Ethan Zuckerman’s online home, since 2003
Currently Browsing: Media

A public apology – on screwing up by not questioning assumptions – my talk at #BIF10

I spoke this morning at the tenth incarnation of BIF, Business Innovation Factory, an annual conference in Providence, RI that focuses on storytelling. It’s got a lot less product promotion and self-celebration than many conferences on innovation, and more personal stories, which is why I always enjoy attending. So I thought I’d share a personal story that I’m still digesting and the questions it raises for me. If it reads a little differently from most of my writing, the context helps explain why.


About a month ago, I wrote an article about a simple idea. I asked whether anyone really believed that advertising should be the main way we supported content and services on the internet. Given how poorly banner advertising on the web worksGiven that nobody likes banner ads, and given that the current system puts users under surveillance, which in turn seems to inure us to government surveillance, I wondered whether there might be a better way.

Initially, the piece did what I hoped for – it sparked a lively conversation about business models for the web, particularly about business models for news. My friend Jeff Jarvis, whose faith in Google inspires envy in leaders of the Catholic Church, reassured me that once advertising got a little better, I’d like the ads I was reading online, really! I heard pitches for ethical advertising, for different approaches to subscriptions and to micropayments. I’d gotten something off my chest, had sparked a good conversation and was learning from the responses. As a blogger and a media scholar, life was good.

But something unexpected happened halfway through my day in the sun. For a brief and uncomfortable time, I became the most hated man on the internet. Here’s how that happened.

In writing about advertising, I wanted to talk about mistakes we’d made collectively, as an industry, not just beat up on Mark Zuckerberg or any other individual. So I took responsibility for my own small contribution to making the world an ugly, ad-filled place. I admitted, somewhat sheepishly, that almost twenty year ago, I invented the pop-up ad.

My intentions were good. (Please stop throwing things at the stage.) I was part of the team that started Tripod.com, an early webpage hosting company. My popup put an ad in a separate window from a user’s webpage, which was a way of distancing an advertiser from user generated content, reassuring advertisers that they wouldn’t actually be on the same page as a paranoid essay about government conspiracies or a collection of nude images. I thought it was a pretty good solution. I’m really, really sorry.

My editor at the Atlantic, Adrienne LaFrance, is a lot smarter about her readers I am. She knew that this admission, which I hid as a parenthetical comment, would be far more interesting to most readers than a 3000 word essay on advertising and a culture of surveillance. She did a brief interview with me about the process of creating the ad and wrote a 300 word story titled “The First Pop Up Ad”

And then things got weird.
Yes, that’s Jimmy Fallon and Conan O’Brien making fun of me. And yes, that’s one of the strangest things that’s ever happened tome.

Adrienne’s Atlantic piece, lightly rewritten, appeared in about forty news outlets over the next 24 hours. I got loads of interview requests and I did one of them before realizing that this was a very bad idea, and that even normally staid news outlets like the BBC would be far more interested in my “confession” than in the broader argument about advertising and surveillance. And then the emails and the tweets started to come in, first cursing me out in English, and then in Turkish, Portuguese, Chinese and Croatian as the story spread globally.

Andy Warhol predicted that, in the future, we would all be famous for 15 minutes. There’s another possibility: on the internet, we will all be intensely loathed for about 15 seconds.

Let me just pause for a moment and mention that my no good, very bad day on the internet is approximately equivalent to what many of my friends refer to as “Friday”. Which is to say, if you are an opinionated woman who writes online, you likely face ugly, misogynistic bullshit on a daily basis significantly more acute that the tongue in cheek death threats I received once this story took off. I don’t in any way mean to compare the mild abuse I faced at a moment of extreme visibility to the sorts or routine, everyday harassment many smart women face for merely expressing themselves online.

That said, I do now understand why someone would choose to go offline rather than wallow in hostility. I ended up taking a week entirely offline while things blew over, until only one in five emails was a death threat.

Of all the tweets hoping for my swift and painful demise, I was particularly struck by one written by a user in southeast Asia, whose tweet read, more or less: “@ethanz I do not accept your apology. This is your Frankenstein monster. You made it, you should kill it.”

I really liked this tweet. There’s something refreshingly simple about the idea that someone – one guy – could be responsible for things in the world that are badly broken. And once we find that guy, we can either string him up or demand that he fix it. It’s the inverse of the great man theory of history, where we declare Edison the great man who commercialized electricity and brought about the modern world, cutting out Tesla, Westinghouse and hundreds of others. This is the rat bastard theory of history, where if only that one SOB hadn’t put pop up ads on user homepages, the world would be a slightly pleasanter place.

The rat bastard theory is helpful because it gives you a single, concrete individual to hate. Back in my darkest, saddest days, when I used to use Windows, I really enjoyed hating Bill Gates, despite the fact that my anger was clearly misplaced. It’s clear that there were all sorts of design decisions that made Windows agonizing to use in the late 1990s, not an evil plan from Bill Gates, who has turned out to be a pretty terrific guy. But it’s a lot easier to see the rat bastard than to see the whole problem.

The problem with the internet in the late 1990s is that we wanted it to be available to everybody in the world. That’s partially because it seemed like the fair thing to do, and partially because we genuinely loved the web and wanted to evangelize for it globally. But most people didn’t know why the web was cool, why they’d want to build a homepage or send status updates to friends, so they weren’t about to pay us to use it. And so we ended up with the only business model we could think of – broadcast advertising. You’ll get our services for free, and we’ll demand some of your attention to sell to advertisers in exchange. Basically, we didn’t know how to pay for the internet for everyone, so we decided to make the internet work the way broadcast television worked.

We had a failure of imagination. And the millions of smart young programmers and businesspeople spending their lives trying to get us to click on ads are also failing to imagine something better. We’re all starting from the same assumptions: everything on the internet is free, we pay with our attention, and our attention is worth more if advertisers know more about who we are and what we do, we start business with money from venture capitalists who need businesses to grow explosively if they’re going to make money.

But there are people imagining something very different. Maceij Ceglowski, whose brilliant talk got me thinking about the broken state of the web, has built a bookmarking service called pinboard that’s really cheap – about $5 – but not free. He did it because he imagined something different, not a site to sell to another company, but a service he wanted to have available to the world that’s been profitable since the day he started it. WhatsApp is a wildly popular messaging service that charges users $0.99 a year, and had over 400 million users before Facebook bought it, an amazing example that people are willing to pay for tools they need and use. I just learned about a social network called Connect Fireside that’s designed for the mobile phone and optimized for sharing photos with your family and close friends, not with the whole world. It’s not cheap – it’s $20 a month right now – and who knows if it will survive, but it’s exciting to see someone build something based on a different set of assumptions.

