 Welcome to the United States Institute of Peace. Thank you so very much for choosing to spend your Monday morning with us instead of facing the hundreds of emails that are waiting for you at the office and other business matters. Emails, USIP, it's not a hard choice. But seriously we know that it's really hard for you all to get this time away from the office, especially half a day away from the office. So we are committed to making it a very productive valuable time for you. We used to say around here, as my colleagues will tell you, we used to say that we wanted the time that you spent here to be at least as productive as the time you spend in your own offices. But then people pointed out to us that that really wasn't a very high bar to be setting. So hold our feet to the fire. You see any USIP staff around that could be helpful to making this a more valuable time. Please don't hesitate to call on them. I'm Sheldon Himmelfarb and I manage our PeaceTech initiative here at USIP, which works to bring both technology and media to bear on the problems of conflict management. And nowhere does this intersection of media and tech come into focus more sharply than in social media. Whose impact on the trajectory of conflicts and political movements is being felt around the world. And that's why we started our blogs and bullets project here four years ago to try and make sense of this brave new world, to understand how online social networks make online social networks, bloggers, user-generated content, and more how this is being used in conflict, and more importantly how they can be harnessed for peacebuilding. So with each year the project has delved into the complexities of this question, using a cross-discipline approach that combined state-of-the-art data analytics with expert social and political analysis. And you will get a good sense of both today, I think, as we unpack the analysis that our first panel has done of about 40 million tweets from the Syrian conflict. And again, that's a combined team from George Washington and AU that is as well-versed in the technology in the online world as it is in Middle East politics and culture. And that, as I'm sure you know from the many meetings on this kind of topic around the this town, is a very rare combination indeed. Each of the blogs and bullets reports has built on the previous one, which has created a nice record of the evolution in perception versus reality when it comes to the role of social media in violent international conflict. So what are some of the things let me just give you a sense of some of the things that we've learned across this series. In September 2010, when we released our first blogs and bullets study, we identified a problem, a real problem in this field that was holding back our ability to understand the relationship between social media and social change. Namely, the increasingly entrenched positions of what we called in that report cyber-optimists and cyber-skeptics. And to get past the sweeping generalizations and the dueling anecdotes, that report laid out a five-level analytical framework for really helping us unpack the relationship between online and offline activity. So how are social media affecting individuals and their actions? How is it driving group action? How is it affecting intergroup action? How are regimes reacting to the activist use of these technologies? And finally, are social media helping to attract international attention? How is it affecting external actors? How are they affecting events on the ground? But that framework really was just the beginning of the story. As I said, the project has consistently tried to use large data sets and cutting-edge analytics to answer these questions. As we all know, big data is a reality that has entered our vernacular thanks to the social media revolution. And our tools for making sense of this has really never have never been better. So in our second report on new media and conflict after the Arab Spring, which we released in July of 2012, we followed our own advice, as it were. Not only did the report apply that five-level analytical framework to its study of social media's role in the Arab Spring, but it also drew on a unique data set to do that. And many of you, I'm sure, are familiar with the Bitly link shortening service. It takes your long URLs and it shortens them so they can fit into the 140-character limit of Twitter. Now, we recognize that data set is only a portion of online activity. But what was unique about it was that it enabled us to see not only who is making social media during the Arab Spring, as most of the other analyses out there were focusing on who's producing it, who's involved in that dialogue. But by focusing on the Bitly links and who was using them, it allowed us to get a better sense of who's consuming the media that was being produced during the Arab Spring. Who is clicking through on these links and what are they clicking through on and why? And it led the research team to some really unexpected findings about social media in the Arab Spring that placed less emphasis upon it as an organizing tool, which we were hearing a lot of in the traditional media, and more emphasis upon it as a megaphone with which to engage the international community. So that's just a taste of some of the kind of counterintuitive findings that we came upon in the first two reports. And I urge anyone, we've got them all outside, along with the report that we're going to talk about today. They're all on the table outside. Feel free to take them and I'd urge you to take a look at those. So that brings us to today's meeting in the third report. And one of the important recommendations in that last report was that future research really needed to delve more deeply into specific conflicts within specific countries. And that's exactly what we are going to talk about today. On this first panel, you'll hear about the new report on Syria's socially mediated civil war from this talented GW AU team that I mentioned earlier. And then on the second panel, we'll go into the Middle East, go outside the Middle East to places like Ukraine, which is on top of mind for everyone in Turkey, where as you know, digitally active protest movements are again working to shape their societies and their politics. So thank you all for joining today. I hope you'll lean in and really participate in the open. We've got mics on each side in the open mic Q&A program, because I can assure you your questions and comments will influence our research going forward. We've been doing this now for four years. And these meetings where we roll out the report based on the research the year before, and then invite your feedback on those findings are really important in helping us direct. What are we going to be? What are we going to focus our energies on in the next research round? Is should we be focusing towards certain geographic areas? Should we be thinking of other platforms? We've gone from using the bitly data set to the Twitter data set? What is top of mind for you? So it's very helpful to have these meetings and your participation really does influence the research going forward. And I should also say, we have a live webcast going on. The event is streaming across the globe. So let me invite the online audience as well to interact with the panelists and each other during the course of the event via Twitter using our hashtag for the day, which is hashtag USIP blogs. And when our moderator PJ Crowley goes to the audience during the Q&A session, he'll also have the option of going here to Anand Varghese, who will be able to talk about the discussion that's happening online and also relay your questions from the online audience to the panelists. So we want to know what you're thinking, and we want you to know that you really are a core part of this event. So now let me introduce our moderator the day PJ Crowley, a fellow at the Institute for Public Diplomacy and Global Communication within the School of Media and Public Affairs at GW University. Many of you know PJ already as the Assistant Secretary of State for Public Affairs in 2009. And then, and this is what's really prepared him for today's role of artfully handling questions of all types, he served as the spokesman at the Department of State from March 2011, where we thought of him as an extremely active Twitter diplomat. Many of you know PJ appears frequently as commentator on various television networks as regular columnist for the Daily Beast, and we are very, very fortunate to have him with us today. So please give me a round of applause to welcome PJ Crowley. I'm going to hand the podium over to PJ right now, but I would be remiss if I didn't also give a special thanks to our partners on this blogs and bullets series, as I said, it's been going on for four years. And they have really shaped this project through a combination of terrific expertise, but also great imagination. The George Washington University and AU team here, special thanks to Mark Lynch, Shauna Day and Dean Freeland, who you'll hear from now, they have been super partners throughout the entire series. And on behalf of USIP, I really can't thank them enough either. So with that, let me turn to you PJ. Sheldon, thank you very much. Thank you all for joining us here this morning. I was kind of there at the creation as we tried to figure out what social media were and how to integrate social media into our public diplomacy program and what impact it would have on the conduct of diplomacy. Obviously, something that former Secretary of State Hillary Clinton embraced and John Kerry, both in his own voice and as Secretary of State, is now the user of Twitter and other social media and recognize it's important. But we are still in the midst of a remarkable of, you know, five year beyond trends in fundamentally transforming yet again, the environment in terms of use of technology, the information environment, and the impact of that is having on various transformative events throughout the world. We saw the introduction in 2009 of Twitter as an element, large or small in events in Iran. Obviously, we have been studying its impact throughout the so called Arab Spring or Arab uprisings and even try to assess what has happened most recently in the dynamics in Turkey and in Ukraine. So we have an extraordinary panel. My job is to largely get out of the way, but we were still trying to figure out, you know, what does this mean? Social media, even of itself, integrated in a symbiotic relationship with traditional media, how populations are reacting to this tool, how governments are reacting to this tool. But first and foremost, as Sheldon said, you know, quite well, have to try to sort out perception of reality, what is actually happening? What's the dynamic? And what are the implications both in the media term and the long term? So we have an extraordinary group here that has been responsible for this report. And we're going to begin our conversation with Dean Freelon, assistant professor, I'm sorry, we're going to begin our conversation with Sean Abe, the director of the Institute for Public Diplomacy and Global Communication. Thanks, PJ. First of all, on behalf of Mark and Dean, I want to thank Sheldon and USIP for what's been a really wonderful partnership over the years. This has been a fascinating project. And we've had a number of collaborators over the years, including Mark and mine's colleagues, John Sides and Henry Ferrell, Ethan Sikerman, and John Kelly from Berkman, and a number of others who worked in one capacitor and other with us or participated in workshops that USIP and GW have hosted. And it's just been a great time to be studying this. If you think about when we first our first report four or five years ago was about the Green Revolution in 2009 in Iran. And since then, not only have new media changed, but the role of new media, their uses by protesters by regimes, the intersection of new media and traditional media, all of these things have evolved so much that it's been a great time to be in the field because we are constantly updating what it is that we think we know and what we know we don't know anymore. And that's been really exciting. How people, protesters, regimes, etc. use the same media over the course of this time, whether it's Twitter or Facebook or traditional media has changed and is changing. And I think one of the things that our reports have always tried to do is say, we aren't giving a definitive answer that will hold forever. What we're trying to do is say what it looks like now, and where it might go. And I think this report more than ever maybe takes that spirit at hand, because what we saw with this report, because we looked at over a long period of time, Gene will be talking more about the specifics of the data in a minute. But we saw how YouTube videos, for instance, the use of them by protesters and activists and opposition forces and by the regime and by traditional media changed as the nature of the conflict changed as the nature of YouTube itself changed as Twitter changed. We saw changes in the uses and effects of all of these different media and the intersection there. And that's been really fascinating. And as we look forward in our research program, it's a really exciting time to start going back and relooking at some of the things we looked at before. Sheldon mentioned the Arab Spring, we did a report on that a year ago. But the Arab Spring is obviously a very different story now than it was in 2011. And it's a very exciting time as researchers to go back and look at what went wrong, what's gone right and what that role of media has been. There was a very heady time, if you remember, when people were very excited about the role of the potential role of social media in particular in helping foment and peaceful revolution. And yet obviously things are very different now. But one thing that remains the same and presumably always will is that media, new or old, digital or terrestrial, whatever kind of media we're talking about is still the primary lens by which outside public's witness our country's internal struggles and also increasingly how people inside a country share information with each other, see things that in the past they wouldn't have been able to see, whether that's regime abuses or really great empowering examples of other people standing up to those abuses. But we're also seeing other things. We're also seeing ways in which, and this will be I think part of the story of the second panel as well, we're also seeing ways in which social media in particular, in part just because of its ubiquity, its ability to get around firewalls and its use in traditional media is also a way in which protesters in one country are able to learn from protesters in another country. We tend to frame that discussion in terms of a positive, oh they're learning how to protest, they're becoming a power. This was part of the story in early 2011 that we looked at in our report was a claim that there was sort of a domino of protest that happened from Tunisia to Egypt and on. We didn't find a lot of evidence of that per se, but it's worth pointing out that there's at least some evidence coming out over the last few months, especially the last month, that protesters also learn other things. They learn not only nonviolent means, but they learn some ways to have violent protest, how to create a proper Molotov cocktail for instance is something that we've read about over the last few weeks. So there are a variety of ways in which the story is far more complex than sometimes we'd like to discuss or admit and far more complex than we see in the mainstream media. One of the things that is interesting about the Syria case that we report on here is that we saw a dramatic change during the time period that we looked at in traditional media, mainstream media's access inside Syria. So as researchers it's interesting because it's like a quasi-experiment, right? So we're able to see a period in which mainstream media journalists were by and large not able to get inside the country. There were some that did, there were some who died doing so. But generally speaking they didn't have the kind of access that they did in say even the Iraq war or some other traditional environment or in the other instances of the Arab Spring. Later they gained access but it was a limited access usually under escort from certain opposition groups to battlefields or areas of conflict. What we were able to see in our data is how that changed the story in traditional media but also how it changed the interaction between traditional media and online media. For instance one of the things we found was that when traditional media were outside they were much more reliant on the kinds of citizen journalism, opposition created journalism online then they were later when they could use their own eyes to do their reporting. Obviously reporters prefer to be able to report on what they see. It makes it seem more real to them. Issues of verification seem easier but there are other ways in which this is a kind of bias. It's a kind of bias that really fits with certain news norms that we see in other other sorts of situations, other wars, conflicts, etc. In Syria case it was interesting because when journalists didn't have access that really empowered activists and citizen journalists and opposition groups to get their information out. But it placed a premium on the traditional journalists in establishing legitimacy and veracity of those online videos let's say. Which is very difficult to do when you're not there. How do you know what the legitimate group is? On the other hand you because you weren't in a town where you could see with your own eyes what was happening and you had the bias of reporting on what you could see in some ways the story was broader because you were seeing a broader field let's say. Right? The battle space was bigger but so was the protest space. You saw a lot more nonviolent protests. There were more of those sorts of stories in part because that was the nature of the conflict early. There was more of that but also in part because journalists weren't only seeing conflict when journalists gained access it was typically a at a stage in the conflict where there was just more conflict but also it was typically under escort we would in the Iraq case would have said embedded in a way and in some ways it resembles the embedding story in terms of the nature of the coverage but it was a more limited view. We see evidence when that happens of journalists doing less to check the credibility of the sources that they're using the online sources etc we also see them reporting on the narrowly on what they see as opposed to what is out there right so the story doesn't necessarily change in terms of what's actually happening in Syria what changes is what reporters report on and that's not that's not necessarily a lie it's not a lie it's not necessarily wrong but it is something that's different that's not exactly explained by events right one of the stories we heard at a year ago or earlier when we had our last big conference here at USIP was Deb Amos of NPR pointing out that when journalists gained access inside the country they stopped reporting on certain cities where there was a lot of conflict and and that's not because the conflict stopped it's because they weren't looking at it anymore how they looked at it before was virtually but now that they could look at things with their own eyes that changed at a later workshop that we had at Stanford as part of this project Ivan Siegel from Global Voices pointed out that there was still a large protest nonviolent protest movement going on in Syria when most of the reporting was all about conflict and fighting and so it's a different story and it changes the intersection and the way we understand the blending of new media and old media and one of the things that we found over the course of the time of this research project is just that that those boundaries in between old and new media don't exist the way they did even just four or five years ago and understanding that intersection whether it's the Al Jazeera integrated social media during the Tahrir Square protests in 2011 or what we saw during our Syria study that's an important part of the story as well I want to turn it to Dean to talk about the research side sort of the data and and how that went and then to and then to Mark certainly I need to come up here to the podium so that I can advance the slides here I'm going to talk a little bit about the highlights of our report you can read the full data analysis in the report itself so first you know I want to say that we bought our data from a twitter authorized vendor which was able to provide us with complete data for the following search queries which was we looked for the term Syria it's equivalent in Arabic between January 1st 2011 and April 30th 2013 we later after the report extended that data frame time frame until the end of August as you'll see on the next slide but that's what we got it's complete data for those particular terms which lifted up above some twitter analyses that only use publicly available APIs which limit the amount of data that you can pull out from twitter and we also got full metadata for that which means that we captured information such as the tweets author it's date and time the application that was used to post a tweet as well as the predominant language that was used in the tweet okay so what you're looking at here is a chart of the percentages of English and Arabic tweets over time with English in red and Arabic in blue and you can see early on if you look at the left that English dominates the Syria conversation until about June 2011 where they reached rough parody after which Arabic dominates to the tune of above 60 percent all the way through until August 2013 where it reaches just about parody again that's when Obama was considering a military strike against Syria during that month and so the point I think that this graph shows is that the the Arabic spring narrative that twitter was mostly used by far-flung English speaking users rather than Arabic speaking users who are physically and emotionally closer to the action evaporated almost immediately after the initial events of the of the Arab Spring so things right after we published the results of the blogs and bullets to report the situation on the ground as it were changed and and only a major news