letting you create a different view on that news feed and then similarly this was the New York Times had a public financing API where it actually tracked where electoral campaign contributions were coming from in the country and then plotted them according to Republican and Democrat and was able to let this was an app that was letting you see where money was coming from and going to again these are all sort of very simple examples but it's a case where the news organizations through their APIs are making this toolkit and saying here's a bunch of hooks and levies that you can use to access our data you can make something new with it so what I did was I personally built one app with the NPR API but I experimented with all three of the apps did an analysis of several hundred public documents and speeches and stories about the apps and then did a review of systems that were built with these apps and what I again my goal here was to read again go back to this word infrastructure for a second was to read these infrastructures and say where is press freedom being worked out in this infrastructure this API is a place where the press is distinguishing itself from other people so I did this one task here which was with the NPR API I said all I want to do I want to pick a two week period this is at the end of September being of October of the previous year and I want to build a news feed of NPR's health care coverage it's all I want to do I just want to you know something that I can put on my site that I could say here was you know two week period of what NPR health care coverage meant let me do this to sort of experiment with what this looks like me as programmer public I have some programming skills I'm not the best programmer in the world at all and I also made a purposeful decision to try to use the publicly available there's an interface where you can actually it's just checkboxes and you can build a query of an API based on clicking some checkboxes it's not a very heavy programming activity if that didn't make any sense don't worry about it but what I want to do is I had this goal right this goal is to start to use this to figure out what it was like the first thing I do is actually just build a query based on health care and the US these are just two keywords I did to search health care in the US again the details of those boxes do not matter but they're just showing the output of what I found I found that I actually got too few results I actually only got eight results from this this is a two-week period this was the height of a health care debate in this country and I got eight results from searching on health care and the US seemed kind of strange so what I did was I ended up taking out the US because you know I thought I'd broaden the search a little bit and find out what I ended up getting was that a lot of the stories that resulted out of this had a US focus to them but I actually lost some of the stories that I had had in the first query so there was this weird sort of finding or this weird result where it seemed like stories were being sort of metadata indexed partly through US metadata partly through health care data but there was this strange I would have expected to get more results but not radically different results from what I did so the next step was actually to say well I want to actually understand this category of health care and the US because it seemed like that was the difference between step one and step two there's something about indexing US health care stories that's different so I combined it with the world instead and to sort of say well maybe NPR is organizing its stories in this way that they're indexing them through international coverage more than US coverage I'm not sure I got one story returned in that time over a two-week period and it was a story about the US was found to have inflicted syphilis on Guatemalans in the 1940s I was really in two weeks this was the only thing that I got it was a very sort of strange result and it was actually inconsistent with the previous results so I then actually discovered there's this other category called world health which is somehow different from health care plus world and ended up only getting three stories this is still not getting me in my goal here of making a newsfeed I then also discovered that you can actually split it up by world and health which is somehow different from world health as a key word I got a completely different set of stories then that were not related so these stories were not related to these stories at all and at that point I stopped because I really was discovering that there was this real challenge in trying to use the API so again you have to understand here the NPR is actually opening up its whole data source to you and it says you can build a tool, you can do whatever you want with it but the actual act of building this app was an incredibly sort of complex ontological one if I went to a step you just went through my conclusion would be that their API was very buggy or that NPR didn't understand billion operations so keep that point actually so finding is great inconsistent category interactions good point so we have these again so me as news programmer now trying to interact with news system this is this new moment of press public relationships we have some inconsistent category interactions going on between world health, healthcare, world health, US can't recreate something that makes sense at least to me and to your point exactly what I wanted as an ideal here so Boolean logic wasn't built into the system for programmers it's not a big deal you can rebuild the Boolean logic it's not a huge deal to do with the data source but it wasn't available through the public tool that supposedly anybody can go and use and interact with we interpret it as a Boolean and no no or whatever right so there's something here and this is this led to this last conclusion is