Once you get rid of those assumptions – everything is free, the user is the product being sold to advertisers, and the goal is to be a venture-backed billion dollar business – and lots of things are possible. My friends in Kenya got sick of losing internet connectivity every time the power went off, so they went onto Kickstarter and funded BRCK, which is a portable internet router for the developing world that relies on the mobile phone network. The people who paid for it were people who wanted and needed the device, so were willing to pay for it to be built – crowdfunding, as a model, asks that you give up the assumption that people have to instantly receive the things they buy. Turns out that sometimes we’re willing to pay for something in the hope that it might come to pass, someday.

Turns out we’re also willing to pay for things because we love them and we want them to exist. I’ve become a massive fan of a deeply strange podcast called Welcome to Night Vale, which might be described as a cross between A Prairie Home Companion and the Twilight Zone. The podcast is free, but listeners are encouraged to donate $5 a month, and many do… Many more come to live shows, buy tickets and merchandise, which has allowed the creators to expand the podcast, bringing in new actors and musicians, launch a 16 city tour of Europe and write a book about the Night Vale universe. Putting something beautiful and strange out into the world and hoping people will love it is not the most reliable business model, but it sometimes turns out to be viable.

Remember my friend on Twitter who wanted me to kill Frankenstein’s monster? He got part of it wrong in assuming that I’m the rat bastard solely responsible for the situation we’re in today. But he got part of it right. He’s right that people like me can and should be trying to fix things.

The things that are broken about the internet today – and there are a lot of them – are the product of design decisions that fallible, mortal human beings have made over the past twenty years. Twenty years is not a long time. I have t-shirts older than the world wide web. The web wasn’t built by enlightened geniuses whose trains of thought we could never comprehend – it was built by idiots like me, doing the best we could at the time. It’s possible to look at every technical and design decision that’s led to where we are today and make different choices than the people who made those bad choices.

There’s a lot of things I don’t like about how the web works today. I don’t like that our attention is constantly for sale. I don’t like that public sharing is the default. I don’t like that the web never forgets. I think the always on, inescapable nature of the web is proving really exhausting for us as human beings. And I’m scared that the new public spaces, the places where we come together to debate the future, are owned and controlled by a few massive companies that have enormous potential power over what we’re able to discuss online.

But the mistake would be to assume that these shortcomings are inevitable, that they are simply natural consequences of how people interact online. The mistake is to stop questioning the assumptions about how the world worlds, to stop imagining ways things could be different and better.

I should probably mention that this is not a talk about internet business models.
Actually, it’s a talk about civics.

If you think the web is broken – and I do – you should take a look at American democracy. Here’s another case where fallible humans made design decisions that seemed like a good idea at the time and now have had some clearly disastrous consequences. Sure, having professional representatives deliberate together and govern a nation is a pretty cool hack when the dominant governance technology of the time is feudalism. One person, one vote elections – once we finally got around to fully implementing it? Cool idea. Freedom of speech for individuals as well as organizations? Seems pretty important for the rest of the system to work.

Put these things together and we’ve somehow ended up with a system where the Democratic Congressional Committee recommends to new congresspeople that they spend 4-5 hours a day raising money, and only 3-4 hours meeting with constituents or working to craft legislation. It’s a system that constituents hate – it’s part of the reason Congress has a single-digit approval rating – and Congresspeople hate, too, which is why some incumbents are choosing to leave office. But it’s not hard to figure out how it happened, to trace the decisions that brought us to a deeply undesirable place. We can invoke the rat bastard theory and blame Chief Justice John Roberts for Citizens United decision, but we’re following the same fallacy – this is a failure of imagination, not the failure of any one person (or even five supreme court justices)

It’s possible to imagine something very different. Larry Lessig is working very hard to bring a new model to life, where every voter gets $50 of government funds to give to politicians who can use the money to campaign. To imagine that model, you have to give up a bunch of assumptions: that you have a right to spend money to influence politics; that governments should try to do less, not more. And you have to be ready to cope with all the unintended consequences of the new system as well, of people selling their government funds to politicians through brokers, of social media becoming an endless campaign battle ground.

We need a ton more of this, people questioning assumptions that representation makes more sense than direct democracy, that decisions at the federal level are more important than those locally, that money is convertible into power. We need lots of Lessigs taking on these sorts of imaginative experiments, because most of them are going to fail.

Here are two really big assumptions I want to question:
– that participating in representative democracy is the core act of what it means to be a good citizen
– and that everyone is going to participate in our democracy in more or less the same way

You know what it means to be a good citizen – you’re supposed to read the newspaper and keep up on events locally and globally, and vote every two years and maybe call your congressman if something really pisses you off. And you probably know that this isn’t going to make very much of a difference. And that your congressperson is heading into a paralyzed institution that’s rarely able to pass legislation.

So let’s question the assumption that fixing Congress – or even more ambitiously, fixing politics – is the most important part of fixing citizenship. In the 50s and 60s, people figured out that, when you’re prevented from voting by law, public protests – marches, sit ins, boycotts – are a critical part of citizenship. As Congress starting passing civil rights legislation, activists learned that lawsuits were a critical tool to ensure that laws applied to everyone and enforced fairly. We rarely think of suing the government as a way to be a good citizen, but it’s been critical in building the rights-based society we have today.

There are at least three ways I think people can be civically active even if they’re frustrated with paralysis in politics. Make media. When people saw tanks in the streets of Ferguson, Missouri to counter peaceful protests, they made media and started a conversation about the militarization of America’s police forces. When they saw a newspaper picture of Michael Brown looking threatening, they started tweeting #iftheyshotmedown, asking whether the media were being fair in their choice of picture to portray the victim of a police shooting. Media is how we understand the world, and we can shape our public conversation by building civic media.

Make new things. I’m deeply frustrated that my government is surveilling my communications and those of millions of others worldwide, and I’m angry that so little legislative action has been taken in the wake of Ed Snowden’s revelations. But I’m really happy that software developers like Tor, Mailpile, Redphone and Mailvelope are working to make it easy to encrypt email and phonecalls and make it harder for anyone to surveil our communications. Make businesses. I’m concerned about climate change, I want to see a carbon tax, but in the meantime, I’m excited to see Tesla making electric cars that are sexy and for people as broke as me, excited that I drove here in a diesel car that gets almost fifty miles to the gallon. Building businesses that do well while doing good is one of the most powerful ways we can engage in civics today.