event like Obama's consideration of a military strike on Syria could was sufficient to change that ratio to a significant degree okay so this next chart is a chart of the number of tweets in English and Arabic over time and it clearly shows that the event driven nature of communication activity about Syria on twitter notice the difference between the events that create spikes in Arabic language and English language tweets they both spike when events circumstances involve the the US but the latter that is oops look at this come back oh there we are but the Arabic language tweets spike were major events occur on the ground that don't necessarily involve the US so near the beginning of the duration we see double spikes in both English and Arabic when Obama calls on Assad to step down and in February 2012 when the US shuts down its Syrian embassy and there are also spikes in both English and Arabic toward the end for two events involving Assad's use of chemical weapons but if you take a look at the the two massacres in the middle in May 2012 and in July 2012 you see major spikes in Arabic but not in English indicating that those events are closer to the emotions as well as also you know the physical locations of the individuals who are tweeting about that in Arabic as opposed to the English language so in addition to this language analysis we also clustered twitter users based on who they retweeted so what you're looking at here is a group of network clusters from March 2013 so this is just an example which shows which groups of individuals attracted the most attention during that month English-speaking journalists lead the pink cluster at the top the colors and the positioning of this are arbitrary but the distances do represent the degree of separation between these clusters so that's the English journalist cluster at the top while the other clusters all of which predominantly speak Arabic are much closer to one another you'll note also that only one of these Arabic-speaking clusters is Pro-Assad and that the others support one or more factions among the opposition there okay last slide so let's take a look at how these clusters changed in terms of how open they were to outside information so what you're looking at here is a chart in which the lower the line on the lower the lines on these chart drops the more cluster members retweet only one another as opposed to others outside of their cluster so the major finding here is if you look toward the bottom the blue line that slopes down that's the English journalist cluster going increasingly isolated over time um but you'll see that among the Arabic language clusters there's a lot of variation so the all basalt cluster which is based inside of Arabia kind of goes up and down as does the one for the Syrian opposition Kuwait actually stays fairly even right between point two and point negative point two and negative point four Al Jazeera has a little bit of a decline and if you look at Pro-Assad down there at the lower right it sort of starts low in September 2012 and stays fairly low in terms of being very isolated and insular now these patterns are almost certainly a consequence of both language and ideology and we're currently digging deeper into the data to discover what else might play a role in determining how the parties to this conversation direct their attention so I'll hand it over to Mark all right is this is this on can you hear me okay so first off I want to to really give a shout out to Dean Freeland who's been the backbone of a lot of this research and I think it's I think one of the things that we've learned as we've gone on with this is that basically to do the kind of research which I think really needs to be done you need to have partnerships now you need to people have people with unique skill set Dean able to do the hard the hardcore data analysis people speak Arabic or the local languages people who understand the you know the linkages and the the media communication side I think that in the past I've always been more the lone wolf on research and you can't do it anymore and I think that this has been I've learned an enormous amount from working with Dean over the last six months what he presented is what was in the blogs and votes report what he didn't tell you is that this is based primarily and we did a kind of early snapshot slices of six month intervals now we're doing the full data set all 32 months and we're also going a lot deeper and so the qualitative analysis that you saw the hints up there was based on my qualitative coding of 200 the top 250 most retweeted tweets so now he's gone from the top five clusters to the top 10 clusters and all kinds of fascinating new information comes out and I've gone from the top 250 tweets to the top 5000 tweets and we've learned an enormous amount if you by looking at the top 5000 most retweeted tweets that captures about just over 10 million tweets out of the 30 million roughly that are relevant and so that's what a lot of this is based on I just want to bring out there's a lot of things that we could talk about and I think that some of the things that Sean and Dean have mentioned are at the core of it I want to focus on just a couple of the what I see is the key takeaways from this deeper analysis a little bit going beyond via the short report you read the report there's a lot of important stuff in there we've gone up beyond it a little bit on the isolation of English journalists I mean this is a very interesting and important finding of course that the vast majority of what we we being the collective American policy community look at is you know the the discussion among a relatively small group of English language journalists and analysts and the and then English speaking Arab bridges who then translate the the the discussion on the ground or the discussion in the broader Arabic-speaking community out for that audience and I think that one of the important points to flag from Sean is that there's absolutely no reason to assume that this is an innocent or a non-biased bridging function that people bridge because they want to accomplish something if you were a Syrian activist and you were trying to build support for international intervention in Syria or the funding of the Free Syrian Army of course you were going to highlight the peaceful pro-American nature of the Free Syrian Army and downplay sectarian acts why wouldn't you I mean this this should be built in to to people's expectations one finding when we first presented the finding about the increasing isolation and marginality of the English-speaking Twitter community people's I think a lot of people came back at us with the comment that wow the the English speakers are really you know they're really unimportant that we shouldn't even be paying attention to them and I was entirely wrong because of course they are within the English-speaking community they are setting the agenda they are reporting things they are helping to determine what information filters from the online stuff into mainstream media the one I think one of the key findings here is the enormous power which you have if you are one of the privileged bridges between clusters and that doesn't only go for the Arabic into the English it also goes between the the Arabic-speaking clusters the research that we're doing now we're trying to identify specific individuals specific accounts that have a disproportionate impact on the understandings going across these increasingly insular and increasingly clustered dialogue communities and now one of the things that this allows you to do to the extent that you have highly insular insular clusters is you would think that it allows you to narrow cast as they say in other words one message to one community a different message to another community and so for example maybe you want to be able to show your English-speaking potential backers in the west that you are a civic nonviolent democratic movement but you want to signal to your your potential financial backers in the Gulf that you are you know a good sectarian Islamists willing to go out there and fight for the Sunni community as long as those clusters are completely insular you can do that and you can send one message here one message there but as soon as people start noticing things across those clusters your messaging starts to break down and actually that's one of the big things that happened over the course of the last year is that the pressure on actors who are involved with the Syrian opposition who are increasingly focused on attracting financial resources and political support from the Gulf they needed to highlight their military effectiveness and their ideological purity in order to attract funding from the Gulf the more that they did that the more resources flowed in but the more it complicated their messaging to the West and their messaging to the international community and I think that contrary in a sense you might think that they that as we see this increasing clustering they would be more they would be able to maintain this but in fact I don't think they were you might you all might remember the the famous thing we reported in FP and it was everywhere though the the heart-eating commander actually was a lung-eating commander showing that wow we thought these rebels were kind of freedom fighters and look this guy's crazy he's out there eating the hearts and lungs of his opponents I think this is a classic example of the breakdown of narrow casting something which was created for one community and was effective in demonstrating both military effectiveness and ideological of purity but when translated over into another discourse community looks insane and really undermines the broader perception and broader beliefs about what the about the nature of the opposition so that's one of the big things that that we found one other thing is that we here in Washington especially we we we think that the United States is really important right we think that American policy is the single most important thing about Syria and what Obama does is what everybody wants to know about you might be interested to know that with the exception of August 2013 when when we actually we're talking about that military intervention and everyone in the United States started paying attention again if you take that out the 93% of all references to Obama are in English 93% the Arabic speaking community does not care about Obama for the most part they don't think American policy is all that important they're focused on other things and this includes during the election months where the you know to the extent that Syria emerges in the English speaking Twitter community it's about you know the election and Romney and Obama debates and stuff like that doesn't even merit there's actually one of the months I think it's October of 2012 just before the election guess the number of Arabic language references to Obama that appear in the top 5,000 most retweeted tweets zero that tells you the gap between the way people in Washington are talking about this and the way people in the Arab world are talking about it okay just a few more things before we off the stage one of the things that we find which I think is we've documented this I think in a way that no one else has documented it before is that there really was an early early on an Arab spring frame there was something very real about it and we can track its degradation and its disappearance early on in those first about roughly the first six months or so up through about July roughly of 2011 there were very clear clusters of Twitter discourse that linked together different issues and different different communities there was an Al Jazeera cluster which linked together Egyptian activists Tunisian activists yet all activists from all over the place which actually had a positive insularity which is really hard to do that means you're actually having more connections outside of your own community than inside your own community Dean's going to tell me I'm wrong about my I'm not I'm not I'm not I'm not I'm basically right but but basically it's hard to do but it tells that this was a very non-insular kind of discussion there was an Egyptian activist cluster that was talking a lot about Syria the main Syrian cluster opposition linked Syrian opposition had names like revolution Syria and Syrian Jasmine was a really big account back then explicitly linking Syria to the Tunisian revolution and there was an Arab Spring cluster English translators Sultan al-Basimi Andy Carvin Blake Hounshaw were really key on that they shared content with all the other clusters we can show this all statistically by about the fall of 2011 they all disappeared completely completely Al Jazeera goes from a positive insularity to becoming one of the most insular communities by about the what the fall of 2012 roughly its insularity goes like this all the way down it crashes completely the Egyptian activist stopped talking about Syria probably because they have other things on their minds by that point that you know with their own revolution falling apart the the the Syrian opposition cluster fragments just like it does in real life into a free Syrian army cluster a Salafi cluster a whole bunch of different things so there's an early Arab Spring frame that was really powerful you can really see it clearly and then it disappears and then it goes in different directions what it turns into is really interesting and this is kind of the research that I've been doing recently that we haven't published yet but we will soon in the west in kind of English speaking community it's largely replaced by what I would call an R2P frame responsibility to protect frame focused on the question of of intervention of western intervention with lots and lots of efforts to try and highlight atrocities in this kind of regime killing innocent people pushing for no fly zones pushing for active you know various kinds of hashtag activism showing pictures of dead children hashtags help for Syria you know things like that and I think that this actually had a counterproductive effect in the sense that the more that people talked about intervention in the United States the more this triggers an Iraq frame rather than an Arab Spring frame the minute you say the word Iraq Americans run screaming and you see the public opinion polling goes along with that and again I don't think that the online stuff directly is the only thing explaining things happening in public opinion around the ground but I think it's pretty clear you talk about Arab Spring the more you saw an interest by Americans to get involved the more you had this Iraq framing the more you the more you didn't and think that that's something which really the more intervention talk the less American support you had you also had these credibility issues that Sean was talking about as people began to see things which didn't fit the frame the more they kind of lost I think confidence in the information they were seeing and so I think in the West I think the Syrian opposition I wouldn't say that they lost the media war or the framing war but I think that the shift to an intervention frame hurt them badly with key audiences at least outside the beltway on the other hand in the Gulf it went in the entirely opposite direction in these Gulf the Arabic language Gulf clusters that we're talking about the focus was not upon trying to pressure governments to intervene it was on trying to get individuals to become more actively involved in specific things that they could do to support the Syrian people and Syrian opposition factions fundraising appeals calls to demonstrate outside of the Syrian embassy calls to boycott a local Syrian business calls to retweet a particular image or a particular sermon there was a lot of imagery and rhetoric and discourse which was not just religious in focus but in but intended to trigger identity and this sense of not only that my people are being killed but I can do something concrete to help them and this is something which is just fascinating to watch as it goes along by far in the middle period after that fall 2011 by far the most frequently retweeted and most influential accounts are these Kuwaiti Islamist people like Nabila Dawadi Haja Rajmi those are two that really jump out there's a bunch of others but these guys are out there retweeting saying I'm going to be outside this mosque on Friday come by and bring clothes text money to this bank account you know things like that they were specific things linked to a common communal appeal and it was extremely effective now the effects on the ground were often that it tended to promote fragmentation among different rebel group because they're competing with each other for access to these outside resources I think it helped to drive sectarianism because these appeals often took an anti-Iran and anti-Shia tone very different from what we saw in the English language discourse far more sectarian language far more appeals to Jihad to anti-Shia rhetoric and the like and you know we can see this very clearly emerging the last point I want to make and then I really will stop is I didn't have time to turn this into a slide because this is active stuff we're doing but I broke this down into three three basic phases in the Arab Spring phase April 2011 to November 2011 a kind of a mixed phase December 2011 to July 2012 and then the full insurgency phase of July 2012 to July 2013 I left out the August 2013 month because it's so skewed by the debate about American intervention that it swamps things and it's not useful and this is I think this is really interesting in that first period this is the analysis of the top 5,000 most retweeted tweets free on a monthly basis in the Arab Spring phase Obama is mentioned a reasonable 328 times 328 tweets with the term Obama and English or Arabic Jihad is mentioned only 38 times out of 145,000 potential observations Jihad is only mentioned 38 times Iran 11,143 calls for donations 85 you know very little and regime which is a proxy for Syrian opposition activism basically attacking the Syrian regime was by far twice over 2,000 individual tweets were retweeted by far the most that's the Arab Spring phase you go from the second phase and Obama basically disappears nobody talks about Obama almost at all between December to December 2011 and July 2012 Jihad jumps up to 287 which is something but not a huge amount Iran stays constant regime drops in half down to 1400 as kind of straightforward anti-regime the activism goes down and calls for donations jumps up to over 1500 it's the single most retweeted phrase keyword that I searched in that middle period calls for donations finally you get to the full on insurgency phase and Iran jumped triples in mentions all the way up to over 3,000 mentions Jihad spikes up to over 1,270 references and regime stays constant donation cuts in half down to 722 in other words by tracking this qualitatively and doing it like this you can actually see the trends that we're all talking about I wish I had made a slide but I'm not that organized the last point I want to make just to link this for this panel to the next panel is the inspiration for doing this actually came from Xana was sitting in the second row who basically came back at us and said this is wonderful quantitative analysis but you're not really capturing what people are doing with the retweets or what they're actually saying to each other so now we're trying to answer Xanaab's critique and so we'll see if we succeeded later okay wow a lot to chew on there let me ask the panel kind of a combination of two or three different questions perhaps each of you can take one is there is there anything fundamentally new about this if you think about Robert Edmonds you know cascade model where you you have this dynamic between policymakers and elites that that drive you know policy formulation and public opinion as it relates to policy formulations or are these bridges just a new form of social elites that will play a role in the future secondly Sean you mentioned Ethan Zuckerman of MIT and his fine book Rewire and Ethan talks about the potential that social media have for this kind of global conversation but when it breaks down into your self-selecting clusters what do these slides represent is it a global conversation with threads that cut across that ink splot there or is it just a combination of isolated conversations with very little in the way of of overlap I'll take the first part of that since Bob's actually a colleague of mine at SMPA of ours at SMPA at GW and I've talked about this with him a lot so I can give you the Annie Hall answer No, that would be if he's here That would be if he's here That's true Right I will misrepresent him since he's not here he can't No, so one of the things that I know that Bob's very interested in doing since he came out with this cascade model which some of you may or may not be aware of but which PJ accurately distilled right there is how to integrate outside particularly outside the U.S. international sources non-state actors etc and so what we're doing I think is showing one of actually several different ways in which the model can be expanded to include pressures basically what the model's trying to do is explain this the cascade is basically from foreign policy elites down to publics and the various pressures and feedback loops that help explain how certain issues get on the policy and news agenda and public agenda how they stay there why they do why they don't and this constant reframing of issues and who has the power and generally speaking what we find in various kinds of research is that foreign policy elites especially the White House in the United States context have all of that power but what Bob shows is the various ways in which pressures are put on what we're doing is I think adding to it in a couple of ways at least I'll just focus on a couple real quick one is the way in which international non-state actors and other groups influence the foreign policy decision-making and framing of foreign policy issues in the United States and elsewhere so that's one two across cultural approach we're able to show I think with these data especially with what Mark was just getting at the ways in which how the United States might be talking about something or conceiving of it or framing a conflict or a crisis is completely different if not totally alien from the way others are in ways that