that I as member of public trying to interact with this system I kind of reached this what I was calling an ontological black box I was kind of reaching this moment where I actually couldn't look inside the categories that NPR was using to create its news stories and I couldn't ask questions I couldn't debate them I couldn't say I think I want to relabel it like this or I think that I want to actually do some hand editing it was a difficult API to work with and I do not fault NPR for this at all because I think this is actually an incredibly challenging moment to actually take internally consistent logics that people have a tacit understanding of how to work with over years of being employees there and figuring out how to work with metadata The New York Times has had a department since the late 1800s exactly on dealing with metadata in their news stories this is a long-standing problem in news organizations but I use it to point out this idea that if you are a citizen programmer member of a programming public trying to do this you're going to run into some real I would argue sort of ideological moments where the press is going to say no that's our job to manage the metadata it's not your job sorry this is a moment we're going to do the description you can do the repackaging so what I ended up finding was there's these moments in the interest of time I want to just blow through these and not linger on them but so this is not just this experiment this is sort of one example of one that I did but looking over across all the documents and across the other experiments that I did what I found was there actually three different ways I would argue that news organizations are mediating or regulating these press-public interactions there's three ways that the press is distancing itself today from publics and we can look at each one of these places to see whether or not we think our ideal of a press freedom or freedom to hear is being instantiated so the one is just this idea of regulation so this is tiered levels of access for instance so the guardians API actually you can get different levels of access to it depending on whether you want to pay pay the guardian directly to use it whether you want to let the guardian put ads on the site that you might create with the API and what kinds of ads that you're willing to accept on the site so this is not sort of an open completely equitable use of the API we have a moment of differentiation here where if you pay you get a different experience with the API that's a decision that the guardians made NPR for instance doesn't do that so this is again a differentiation moment limits on revenue generation this is actually just this moment where if you put ads on your site for instance with the guardian you have to engage in sharing arrangement with the guardian because again one of the ideas is news organizations want to drive traffic back to their sites right this is totally makes sense that they would offer API to let people build apps with them but the tacit deal that you're making is that you're also going to engage in some kind of revenue sharing based on the traffic that's being driven back to the news organization sites and the last one we see is this idea of there's a prohibition on archiving so you have to go back to the API it's different in different systems but once a day to actually get new copies of the story so again this should be another moment where we say so me as citizen programmer I don't actually get to say I think maybe this story is done on my news feed I don't get to say I want to archive this story do a comparison between this story and the story that came after it I'm contractually obligated to go back to the news organization and refresh my story constantly and again I get why news organizations would do this it can be really messy to have multiple versions of the story floating around but this is a moment where the press is again asserting itself it's sort of saying no we're going to keep track of what the version of the story is in this moment that's not for you to do you can republish for us again it makes sense but it's if we want to ask this question about what press freedom means that's where it is second this idea of architecture so the first one is regulations sort of the terms of service and the kinds of agreements that you enter into when you use these the second is sort of the code as regulator so for people following Professor Lessig's work this should be really familiar this idea that the code is doing architectural work it's regulating the kind of access and the one you know it's a rather simple point is this idea of key control and personal information so when I use an API I have to release a certain amount of personal information I'm issued a code or a key that key is my access into the API at any point my access can be revoked through the key if I'm breaking terms of service or I'm doing something that's not not good or not open it's a moment where the code can actually be a regulator and again this should raise a little bit of flags here when we think about who owns the stories that are produced by a news organization think back to the idea that there's at least in theory or ideal there's a notion of a constitutional protection here for the news that's generated here's a moment where a news organization can say well we're going to let you have it under some circumstances but we're going to revoke that access if you're not following the terms of service the architecture point here this internal indexing and vetting systems this should raise flags with the NPR one so basically there's a whole bunch of ways that news organizations are generating and organizing stories that are not accessible to you as programmer finally this idea of API is organized into beats remember one of the points that I had these historical things where news organizations have real estate sections and automobile