Here’s what’s tricky about expanding civics to include making media, making code and making money: it’s not a level playing field. Voting is something everyone can do – right now, making code or making media is easier for some people than others. We’re going to need to think hard about how we prepare the next generation of citizens for a world where their power comes not just from voting but from making, and where someone’s civics unfolds in the marketplace while someone else’s unfolds in Congress.

If you want to overcome failures of imagination – accepting that the web or our politics are inescapably broken – you’ve got to try something new. You’re probably going to fail. And when you do, I recommend that you try again. But first I recommend that you apologize. It feels really good. I’m sorry, and thank you for listening.


Self-segregation on social networks and the implications for the Ferguson, MO story

Michael Brown was fatally shot by police officer Darren Wilson on Saturday August 9, 2014. After his body lay in the hot street for four hours, Ferguson residents took to the streets to protest his killing. Brown’s death, the ensuing protest and the militarized police response opened a debate about race, justice and policing in America that continues today.

I heard about Michael Brown’s death for the first time on the evening of the 9th, through Twitter. Sarah Kendzior, a St. Louis-based journalist, was retweeting accounts of the protest, and other Twitter friends who write about racial justice issues, including Sherrilyn Ifill, the director of NAACP’s Legal Defense Fund, were discussing the implications of Brown’s killing and the community and police response. I wrote a quick blog post, noting how little mainstream media response there had been at that point, and how much of the coverage had focused on the anger of the crowd, which one outlet termed “a mob”. At that early point, there was not a dedicated Wikipedia article focused on Brown’s death, nor had major national newspapers, like the New York Times, written about Brown. For many people outside of the reach of local St. Louis media, Twitter was the medium that introduced people to the city of Ferguson.

The Pew Research Center confirms this picture. In a thorough analysis of early media coverage of the events in Ferguson, Paul Hitlin and Nancy Vogt of Pew note that cable news didn’t report the story until Monday the 11th, two days after the shooting. Cable news attention peaked on Thursday, August 14, after Ferguson police arrested two reporters and President Obama interrupted his vacation to report on the situation. Twitter attention peaked at the same time, with 3.6 million Ferguson-related tweets on the 14th, but it’s worth noting that Twitter showed a steady and growing interest in the topic from Saturday the 9th onward.

This pattern of attention to Michael Brown’s killing contrasts sharply with early attention to the killing of Trayvon Martin by neighborhood watch volunteer George Zimmerman on February 26, 2012. My students Erhardt Graeff, Matt Stempeck and I studied media coverage of the Martin killing using our Media Cloud tools and found that they key factor in expanding the conversation about Martin’s death beyond Central Florida was a careful PR campaign orchestrated by Benjamin Crump, an attorney retained by Martin’s parents. These PR efforts led to national television coverage of the story, which in turn led to a surge of interest on Twitter and in other participatory media.

In the case of Michael Brown, over 140,000 Twitter posts mentioned Ferguson on August 9th, the day of the shooting. Twitter led other media in 2014, while broadcast media led in 2012. When Crump announced he would be representing the family of Michael Brown on August 11, two days after the shooting, there was no need for a PR campaign, as the story of Brown’s death was already gaining traction in the media.

Pew’s analysis of Michael Brown versus Trayvon Martin shows that attention to events in Ferguson built more quickly than attention to events in Stanford, FL. There are several reasons for this. The police response to protests added a dimension to the Ferguson story that was not present in discussions of Trayvon Martin – many of the tweets on August 14 focused on the arrest of working journalists and the use of military equipment by Ferguson police. (When Missouri Highway Patrol Captain Ron Johnson took over control of law enforcement response in Ferguson on August 14, ordering officers to remove riot gear, attention on Twitter dropped sharply, suggesting some of the attention was on policing tactics.)

Twitter’s reach has grown since 2012, and the power of “black twitter”, which Soraya Nadia McDonald describes as both “cultural force” and social network, to challenge racist narratives has grown. Outrage on Twitter over a book deal for a juror who acquitted George Zimmerman led to the deal being withdrawn. Noting the power of outraged channeled through a hashtag, some African American Twitter users began posting contrasting photos of themselves under the tag #iftheygunnedmedown, suggesting that media were emphasizing the narrative of Michael Brown as “thug”, by using a photo in which Brown is displaying a peace sign (which many have read as a gang sign), rather than other photos in which Brown looks young and entirely unthreatening. The hashtag allowed Twitter users to participate in the dialog about Ferguson in cases where they had nothing to share about the situation on the ground by participating in a conversation about larger issues of structural racism in American media.

While events in Ferguson received widespread attention on Twitter, some observers saw very different behavior on their Facebook feeds.

John McDermott, writing in Digiday, offered this elegant formulation: “Facebook is for Ice Buckets, Twitter is for Ferguson”. Using data from SimpleReach, McDermott suggests that there were roughly eight times as many stories about events in Ferguson posted to Facebook as stories about the ALS Ice Bucket challenge, but that the stories received roughly the same exposure on Facebook (approximately 3.5 million Facebook “referrals” – appearances on a Facebook user’s screen – for each story between August 7-20th.) In the same time period, Crimson Hexagon calculated that there were roughly 1.5 times as many tweets about Ferguson as about the Ice Bucket Challenge.

The near-equal attention to Ferguson and the Ice Bucket Challenge that SimpleReach’s numbers suggest hardly seems like the substance of a controversy about algorithmic censorship, but it is worth noting that vastly more stories about Ferguson were posted to Facebook than stories about the Ice Bucket Challenge. The average story about the Ice Bucket Challenge was much more heavily promoted by Facebook’s algorithms (a factor of 8x) than the average story about Ferguson. Sociologist Zeynep Tufekçi suggests that this disparity may have been more significant early in the Ferguson story, which made the disparity between Facebook and Twitter more dramatic.