are really really relevant for understanding how this policy is going to play out including the foreign policy choices and options of the United States and third finally understanding that you know a key part of any mass communication model about foreign policy-making and its relationship to media and public opinion involves media and what we're showing most of that research of course is in the U.S. context and what we're showing is that in a lot of very important ways the English language media the Western media was divorced from at least our reality that was particularly salient within the country of Syria within the region within the major players in the area in ways that really make you question so what version of this is true and also what's the influence of this on policy-making Mark mentioned R2P R2P is one of the more important aspects of foreign policy-making particularly in this administration well a lot of that at least in theory is going to depend not only on what the administration knows which is more than media but also on in theory at least public preferences well the public preferences are probably dictated to almost an entire degree by what they're seeing in media what how else do ordinary Americans learn about what's happening in Syria but what they're learning about in Syria is completely different maybe even than what's actually happening in Syria and I think that's a real contribution for helping us understand this foreign policy process I can take the question about what do these clusters represent so the clusters are based on retweets and unlike a lot of other research on Twitter the implications that we draw from these clusters are not dependent upon assuming that retweets mean anything in particular so a lot of research will say okay well we're going to assume that retweets mean endorsements or that retweets mean authority or what have you we assume only that you know we look only at what a retweet actually is so a definition of a retweet so what does a retweet do it redistributes information from one source to another and so we consider that a retweet is a social signal that reveals how a group of people are directing their attention and so in Twitter the way that retweets typically work and the way that the clusters in this research operate is that they are very long tailed so in other words you've got a large number of people who are in other words the leaders of the cluster right so you know a lot of leaders of the English journalist cluster and where can speak with some of the leaders of some of the Arabic clusters are very famous journalists they're on the official accounts of outlets like Reuters and New York Times and The Washington Post and CNN and everything else and so these leaders kind of constitute you know a group of people who are sort of in the long tail and retweeting them and so the clusters themselves represent what we call I mean this kind of technical term Mark actually mentioned something very similar when he was talking this idea of discourse communities right so if you think of something like a good example of this would be you know for Fox News you know the Fox News Station or for the Kohl Paribor the Kohl Baribor or the Kohl Bear Nation it's a group of people who are all paying attention to a very small number of individuals who can be labeled in an intuitive way so English journalists you know Kuwaiti preachers of religious figures the Syrian opposition all of the leaders of the vote or the vast majority of the leaders of these clusters fit under this umbrella and so over time we can see how the leaders of these clusters how the attention of their followers are distributed and the extent to which you know the the retweets in these clusters are mostly within the clusters mostly within that sort of discourse or epistemic community is no term that we use and to the extent to which the lines of retweeting and the lines of communication by extension extend to other clusters and critically this is something we didn't talk about in the talk whether the clusters that are connected are primarily clusters that are allied or clusters that cross linguistic lines or clusters perhaps you know maybe there's some sniping between clusters that oppose one another so so that's just a little about what these what these clusters actually do represent Dean has also calculated something called the IO ratio which it's just too complicated even talk about but basically what it's trying to get at is how much information is shared across clusters exactly your question and which clusters tend to share information with each other one thing which we have not been able to deal with and this again gets back to say enough to fetch key we'll be speaking later is the sub tweeting phenomenon you know when people are like saying things without directly linking or retweeting them and so there might actually be a dialogue going on Jake you're probably talking about that when you're talking about Turkey but there's like a dialogue going on we might not be able to capture and so that's one of the things that the social dynamics of retweeting that we haven't really solved for yet actually one more thing about retweeting it's really important to understand that within each cluster there's information that goes out of each cluster and information that comes into each cluster so one thing that you tend to see with the English journalist cluster especially in the beginning during the Arab spring phase is that the English journalist cluster is the source of a lot of information that people that other clusters take from them so in other words other clusters are heavily retweeting them but the cluster itself is not heavily retweeting those other clusters so you have different clusters that serve these different roles some provide primarily information to other clusters others are aggregators of information that come from other clusters and so that's another thing to keep in mind so I'll ask one more question and we'll open it up but in the the headline of your report Syria has been the most socially mediated civil conflict in history is this another form of curation in a sense or are these individuals that can bridge across clusters are they the new Walter Lippmanns or Henry Lusses who have obviously in the context of traditional media always been able to shape and influence public opinion in one direction or another in the context of war or protest I can't speak for Dean Arshad but my take on the socially mediated nature of Syria's war is that this is an environmental rather than a straight causal kind of argument where basically it's like this is something which makes Syria different from anything which has happened before that you have just so much information being recorded on camera phones and uploaded and social networks being connected and information being shared I think it matters in all kinds of ways that are often unpredictable and these are genuine network effects and so what you're describing might be one of the effects but I'm not sure that it's even the most important of them because really if you go back to like you know you rock in 1990 1990 1990-1991 during that war and think about like the information black in blackouts the coalition was able to mid-sustain think about how the Saudis were able to prevent their own citizens from even knowing that the invasion had happened go back to like you rock in like the mid-2000s you see kind of a proto type for what we're seeing now but it was still much much less than we than we have now now it's like full-blown I think all future conflicts are going to look like this and so basically you look at Syria you say environmentally this is what simple wars are going to look like in the future and so then you try and figure out what are the various effects of that might or they're going to look more like this and what it blossoms into than like the old model mm-hmm mm-hmm mm-hmm mm-hmm mm-hmm uh how should we work the microphone yeah trap people lined up with the microphones or someone pouring them around or step out step out of the microphone mm-hmm mm-hmm mm-hmm mm-hmm smile the camera on you yeah thank you very much my name is Faribo Parza George Mason University um could you tell us who are dominating groups using social media did you uh did did you make any research on about gender and political opinion my interest I'm working at George Mason as a women's activism on social media and women are very significant because in uh Middle Eastern countries women are excluded from political decision making so their activities on media it's so important so did you have any research of based on gender or political opinion thank you I can speak a little bit into that gender is tough is a tough not to crack you know we in some in some earlier research of blogs and blogs too we were we had some very nice location data where we were able to see where people were were clicking on some of these bitly links from and that was that was information that had a very high level of validity to it because they were they were pulling that straight from the the IP addresses and they had that down to the country level which was great doing gender at scale is tough for a number of reasons first of all not everyone uses their real name on twitter so you know if you're trying to judge based on the names of the individuals that might be a way that you could get some people but that would introduce some bias into the research because what if the people that there's a major differences in people who use their real names and people who use pseudonyms or who tweet from institutional accounts and it's very difficult to to figure that out now one way that we could do this and if we if we have time we might be able to do this is we've actually found it very fruitful to look at the the most retweeted tweets so these are the tweets that they have different disproportionate purchase and disproportionate degree of spread and so one thing we could do is we could say okay let's try to qualitatively analyze and code each of these names of the people who tweet these tweets and figure out how many of them are our women so you could say okay well there's a high proportion of women represented among people who make these top tweets or low portion so that would be one way of getting at that I would just add there's a couple of a couple of women who are very prominent and they show up very frequently in our month by month re-enact like Refugee Jati who's a spokesman for the LCCs Razeniat who is a prominent Syrian based activist but I think usually it's not because of their gender it's because of their institutional position within the opposition and that sort of thing and you remember the gay girl in Damascus Fiasco and you know the issues that that raised with identity and gender so I agree with Dean that this is not something that you can solve quantitatively I mean I think you'd have to actually know the identity of the accounts and dig in to specific things but there's a group over at Keras they've just done this amazing study of Aleppo and basically going household using a lot of social media data but also a lot of on the ground data and that's I think what you would have to do is to be able to dig in at the neighborhood street level to figure out who's doing what what the role is in civil society of service provision and and I think that's probably the model for how you do that we probably only have time for these four questions we'll go ahead thank you Williamans from George Washington University first of all thank you for the great report and reports and I really look forward to seeing what you guys do next with the with the more in-depth analysis I want to go back to one phrase that Mark used in particular which is in real life and my question is is there another kind of direction or another component of research that's necessary here to make the case that social media can be more of a proxy for sort of what's really going on and so when we talk about the role of social media in war and protest how much do we have to start looking at you know different sorts of methods like ethnographic research or interviews to see what's actually happening kind of on the ground and then pairing it up better with the social media stuff I know there's a lot of papers coming about the potential predictive power of social media for sort of real events but is that sort of the next step in your work and isn't that a necessary component and maybe you'd be a great book to make a better case for the role of social media in actual war and protest thank you yeah I just want to say really quickly I think that methodologically you're absolutely right on point we need ethnography we need people talking to folks on the ground to figure this out that said I don't really think that the real sort of not real distinction is super helpful I think what we might want to talk about more is this idea of distance to people who are closer to the action on the ground versus far away I think that if somebody you know if a lot of the news junkies who are in the long tail of some of these of the English journalist cluster over time are retweeting and talking about this is not that it's not real it's just that these people are not directly involved in the action they are spectators rather than participants you know likewise I think if you're looking at the Middle East you've got people who have regional concerns people who may be in Syria itself I think that their participation is real in a different way and I think that they you know what they what they have to say will be very different and could be investigated by many of the techniques that you talk about but I think we should you know the idea of distance you know further and closer I think is a sort of a better analytical place there Terry why don't we do this why don't both of you step up and ask your questions together and then we'll wrap up with one more here I'll answer you later well okay I have two questions and they're relatively broad in scope firstly you've got a lot of great data on activity so tweets retweets what's in them who sees them you know all this great metadata to I'd be interested to hear your inferences and your thoughts on not activity but perhaps inactivity so you know what could be revealed by sort of you know when there's a lapse of you know one of these key players and and you know what that might signal about you know influence between or within groups and and my second question is related to that in the sense that it was really interesting your idea of of what I would call net importers or net exporters of tweets you've got you know some people are are retweeted very constantly and then you know others they always you know take as opposed to give and I'd be interested to hear your thoughts on what that might imply for relative influence within or between clusters because certainly with you know with Syria it's you know there are a lot of these big issues of who has influence at any given moment and who's taking it and who's getting it sure thank you my name is Anna Tsaragusti I'm the director of Bitsel M.U.S.A. which is a human rights organization but I'm also an ex journalist and from what I know I mean journalism is about fact finding so how does that correspond with your findings I mean what I hear is a lot of chaos from coming from the Arab world and we don't really know what's going on there so correct me if I'm wrong if this is my impression if you have any you know findings what is really going on there well thank you maybe I'll go first because I can tie them together a little bit to set up a dean of Mark for more in depth but you know one of the important things about understanding journalism is of course the gatekeeping function of journalism an important part of that is who journalists decide are legitimate sources and that's really the one of the undercurrents of of this report is the way in which certain people or organizations became considered legitimate sources and because mainstream journalists are really the gatekeepers to the broader western world at least in this case it became very critical to understand who those people were and how those people had moved us how that evolved over time and we have you know anecdotal examples of the report for instance of something that Mark alluded to briefly earlier of you know some of the more important institutions or individuals that that knew that mainstream reporters listened to them and thought they were credible you know getting video from the ground and realizing that it didn't look good because it was sectarian for instance and not not putting that up on their site and so it got it did show up on youtube but on a site that nobody went to because they didn't consider it credible so no one saw it that telling the people they got this video from you know next time you got to make sure you guys don't have any sectarian chance so the next time there was a protest video they didn't have the sectarian chance but that doesn't mean they were any less sectarian so and that got published and made it into the mainstream press so you know it's a very difficult job you know your former journalists it's a very difficult job to understand who these people are and to make sense of it and we saw you know a lot of very important curators on the the mainstream media side going to really great effort to try and do this to establish credibility to to be clear when they couldn't and that sort of thing whether or not we're getting the truth on the ground you know as a relative terms gets sort of at what will was saying but it's I think what's interesting is that we're getting a lot of different truths and and there are I mean I don't mean to sound postmodern matters I don't mean it that way I mean that there are different things that are happening and I think it's far more interesting in a way again getting back to Will a little bit that what we see in the English language media particularly over the course of our study really becomes not only insulated but very different but it's not it's not fake it's a real thing that they're showing us it's just not this other thing and everybody else in the region seems to be looking at the other thing and that's sort of understanding that daykeeping role I think is it's really critical yeah can I do one of the things interesting about the Syria kind of the Twitter the online universe around Syria compared to say Egypt is that it's far more curated and far more organized I mean in Egypt you had a lot of people who were out there kind of as individuals who were so if you wanted to know what you know Hussam Hummelaoui is doing or the April 6th movement is doing in Egypt well these are the founders and organizers and key activists of those movements and they're telling you what they're doing and in a sense there I think there was there was a less intermediation there whereas in the Syria what you get is I think a far more organized campaign type of curation of information like Sean was talking about and so for example you see one of the most frequently retweeted tweets across the entire again and again and again it's just a very basic these accounts they'll say you know the the Twitter army you know please help us in spreading this information and this will be retweeted thousand just the appeal for help will be retweeted thousands and thousands of times and then they'll choose a hashtag and they'll choose the information and they'll put it out very different from Egypt which is I think far more anarchic and when I was a skender is here he can talk about it but it's it's very different so from a journalist point of view I think you're getting from from the Syria side you're getting what various organizations want you to get for the most part whereas in Egypt you were getting a whole blanket of this kind of chaotic everything is happening on on on Tarfur Square this also gets to Will's point a little bit about kind of the what's happening in real life I would only add to Will Thammy and ask her now is a different question you're asking because a few weeks ago I was printing a paper based in part on on on Dean's and network data to an audience of kind of civil war political scientists and they came back and they said well that's really nice but what does any of it have to do with the civil war and it's like okay well if what you're trying to answer is what's the network structure of Syrian discourse online Dean's study is really relevant if what you're trying to answer is what explains rebel fragmentation in Syria maybe not so relevant so I guess the question is really disciplinary and you know what question you're trying to answer can I speak very quickly to the center perceived question between the clusters and what that implies for influence and all of that so there's trade-offs right so if you have a cluster for example that primarily is providing information to other clusters and other individuals outside of that cluster that tends to look more like a mass media model of something that pushes out information but does not pull much in sort of the it's going out you know you're watching television in that model it's going out but nothing can really come in on the flip side of that when you've got people who are pulling in information but not putting much out if you're doing it from a number of diverse sources what you're looking at is a model sort of the ideally informed citizen people who are getting information from a lot of different sources but they're not getting their message out now I think whether that's good or bad depends on what your goals are but but I think that there are there are trade-offs and of course you can have perhaps they would be both having your information but go out in large volumes and come in in large volumes or sort of at the other you know if you could think of that this is sort of a grid on the other corner you've got very little information that's going out and very little that's coming in and that's sort of maximum celerity so so thinking about it in those terms I think helps to tease out some of the normative and is there a hybrid in between? well so right so the the hybrids are you know so you think about the I should have put this up but you think about it in terms of many in few and then sent and received so you've got many sent many received that's you know you've got you're you're being well informed as well as you're projecting out when you've got sort of a lot