sections and house sections but no labor section or often no environmental section so we see this recreation of those logics in the organization of the APIs it's up to you to use the keywords to recreate your beats if you want to but the news organization is not going to do that for you and finally this idea of community I think is this last one so what I also did was I attended an analysis of several of the different hack days and different kinds of places where API programmers are coming together with news organizations to create new systems so the New York Times runs one, Guardian runs one NPR has done some smaller ones at sort of a local level with these hack days where you actually get together and you work with other programmers and you make something new with the APIs and what ended up happening there a lot of times was so even though we have these regulations and these architectures the reason I say this idea of personal and informal relationships and certain regulations tons of examples of people at the news organization saying I know you're only supposed to access this thing like you're only supposed to make 5,000 calls a day to an API you're only supposed to trade data with it 5,000 times and you're supposed to go back and you're supposed to refresh the story once a day and you're supposed to do this revenue generating thing and you're supposed to put these ads on here a lot of that is there true but if you're doing something that's really really cool that's driving a lot of traffic to our site we'll talk and there's sort of this moment where some of these rules and regulations and again totally makes sense these are humans, these are people who are sort of trying to create some system that has economic robustness to it I get why it's happening but again as member of programming public you need to understand that simply accessing the APIs is not giving you the same sort of exposure to it the exemption that you might get from the API who runs the New York Times API really well who's just going to sort of nod and wink and look the other way if you're doing something that breaks the terms of service but that's really cool so that's this moment and so finally one of the key findings from this work is this I think this emergence of programming publics so these are people think about theories of public sphere and theories of publicness from people like John Dewey who talk about the public as this shared set of constituents so we all come together and we all need to have to experience the same thing we can't separate ourselves out from public goods I'd argue that actually we're seeing some programming publics emerge these are people who are creating the conditions under which we're experiencing news and I think these are a different set of actors than just the news organizations and just the social media companies this is a set of hackers and a set of I mean hackers in a good MIT way okay again good in MIT but I mean that as a sense of these are people who are actually creating the conditions under which you're experiencing news this I think is a new place and an important place to look at press freedom this is a place to look at and say what are these people's conceptions of a public right to hear versus an individual right to speak what are these people's conceptions of what constitutes hearing diverse information in a timely fashion these are a lot of programming types decisions that get made but that I think have not been scrutinized in terms of sort of a journalistic almost civil rights value so I think I want to end here I think this is what I've been trying to do in this project is one just provide this definition of network press freedom to say that press freedom isn't this thing that lives in any one news organization that we can point to necessarily it's not a singular solitary profession that's responsible for press freedom if the press itself is distributed and is existing in all these different kinds of places then the notion of press freedom is also distributed and existing in multiple places and the challenge is to sort of trace an idea of press freedom among this really complex changing ecosystem and figure out when do we encounter a free press we can't just look to a news organization to find it we have to look to this this ecosystem and so that yeah I'll just skip to the end there's been a lot of debates around this issue of some of these old tired debates around is a blog or a journalist is a blog or not a journalist and I think that some of those things are boring and not terribly interesting to engage with what I would argue is actually bumping it up a little bit in terms of abstraction and saying when do we encounter a press that we want when do we encounter the press that exists with the normative ideals or the values or the ethics of the democratic press questions of whether a blog is a journalist or not kind of go away then because we don't worry about it we say if we want to accept my notion of democratic freedom as this thing that means freedom from and freedom to there's this positive notion to it then we can I think look across the press and say if it's enabling this public right to hear if it's enabling a kind of listening that we can't do as individuals then maybe we don't worry about whether somebody is a blog or a journalist but we worry about their role or activity within this news ecosystem so I'll stop there because I think we'll then have 15 20 minutes for questions thank you and I'm going to sit down because I've been talking sorry yeah I I think I saw your hand first but I'll go around and get the others after me just on the question of APIs and searching searching is very difficult has a lot to do with taxonomies very complicated things so I'm not surprised that even a sophisticated news organization would have trouble turning up things that you search for where I had trouble with the Boston Globe in the same way but