Tufekci goes on to observe that while Twitter (at present) is a “neutral” provider, simply delivering you tweets from the accounts you’ve chosen to follow, Facebook algorithmically “curates” your feed, offering you a selection of content that it believes you will find most interesting. She points out that this is a serious danger to democratic discourse. Facebook could censor our information flows and it would be difficult for us to determine that such censorship was taking place. The algorithms Facebook uses are opaque, which means conversations about how the algorithm work sound a bit like speculation about divine will. Tufecki suggests that Facebook may have altered their algorithm overnight to give Ferguson stories more prominence, which may be true but is impossible to verify. Others have speculated that Facebook tunes its algorithm to favor “happy news” and discourage serious news. Again, this is possible – and could even reflect lessons learned from Facebook’s mood manipulation experiments – but unverifiable.

Here are some of the things we do know about Facebook, which may explain why news appears differently on different social media platforms.

It’s hard to search Facebook. When news breaks, Twitter users enter search terms and hashtags and receive an updated stream of tweets on the topic in question. On August 9th, I was able to follow reports from Ferguson and quickly choose Twitter users to follow who appeared to be at or near the scene. On Facebook, searching for “Ferguson” gives you the names of many people named Ferguson who Facebook thinks you might know. (You can, conveniently, choose to “Like” the city of Ferguson, which is probably pretty far from what most people searching for Ferguson want to do.)

Non-reciprocal following makes it easier to diversify who you follow on Twitter. If I see that St. Louis-based rapper Tef Poe is reporting on Twitter regarding Ferguson, I can follow him, and I will see any public tweets he makes in my timeline. On Facebook, I need the approval of someone to follow their personal feed – they need to confirm our friendship before I can read what they’re saying.

It’s typical on Twitter for people to start following accounts of people they don’t know personally, especially when those people are eyewitnesses to or well-informed commenters on an event. I followed a set of users in Egypt while the Tahrir protests were taking place and stopped following many afterwards – I’ve done the same with Ferguson. This is much more difficult to do on Facebook – the platform is designed to connect you with friends, not sources, and to maintain those friendships over time.

People have a relatively small number of Facebook friends – the median figure is 200 – and many white American Facebook users likely have few or no African-American Facebook friends. This isn’t a phenomenon specific to Facebook – it’s a broader reflection of American demographics and patterns of homophily, the tendency of “birds of a feather” to flock together. Robert Jones at the Public Religion Research Institute used data from the 2013 American Values Survey to estimate the racial composition of people’s social networks. He projects that the average white American has only one black friend, and that 75% of white Americans have entirely white social networks. (There’s an interesting debate about PRRI’s methodology, with Eugene Volokh arguing that it’s hard to estimate weak social ties by asking people about people they confide in.)

We can build on these numbers to speculate about what might be going on with Facebook and stories about Ferguson. Most people who follow Facebook believe that the company’s algorithm favors stories likely to be shared by a lot of people in your social network – Facebook favors viral cascades over surfacing novel content. If few users in your circle of friends are sharing stories about Ferguson – which is a distinct possibility if you are white and most of your friends are white – Facebook’s algorithm may see this as a story unlikely to “go viral” and tamp it down, rather than amplifying it, as it has with the ALS stories. On the other hand, if many of your friends are sharing the Ferguson story – likely if you have many African-American friends or if you follow racial justice issues – your corner of the Facebook network could be part of a viral cascade.

A new paper from Pew and Rutgers suggests another reason why Ferguson might have had difficulty spreading on Facebook. Keith Hampton and the other authors of the study found that people are acutely attuned to what conversations are happening on social networks and are unlikely to bring up topics perceived to be controversial. The users the authors surveyed were half as likely to bring up controversial topics on Facebook or Twitter as they were in face to face conversations. The exception was when they believed their friends on social networks agreed with their positions, in which case they were more likely to bring up the topic on social networks. In the case of Ferguson, this suggests that a Facebook user, unsure of whether her friends share her opinion that the police overreacted in Ferguson, might be more reluctant to post a Ferguson story on Facebook than she would be to bring up the situation in conversation. If there were a cascade of stories on Ferguson, she would have a social signal that the topic was acceptable and would be more likely to participate. On Twitter, where it’s not uncommon to follow dozens of people you don’t know well, it’s easier to interpret those social signals than on Facebook, where you are more likely to know an ethnically homogenous set of friends.

I’m far from immune to these echo chamber effects. One of my students surprised me with a link to a USA Today article reporting that more money had been raised to support Darren Wilson, the police officer accused of killing Brown than had been raised by Brown’s family. Few of my Facebook or Twitter friends were raising money for Officer Wilson, and so I was surprised to discover that a successful campaign was being held on his behalf. My ignorance of the campaign to support Wilson suggests that I have my own diversity issues in social media – if I want a more nuanced understanding of Americans’ reactions to Ferguson, I need to increase the ideological diversity of the people I follow online.


Special bonus: Jon Stewart takes the Ferguson challenge, which turns out to be utterly unlike the ALS Ice Bucket challenge.


When Being Happy is Political

The Atlantic was kind enough to run a lightly edited version of this post. I’m posting here after their publication so that this remains in my archives.


Pharrell Williams is a happy man, but he’s crying. He’s one of the most in-demand record producers in the world, and had a hand in the two hottest songs of 2013, “Get Lucky” by Daft Punk and “Blurred Lines” by Robin Thicke. While those songs were inescapable on radio and television last summer, Pharrell’s most recent hit, “Happy”, has taken a different path to prominence.

French director Yoann Lemoine and production team We Are From LA worked with Pharrell to create a unique video for “Happy”. The video is 24 hours long, and was shot all across Los Angeles, featuring dozens of celebrity cameos interspersed amongst shot after shot of people dancing happily. It took 11 days to shoot the video, though many of the shots were single takes. The video follows the course of the day in LA, with footage from dawn to dusk and through the night, with Pharrell appearing each hour.


The original “Happy” video

The video quickly spawned thousands of fan remakes, featuring workplaces, business schools, college dorms who are all happy. Faced with a viral hit, Pharrell’s label, Columbia Records/Sony Music, has turned a blind eye to possible copyright violations, and one can easily spend hours on YouTube flipping from one fanvid to the next.

There’s a special subcategory of these videos that I think of as “georemixes”. The georemix builds on the idea that the original “Happy” video is a love letter to Los Angeles, a portrait of the city’s architecture, landscapes, people and spirit, and moves the party to a new location. More than a thousand georemixes of “Happy” exist, and they portray happy people on all six continents.