that is sent a few that are received you've mass media when you've got lots that are received and few sent you're sort of the ideal citizen playing information and then when you've got the last one that is very insular very little going out and very little coming in great thank you very much my name is Maria Stefan I was formally at State Department working on Syria now here at the Institute as a fellow my question is about bridging clusters in the Syria context not bridging between Arabic and English but the bridge that I mean is pro-regime pro-opposition did you see any such clusters in your research how large were they what were they saying the Zainab question did they translate into any action was there a difference between such clusters if they existed when the conflict was principally nonviolent versus violent and what can we learn about bridging clusters and other highly polarized conflict societies thanks sure I'll start with that one the assigned cluster actually I didn't back up so our analysis looked at the top 10 most retweeted clusters or actually the top 10 most populous clusters I should say within each month and so the way the algorithm works is it decomposes a large number of clusters over a well over a thousand in the smallest month and well over 10,000 in the largest months and so the way it works is you have a very small number of very large clusters and a very large number of very small clusters and so cutting it off at the top 10 means that we're looking at the the largest clusters within that month and so within that framework the prophesied cluster emerged really late in the analysis I think it was late 2012 in which we got where we first saw it so the probably was a lot of prophesied clusters that was happening activity that's happening below that cutoff but it really only rose to that top 10 arena in late 2012 a second thing I want to say is that we haven't looked at the bridges there yet but I just wrote the code last weekend to analyze these individuals and pull out these individuals who are heavily retweeted by multiple clusters when we break that down we can see whether the assigned cluster is heavily linked to other ones which ones so far we've seen that the assigned cluster is very very insular very closed in terms of how it interacts with other ones so there may be very few individuals who are able to bridge to bridge those in a substantial way let me just say qualitatively about this about the prophesied cluster it's not a regime cluster per se and what's fascinating about it is it links together when it emerges identifiably Shia individuals and figures from across the Gulf Lebanon and Syria along with regime supporters and I would actually go so far as to guess that a number of people who are like mainstays fixtures of that cluster would react violently in the suggestion this was a pro-assad cluster in other words they would say that we are anti-opposition and that but we are not pro-assad and so it's very interesting the dynamics of that of that cluster but it's extremely insular and to the extent of this interaction going on between them when I see it it's almost always in this kind of hostile interaction kind of way like making fun of each other it's like oh look what they're saying now can you believe these people like that sort of thing we're close to intruding on the start of the second panel but we've been monitoring this on yeah I'm just I'm here and I'm here on behalf of the online audience extremely active audience and I'm just gonna ask one question about the English language journalists and how insular they are one are they really aware of this insularity and two is this having any impact on policy and what impact is that having right just to just really quickly on this this is some late-breaking analysis that I just completed last week basically what happens what happens with the so the English language cluster becomes more insular over time what happens is that actually coincides with a dramatic increase in the number of Arabic speakers on Twitter so what ends up happening in this in the Syria conversation is that from the start of the era of spring to the end of this time period the number of Arabic-speaking users they're talking that's specifically they're talking about Syria explodes and so what happens is the English cluster you know gets busy gets crowded out of the conversation and they get sort of corralled into their own little cluster until that last month of August 2012 when what happens is a bunch of other people that aren't normally talking about Syria all of a sudden start talking about it so in that August 2012 cluster whereas all along you've had this English journalist cluster you've got American conservatives you've got American liberals you've got some British folks at a couple and I think one other one and so what ends up happening is that a lot of the these sort of interconnection and inner conversation happens on the Arabic speaking level the the English accounts that tend to break through through a lot of this are like Sarah Palin Glenn Beck it wasn't Rihanna it was so Nellie Portado or someone like that when she retweeted on it like things like that are when it tends to break through for a lot of that period and it's often in a very partisan way there's one very popular tweet that was you know about Obama running guns to the Syrians just like fast and fury like very partisan like a lot of people doing the exact saying tweets again and again a lot of stuff like that in English speaking in that little that little cluster yeah I don't think English language journalists are because of what Dean said about this coinciding with the you know influx of the Arabic language tweets I think they're aware of what is that they're not aware of I think they just keep doing what they're doing how it affects policymakers you know I presumably policy makers have access to intelligence and all sorts of other things that influence their decision making but to the extent that public opinion has some influence we don't want exaggeration but has some influence on policy makers that the public is going to get that information and process it through of course their partisanship and everything else but this is an issue that really divides easily on partisan lines in the United States and so the information that they're getting from the media are particularly maybe more salient than in normal formal policy crises so it's all the more interesting and important to understand what that story is that they're seeing and how it may or may not be different from the story's other foreseeing and then can I say one other thing on the policy question is that there's two actually two different ways to read the data that we have here one would be to say that you know the English speaking journalists and the Washington policy community are convinced that American policy is the most important thing that matters in Syria and yet the data shows that in fact very few people actually care about it all outside of Washington that would be one way of telling the story the other way of telling the story would be looking at that August 2013 spike when Obama talks about bombing Syria and suddenly it goes through the roof and then the second way of interpreting the data would be to say America was absent it wasn't leading therefore nobody paid attention to it but as soon as we talked about intervening suddenly it becomes the main story again so I don't think there's any obvious I mean there's multiple ways you can interpret the politics or the policy impact of America's relative absence from these Twitter discussions well the great thing about the great measure of a panel is that you're forced because of time to end the panel before all the questions have been answered but join me in thanking Sean, Dean and Mark for a wonderful panel discussion we'll take a 10 minute break and then reconvene all right we're gonna get started now I'm gonna turn it over to our moderator PJ Crowley for another great discussion I come cheap so if we have some fast in perspective on the Arab uprisings at now into their third year we can we can add to that perspective but also compare what we've seen over the past three years what we've seen over the past several months the great thing about being professors in the space is that you have no shortage of case studies so here we have Adele Escondaire adjunct instructor communication culture and technology at Georgetown University we have Zainab Tufeki assistant professor University of North Carolina Chapel Hill a Duke fans in the audience you can get an early line on the March Madness and Joshua Tucker professor of politics at New York University so we're going to do Egypt Turkey and a most significant and topicality Ukraine so I'll take it away thank you well first and foremost thank you thanks so much for for having me it's a pleasure and a delight I'm honored to be in the company of colleagues and it's it's always difficult to talk about Egypt and the social media largely because Egypt gets a significant amount of attention and and fanfare when it comes to whether it's you know discussions of Twitter and Facebook Egypt is is heralded as a success story or at least had been for a significant period of time and so much has happened in an extremely short three years and and if one were to look at Egypt as as a case study it's a case study for both the social media flourishing as well as faltering at the same time so we have sort of moments of jubilation and moments of absolute melancholy and and real sort of disconcerting evidence I of course I don't do sort of the kind of research that that that Mark and Dean do so I can't really speak of the actual sort of patterns of distribution and and content creation but I can tell you a little bit about how the political quagmire in Egypt helped precipitate shifts within the social media and vice versa like the sort of the strange homeostatic relationship between what happens online and what happens offline so I'm going to skip over what you probably know and have been following in the first two years from 2011 to the the current moment and we have some pretty startling transformations and in both patterns of use in the social media as well as the political climate that underlies the first first and foremost we often talk about the Egyptian social media scene as being sort of indicative of a very sharp digital divide that's a significant proportion of the population is disconnected from from online spheres that are not part of the dialogue and and therefore they're completely marginalized this is starting to shift significantly as we see a sharp rise and a proliferation of use of of the internet in about a year and a half so that of course is very startling but what that does is to a large extent open up and in a sense also egalitarianize the distribution of information online by egalitarianizing I have you know I mean to an extent that the the the voices that we typically used to hear from online are also changing they're changing because their convictions their views their perspectives on the political circumstances are changing as politics become much more polarized and much more and much more contentious but but egalitarianize also from the sense in the sense that we have entire constituencies entire sort of social classes within Egyptian society that are making their way online they may they may they may not have been the early adopters but by the time they they become entrenched in this community they are providing sort of a whole new sort of lens analytical lens on what's happening in the social media so there so the social media environment in Egypt is becoming much more diversified and it also sort of shifts the bias if you will that we typically associated with the Egyptian social media we often said the social the social media in Egypt is predominantly activist predominantly revolutionary that is no longer the case in fact it's one might argue that it's quite the contrary now the the cluster the major clusters are the major nodes that we're accustomed to hearing from they were sort of the key figures the informants if you will from Egypt the prominent journalists and and reporters and and activists are still the individuals that we glean our information from here in the West the folks who are actively actively translating content into English to make it comprehensible for Western journalists and Western media and Western policymakers they're not necessarily the trend centers in the Egyptian social media seen anymore quite the contrary where now there are there's a whole sort of parallel realm another echo chamber if you will that is proliferating and growing quite significantly that doesn't necessarily care to speak to those communities or to send those messages internationally and those and this community is also diversifying and it may in many instances be very counter-revolutionary and very critical of revolution the other sort of complicated dynamic is that we've seen a major shift in the way in which discourse is expressed online whereas in the past or at least in 2011 we associated most voices online to be dissident voices today they are the dissident voices may be the much much more silent or much quieter so today's sort of government propagandists are sometimes tomorrow's dissidents and today's revolutionaries are sometimes tomorrow's apologists for government and that happens in the ebbs and flows of political transformation from a time when the supreme council for the armed forces was in power and we knew who the contrarians were against them and then of course and we knew who the apologists on the behalf were and then again when with the muslim brotherhood in power we have we see that shift again and with the removal of morsi and the muslim brotherhood from power on july 3rd we see another sort of another page turn and so it's extremely important to note that any sort of transitional interpretation of what happens in the egyptian social media scene has to be historicized you have to understand that this is happening at such moment such and such individual said this on this day and it has to be sort of contextualized within that realm because a month and a half later we can see a complete transformation not only in this individual's expression but the implications the political implications of this expression there is a significant at least nowadays we see a significant decline in in in politics on online politics I mean politics in the way that we grown accustomed to it in the year in the in 2012 in the year when the muslim brotherhood were in power and it was a significant level of opposition the sort of the contentious nature of online politics led to a real sort of burgeoning space for expression today that space is narrowing and it may be narrowing because of a sort of a spiral of silence to use sort of communication scholarship that those who feel that their opinion is in the minority tend to be quieter and then there's a sort of a chorus of criticism that shuts down particular spaces of these online online discussions what I'm particularly interested in if we're if we're not going to look at dissidents necessarily in the way that we used to think of it which is you know there's a protest an X and XYZ location it leaves in such and such time we're all going to mobilize from from here to there and and the use of social media as a as a portal for for collective collective action there's there's another form of dissent if you will that I think is much more creative which is basically a shift from looking at the use of social media as politics versus leisure I mean we often think of how you know we use social media here or the majority of the population use social media here versus in Egypt and there's a tendency to look at the Egyptian social media users as highly politicized so there's always this sort of challenge politics versus leisure but now I think there's politics as leisure most of the critique against the sitting government now and arguably the one that preceded it that of the Muslim Brotherhood or the military back government now actually uses humor and I don't mean humor in the sort of in the cliche sense that all Egyptians are funny and they all crack jokes and we all know about you know the Egyptian sense of humor all of that is is a given I I expect that to be the case but more importantly Egyptians are are effectively leveling some significant political critique in the form of humor and this humor manifests in very interesting ways whether it's in the form of sort of user-generated content like memes for instance which distribute very widely I mean they're not you know they're not frivolous it's not frivolous you know interventions these are really significant political interventions that that can be extremely damning for the institutions in power but the question is who's laughing and that's I think the critical point so even though humor and satire and parody is being used as a form of political critique the compartmentalization and the real polarization in Egyptian political scene means that a joke can be funny for one community and really just an extremely abrasive and offensive for another and so to look at the sort of the landscape and the spectrum of political expression online today especially in the context of humor we can really understand what all of this means I'm sort of you know in a very crude way a very simplistic way I think of the Egyptian political scene online as sort of a sort of a very very large but you know extremely simple because I'm not a a quantitative researcher and I don't really understand mathematical models very well but like a Venn diagram is like great school right but like a really really big Venn diagram with various sort of overlapping circles and each of those communities share something in common both politically and in terms of sort of and in terms of expression what they find funny or what they find interesting or what they find sort of subject to critique but the one thing that most most of them sort of have in common is that little that little sort of central coordinate of overlap which is that each member of a community online believes with conviction that they are part of a revolutionary movement whether they are you know members of the April 6th movement youth movement or if they believe they are members of the Muslim Brotherhood who are actively pushing against against the coup in Egypt or if they're even sort of former Mubarak supporters in Egypt people call the Falul they too participated in a sort of an act of political populism against the sitting government and they too believe that they are revolutionaries so revolutionary discourse effectively subsumes all public expression in social media but the the real point of contention is who's a revolutionary and that I think is speaks to the core of the polarization that exists in Egypt today and that's what makes it extremely difficult for any for any one community to really speak on behalf of a revolution versus a counter-revolution the state versus versus non-state actors civil society versus institutions of power so this is this is the realm that we're in today which is extremely different from what was an extremely it was a very very simplistic dichotomy a binary between you know good and evil as in them the state versus versus non-state actors civil society versus versus the military and what have you so unfortunately that's where we are but it also leaves room for some really interesting analytical and contextual work on how communities express themselves and what they mean by these expressions and to what extent they can they can further change or push for change on an incremental level both online and effectively or eventually offline I have a few slides so what happened on social media and offline also in Gezi a protest in Turkey and I want to also bring it back to a few of this methodological questions we're all trying to grapple with you know what happens when you get Twitter big data sets and how do you think about them so in Turkey was what is referred to now as penguin media is that there is massive censorship in television in mass media in broadcasts it's a very complicated form of censorship it is a mix of a lot of large conglomerates basically buying big media mass media as a ways to grease their contracts in other sectors of the economy with the government so any energy conglomerate will go by a major television station which will likely be a loss making it won't make money for them but they will use it to do very pro-government broadcasts which then they look good to the government which then helps them get contracts from the government in the other sectors they're active in and there's also new sort of reports coming out of direct government intervention so but it's a complicated scene right so the government intervention and self-censorship by mass media creates an environment in which a lot of sensitive news did not make it to mass media first it made it to social media first the gezi protest was one of multiple such events so it didn't just come out of the blue there had been the military had accidentally bombed Kurdish smugglers in the southeast region and the news first broke on social media even though the television newsrooms were aware of the news about a year before gezi and they were the newsrooms the journals were following and they were kind of unclear on whether they should go ahead and broadcast this news which of course is a major event right 33 36 people died it was a very unfortunate incident because the news first broke on social media people started using social media as a way of getting around the television the mass media sort of self-imposed sort sometimes and sometimes government intervened blockades on news so when the gezi protest started as a very small incident overall spark that's not that big a deal but it became symbolic and long story short a lot of people started hearing that there were clashes in this region in this biggest square in istanbul basically one of the central squares but that wasn't on televisions the clashes start grow grew and grew while it still wasn't on televisions if you turned on your television you weren't getting the news that in the major you know in the biggest city in the country in the biggest square there were 24 hour 36 hour long non-stop clashes with the police and protesters which is major news so what happened was the cnn was showing