there are some systems that do very sophisticated natural language processing both from Alpha with underlined Siri and IBM Watson which we all know from Jeopardy and these things are actually quite opaque they're proprietary and even if they were opened up most of us could not understand the algorithm so I just want to throw that out to you as a yeah I think it's a good point in that so the Guardian especially has been moving because yeah another thing I didn't really talk about was the proprietary or open source nature of the software that these news organizations are relying upon to build their infrastructures so sometimes you're right it is these proprietary systems but other times and the Guardian I think has done a good job of this of making a concerted decision to try to use more and move to an exclusively open source infrastructure partly it is this ideological decision where they're saying we're a news organization who's trying to be open in both not just what we publish but also how we build so I think it's a good critique to make of these systems is that I think I saw your hand first and then I thought this was a fascinating talk and it's a very interesting framework to think about a whole host of issues oh this was a fascinating talk and I mean a framework to think about a whole host of issues one thing I sort of thinking about as you explained sort of the network infrastructure was WikiLeaks which is sort of our bloggers journalists were WikiLeaks journalists and let's presume they are I mean they ran into a whole other set of tensions which you didn't really explore their hosting service deciding they were politically incorrect their payment processor deciding they were politically incorrect etc do you have any thoughts on that? yeah no I think I've written a little bit about WikiLeaks in other context but I think it is it's an example of a moment where for me it depends on what you want to mean by a free press right so instead of me taking a stance on saying WikiLeaks was you know X or not X I actually see it as a great moment where we saw coming to the surface a whole bunch of different exceptions of what press values are so and I think for me that was the real value of WikiLeaks is it's this moment where we get to debate what is privilege information who gets access to privilege information what is the gatekeeping look like on that what are the background reliances on you know whether it's the credit card companies or whether it's hosting services such that those are the conditions under which news is made and for each one of those organizations they're kind of getting sucked into this I think nicely sucked into this problem of what do you mean by a free press and for me that was the value of WikiLeaks was to say we just made that conversation a lot more complicated and that's wonderful that's great because it actually brought to the surface a lot of dependencies a lot of infrastructures that's why I use the word infrastructures it's that invisible stuff that we don't get to see so WikiLeaks moment is a moment of of seeing it which is great sorry I think the gentleman with the beard was next and I'm going to try to remember but also please jump in if I've forgotten three data points which may have nothing to do with what you're talking about Marcy Wheeler reported out of press conference I believe with the president but maybe with the national security advisor or someone like that basically speak up a little bit basically saying that oh we don't need to take take reporters to grand juries to find out their sources we already monitor electronic communications and we already know that information we're storing that at the NSA right you know so it's one the second is what's happened in Chicago over the weekend we have people like Tim Poole who's been live streaming occupied with two cohorts also citizen journalists being stopped without warrants by the police and interrogated at gunpoint even though they knew that these guys were journalists or these people were journalists and the third is a recent move within the congress I guess within a defense authorization bill to say let's take away the prohibitions against using propaganda on the public the united states public so here you have three sort of forces of the state coming in you know and it's getting rid of all that freedom from basically so it's a really great example of how it is more complex and that's why I don't I never want to argue that we don't need the freedom from part of this at all I absolutely keep those together and hold them up together so I agree with you how do you see that dealing with the freedom too the stuff that you've been discussing here in those cases is exactly exactly but there's trends so I mean one I think is this idea of again I put it to sort of a structural question so we sort of say it's not about any particular moment making a decision based on a freedom too but it might be something like you know if we're able to for instance there's a series of supreme court decisions that actually took away the fairness doctrine the right reply thing because it said technologically this doesn't really matter much anymore because it used to be when spectrum allocation was limited then there was a reason there was a logic for having a right of reply but there was a series of cases that said for instance in the cable industry there was a judgment that said well cable is this sort of infinite thing so there's no technological rationale for why we might have a right to reply there so we saw Justice we almost had Justice Kennedy who actually said well but if there's an economic monopoly that's created then maybe that's a scarcity issue so we might want to regulate that because even though it's technologically endless maybe there's a market issue so the way I would add to this actually is more of an attention based one so I think attention is actually sort of this scarce resource