Pharrell, on Oprah, crying over “Happy”

Pharrell is on Oprah, watching a compilation of these remixes that bring his song around the world, from Detroit to Dakar. In the 30 seconds of the video Oprah shows, we catch glimpses of happy Taiwanese women on a spa day, Icelanders dancing on a glacier and Londoners strutting with Big Ben in the background. Pharrell’s reaction is the one many of us have had to the remixes of his video: he cries for a long time, overwhelmed not only by his success but by the experience of watching a simple idea – film yourself being happy – as it spreads around the world.


“Happy” is not the first video that’s been georemixed. Last summer, I gave a talk at the MIT8 conference focused on remixes of PSY’s Gangnam Style and Baauer’s Harlem Shake. In researching these localized remixes, my students pointed me to Jay-Z and Alicia Keys’s “Empire State of Mind”, remixed in remarkable fashion into “Newport State of Mind”, by comics M-J Delaney, Alex Warren and Terema Wainwright. (The parody was further parodied by Welsh rappers Goldie Looking Chain, who complained that the Newport parodiers lacked local knowledge and cred.)


The original “Empire State of Mind”


“Newport (Ymerodraeth State of Mind)”

The georemix dates back at least as early as 2005, when Lazy Sunday, produced by The Lonely Island (Saturday Night Live’s Chris Parnell and Andy Samberg) was remixed into parodies like Lazy Muncie, showing midwest pride, and Lazy Ramadi, which replaces a search for cupcakes with a confrontation with Iraqi insurgents.


Lazy Munzie


Lazy Ramadi

The Lazy Sunday georemix was born out of a mock East Coast/West Coast rap beef, which quickly set the tone for georemix videos. Each response is a retelling of the core story, transposed to a new location, bragging about local landmarks and habits. While the braggadocio in these remixes is pure parody, there’s a sense in which each of these videos makes a claim to share the stage with the original. YouTube’s related videos feature means that there’s a chance that some of the 2 billion viewers of PSY’s Gangnam Style video will encounter Zigi’s “Ghana Style”, a georemix that relocates Seoul to Accra and replaces PSY’s horse dance with Ghanaian Azonto. (And if not through YouTube, viewers may encounter Zigi through the hundreds of listicles that advertise “10 Best Gangnam Style Parodies”)


Zigi’s Azonto version of Gangnam Style

I think of the georemix as a claim to attention, a way of demanding part of the spotlight that shines on a popular video. It’s a very basic demand: accept that we’re part of this phenomenon, too. In remixing Gangnam Style, Zigi sends the message that Ghana has YouTube, is clued into global cultural trends, has its own distinct sound and dance style to share with the world, and can produce videos as technically proficient as anything coming from other corners of the world. To me, “Ghana Style” reads both as lighthearted celebration of a catchy tune that truly went global, and a political statement about a world where culture can spread from South Korea to Ghana to the US, not just from the US and Europe to the rest of the world.

Of course, the georemix can also be purely political. Ai Wei Wei’s Gangnam style, titled Grass Mud Horse Style, moves the dance into his studio in Beijing and is filmed almost entirely within the walls of that compound, alluding perhaps to the artist’s frequent arrests and detentions. (If the location doesn’t set the theme, his appearance a minute into the video, spinning handcuffs certainly does.) Other georemixes take on specific issues explicitly. Consider Dig Grave Style, a protest video made by students from China’s Henan province, in which dancers rise from the earth to protest the moving of graves from villages to open land for real estate development.


Dig Grave Style

Remixing a video is a shortcut to creating original content. The script is partially written – the creativity comes from changing the lyrics and the setting. The popularity of the Harlem Shake meme (which was georemixed around the world, and saw political georemixes in Tunis and Cairo) came in part from the extremely low levels of effort required to participate in the phenomenon – simply film people behaving in an ordinary way, then dancing like madmen in strange costumes and you’ve got your localized Harlem Shake.

“Happy” benefits from this low barrier to entry. There are Happy remixes that function as shot-by-shot remakes of the short, official Pharrell video, and there are vastly more that adopt the spirit of the video and transpose it to a local context.


Loïc Fontaine and Julie Fersing deserve much of the credit for the georemixes that made Pharrell cry, though neither has made a video. Fersing, an interior designer in Nantes, began collecting georemixes of “Happy”, searching YouTube to find new material. When she’d located 21 of the videos, she turned to her husband, Fontaine, who’d begun a career in website development nine months earlier. Together, they launched We Are Happy From, a portal that now hosts 1682 videos from 143 countries.

We Are Happy From front page

We Are Happy From front page

Once the site had attracted about 50 remixes, Fontaine contacted the We Are From LA production team, who gave the project their blessing. While Fontaine had not spoken to Pharrell when I interviewed him a month ago, he felt quite confident that the project was consistent with the artist’s wishes and would survive, pointing out that Sony had not taken action to remove the vast majority of remix and parody videos posted online. Indeed, the success of the song has likely had a great deal to do with the widespread participation online, giving “Happy” an online life and prominence that no amount of radio payola could provide. (Pharrell has embraced the notion of the georemix, urging people around the world to produce their versions of the video as part of a UN-sponsored International Day of Happiness.)

We Are Happy From is simply an index, pointing to videos hosted on YouTube, Daily Motion and other platforms. While the videos have a consistent look, usually opening with a black on yellow title screen (as Pharrell’s video does), Fontaine doesn’t provide any production help or guidelines. Still, the videomakers are clearly conscious of We Are Happy From’s role in promoting “Happy” videos as a global form, as many videos feature a screencap of the We Are Happy From map.

While anyone can submit a video to We Are Happy From, not all videos appear on the map. Fersing is the curator, and she watches all videos before adding them to the map. (As of April, the couple were receiving 20-40 videos a day.) Videos that are overly commercial or connected to political or social causes don’t make the cut. Fontaine explained that some French political parties produced Happy videos as campaign materials – We Are Happy From chose not to feature those videos. An Italian version of Happy with an environmental message was also not included, nor was Porto (un)Happy, which features activists dancing through unfinished construction sites in Porto Allegre, Brazil, along with subtitles that critique government spending on public works projects. (Manaus is unhappy as well.)

I asked Fontaine why he and his wife had chosen to become active curators of the project. It was a practical decision, Fontaine explained: “They say it’s black, someone else says it’s white. How am I to judge?” Rather than evaluating the validity of political claims, he would rather focus on what he sees as the core message of these remixes: “We Are Happy From is purely about the happiness. We just want to show a simple message about being happy about where we live.”