penguin documentaries at the same time as the clashes had gotten so tumultuous that cnn international was broadcasting live from them so somebody put their television side by side and said here's cnn turkey here's cnn international and that became the symbol of this unlikely symbol the penguins in fact the next week you know people started calling cnn and requesting documentaries for penguins as the city was trying to say so documentation of the site documented multiple times there would be something major going on which would be considered news in any other context and you would turn on television and you wouldn't see it on your television so and the you would watch I watched personally interviews with major government you know officials on Turkish television where they wouldn't be asked any serious questions all the questions were softball so that created this distrust in certain sectors of society that television wasn't covering the range of news that you would normally be seeing so non penguin media makes this resurgence this is occupy gezi pain these became sort of symbol symbols that you saw everywhere and remember Turkey is a very wired country like unlike sometimes in Egypt when we're talking about the Egyptian social media scene there's a lot of sort of like how many people are these while Turkey is a pretty wired country health to countries online and this is another misunderstanding because our prime minister made the statement that Twitter's a menace to society he made it in a couple days into Gezi people sometimes think social media is only anti-government people absolutely not pro-government people anti-government people pretty much a large section of the politically engaged society in Turkey is online is on social media it has one of the highest use of Twitter in the world so if you have this notion that the ruling party is you know the major main people in the ruling party are not online are not using social media are not up to date on how it works you're wrong there's very much a lively public sphere of its own kind there so so this also became like this the symbol this is of course Bansky made into the Turkish version it says the diran is a Turkish word that's came to be used like akifay so this the penguin media and the non-penguin media became a major source of contention now the data so I went there I didn't just want to watch it online and unlike scraping I needed a helmet because there's a lot of tear gas canisters being lobbed into the park occasionally and tear gas well awful is not likely to kill you unless it hits you in your you know in your head and since I'm an academic that's the part of my body I really need so I only need a tape recorder and my helmet interviews for well basically till the park was evacuated by police force at the end of that I was there and the whole thing like Egypt where the internet was unplugged in fact the internet was functioning very well during the park the whole time now this might have been also due to surveillance concerns you know the government because in Turkey if you need to get a cell phone which my cell phone is a Turkish cell phone you need to give your citizen ID so there's a sort of the flip side might be that it's a surveillance environment but at the whole time the internet worked really well people were able to sort of use it and it didn't happen in Egypt happened in Turkey for example people don't think that people used Twitter to coordinate action on the ground like during a protest they did I mean I was there and people would very much because partly because again some more modern country better off country a lot of people have for smartphones so what would happen is there would be a massive amount of on the ground real time coordination people did using social media and so sometimes I would get people from abroad saying what's happening in Turkey they would ask me the prime minister would make a statement and they would ask me what's going on and I'd say I have no idea I'm on the ground I don't see we look on Twitter I mean look on some place else there's really a very different view from being in the middle of a sort of a protest area versus being on social media looking from top down and which is something that I want to say a lot of the sort of the analysis for example that and I love that paper I love the analysis but you have to remember there is not the team you just saw speak who's looking at tweets using the word Syria there is no eyeball that is looking at it as a whole it's good for analysis but if you're actually going to look at how the social media impacting people you have to start from the people up you can't have big data but you still have to start looking at people up because for example there's I mean I saw I heard the word crowding out and I don't think there's any crowding out really in the Syrian space what you have is different audiences that get bigger and lower smaller because there's nobody once again looking at you know there's no sort of this bird's eye view except the researchers themselves people are sort of partial everybody's seeing small portions of it it depends on who they're following what they're thinking so the story I heard is that English speaking audiences were largely tuned out besides the journalists and then they tuned back in when Obama made you know sort of statements that there might be intervention possible and they tuned back out they did not push the Syrians conversations on or off the sort of the space because they're not in competition for a limited space what's happening is from the eyeballs so I went and I interviewed these protesters I also looked at the big data sets this was quite interesting because I'm going to highlight a few things about sort of this online offline inflection because I think it sometimes gets into the nuances of how to do that so if this is from the same source you guys have actually because this is somebody else that published this is the top 20 hashtags that were associated with the Gezi protests if you look at them on you know you see this huge spike when people start using the hashtag and then you see this huge fall if you just looked at this you might think that around June 3rd the protests had died down what had actually happened was everybody stopped using the hashtags or even using the word Gezi because there's nothing else you could talk about that's just everybody was so focused on that in fact there were jokes on you know when is it going to be safe to talk about you know stuff again the turkey's onion was joking that you know the sort of the fledgling few people who wanted Instagram their food were wondering when it was going to be safe again because you would get run out of the social media room if you talked about anything but Gezi and this is true not just you know this is true for I follow a wide range of people I follow Pergaran people I follow people who are in the Gezi I have friends on sort of across the political spectrum this is what everybody was talking about therefore the word Gezi just didn't appear the word occupy Gezi didn't appear the word didn't Gezi didn't appear but you don't see that right so this is one thing another thing see every social space it has social norms I've not seen this elsewhere and I'm super curious if you guys see this elsewhere this is what Turks do this is she's a pro-government influential columnist and what she's doing is she's criticizing some people but she's not mentioning them she's not linking to them she does these screen captures so this is algorithmically invisible to any big data set now how common is this this is such a hard thing to capture what I did is I have a diverse sort of data set of people I follow and I've occasionally done things like when people discuss a lot of politics on social media which is like Egyptians it's late at night and if you you want to ask an Egyptian something 3 a.m it's usually a good time it's sort of like that in Turkey it's like you know 12 1 a.m. like chance of people about 200 people hundreds of such screen captures in an hour period so there's this enormous actual quoting talking that is algorithmically absolutely invisible and in the Turkish political social media sphere this is super common this is very very common common enough that there's no analysis possible without it it may not be common elsewhere I'm just sort of trying to point out and this one you couldn't even OCR I mean Dave you see this is kind of twisted so it would be a little hard subtweets I put an English subtweets so you guys can figure this out this is Blake who got quoted in the last panel too as a one curator he used to be stationed in Qatar and now he's the person and he's probably tweets 55 favorites you're like what the governor Chris Christie's press conference on where he blamed every aid in the building and you know their relatives for what had happened with that the scandal over traffic and bridge and everything so all Blake says is getting crowded under that bus and everybody knows what he's talking about right you have 64 people retweeting this this is a subtweet this is what they were talking about now again how common is this thing is a real question in some social media places this is very common some they're not so in Turkey this is a Turkish example that's why I showed you guys the English one we're talking at each other if you speak Turkish you know they're referring to each other maybe if you know the context but it's so hard to tell there are many times when I see these communities talking at each other subtweeting and I know they're seeing the other one because half an hour later comes a response and then the other person snipes back and then there's another one this is this is really engaging in fact I picked this one because this was so obvious to everyone not just the thousands of retweets this guy mentioned the mainstream media as them responding to each other if you knew the context and if you knew who these people are and if you read this thing you immediately knew what they were referring to but there's no names or links or you know it's kind of totally gonna see it but you are not gonna give them the link or the mention it's kind of like talking it's a new mixture of things again I have no idea how common this is in the Syrian social sphere very common in Turkey gaming hair of Ankara he is oh he's very active on Twitter again 3 a.m. you go I don't know when he sleeps right he's like constantly tweeting he's tweets himself by the way he's the pro-government in his AKPs leading people and what he decided to do was in hashtags that would trend now trending is something that Twitter doesn't tell you how things trend but people have reverse engineered it it's a inverse document frequency kind of algorithm in any case what you need to do is have a hashtag that is not a common word keep it close to and then everybody use it all at once that spike is what produces the trend otherwise Justin Bieber would trend all the time they're trying to get rid of such things so what he's doing see he there says hey hashtag stop lying CNN it trended worldwide I had people asking me well what you know because that I had friends asking me why is this trending online because it was it was a game space so because it's a human space what you are watching is a very deliberate reflexive action by people so what does it mean to look at trending so currently in Turkey we have a very active politicized social media environment once again pro-government anti-government now coalition is kind of splintering so there's like multiple factions all of whom are battling it out on social media there's a potential mass media social media divorce I'm trying to get on their way a survey that gets at this because half the population is not online and it's unclear what they are seeing because mass media still remains fairly uniform in what it does and does not cover there's some governments attempt to bring social media under control it's not just beat it's beat and join they're there they talk they're very active but there's some legal in legal changes that bring a lot more user surveillance that's the most worrisome part and some some very complex uncertainties what I'm trying to say you know to end is that data I mean I when I look at the Turkish big data Twitter big data it's really striking but you have to triangulate like any other method you have to get offline data you have to get mass media data you have to get other kinds of data put it all to clay you have to start from the people up even if I mean from big data sets you really want to see things which look at what are people seeing in a big data way I mean a lot of our research is what are we able to see and what we are able to see as researchers is not as consequential as what are everybody whose individual eyeballs are seeing different sort of worlds and we could also aggregate that and have a huge big data set that looks at what are people seeing or are they following what are their streams are looking like and that is all I will say before coming to the other current events this will never end crazy couple of years for this area of study thank you yeah where do we go is it in the same folder all right great thank you very much to Mark for inviting me to participate here it's wonderful to get a chance to be here I want to start off just by saying that the research that I'm about to show you is from the NYU social media and political participation lab that we've set up this is generously sponsored by the U.S. government's National Science Foundation so your taxpayer dollars at work in supporting really important academic research the goal of our lab is to do two things simultaneously one is we're trying to understand social media as a new variable the world's changed there's social media in this world we want to understand how it affects political participation the second thing is that as social scientists as policy analysts as anyone who's interested in how people around the world interact with the political sphere actually interact with all phases of social life we have an enormous new data set at our exposure enormous new enormous amounts of new data available to us to study individual behavior so our lab is trying to also develop social media as a form of data to make it accessible to develop tools that we can use to study how people behave so sort of two overlapping tasks I'm going to talk to you here I've been very I was invited here to talk to you about social media in Ukraine a couple of you may have been at the talk that I gave at IFIS a couple weeks ago well and so obviously I had the talk ready a couple weeks ago I could just you know use the exact same talk today and there'd be no problem right okay so Ukraine this is going moment by moment as we talk so let me just quickly talk a little bit about what we've learned about social media usage in Ukraine and maybe what you social media has taught us a little about what's going on there all right so we had talks before about people using social media to organize them and I'll show you we were doing this as well we're using five talks before about people using social media to contact the West to be in charge to touch with the West right Ukraine may have been the first time that we've actually seen negotiations taking place on Twitter right so this was earlier you know way back in the early days of 2014 January when there was a negotiation going on about what was going on between the government and the opposition in Ukraine this was Arseniy Yatsenyuk who was one of the opposition leaders there was a deal offered by Yanukovych this is obviously as it becomes a recurring theme in what's happened in Ukraine but there's a deal offered by Yanukovych to the opposition how does the opposition reject it the opposition rejects it on Twitter no deal and this is actually tweeting at this was the president of Ukraine's Twitter handle so this is a tweet at the president we're finishing what we started the people decide our leader's not you now as we've been coming a theme that's been going on throughout everyone's been talking about here note that this tweet was in English now here's a whole bunch of Yatsenyuk's other tweets some of them are in English but some of them are not in English a number lots of other tweets are in Ukrainian right what's the result the result is that within hours you see on the New York Times unmediated transmitted to the world exactly what Yatsenyuk said right no deal we're finishing what we started the people decide our leader's not you this is a really valuable tool to an opposition leader in the middle of a protest situation in a square where there's violence in square right but you don't have to have a press conference you don't have to worry about getting mistranslated you write it in Twitter it shows up in the New York Times right yes to three hours ago as of when I was on the train this morning grabbing this thing so this is now a whopping six hours ago which probably means is completely out of date but announces that criminal proceedings are being brought against Yanukovych how does he announce the criminal proceedings this actually says Yanukovych is disappeared right later on they announced that they're announcing criminal procedure how do you do it this is on his personal facebook page what do we see again within hours on the New York Times Ukraine's acting interior minister issued a warrant for the arrest how do we find out about it Avakov made the announcement on his official facebook page right so this is a new way of people in conflict situations in civil service of civil unrest of communicating right you can as an opposition leader communicate directly with the outside world on where you're going to do this and I think we're going to see this more and more being recognized as a top as a tool of people in these kind of civil things all right what else can we say about stuff twitter usage in Ukraine has increased over the course of the protests this again was the figure I showed you know what we showed way back when that our lab produced for a post at the monkey cage which is a blog that I write for at the Washington post along with a number of George Washington University professors this is the first thing we showed up and we compared it to research we had done on turkey these were much lower numbers so we were seeing much less twitter usage in Ukraine 1500 tweets per hour a thousand tweets per hour maybe 2,500 tweets per hour and we were trying to figure out so what was twitter being used for it was less popular than what was going on before now look at twitter usage this month okay as things broke on the 18th of February which was the biggest day and then 21st of 21st now we're seeing 10,000 tweets per hour 20,000 tweets per hour 30,000 tweets per hour all right so this is in the course of a couple of months same set of collection words keywords that we're using here you're seeing dramatic increase in the amount of usage of twitter and twitter in the Ukrainian context interestingly enough has been serving multiple audiences we've seen tweets in English as I showed you before we've seen tweets in Ukrainian we've seen tweets in Russian what's interesting though is that in these absolute moments of crisis and we have January 19th in here and then we also have the events of the last couple of days huge spikes in the amount of English usage right on these keywords some of that is coming from outside of Ukraine obviously there's just much more international attention people are using the hashtag zero my down to be talking about this but some of it's obviously the kind of things that we're showing before attempts to directly communicate with the outside world using these different using social media as a means of communication now what we showed but the other thing you can pick up here which I'll come back to in a second is obviously Twitter usage is responding to events we can pick out the big days right January 18th in here of February the last few days in February in December when the squares were clean we can pick out all the sort of big events in the Ukrainian crisis we can if we had all this data you can go back essentially and recreate what happened by seeing so Twitter responds to what's going on Facebook also responds right so both Twitter and Facebook what this is is shows a graphs of sort of likes comments and shares across Facebook posts but we see it every moment of the sort of main moments in the crisis we see these dramatic increases in people liking Facebook pages becoming signing up to follow Facebook pages and sharing what's going on in Facebook pages all right so what's happening here Facebook what's interesting is that as time has gone on more and more of what's been appearing on Facebook has been primarily in Ukrainian right so here's a picture from earlier on in the protest this is still in December these are needs on them things we need on the Maidan right these are actual coordinating devices here are things that we see more recently right just in the last couple just in the last couple last few days Maidan doctors of the Maidan right this was a Facebook page that was designed to show where there was need for healthcare on the Maidan right this is Euro Maidan SOS this is a thing that was set up to coordinate emergency responses on the Maidan right this is the actual one of the Euro Maidan the actual sort of Facebook page that they have again we see this all sort of in Ukrainian all playing these kind of strong coordinating roles for what's been going on here now the other interesting thing that's going on with Facebook and this comes from survey data that was done by Olga Unik who's at Nuffield and Harvard this year she ran a sort of rolling survey of people in the Maidan and asked them questions about a lot of different questions but asked them questions about how did you find out when and where you should go to join in the protests right these were people who are at the Maidan what is interesting about this is that you know we do see that the top choice here is I saw on TV that people were protesting so I joined in but if you look at what I've we've got circled here right my friend family members said they were going on Facebook 40 percent right now this is much lower much much lower down here I received a Facebook invite from a civic organization right Olga and I had a back and forth about this she sent this to me she ended up doing a guest post about this