that hasn't made its way into a lot of these structural conversations for freedom too so on a really sort of basic level if as those stats are showing if we're increasingly going to start encountering a lot of our news through what friends and family show us on Facebook for instance if that's a major site if we start to see that happening I think there could be a case to sort of make some structural changes to what kinds of information I attend to or don't attend to and maybe it's a design decision to say you know you've kind of been paying attention to these people an awful lot here's a moment where I'm going to interact or interject as Facebook or as Twitter or somebody else in a journalistic like way and say we're going to give you something different than what you had had before I don't think that really answers your question I'm sorry but it's that's how you answer it this is the best kind to try to answer then I think there was one more here and then and then you were there going back to the history public relations is not journalism but journalism is public relations correct I wouldn't I would say it depends the president of the public relations firm that I worked for 11 years ago that's what he would say so the real question is like what is the relationship now because I know that it's all in some places and rocks all at certain times and with the growth of interest in business and finance reporting that was a big as well as technology was a big bone for public relations people what is the relationship like or is it all behind curtains yeah it's a good question I haven't thought a ton about PR in the sense I mean the most experience I have with studying PR is mostly in this historical sense where one of the reasons for instance one of the reasons the Columbia journalism program was was founded was this desire to distinguish journalists from public relations officials and that was the or professionals and that was this moment that was made there was another moment in history when opinion polling was sort of this increasingly used thing where opinion polls were coming from both market research organizations and political opinion polling organizations and yeah and there's a sometimes there's a desire to distinguish between those different types of polling and on part of journalists to report on those polling differently because one but I don't have a lot of great insights into contemporary public relations stuff but my I would again sort of put it through this framework of not to take a stance to say public relations people you know are or not journalists but to say it it probably depends in a particular moment of what's driving their production or what's driving their their participation in this it definitely had an effect over the behavior yeah I probably right for good or bad yeah you're probably right uh yes at the end um so I have a couple questions we're sort of tying together so first of all it sounds like you had not just a question not of open data but open metadata question of how to so have you considered using sort of third party apps like semantic tools like say Kalei open Kalei for for journalism in order to more meaningfully categorize these stories yeah sure so I mean in this case though I was relying upon the metadata of the news organization so I I think that would almost be more of a question for them to say how are they internally generating the metadata that is then used to describe the stories but I think that would be an interesting move for them to make but it also then is this reliance on another piece of software another piece of technology that's doing some of that logical work as well so I I would want to see more about how those pieces of software making decisions around metadata and I my hunch is that news organizations are not going to want to sort of completely give over their metadata generation and maintenance I know New York Times has this you know incredibly complex set of heuristics and set of rules around metadata NPR has this ingest system which is doing this really complex work of trying to make sense of what member stations are sending it so if you go to NPR.org some of those stories are coming from you know WBUR or KQED there's sort of a lot of these data is flowing in and there's this thing called the ingest system that's trying to make sense of all this data and trying to square up the data that's there and that depends highly on how different member organizations are doing it and that's a hard piece of work so I would honestly doubt that they would be willing to sort of give that over but I think it's an interesting question to ask of them. My question was whether you sort of feel like we talk about the freedom of the press not just freedom from but freedom to but also the freedom of the consumer so something something I could sort of see worrying about is like when you try all these searches and you find wildly different results that you're sort of getting maybe like force fed into stories that you may not necessarily want to read if you want to so it's sort of a question you do does the consumer have a right for a meaningful search or for meaningful for meaningful topics that they articles of their read and not being like you know sort of pulled into something that they may not have been looking for. Totally yeah so I understand it I think so I think there's two things to tease apart there one is the generation of the stories that appear in the search and then the other is the logic driving the search or structuring the search and I think that those are both places to ask around what what consumer choice means I think that this sort of the philosophical theory that I presented there's a lot more detail about even the illusion of consumer choice in a sense or what it is that you may or may not want has also been rather conditioned and structured upon what you've encountered previously and so this is why