For me, as a student of civic media, the dissident videos excluded from the We Are Happy Map are the most interesting ones. Fontaine has kindly shared the list of rejected videos with me, and I hope to spend some time this summer watching those 500 remixes in the hopes of developing an understanding of how “Happy” can work as a script for advocacy (or how videomakers think it might act as that script.) But for Fontaine and his wife, the mark of success wasn’t raising awareness for a cause or an issue – it was documenting the spread of happiness globally. When I interviewed Fontaine, he was celebrating the spread of “Happy” to Antarctica, with a video from French research station Dumont d’Urville.


The 1600 videos on We Are Happy From may not advocate for a political party or a cause, but they are “political”. When the residents of Toliara, Madagascar make their version of “Happy”, they’re making a statement that they’re part of the same media environment, part of the same culture, part of the same world as Pharrell’s LA. This assertion isn’t quite as anodyne as Disney’s “Small World After All” or the “I’d Like to Buy the World a Coke” campaign. Even with Fontaine and Fersing’s curation, we get distinct glimpses of how different it can be to be happy in different corners of the world: Happy in Damman, Saudi Arabia features wonderfully goofy men, but not a single woman. Beijing is happy, but profoundly crowded and hazy – intentionally or not, the video is a statement about air pollution as well as about a modern, cosmopolitan city.

A few weeks ago, We Are Happy From added a video from Tehran, Iran to the map. If you don’t know where the video is from, it’s unremarkable. A dozen twenty-somethings, men and women, dance on a rooftop, wear silly outfits, and wave their legs while lying on a bed. It’s remarkable only if you know that women in Iran are forbidden to appear in public without their hair covered and that men and women are prohibited from dancing together in public.


Happy in Tehran

Given context, the video is an incredibly brave statement. The young women in the video covered their own hair with wigs, keeping themselves technically in line with local Islamic law, and kept clothing around so they could cover up if seen from neighboring buildings. One of the videos stars, identified only as Neda, said, “We were really afraid. Whenever somebody looked out of a window or someone passed by, we ducked behind a door to make sure we were not seen.”


The makers of the video, forced to apologize on state television

Neda and her compatriots were right to be afraid. Six people involved with making the video were arrested and forced to appear on state television, testifying that they were tricked and duped into making the video. It is unclear what consequences the filmmakers will suffer beyond public humiliation, and a hashtag, #FreeHappyIranians is emerging to protest their detention. Pharrell, to his credit, has tried to call attention to the situation:

It’s clear from Neda’s interview with Iran Wire that the intention behind the video is precisely Fontaine and Fersing’s intention. ““We wanted to tell the world that the Iranian capital is full of lively young people and change the harsh and rough image that the world sees on the news.” They chose a middle-class Tehran home to make the point that ordinary Iranians, not just the elite, were happy, creative, modern and globally engaged. And the video, with subtitles and credits in English, was clearly created for a global audience, designed to be part of the International Day of Happiness, though it was turned in too late for inclusion: “We want to tell the world that Iran is a better place than what they think it is. Despite all the pressures and limitations, young people are joyful and want to make the situation better. They know how to have fun, like the rest of the world.”

Perhaps a video that asserts that you and your friends are part of the wider world is political only if your nation has consciously withdrawn from that world. Perhaps it’s political any time your city, your country and your culture are misunderstood or ignored by the rest of the world. We Are Happy From is cultural politics in the best sense of the word, a good-natured assertion that what brings us together is more important than what divides us. That the Tehran video has led Pharrell to a different type of tears is a reminder of how powerful and threatening this sort of statement can be.


Water monitoring in China, and the changing role of citizenship

This January, a few hundred employees of Alibaba, the massive online retailer and digital payments company, participated in an interesting experiment. Like many Chinese, they traveled home to celebrate the Lunar New Year. While at home, they used inexpensive water testing kits to sample water in their villages and uploaded their findings via smartphone to an environmental mapping website, Danger Maps. Employees measured water quality in 420 locations across 28 provinces, testing open bodies of water as well as sources of drinking water.

The experiment was a trial run for a much more ambitious rollout, announced this week. Jack Ma, Alibaba’s billionaire founder, announced that water testing kits would be sold through Taobao for between 65-80 yuan ($10-13) and invited the public to join his employees in becoming water quality monitors. Yang Fangyi, one of the managers of the Alibaba Foundation, explained that by mapping areas of poor water quality, the Foundation can work with local environmental authorities and NGOs to work on cleanup plans.

watermap
Test results posted to water.epmap.org

Environmental degradation is one of the most serious problems facing China. A report from the Chinese Academy for Environmental Planning suggests that China lost 3.5% of the nation’s GDP in environmental damages in 2010. Air pollution contributed to 1.2 million deaths in 2010, and journalists have compared Beijing’s air quality (unfavorably) to that in airport smoking lounges and industrial London on the most polluted days of the mid-20th century. Maintaining and improving air and water quality while continuing to rapidly industrialize are huge challenges for the country. Environmental issues are also an area where the Chinese government has been comparatively open about discussing problems and seeking international cooperation; Premier Li Keqiang addressed environmental problems in his address to the National People’s Congress last month, and US organizations that work with China report that it’s far easier to cooperate on environmental issues than on more sensitive issues like human rights or worker safety.

textkitpackaging

testkitcontents
The kit, and its contents

The little blue kit, manufactured by Greenovation Hub, may test China’s openness around environmental advocacy. Inside are tests for ph, Phosphates, Ammonia, Chemical Oxygen Demand (used to indirectly measure organic contaminants in water), and for five heavy metals, including cadmium and zinc. It’s more home chemistry lab than slick, sophisticated sensors – you’ll be dipping litmus paper into a stream and measuring the color that results, then entering the data into your phone if you participate in the project.

It’s unclear how many of Alibaba’s 500 million customers will purchase water quality kits and start uploading data to Danger Maps. Even if only a few participate, the implications could be very interesting. Land use issues are a major civic flashpoint in China. If farmers are able to document damage to the local watershed from a new factory, for instance, it might change the dialog, bringing nascent environmental watchdog organizations and government departments into the debate over land use.