survey on the monkey cage and she said oh I want to push back on what you guys are saying about the importance of social media here because only four percent of the people say oh it's because they got a they saw that read about it on Facebook from a social from a a civic organization but my response to her was say I think actually it's the opposite you may have here the sort of greatest the single largest documented effect I mean and we don't know if the survey is representative we don't know anything you know let's put all the caveats about running a survey but they're trying their best to get a sort of random sample people around the Maidan right if this is correct this may be the single greatest effect we've ever seen documented of how social media and involvement in social media led to people actually participating in political action right now the key thing here which I think is important is that it combines the two things that theoretically I think make social media most interesting and in our lab we sort of it's an interdisciplinary lab that we've set up and we have political scientists who tend to focus a lot on information and utility calculations and cost benefits we have lots of models about whether you join protests based on your expected benefit from being in the protest the cost of being in the protest in that sense we think social media is crucially important because it may affect information right it may affect what you know about the costs what you know about the expected benefits but the social aspect of it and we also have a social psychologist on the project that's one of the co-pis of the project co-directors of the lab right the second thing about social media that's so important is obviously the network effect and we've seen graphical displays of the networks well here's another way we see the networks right it's that you're hearing from people who are in a network that you've self-selected yourself and to join these are your friends on Facebook right they're finding out about this on Facebook so Facebook is providing a way in these protests to facilitate communications between networks that you've previously decided to join right and according to this particular piece of survey evidence this seems to be very influential in helping to get people onto the Maidan now what are other things we learn about this we've learned from looking at the social data here social media data as I said Twitter clearly responds to events on the ground right so this is something that is now being used Twitter is being used we think because of the English aspect to spread internationally what's going on on the ground we also know it's being used for coordinating roles on the ground but it clearly is jumping when you see big events taking place however and here's something that's really interesting that much less has been said about what we've also seen in Ukraine is that the number of people getting on Twitter also responds to events in the ground now this is a more compressed this is a more compressed because we wanted to show more time but these are again from our data set of people who've tweeted about events in Euro Maidan use these hashtags that we put to to collect this data set when their new accounts were created and what we see is that every point where the protests heat up in early December in January and then bam in the last few days more people go sign up for Twitter accounts when these big political events are happening so there's an endogenous relationship here right Twitter or social media is helping to inform people about political events but at the same time political events are driving people to social media right so the relationship is reciprocal and we this is we thought was fascinating when we first sort of saw this the other thing I just want to stress is that we tend to think and there's been a lot of discussion that they're all at the same affair we want to we always want to dig deeper right this is the tension with big data we end up with data sets that are too big to hand code I have we have a data set with this Turk with Turkish Twitter data that's got 93 million tweets in it right there are not enough undergraduates in the world for us to pay them to code all all right so we have to go back and forth but the danger of course is that you go straight to the big data and you don't know what's lying underneath there right so I'm you know going back when these things and reading and if I was here giving Zana is on the panel for the Turkey if I was here giving the Turkey talk I would show you a whole bunch of tweets that we translated that actually illustrated a lot of things Zana was talking about tweets in Turkey that said you know really specific effects of peer pressure like get out of the bar and come to get it you know like very specific at people's friends and stuff like that Ukraine we're not we're a little bit behind that but one of the things that I want to say is so I think there's always this tension and we want to get in and see what some of this stuff is but we want to get in and see what some of this stuff is in part so then we can then teach the machines to how to code this stuff because we can't code this stuff by hand so it's not just we want to read some of the stuff because we want to know what we're talking about but the what we're also really trying to do in our lab is then use that information to then go and train the machines to go extract the information that's important so one of the things I want to say is that we think of Twitter right Twitter's this 144 character way of conveying information right we tend to think of it is so limited these short verse of text but Twitter much as Zana was showing where you could do stream captures right Twitter's visual right and so one of the things that we can do here is we can see what it was that people were seeing on the days of these protests right so this was according to our calculations based on the data said we had this was the most retweeted picture on February 18th the day that sort of Kiev went up in flames and people started getting killed right this had this was this particular picture was was included in tweets as a retweet over 1500 times right this picture got tweeted over a thousand times this is the one where lock and loaded people were going in with a clutch to cause and one of the big stories in Kiev became about the change over from rubber bullets to real live ammunition right and the question whether there were snipers this picture of of their coup guards who were the sort of elite security forces of Ukraine walking in with these locked and loaded collision costs went all over and this picture you may have seen as well right so this is what Twitter was actually communicating to people and we're looking each of these has over a thousand retweets and that's a bare minimum right because that means a thousand people saw this and sent it on to their followers this is a multiplier so you're talking about images that can easily be seen by a hundred thousand people this one then gets picked up by the news some of you may have seen this in in mainstream media as well pictures go up and throw these things right what are the pictures from yesterday right from sunday well do you know what this is this is the boat that Yanukovic has installed on his private residence that functions as a restaurant okay this is a gold chair from inside Yanukovic's palace right and this is Yulia Timoshenko who is now flying on the way being flown back to the Maidan after having been released from prison the opposition leader right so the dramatic difference between what the tweets are conveying and the information that this is being made out there and all of these pictures again are getting massively retweeted and disseminated and the speed at which this occurs is just sort of stunning right like nobody had seen the inside of Yanukovic's palace before what Saturday right Saturday morning Sunday afternoon this picture gets retweeted 1500 times I mean I'm sure you also the pigs on the wild boars and the goats and all that sort of stuff on the news too as well but this is this is just sort of an incredible ability for this stuff and and one of the people who's reading you know there's also a picture of the cars in Yanukovic's garage and like you go back to Tunisia right and Ben Ali and the private cars and stuff like that this is something new right where the way that these images can yet spread and it's almost like this is becoming a new phase of the revolution right the spreading of the pictures of the opulence of the deposed leaders so in summary what we found is that you know and I would go so far as to say it's becoming increasingly difficult to imagine you want it to what's the takeaway right like it's becoming from today it is becoming increasingly difficult to imagine protest movements that don't utilize social media so if you want to understand protest moving forward what leads to protest the dynamics of protest you have to get a handle on how social media impacts protests and this is one of the things we're trying to do we're writing papers about testing hypotheses about these kinds of things but one of the interesting fun things about studying social media is that if you go way back in the dark ages right to 2009 when you had the first Twitter revolution in Moldova which then turned out to be like well maybe there were only seven people with Twitter accounts in Moldova although that's also wrong too right but the sense that okay so these things you get you get pushed back and say oh it wasn't really a Twitter revolution or you get to the I was saying oh people weren't as time goes on right anyone who pushes back and says this stuff doesn't matter well the next one it's just there more and more and more and by the way I could have given this talk about Venezuela this morning too right like that broke in the last couple of weeks while we were following this and our Twitter feed on Venezuela went crazy over the weekend also right so this is I think crucial for those of you in this room who are policymakers who are thinking about how we anticipate responses how we estimate political unrest around the world getting a handle on how social media impacts this it's not I don't think it's hard for me to imagine anyway and I'm interested in hearing opposite perspectives but it's hard to imagine this is going backwards right that we're going to be in less network worlds and where people have less access to instant information transmitted to them via networks in these kind of protests all right so what do we see in Ukraine social media is being utilized in in the euro was is being was being whatever it is right now social media has played a big role as was playing a role in the Euro Maidan protests Facebook we saw a lot of things about organization but it was also a way for probably social pressure to be brought to bear on people to participate or to learn that your social networks to put a less pernicious term on it Twitter has played roles in international outreach and undoubtedly in real time communication very interestingly I think we want to take away from this is two way effect that social media may facilitate protests but protests may also bring people to social media and social media I think is more than text and I think we want to keep reiterating that even if it becomes even that much more difficult to study things that aren't text but it's a crucial piece of the puzzle so thanks very much and looking forward to the discussion so to start off in the summer of 2009 Maxwell Malcolm Gladwell famously wrote the article the revolution will not be tweeted if we were updating that article in 2014 would there be a new headline? Well it was wrong even then I want to say just in Egypt I have a publication with survey data once again it's a protest survey data so how representative it is is always shaky because it was the survey was done at the height of the clashes in the 2011 and there 28% had heard of the protest first on Facebook and the sample size similar to about a thousand people and this is Egypt which is much much much less connected there's fewer people so even in the Egyptian situation partly because of the sort of media censorship versus what's going on on the ground that was important now Malcolm Gladwell's argument was that this is all weak ties and you can't do a protest with people's weak ties so that was the crux of his argument well it's wrong in multiple levels and he gives the civil rights era as an example first thing social media is not just people's weak ties I mean that's I've never seen empirical research there's some weak ties some strong ties places like Facebook you tend to have more some depends right teenagers have their strong ties on Twitter and their weak ties on Facebook where because everybody they know is there it's like a friend of mine Nathan Jurgens they call it decoy social media they help little old ladies cross the street twice a day on Facebook because their parents are on Facebook whereas on Twitter which is very public they have their close friends because they're strong ties so social media is not just weak ties also there's this idea that social media is this virtual world and it has you know some weird effervescent presence that's not true even in Egypt when I interviewed people who told me like I'll tell you the story as it was told to me a political person in a family that doesn't speak politics and is from a relatively well off family is like young goes on Twitter says I want to find people like me this is before the you know first uprising and that finds people and then gets involved in the protest and ends up in Tahir for those 18 days so those weak ties can be turned into very strong ties so that was the case I think Malcolm Gladwell kind of way of looking at the world is both misunderstanding how protests occurred in the past if you look at civil rights movement what was the key thing that spread the sit-in counters Fresno from Greensboro to elsewhere it was television it was not people's strong ties college students saw it on television and said I can do this I can do this I can do this so if we were updating it of course it'll be tweeted what else but it's not ever just tweeted because it never was just tweeted it always has been online offline all mixed together and I think you made a great case what you see on social media since they're your friends it has a pressure on you if everybody you know is out there you know getting beaten up and you're Instagramming your lunch that is not going to work so I would say he was wrong then he's wrong now yeah I mean I think there's two ways to think about this one way to think about it is that it is technology is moving forward and regardless of the argument and it was sort of fun people jumped ahead and said these things mattered a tremendous amount in Moldova and then there's pushback and in fact but no matter what happened in the past it's more and more in the future and every time we see movements emerging right now there are these social media components to it now of course it's important not to confuse that with saying there's no other components to it which is I think a point that we're all in agreement with right but to ignore it and say that you have these movements that come about and that they're not that social media is not playing a role in the ones that are successful I think is a mistake but the more interesting question I think to be asking is to go back to at least the sort of theoretical argument which he's making which is to say instead of splitting it in such black and white terms right and saying there are good social movements and there are bad social movements right the question is what are the different dynamics of social movements that emerge over social media one dynamic I think and to say how are they fundamentally different from movements that may have emerged through sort of top-down traditional organizational structures right one thing about it clearly is that these things can come out of left field in a way that you know might have been harder before you might have been able to be a policy analyst and you look and say oh is there going to be instability let me look at the organization let me look at the cells let me look at the groups right one of the things we've seen recently and it wasn't just by the way it wasn't just I mean the speed at which this has happened the indignatios to turkey to brazil to ukraine right like the speed at which these things can happen in ukraine and eric correctly if I'm wrong but if you had surveyed ukraine experts right on december 1st and said are we heading for massive upheaval right in the next three weeks it would not have you would not have thought this right but you get one event you get one reaction you get a bad reaction to it so I think the question so one thing I think that's obvious is that these things can emerge that can emerge now another question is going to be right do movements that emerge protest movements that emerge in these kind of faster ways over social media do they end up having different sorts of demands do they end up having different rates of success are they more or less likely to to turn violent you know these are all I think very fundamentally important research questions I think gladwell's attempt was to say they're all just going to fade right that they're not going to be real because of these strong ties right well it might be sort of that might be I mean one thing to look at is to sort of demands that come out of it and I'll stop I'll wrap up with this but if you look at the indignatios in Spain and people correct me if I'm wrong here but in a sense the indignatios never were able to get a concrete goal a concrete demand out of that protest movement right if you look at however what happened in Ukraine that even though this movement may have emerged with a sudden speed and with a sudden reaction in a similar way there was this agreement over the resignation of Yanukovic and that Yanukovic had to go right and that may have been new for this kind of this kind of an approach and so we may see like things like a central focusing idea whether you have central focusing ideas become much more important because in older days you with an organization the organization itself wants to perpetuate itself people are running that organization so you have these you know you have people who are invested in this for it so I think these are the kinds of interesting questions should be asking it's going to be tough to follow but let's see I'm just going to be a little bit of a contrarian and say that again I'm a huge advocate of historicizing but if we were to look at that particular moment in time when Malcolm Gladwell put out that that article we have well first we have sort of a head-on collision between him and and Clay Shirky from the standpoint of you know it is purely technologically determined versus you know socially determined and I think that that speaks to a juncture where the the technological innovation in and of itself had very limited capacity but very limited reach and to sort of play off of what what Zadab said even discussion of clothing you know weak ties and strong ties factors in into the research I'm looking at today I mean Josh's sort of notation about the finding from the Ukraine protest where what is it 40 percent of the of the of the people surveyed said that they decided decided to go up down to the the Medan Medan specifically because they'd seen a message posted on Facebook well if you look at it from a purely sort of instrumentation standpoint that it affirms the role of the social media as a technological tool to draw people into the square but if you were to look at it from Malcolm Gladwell's sort of social psychological interpretation what says the first part of that statement says friends or family members posting on Facebook so it is the level of of social proximity that people have to it now Egypt you know Egypt the reason why Egypt became so so so strange and so complicated is because as Zainab said the initial sort of impetus for participation was largely out of loose ties or weak ties when I came out on January 25th I didn't know a single person who who was going out to the square I mean it was basically these virtual spaces of people who had you know ideological agreement or sort of a general conceptualization of what might be problematic in the country politically and the call in and of itself was generated entirely online versus anything that may have been made its way or seeped in through mainstream media or traditional media or various other tools yes there are people who were coalescing and organizing behind the scenes but as far as the mass distribution of the message it existed largely online and so I think it's it's partially unfair to sort of discredit the argument in its entirety today given what social media has become and its level of breadth and depth I think it was too hasty of an argument one should say but at the same time I think if we want to understand how all of these dynamics work we have to also look at it from from the standpoint of basic interpersonal communication and social psychology as well which brings in an important argument that exists within Malcolm Gladwell's general conceptualization he may not be right about innovation he may not be right about the role in which technology is used as a catalyst for social for social mobilization but he has something interesting to contribute that I think we'll live on and we will begin to discover as we triangulate and as we become much more interdisciplinary in the study of of you know big data and everything else I think in one sense Gladwell made two arguments I think one was discredited which is loose networks impel through social media in to some extent could not upset the status quo obviously on that he's profoundly wrong he may prove be proven right in Ukraine and that that loose networks may not be able to translate a protest momentum into power in Egypt they weren't interested and you already already see the implications in Ukraine about the return of Timoshenko and what that means and so on and so forth we'll do one more question here we'll do one question from online and then we'll open it up you know to the audience so if we come back to Egypt for example you had a career and then you had Tamarod what's the difference was there a difference in the role of social media