it's kind of a it's a normative departure is to say there is a difference between the news you might need versus the news you might want and that's why it's a structural question it's not about me you know preventing you from seeing something or giving you bad search results it's actually about me having a different value that goes into both the construction of the stories and the structuring of the search that says I'm not it's not just about fulfilling your consumer desire and I know that that's a really controversial sort of thing to say because you might be unhappy I totally get that but you might be yes I think of you and then you sorry you already asked kind of building on what he was saying it feels like from the fairness doctrine and the writer reply we sort of did a shift from government oversight of what the media's responsibility is to the public to almost a commercial oversight where you see like in WikiLeaks where the ISP decides what goes out and what doesn't go out and so like the commercial role within a media presentation or availability has gone sort of from like a single booster to a potential signal squashed and so I'm wondering if you see like a potential for another way that government could be involved going forward or if it would have to come from a totally different quarter so you briefly talked about like attention as a scarce resource and I don't think anyone really wants government mandated attention so that would you know like it couldn't be the same kind of thing as like you know AMFM dial resources being allocated but I'm wondering maybe it's like like the best thing would like more of a NGO kind of a role. Well I think so one place I'd look is there's a lot of so even tax law stuff around you know funding or not funding but calling news organizations as nonprofits and there's sort of you know there's been in the last probably six months some really nice stories and documentation around how difficult it is sometimes for news organizations to get themselves classified as nonprofits and folks in the way you guys can talk a lot about that or what the challenges are around that so there's a case of you know through the tax code there could be a government role for supporting journalism but that's not so there's a bunch of things to separate out here there's not I'm not talking about government having voice so I'm not talking about creating a you know state news channel that has to get your attention that's not yeah they're okay right yeah so I think that it's it's not about giving government a voice and there's Owen Fiss has a really nice piece actually if anybody's interested on different types of state roles that can be in a in a media system and the idea of government speaking is actually a sort of a small one that's not really a lot of it's about structural stuff in terms of you know tax code might be one instead of being Comcast like you know as much as I mean distrust government to be involved in detailed ways in certain ways I trust Comcast less because I can't really out vote Comcast even you know that's even more impossible than you know like I'm on the phone longer waiting for Comcast for my congressman yes I can totally yeah I can totally see that yeah yeah more arbitrary not as a voice yeah yeah not as well I totally agree and the other thing is the I think one thing you know the state can do as well is sort of provide either an investigative role or provide a role of orienting people to even the problem existing as well so it's not necessarily that you know the state is going to fund some news organizations and not others or have a voice but it's going to say maybe there's a problem here that could be addressed through private markets through civil society through lots of different things sort of one more okay you guys you both spoke before and you both have much to do can we combine them can we combine them I'm not a question okay then can we hold yours is a question okay let's do the question first it seems like a proprietary API is only as good as the organization is willing this to support an ability to support it when you encountered it you know when you were totally baffled by NPR's idea of Boolean logic was there anybody at the other end who could tell you you know what either what you were doing wrong or what they might have been doing wrong that was causing these problems yeah so so the short answer is no exact but there is a really nice sort of user community of people who are trying to build with NPR and this this is true for for the New York Times and for the Guardian as well so one of the but I think those user communities are are they're doing really great work but they're often sort of small and you know this is relying on people sort of volunteer labor to do this stuff I know the black box yeah but NPR is actually I think really good about they're really good about saying they want to support this infrastructure so everything I'm saying is not a criticism of NPR because it's actually I mean what they're doing I think is really incredible with the small set of scarce resources and people are doing it baffling it doesn't make sense to me I'm on a bunch of the mailing lists and there are people who will chime in and help you out with that so it's there but it also requires you know some to ask that question yeah sorry then yes to the comment about the user's right to our fair search I think it's interesting that Google has started to assert that the results of its search is protected speech in and of itself that it represents it's opinion about the various sources yeah I mean I saw that article I think that's a really tricky tricky notion of free speech that it's no first amendment right for them to assert their opinion that this is you know the most important site in the world about Google yeah no I think that's true I'm getting a look from MR that I think we're running out of time but thank you very much