Groups like Public Lab in the US and Safecast in Japan have been using crowdsourcing models to document environmental issues, monitoring water quality and radiation levels. Their work raises questions of whether we want citizens to be cooperative sensors, or citizen scientists. The latter is a high bar to cross – we need citizens not only to collect data but to formulate and test hypotheses. What we gain in exposing participants to the scientific process, we may lose in terms of data quality and believability. Safecast has traded accessibility for accuracy – their bGeigie geiger counter is pretty expensive in kit form, but is a lab-quality instrument, which allows Safecast to use the data collected to engage the Japanese government in dialog about post-Fukushima reconstruction. On the other hand, using a Safecast counter, it’s easy to feel like your job is simply that of a data collector, not someone figuring out the complex puzzle of when towns and villages will be safe to inhabit. (Safecast describes itself as a global sensor network, acknowledging that it’s strength is data collection, not the broader issue of citizen science.)

There’s a balance between accessible sensors, high-quality data and the ability for users to formulate and test hypotheses that crowdsensing projects need to wrestle with going forward – based on some of the results thus far, it seems like the Greenovation kit favors access over accuracy. (I suspect there’s not really that much standing water in China at ph10, despite reports on the map.) But it’s possible that communities affected by industrial pollution might purchase multiple sensors, organize testing plans and oversampling to improve accuracy. They might also look for sources of industrial runoff and test hypotheses about how industrial development is affecting their community. Consider a project from CMU called CATTFish. It’s a water monitor that sits in your toilet tank and measures temperature and conductivity to sense possible changes in groundwater quality. It’s designed for communities concerned about pollution from hydrofracking – with high quality, regularly updated data from multiple homes, a community could have an early warning system for detecting potential ill-effects from oil extraction. (h/t to Heather Craig, who introduced me to the project.)

I think there’s another subtle change we should watch for as well. Environmental crowdsensing is a form of monitorial citizenship, an idea we’ve been discussing a lot lately at Center for Civic Media. John Keane uses the term to describe the non-governmental and civic organizations that act as watchdogs, keeping governments honest and, sometimes, in check. Inspired in part by David Ronfeldt’s work on tribes, institutions, markets and networks, we’ve been looking at ways networked individuals can have similar monitorial power. The work we’re starting with Promise Tracker begins with asking citizens to monitor issues in their communities using mobile phones and will likely expand to asking citizens to use sensors to monitor water and air quality.

In our experiments with Promise Tracker in São Paulo and Belo Horizonte, using mobile phones to document community problems and governmental and community responses to them, we quickly learned that many people don’t just want to collect data – they want to use data to tell stories and to advocate for change. Will citizens become sensors or scientists? Participants or activists? This may also have a lot to do with whether Greenovation Hub wants to build a business model or a movement, and whether a powerful, visible figure like Jack Ma is willing to have Alibaba become the nexus of an emerging environmental movement. That might be more potent and less dangerous than having individual groups organize to address water quality issues on a small scale and face potential backlash from local authorities.

I’m interested in monitorial citizenship because I see monitoring powerful institutions – commercial, governmental and otherwise – as something one can do every day as a citizen. Elections come around every few years and get all the attention, but it’s possible that the real power of citizenship comes from the monitoring that takes place between the elections. In a Chinese context, where power doesn’t come through electoral mechanisms, monitorial citizenship may have even more power – it may be a more genuine, authentic, believable path to political power than others available to most Chinese citizens.


Susan Benesch on dangerous speech and counterspeech

Susan Benesch is one of the leading thinkers on countering hate speech online. She’s a fellow at the Berkman Center this year, and I’m terribly sorry to be missing her talk at Berkman this afternoon. (Instead, I’m watching from home so I can be primary caretaker for my son for a couple of weeks while Rachel gets to travel.) She teaches international human rights at American University and is the founder of the Dangerous Speech Project, which tries to understand the spread of speech that incites people to violence.

Susan’s talk is available online and I’ve tried to blog it remotely while cursing my inability to teleport across the state. The talk is wonderfully titled “Troll Wrastling for Beginners: Data-Driven Methods to Decrease Hatred Online”. Unlike most conventional online wisdom, Benesch believes you should engage with the trolls, in part because it may be the most successful path to countering dangerous speech. The approaches states have taken to dangerous speech – punishment and censorship – don’t work very well, and some evidence suggests that they work even worse online than offline. She suggests the case of Anwar al-Awlaki, who was ultimately killed by a drone strike – despite being punished (via summary execution from a US drone), his online speeches continue to be influential and may have influenced the Boston Marathon bombers. Censoring that speech doesn’t work well in an online environment as it’s likely to move onto other platforms.

So what about “don’t feed the trolls”? Benesch points out that there are several implicit assumption in that advice. We assume that if we ignore a troll, they will stop (which, in turn, tends to assume behavior that’s only on a signal platform.) There’s an assumption that online hate is created by trolls; in the few experiments that look at racist and sexist speech, at least half is produced by non-trolls. We tend to assume that all trolls have the same motivations and that they will respond to the same controls. And finally, we assume that the trolls are the problem – we need to consider effects on the audience.

(Benesch doesn’t define trolls until pressed by the audience and points out that it’s a term she uses with tongue in cheek, most of the time – she acknowledges that different trolls have different motivations. Her goal is to move away from considering trolls as the problem and towards understanding dangerous speech as a broader phenomenon.)

One of the benefits of online speech environments, Benesch posits, is that we can examine the effect of speech on people. In offline environments, it’s very hard to measure what reactions dangerous speech leads to – in online environments, it may be possible to track both responses and effects.

Benesch’s suggestion is that we should approach dangerous speech through counterspeech, in effect, talking back to the trolls and to others. In explaining her logic, she notes that the internet doesn’t create hate speech – in some cases, it may disinhibit us from speaking. But more often, the internet creates an environment where we are aware of speech we otherwise wouldn’t hear. Most of us wouldn’t have been aware of what speech is shared at a KKK meeting, and many of us wouldn’t have heard the sexist jokes that were told in locker rooms. Now speech is crossing between formerly closed communities.

This is a new feature of human life, Benesch suggests, and while it causes a great deal of pain, it’s also an opportunity. We can “toss speech back across those boundaries to see what effect it has.” For the most part, we don’t know what will happen when we expose speech this way, and it’s possible the effects could be very positive. She asks us to consider norm formation in teenagers – most 16 year olds, she argues, have historically developed opinions from a small, homogenous community around them. That’s no longer the case, and it positive opportunity for teens to develop a broader and more nuanced worldview.