from one to the other I'm really glad you asked that question I think of course social media sort of scholars are less interested in Tamarod because of the the kind of the channels of outreach and mobilization for the most part the epiphany that happened I would argue the moment that Egypt had its first presidential election specifically the second round of presidential election went 49 percent of the of voters came out and supported the member of the old regime it was this sort of wow we didn't realize that a substantial proportion of the Egyptian population supported the regime with everything that it represents so it raised the question of whether or not you know the the whole sort of historicization and discussion and the discourse and the narrative of revolution if it really sort of exists in a tiny little crucible called the social media and the the really sort of entrenched sort of independent and private media that had something to gain in this or various political interests and in the case of the Muslim Brotherhood and various Islamist groups and and sort of secular progressive groups that where where is this where for the first time we're concerned about public opinion and so when Tamarod first started the intention was to go beyond the 32% who have internet access in Egypt and really sort of turn it into a grassroots movement across the governor and keep in mind that the entire revolutionary movement in Egypt was largely urban I mean at the end of the day we were focused on Tahrir Square and we're looking at other sort of parallel squares in other sort of metropolitan areas like Alexandria or Suez or Bursaid but the majority of the population in Egypt exists in areas outside of these urban metropoles which means that where do they stand in all of this and so when Tamarod Tamarod's success is largely a result of an extremely basic you know mobilization which is to take a piece of paper photocopied by ordinary people you know at their own expense in most instances and going sort of door to door to try and get people to sign it with very very simple demands now the social media dimension may have been may have been large but at the very but it was the tip of the iceberg below which there was a real sort of mass movement the protest movement of January 25th was completely the contrary when there isn't a capacity to utilize public space for mobilization and you're really sort of communicating to a small constituency of like-minded individuals you get a completely different realm so the total the total participation in the January 25th revolution and all of its 18 days is probably again you know the estimates and the you know and one could do the arithmetic and geospatial sort of arranging this of course becomes a very contentious issue in Egypt because it's a competition between different revolutionary movements but the total capacity or total participation in the Egyptian revolution of the 18 days could possibly be in the range of like one to two million again I'm throwing I'm throwing numbers out there but the total capacity in terms of the distribution of information and the knowledge and the anticipation of the June 30th protest movement because of the the capacity to you or the decision to use sort of a paper you know paper petition if you will expanded the both the breath and the depth of the movement in such a way that it is fairly conceivable that we could speak about at least 10 million people participating in the protest and both through you know you know anecdotal and in terms of the breath the geographic breath of the protest was very very sizable so does that mean that the utilizing sort of non-digital mobilization tools trumps the social media in the case of the June 30th protest I think it was an extra and it was a combination of the political circumstances the general climate in the country and more importantly the decision to to utilize all tools necessary so the modern movement represents sort of a maturation of the relationship between the online offline rather than looking at them as two disparate realms that don't interact but really sort of bringing them together in such a way so that it becomes a sweeping movement but again keep in mind that there were always sort of the there are always fissures and there are always sort of the the blind spots and so there even when it comes to tomato there are always blind spots what was missed you know where was the support was the state involved in any way and all these things are are extremely interesting to keep in mind nevertheless well I want to pick up on something that's sort of floating out there there's two things and I really sometimes I get tagged as a cyber optimist and a cyber pessimist at the same time and that's usually my take is you know yes I think there's one key way I've written this article in digital media and learning published that is a MacArthur fund that's sort of this blog there's a boom bus cycle to these right there's a boom bus cycle to these social media field events that come out of nowhere M-15 Turkey is going through that Egypt kind of had this Occupy kind of had this there's a sort of great boom bus in part and so what's going on here is one thing and my take is that the one thing that we thought was great about social media lower coordination costs actually has some negative side effects and I think the one thing that everybody dismissed about social media slack activism is clicking unlike I think that might it's the strongest effect so very briefly so in the past if you needed to organize a march on Washington you needed to organize logistics you needed to organize buses if you wanted to do the Montgomery bus boycott you needed to organize carpools I mean they met every single night if you read their memoirs I mean it's this intense amount of work just to organize the carpools because these are for black people they're not all going to boycott going to work right they are going to go to work they're going to walk to work they can't always walk to work the amount of logistics that under to do a protest has gone down dramatically and in our theory it says it's usually cited as something that's going to increase social movements which it does I think you can organize gives a protest like this without having this organization that had to do months and years of logistic work but all that logistic work left you with an institutional and organizational capacity that was crucial to the second step and the third step and the fourth step so now that you can do something easier I sometimes likening to climbing Mount Everest by being helicoptered into base camp and then being you know you use you don't build certain kinds of muscles that were teased in awful and horrible to go through and you can do the street protest spectacle which can have impact in Ukraine inside impacts in Egypt it's that impact so I'm not dismissing the street protest spectacle but let's not be asphalt fetishes either right I mean where does it get us to be able to put on spectacular street protest sometimes it gets us a lot sometimes it doesn't in the case of Turkey the political opposition is weak and incompetent and fractured long political story it basically means that the spectacle of the street protest is not communicating to other political processes because they could do it without building the muscle necessary to do that street protest and also think of the to bring it back to the U.S. think of the first internet protests the the sopa Pippa the blackouts that were so influential because a lot of congressmen we saw like went from I support this to no I don't support it what happened then if you remember there was this going to be this pretty awful law it was unenforceable probably and Google and Tumblr and a lot of sites connected people who landed on its website directly to congressional people so congress people were getting 10,000 and 100,000 funkos and they penned because in old deluge signaled an organization behind it that was maybe going to challenge on a primary street protest no longer signals the same organizational capacity so for a government it is now a question of can you write it out because in a month later you there's no organizational capacity to challenge you if you're able to write it out you may not be able to write it out I mean it depends right so I think the spec the lower coordination costs are actually depriving movements of muscle building that wasn't always fun or great on the other hand I think what people derived as lack of us the sort of symbolic part of social media I think that's the long term most consequential thing I mean to dismiss the symbolic sphere in human life is insane I mean we we started with UNC Duke jokes I mean people die for flags which are you know and language and identity ethnicity most of our motivation is from this hugely symbolic sphere so changing perceptions in a socially embedded way those little likes and clicks that everybody dismisses might well be the longest term impact the Turkish social world is completely changed as a result of what is happening people think differently and those online's the epistemological effects of the intervention online but it's hard to watch right so I think the street protests are weak in policy impacts because lack of organizational capacity that was necessary but our strong and biological impacts if you if you participated in talk or if you participate in those 18 days you're a changed person if you participate in 1968 anti-war movement you're a changed person so biocrat so a lot of real impacts might be long term including the likes and these things so you know I love studying the protest movements there's a lot of interesting stuff those spikes and things like that but I think there are ways in which there's these longer term deeper impacts and they're not going to play out the way people saying where slack activism quote unquote is going to be I think very impactful and I think lower coordination costs are coming back to bite a lot of movements in this boom bus cycle but just to add one dimension to what the senator just said this this ability by institutions to ride out you have two cases now we have one more see you never got barely got to finish the first year of a designated term and Yanukovych who didn't get a chance to finish you know the fourth year of a designated term so it does this speed you know put additional pressure on on much of the term system yeah so just to make three points quickly about this the first thing is I think we need to get again I want I think it's important to get away from the sort of paradigm that Malcolm Gladwell unfortunately put us into in this sense of good protest and bad protest good organization bad organization right what we want to do is theorize about the factors that the dynamics that affect protests as the component of that protest that's driven by social media increases I think that's the most important thing for those of us who are social scientists who want to understand how protest takes place and also for policy makers who want to understand implications of this right and I think there are and like there are all sorts of things that we're identifying in this panel that people are talking about in their research that people are talking about and writing about these things right where where there are dynamics that come out of these things and I think we can start to learn something about what those dynamics looks like but to start saying that the only right the only good protests are the kinds of protests which have large-scale organizational structures behind them right that somehow that's better than other times I mean let's think about it from this perspective right social media may lead to protests where the costs of joining that protest are lower the costs of going to zero to 60 undoubtedly are lower in terms of getting people onto the streets right but that may not lead to the same thing that happened after you had civil rights organizations that organized for 15 years right but there are lots of cases where people can't organize for 15 years beforehand right and this may give people an ability to put things on the agenda of politicians who are either trying to to try to counteract current events or trying to avoid them in the future that they didn't tools in none of the disposal beforehand there are regimes that don't allow people to organize and so the most likely case that you're actually going to get large numbers of people on the street is if you don't have years of small little organizations that the regime can go after but that you have sudden dramatic moments that catch the regime by surprise right there are people who we know have fewer resources to devote to things like political organization people who are single parents who are working to 15 you know two jobs 15 hours a day aren't going to go out and join organizations that involve you know massive time commitments but might they be willing to give a little bit of time in a concentrated period of time when lots of other people are giving time absolutely so the good versus bad thing I think is the wrong paradigm to be in right the paradigm to be in is there are different dynamics and we have to identify what those dynamics are the second thing that I want to say is I digital organization is organization right we have to stop thinking that the only organization that matters long term is a bunch of people who meet with each other in a coffee shop you know to plot strategy quietly while smoking pipes you know in the back room right like an email list of 200,000 people who are all willing to give $10 to a candidate in a local house race that's organization even if you've never met those people before right 100,000 people who sign up to like a Facebook page and then when you post information on that Facebook page it says today's the day and you know it's going to show up at 100,000 Facebook feeds if Facebook puts it there which is another manner entirely trying to figure out what Facebook is doing with its algorithms on who's seeing what and we should not those of us who are studying this should not underestimate we're not sure what people see and I agree there this is a big area for research we've got to figure this stuff out and the reverse engineering things that we're talking about are really important in this regard but that's a real resource that's an institutional resource we had, I was teaching of course this spring in January on social media and political participation and we had a speaker from Occupy Wall Street coming in Occupy Wall Street doesn't think they're in the doldrums right now they think they're doing really well because the minimum wage rising the minimum wage seems to be one of the most important agenda items on President Obama's agenda right now and has a realistic I don't know I'll leave it to you guys whether it has a realistic chance of being enacted but we just saw who was it one of the big companies just announced Gap, Gap old Navy Financial Republic just announced they were going to $10 an hour Occupy Wall Street sees this as a tremendous success what they are doing now is using this the most interesting thing from this talk this was my words not theirs but it was almost like the protests were a lost leader to get a big movement of people who could be coordinated and who go to the same social media sites read the same blogs and can be motivated to get agendas affect the political agenda in the country right so they don't see it as a lack of success I don't know what similar people think in terms of Ignatius so this idea that digital organization is not real organization I think is a misnomer and people are going to be caught by surprise if they think that digital organization is not real the final thing is that I really think that Zaynep is absolutely correct about this thing where getting people who don't traditionally think of being politically active I think this is you know this is what we might call a bit of a tipping model right there are an awful lot of zeros in this thing and then there's a bunch of people who participate and they arrange from you know 0.1 to 1 if we have some scale or index we make of the amount of participation people do but those clicking on likes or joining the move on email signing one of those online petitions and then you get more information about stuff it may very well turn out to be that there is in terms of a kind of internal efficacy mechanism which is thinking you can have an influence on politics that getting involved in just one of these things that it's signing clicking a like on a politician it's a gateway drug in the sense that you click a like here you get on an email list here next thing you know you don't think of yourself as completely divorced from politics and over time I think that may turn out to be and it may be part of the reason why we're seeing more spread of more political movements in more ways in the last few years than we have in previous years because there's a greater and greater base the other thing that I think and this is one of the things that I'm working on in my own research but I think the idea one of the things that's happening with Facebook with Twitter is we are inserting ourselves into networks voluntarily that we feel like we have a connection with but most of those people in those networks we wouldn't normally hear about their political views right if I stop my colleague you know his wife on the street and say hi to her on the street she doesn't normally say to me you know I support Wendy Davis you know she says I'm but if I'm on Facebook with her I'm hearing about that right and you have these networks where more and more people are being exposed and finding out that more and more of their friends and neighbors have political thoughts and the question is do these networks that we insert ourselves into in these social media then create the groundwork for having a belief that we are part of a community and when members of that community urge political behavior we may be more receptive to it or when you finally get onto the street to protest if there are people there who you've been talking about it these issues with on Facebook a year or two years even if it was simple just a little bit of knowing what's going on how we massively expanded the amount of people who we think of as being vaguely politically relevant in how we think about politics and that could also have a long-term effect and I think that could be part of the answers why you suddenly see these out of nowhere movements emerging because we've setting the groundwork for it by allowing these sort of fora where people can discuss what's going on politics and this may be a lot of what we've thought about in terms of social media and protest has been in more authoritarian regimes this may end up being just as important for more democratic regimes as well what's happening online I want to congratulate our online audience who just got spammed by the same people who are spamming I love your selfie and Alec Baldwin if there's ever a sure sign of online success that's it but I want to ask a couple of questions from our online audience Max a question for you do we have any information about how the pro-russian camp is tweeting and whether they're really active online and secondly how insular sort of reflecting Dean's analysis how insular are these conversations are they is there a big pro-western insularity to this conversation that's going on and then sort of to the rest of the panel a question about Venezuela how does that fit these trends that we're seeing on in all of these regions I don't know the answer to the first question I mean we're tracking some websites that were set up or some Facebook pages that were set up originally that were anti the protesters and this is the thing where we have to sort of get in there and start digging and seeing what's going on in there this to me though by the way is the second most interesting question that is that's out there about social media and political mobilization at this point in time which is that how does the regime use social media to try to deal with mobilization and so we are and this is something we absolutely want to explore and maybe Dan can talk a little bit about this in Turkey because Turkey's been the big example where we know the regime went out and hired people to try to get in bombs social media and this is where the kind of network analysis the team was showing us in the beginning as the next steps you have to start taking place and seeing what's going on with these things it will be incredibly interesting right what seems to have here's my hypothesis right we haven't looked at the data but what seems to have happened is that very quickly a lot of Yanukovic's support has dissipated right among party of regions the party of regions which is supposed to be his party came out with a statement blaming him already right the support has dissipated really quickly so what would be interesting to see is if this becomes reflective in the social media networks right if we see a change in the shape of these networks as the questioner asked from sort of more insular within the Russian community within the Ukrainian speaking community within the sort of protesters and again these are very sort of naive broad stroke generalizations about things we'd have to look at what these look like we want to be very careful about ascribing political political positions to people based on language usage but it would be interesting to see and this is one of the really big fun interesting questions to ask about social media that you can actually study because you have the data right is whether it's elite led or mass led and so it would be really interesting to see right after the party of regions starts breaking with Yanukovic in the last 24 hours or so if we suddenly see a correspondingly shift of the masses in terms of who's retweeting whom who's following whom and whether we can we can pick that up afterwards or or even more interestingly if we can see it predating it right we have a paper in our lab where we're doing this with congressional data in the U.S. we're looking at all the tweets by members of Congress and we're looking at all we're looking at tweets by their followers trying to classify them in different topic areas and seeing is it the members who are leading the discussion and then the followers pick up on it or the members reacting to the discussion well we could do the exact same thing in the Ukraine case which is to see maybe part of the reason the party of regions deserted right Yanukovic is because they saw that their own network was crumbling already so those are interesting things to look at both the data are just some of the data I showed you today or from two hours ago so I can't I don't know that yet okay well we'll try to do we've got about 20 minutes left so perhaps to record your question if you can do one member of the panel we'll get that way we'll get more in question yeah and it's a good segue from the previous one Ben Homer I'm a grad student in new school and Josh I'm really interested in talking to you especially because the I'm gonna do work in New York the Venezuelan community oh great so I definitely like that but Zana I'm more interested in the point that you made before Turkey has hired something like six thousand internet monitors basically and it seems if you're a smart government that this is probably you know you've seen what's what's happened this is probably a good policy to have is to have some kind of strategy for dealing with people who could become problematic to you in the future so I wonder what you're what you're so um very interesting things so many questions can I take 60 seconds to say something about one question so the digital organization I'm absolutely what I should have said is that being able to organize using digital media sports different cultural norms like being leaderless is a lot more possible because you can do it on social media and being leaderless movement has different long-term institutional impacts so our communication as an organization I'm totally there with you and I think there's a cultural trend towards leaderlessness never used to exist in Turkey which came up in Gezi and it just blew my mind because Turkey's never had spontaneous leaderless movements and I think the fact that and it's had it and I was like what country is this we were really we know with Turkish it's always hierarchical it's always institutional it's always leaderless leader very factional so I think it leads to different cultural possibilities which then impact in being social media organized so parenthesis in terms of I think the Turkish government buying fake people or not is overrated and important because once again look at it from the point of view of the person if you are tweeting Twitter and a big data it'll look like there's a lot of them but nobody's following them so who cares right if you look at actually from the people's point of view they are following people they know so you can hire a million people you want and if nobody thinks they know them right you don't just look at random eggs you know how Twitter people who don't have on Twitter and all of a sudden decide oh I am wrong you look at you're you know offline and online and you're near twine people are not following random people spam accounts are good for spamming hashtags as we are being spammed right they're good for misleading big data researchers they are not good for convincing people so you can buy all the accounts you want it's not going to get you anywhere what you can do is you can pollute a hashtag you cannot use the word Bahrain hashtag in online social media without getting trolled probably by people hired by the government of Bahrain or something in Turkey though do know that the government has support and those people are also online so the real issue you have isn't the government buying fake accounts the real issue is you have a polarized discussion and this polarized discussion what you can do is you can spam somebody's mentions like you can constantly mention them but their ways users can deal with it so the fake social media matters to the degree that you're not actually looking at people you're just looking at things it looks like a big deal in my experience of the whole thing and I get you know spam bots spamming me and in my experience of what happens to most people unless you're one of the most prominent people in getting your mention spam for an ordinary person they're invisible they don't really matter they only show up in your big data that's kind of my take on two very quick points to add on that I mean I think the hiring of people to do publicizing tweets and there's a difference between hiring people and I think creating bots to do things it is only the tip of the iceberg right there's there's a much larger question which is how regimes respond to this right and this is an ongoing cat and mouse game as one side develops the other side develops is going back and forth just to mention sort of a couple other things right that are possible right so one sense on the other example and you should see work by Gary Kangan Jen Pan and Molly Roberts on China and what China is doing with censorship I mean there are you know this is just about huge things about taking systematically taking huge amounts of content off of the off of the internet the other thing that we've wondered about that's potentially even more insidious about this is that as network science develops the kind of thing Dean was telling us about the bridges right will regimes go after people based on where they are on networks right if you want to take down social media networks if you have to decide whom you're going to target whom you're going to arrest right like when you arrest a blogger or something like that which we've seen get around the world this is not something that doesn't happen right will the ongoing development of better understanding of where people are on networks will that lead to regimes trying to take out right we've seen there's huge costs to trying to take out well we'll go in this case that backfired profoundly absolutely absolutely and all this stuff I think is cat and mouse and it's evolving and what we tell you today three months from now there may be other things but I do think you know we know there are costs to regimes to try to take down the internet entirely that doesn't seem to work out so is there going to be sort of more selective ways of doing this in the future I think is an interesting question but I think the question of how regimes respond is really important just to say what they usually do is what I'm seeing in a lot of sort of authoritarian context is they're not taking the highest profile people because that also creates international media there's I'm seeing second level people who are actually crucial being targeted more by regimes than the most high profile because because their high profile works both ways they get external attention Western media but if you're like the mid-level person you don't get that protection but you are invisible to the regimes I think they're more vulnerable there extremely brief in a venture I just want to add by saying that in the case of Egypt the government slash regime has often or at least in the current circumstance decided that rather than actually mobilize recruits or pay recruits to do this kind of work what they did was they activated what is otherwise described as the couch party you know has been kind of and they do so by magnifying the messages from Twitter and Facebook of prominent activists or revolutionaries or conversely of Islamists to say well look look who's dominating the social media sphere come on do you want to come in you know enter enter the playing field and what and what that does is it all of a sudden transforms the the demographics of the social media discussion and in doing so they recruit without necessarily having to deal with the the paper trail or the track or or to cover up their footprints let me do the pairs again we'll do two questions two questions and then we'll wrap up but why don't we do all of them mine's quick for Professor Tucker I'm Jim Bullock I'm gonna adjunct to GW some of my students international students have made presentations on other platforms there's a there's a Facebook-like platform in Moscow there's a another one in China and my question to you is did those non because we've talked about Facebook and Twitter today but to have those non Facebook, Twitter platforms figured and then getting back to your comment about how the algorithms are going to begin to be manipulable and you know is is these are these alternate flat forms an issue okay all one alter platforms so yeah thank you my name is Faridou Parsa George Mason University Center for Conflict Analysis and Resolution my question is was also about the regime's response and I want to bring the Iranian case more than half of Iranian are online and they are very active in Twitter and Facebook actually Iran is the in the number is the biggest in the entire Middle East but because of many of Twitter people who are tweet against the regime and they have been arrested what has happened itself censorship so they aren't Twitter they aren't Facebook but it's not safe to share their ideas and opinions so what do you say about so so very optimistic picture of social media it became so just scrub thank you let's keep going sure Hi I'm Matthew Lackenback with Legacy International and I think it's been an extremely interesting discussion on so many levels my question specifically is regarding the use of these technologies in building bridges and their value in making connections and fostering dialogue and peace as opposed to maybe the opposite pole which is more revolution and organizing against something as opposed to building those ties but I'm wondering if you can comment on that if there are any markers to look for guideposts that would suggest a value of these tools in promoting peaceful resolution of conflict my question is for you Mr Iskander it seems to me that both Egypt and China thank you for mentioning that are countries with large populations and significant rural populations in particular that both also have governments with authoritarian tendencies and I really liked what you said about how how you need to go back to a lot of traditional ways of getting out the vote so to speak and so I'd like to hear what lessons you think that Egypt might have to offer to Chinese dissidents and ordinary citizens so we have alternate platforms self-censorship technology to build more bridges and to a better organization to okay I'm going to start at the bottom with the question on on Egypt and what can be learned I think the you know the one can talk about Egypt in the context of being sort of a newly politicized country where political participation not only in the context of contributing at the ballot box but also basic political expression is something that didn't exist prior I think you know the the lessons learned are the lessons of January 25th what existed there the use of social media to create sort of a parallel sphere to circumvent you know institutions of power and surveillance and monitoring and what have you today you know we're past that point we're at the point where where political participation is is sort of a title wave is a really significant title wave that is impossible to to silence I mean even when the government has sort of an axe to grind with particular political players it it it can only do so with with a certain degree of efficacy beyond which I mean for instance the Muslim Brotherhood which is demonized and is now listed as a terrorist organization they continue to protest literally every other day or every day and with you know dwindling numbers but still sizable enough that it's impossible to completely silence and that of course is a realization on the part of the institutions of the state that there's nothing they can really do lessons learned for China I mean I I know too little about China to really sort of make that that leap of judgment but I think to a large extent it's the problem is is a is a capacity problem in China where it appears to be the capacity problem in China and also the ability of the state to really sort of subsume political discourse online and and really sort of circumvent any attempt to express dissent but nevertheless I think what's interesting in China is that dissent is manifesting in these really sort of small interstitial interesting ways akin to what I described earlier in the form of humor so what may not appear to be a massive critique of the institutions of the state or the communist party's management of affairs can sometimes lead us to believe that there's nothing really going on whereas arguably I think the extent to which social media penetration in China is increasing it provides a significant platform for these dialogues and these discussions to happen and I would go back to what Zanev said on the issue of symbolism I mean what appears to be like a very very small negligible expression you know online by not necessarily even an activist but an ordinary person who acts as a node or as part of a significant cluster can be much more dissenting and much more sort of complicated and embedded in meaning than then let's say a call to protest on January 25th where everybody says we're going to remove the government so so I think we can benefit from these these minor nuances as opposed to imagining something so substantial which I think was a little bit romanticized to begin with that we can we can look at the sort of the microscopic as opposed to the microscopic view so there's one thing there's great amount of platform politics we now have algorithms as political players which really we need to sort of dive into right what Facebook news feed shows you our trending topics as calculator who gets in where who doesn't are incredible consequence and that's one of the ways governments are trying to intervene right trying to control the platforms there's a lot of contention over that in Turkey who gets in into China is a big deal I mean so that's I think there's a new kind of study of algorithms as political players platforms as political players the in Turkey it's Facebook and Twitter not really alternative sites right now it just depends on where you are on the planet how that plays out in terms of bridges and unity I have to say I mean social media is interesting it's an amplifier for sure so it can amplify many things on the ground it can amplify polarization amplify dehumanization it can create waves of vortex of polarization where people are increasingly instead of coming together sort of able to do this because it changes visibility a comment that might have been an offhand comment an ethnic slur that might have been just said in the quiet can now be visible to larger audiences which can drive these dynamics of polarization it can also unite our you know here are 17 things that will restore your faith in humanity listicles but that's kind of things that we like clicking on that make us feel good it can do both of those things at the same time so I mean a lot of this doesn't do x doesn't do y I keep saying yes I mean it's a more and more situation where the humanity it's probably happening at the same time low level dialogue between people who are politically otherwise in different parts of the spectrum what you said earlier thing is very important that as we connect person to person since people are complex we get more and more diverse views that we otherwise wouldn't have and it sometimes flares up during the Wisconsin union conflict Facebook was a mess if you lived in that state because families were fractured along the politics so in the old way where it was information driven I think it was easier to be filtered by topic you just want people like you whereas right now because you're connecting to person to person because people are a lot more complex than just you know if you believe in this you also believe in that you know there are people who are you know say in any case there are people who can be in and out of state places the China example is fascinating China Gary King's research is a great example of a lot of things that you mentioned the idea that China's politics is not impacted by social media just because China censors in certain ways is nonsense obviously it's had a huge impact you know people if they petition the emperor like they put you know they kind of use social media to petition the emperor in a lot of rural places the way that he used to happen in the middle middle ages because it's a way of getting the power to say this is what we don't like and Gary King's lab shows that that stuff doesn't get censored that's actually Chinese government is responsive you see the similar dynamics in Singapore a lot of authoritarian governments are becoming more attuned to their citizens in ways that are legitimately viewed as participatory but that are not loading you know it's complex what that does to the legitimacy of a government and maybe since we don't have open surveys in China we don't know but it may well be that there's large numbers of people in China in places like Singapore that may prefer this slower gradual political shift then because they're looking at the rest of the world and they're seeing turmoil so I think it's quite complex it's in its impacts and no either or no online or offline I know the sort of that let's look at all the dynamics and their totality okay very quickly on the alternate platforms and on the question of being afraid of being on social media I actually think these two are related right like we if you're going to study this thing I think you have to on the one hand be very careful because there are and it's good to hang out with like 19 year olds and 18 year olds because there are new platforms sprouting out all the time and watching how people I mean did anybody how many of you knew this room to the 450 million people were using WhatsApp I mean like that's a large number of people right like six months ago Tumblr was a new big thing Instagram you know I'm not even sure where Tumblr is right now so there's a limit to kind of as a scholar what you can study in this regard I think for the moment we're fortunate in the sense that a lot of traffic is coming to Facebook and to Twitter and I think in Russia you had to be very careful because there was a lot going on but contact day I think it's more migrating to Facebook now more and more we would like to look at the we will eventually look at the contact day traffic in Ukraine as well and I think there's two ways to think about this one is that you get an incomplete part of the story if you look at different if you only look at certain platforms and I think with all of this stuff the key question we always have to ask ourselves is how biased is our sample and the normal way that we think about bias is how representative is it of the general population but the thing about multiple platforms leads us to remember we also have to ask that question about even how biased is it of the internet population and the digital population and there will be some cases in some countries where you'll miss crucial demographics by not looking at platform X but I think the more interesting thing is the sort of strategic sense of it right like here's the question in the Ukraine right is the contact day where you're going to find more of the answer to the previous question of pro Yannikovic sentiment I have no idea I mean maybe it was maybe it wasn't so I think those but it's also the answer to the fear and the afraid question right like there's constantly new platforms being generated right and that is part of the way and Zina was talking before about having or someone was saying before about having multiple personalities online I mean it's you go to one platform to for your parents to see you go to another platform your friends to see you there may be another platform that you end up going in places to be to be more revolutionary the bigger point here I think on this question of fear of being on these platforms is that for me what's motivated me in the why I got interested in this in the very beginning in addition to the reasons that I mentioned before I was really interested in the question of how the effects so the big picture theoretical question is how does social media affect political participation for me the two big sort of interactive things here is how does it vary across more and less open societies and how does it vary across higher and lower cost levels of political action and I think that this is you're going to get to think that there's going to be a one-size-fits-all answer to social media and political participation is misguided but precisely the reasons that you're that you're identifying here as well well my role is to thank Adelis Kandar Joshua Tucker Zaynep Tufeki for a wonderful panel and then turn the platform back over to Sheldon Heist's Institute peace for hosting us today well thank you it has been a rich and full morning on what's clearly a continuing complex unfolding story and we've got a lot of folks to thank for that first and foremost our moderator PJ Crowley and six outstanding panelists today so second I want to make sure to thank the USIP and the GW staff that have really put today's meeting together Anand Varghese right here and his teammate Julie behind him and Lola where's Lola in the room I don't see Lola but Lola from GW was also invaluable our AV guy Steve back there and the meetings folks Jamie is going but thank you all very very much for helping take this pull this together and also I really want to thank that questioner who stood up and said let's talk a little bit about the peace building application of social media I just came from a fascinating meeting where Twitter it's interesting to see the social media companies themselves doing the analysis and a fellow from Twitter said they had done an analysis of what looked like a onslaught of hate speech online when if you recall an Indian American won the Miss America contest and what was most interesting is very quickly the number of tweets that to your point was in about amplification the number of tweets that amplified how racist that comment was how out of line it was and inappropriate and it swarmed the negative and in fact within 48 hours that voice was completely drowned out so there is no question this cuts in multiple ways and that's exactly why we have this series finally my father was a school teacher he used to say the key to a really good meeting is that the speakers deliver interesting information the audience listens carefully and both finish their jobs at the same time and I think that we met his bar so thank you all very much and we look forward to continuing this conversation in the future