Believing in counterspeech means having faith that it’s possible to shift norms in speech communities. Benesch asks “What is the likelihood an American politician will use the N-word in public?” While there’s a constitutionally protected right to use such an offensive term, the probability of a speaker using the term is near zero. Yet, she argues, 60 years ago there were places in the US where you likely could not have been elected without using that word. “People’s behavior shifts dramatically in response to community norms,” she suggests, and as many of 80% of people are likely to follow the norms of speech consistent with a space and a situation, even trolls.

One of Benesch’s case studies for counterspeech comes from Kenya, where dangerous speech was a key component to violence in the wake of 2007’s disputed election. With over a thousand killed and hundreds of thousands displaced, the 2007-8 unrest was one of the ugliest chapters in the nation’s history, and as Kenya prepared for elections in 2013, many Kenyans were worried about inflammatory and dangerous speech online.

Benesch worked with Kenya data scientists at the iHub and the team at Ushahidi to build Umati (from the Swahili word for crowd), which collected reports of online hate speech. What they found was a wave of inflammatory speech from Facebook, and astonishingly little dangerous speech on Twitter. This disparity is not well explained by platform usage – Twitter is extremely popular in Kenya. Instead, it’s explained by counterspeech.

When inflammatory speech was posted on Twitter, prominent Kenyan twitter users (often members of the #KOT, Kenyans on Twitter, community) responded by criticizing the poster, often invoking the need to keep discourse in the country civil and productive. This counterspeech was surprisingly successful – Benesch tells the story of a Twitter user who posted that he would be okay with the disappearance of another ethnic group, and was immediately called out by other Twitter users. Within a few minutes, he had tweeted, “Sorry, guys, what I said wasn’t right and I take it back”.

This isn’t the behavior of a troll, Benesch argues. If the user in question were simply looking for attention, he wouldn’t have backed down when his inflammatory tweets met with spontaneous counterspeech. This online counterspeech is especially important when online speech is magnified by broadcast media, as it is in both Kenya and the US – it’s possible for television and newspapers to magnify not just the hateful speech but the attempts to counteract it.

By studying successful examples of counterspeech, Benesch is trying to develop a taxonomy of counterspeech and determine when and where different forms are most useful. She takes inspiration from examples like that of a young man in the US tweeting angrily about Nina Davuluri being named Miss America. The young man inaccurately and disparagingly referred to Davuluri as “an Arab”, and was immediately countered on Twitter by people who called out his racism. Within a few hours, he’d tweeted something resembling an apology to Davuluri herself.

Benesch wonders, “Can we put together the ideas of counterspeech and the idea of influencing 16 year olds?” It’s not realistic to believe we’re going to change the behavior of hardcore haters, she tells us, but we only need to influence a critical mass of people within a community, not the outliers.

Twitter and Facebook aren’t the only environments for inflammatory speech online – anyone who’s participated in online gaming knows that there’s toxic and hostile speech in online environments. Riot Games was concerned about the speech surrounding their popular game League of Legends and cooperated with academic researchers to understand speech in their game universe. The study found that fully half of the inflammatory messages were coming from users we wouldn’t normally consider to be trolls – they came from people who generally behaved like other game players, but were having a bad day and lashed out in ways that were inflammatory. They also discovered that very small changes in the platform – changes in language used to prompt players, apparently minor changes like font and text color – could improve behavior substantially.

Facebook’s “compassion research” project works on similar ideas, trying to get people to use Facebook in more pro-social ways. When you try to flag content on Facebook as offensive, Facebook first prompts you to engage with the person who offended you, suggesting language to communicate to the other user: “Could you take this down? It hurts my feelings.” As with Riot Games, they’ve found that small prompts can lead to dramatic behavior changes.

Benesch has been using these insights to consider problems of inflammatory speech in Myanmar (a topic I learned a little about in my visit to the country earlier this month.) In Myanmar, Facebook is the dominant internet platform, not just the dominant social media platform – if you search for information in Myanmar, you’re probably searching Facebook. In this environment, a rising tide of highly inflammatory speech inciting Buddhists against Muslims, particularly against the Rohingya people, is especially concerning. Not only does Facebook in Myanmar lead to echo chambers where no one may be willing to challenge inflammatory speech with counterspeech, but some of the mechanisms that work elsewhere may not work in Myanmar.

In a country that’s suffered under a military dictatorship for half a century, the idea of “reporting” people for their speech can be very frightening. Similarly, being encouraged to engage with someone who posted something offensive when you have reason to fear this person, or his friends, might threaten your life, isn’t a workable intervention. Any lessons from Facebook’s compassion research needs to be understood in specific human contexts. Benesch asks how you should respond to offensive speech as a Facebook user in Myanmar: you can like the post, but you can’t unlike it. If you respond in the comments thread, you’re participating in a space where the page owner can eliminate or bury your comment. This points to the challenge of using a private space as a quasi-public space.

We need more research on questions like this, Benesch offers. We need to understand different responses to dangerous speech, from “don’t feed the trolls” to counterspeech, to see what’s effective. We need to understand whether counterspeech that seeks to parody or use humor is more effective than direct confrontation. And we need to understand discourse norms in different communities as what works in one place is unlikely to work in another. Louis Brandeis advised that the remedy for bad speech is more speech. As researchers, we can go further and investigate which speech is a helpful counter to bad speech.


I’ll admit that the topic of Benesch’s research made me uneasy when we first met. I’m enough of a first amendment absolutist that I tend to regard talk of “dangerous speech” as an excuse for government control of speech. I had a great meeting with Benesch just before I went to Myanmar, and was much better prepared for the questions I fielded there than if I hadn’t benefitted from her wisdom. She’s done extensive work understanding what sorts of speech seems to drive people to harm one another, and she’s deeply dedicated to the idea that this speech can be countered more effectively than it could be censored or banned.

The conversation after her talk gave me a sense for just how challenging this work is – it’s tricky to define inflammatory speech, dangerous speech, trolling, etc. What might be a reasonable intervention to counter speech designed to incite people to violence might not be the right intervention to make a game community more inviting. On the other hand, counterspeech may be more important in ensuring that online spaces are open and inviting to women and to people of different races and faiths than they are right now, even if inflammatory speech never descends to the level of provoking violence.

For people interested in learning more about this topic, I’d recommend the links on the Berkman talk page as well as this essay from Cherian George, who was at the same meeting I attended in Myanmar and offered his thoughts on how the country might address its inflammatory speech online. I’m looking forward to learning more from Susan’s work and developing a more nuanced understanding of this complicated topic.


« Previous Entries

Powered by WordPress | Designed by Elegant Themes