Listen, my children, and you shall hear of the midnight ride of Paul Revere. On the eighteenth of April, in [Seventeen] Seventy-Five, Paul Revere, a silversmith, took off on his horse heading northward to warn American patriots of the impending British advance which officially kicked off the American Revolution. Revere tirelessly spread the word from town to town and convinced others to mount their horses and do the same. The message spread like the flu, and soon all of New England was warned and prepared for the approaching British soldiers. Much less known in American folklore is that William Dawes also hopped on his horse to begin the same midnight ride to spread the word of the coming British soldiers. While Revere headed north of Boston, Dawes headed South. Unlike Revere, however, Dawes was not able to successfully spread a word of mouth epidemic. He lacked either the charm or conviction – or both – to convince fellow patriots to heed his warning and exponentially spread the message by sending out more messengers.

Roughly two hundred years later and Gaetan Dugas, a handsome blond steward for Air Canada, slept with an average of 250 men a year based all over the world. He was later dubbed “Patient Zero” of the AIDS pandemic. According to William Henry, writing in Time, Dugas “could be linked to nine of the first 19 cases in Los Angeles, 22 cases in New York City and nine more in eight other cities — in all, some 40 of the first 248 cases in the U.S.” It’s not that Dugas was the first person to become infected with AIDS. In fact, new research reveals that AIDS has likely been around since the 19th century. But without a connector like Dugas the disease would have never become a global pandemic.

I was reminded of both of these anecdotes while re-listening Malcolm Gladwell’s flat, monotonous reading of his 2000 book The Tipping Point with the hope that it would finally put me to sleep on my insomniac transatlantic flight from New York to Brussels. The Tipping Point is a beautiful analysis of how information spreads and how small “memes” – whether a fashion trend or a news headline – become national or even international sensations.

Information, he argues, flows much like disease and is largely dependent on people Gladwell calls “social connectors.” It took a social connector like Patient Zero to transform HIV from one of millions of memes on the margin to a global pandemic. Social connectors – people with vast networks of friends and acquaintances that span across several distinct geographic and social communities – are incredibly important to the flow of disease … and information. While patient zero – with his thousands of sexual partners all over the globe – was extremely successful in spreading AIDS internationally, Boing Boing is another example of a social connector (actually, a small group of social connectors) who are able to spread bits of information from a single obscure community to a mainstream readership. Without these important connectors information would stay bound in smaller networks without ever reaching the status of mainstream awareness. (Without social connectors hip-hop would have stayed in New York and Converses would have stayed fashionable among a small social circle rather than becoming the uniform footgear of Hipster Nation.)

Social connectors tend to be public in the way they deal with communication and information. While most people pass on information via word of mouth (with single emails to single friends) social connectors tend to send that same information out to their entire networks or on their blogs and social networks.


Malcolm Gladwell isn’t the only one interested in how information flows across networks and how small stories become large stories. Last year I pointed to Evgeny Morozov‘s project Polymeme, which tracks some 25,000 sources of information (news sites, blogs, and information portals), analyzes the content of each article, clumps similar articles together based on keywords, and finds out who publishes a news story first and who links to it after. Something that Evgeny noticed early on is that the majority of stories, or “memes”, originated in the so-called mainstream media and were then picked up and re-hashed by bloggers. There are far fewer examples of stories originating on blogs and then getting picked up by mainstream journalists who link back to the original post.

However, it is important to note that, while every journalist I know researches his or her stories in the blogosphere, few credit bloggers as sources and even fewer link to the original posts. Bloggers are also much more likely to quote chunks of text from source articles while journalists re-word what they see to match their own style and the tone of their publication. In fact, this evening I was having beers with Glenna Gordon, a freelance foreign correspondent here in Monrovia, who said that she often picks up story ideas from the blogosphere, but does not mention those blog posts in her resulting articles. It is much easier for an automated service like Polymeme to discover when bloggers link to mainstream media articles than it is to discover when a journalists discovers a story thanks to a blog post or is influenced by what a blogger has to say.

Yesterday the Berkman Center announced the public launch of Media Cloud, which like Polymeme, uses Open Calais. The project began as a proposed solution to a debate between Ethan Zuckerman and Yochai Benkler over whether or not bloggers and citizen journalists are breaking news stories which then influence mainstream media reporting or if they just echo and opine on what has already been published by mainstream journalists. (Yochai argues the former, Ethan argues the latter.) So far, Media Cloud is pretty bare bones and doesn’t provide you with much information that you couldn’t get from Calais’ other applications, Global Attention Profiles, or Google Trends, which also offers information (collectively and per website) about related keywords, regions, cities, and languages (and, handily for statisticians, is available for download as a cvs file).

(Tangent: Anyone surprised that Miami is the number one ranked city for “anal bleaching” according to Google Trends?)

But what excites me the most about Media Cloud is how it might potentially be used in the future. Ethan and others at Berkman want to use their database of information to track how a story flows and transforms as it spreads across the internet. To speak in biological terms, they are trying to find the origin and the mutations of the DNA of a major news item. This is something that has fascinated me for a long time. How was it, for example, that Octomom became an international sensation and Madagascar’s near civil war is hardly a blip on the radar? How do myths and rumors (for example, that the eskimo have twenty words for snow, that Barack Obama is Muslim … or, for that matter, that bacon-flavored chocolate tastes good) gain the status of accepted fact by large segments of society? Part of the answer, clearly, is that these bits of information reach social connectors where they become contagious, or in Gladwell’s terms, where they reach their tipping point. But I think that there is an important element in how information flows on the internet which can’t be taken into consideration by tools like Media Cloud, Polymeme, or any other.


Above Water Communication

The first type of information flow I’ll call “above water communication”. This is when information flows clearly from one source to another. For example, Osamu Higuchi writes an open letter to Google, which was then picked up by Chris Salzberg, and then by Slashdot, one of the internet’s major social connectors (at least among geeks). This type of flow of information – when articles link to citations or when they quote a chunk of text – is easy to detect.

Below Water Communication

But I believe that the majority of online information flow happens below water where we can’t see it. Any given day one of the top referrers to this blog looks like this:

Each one of those referrals means that someone has sent a link to a blog post I wrote. That is how information really flows – by email, by word of mouth (“try googling ‘so and so’ and you’ll find it”, we tell our friends when we want them to see a video or post that we’ve seen), and by private pages and messages on Facebook that can’t be detected using Media Cloud or any other implementation of Open Calais.

Ethan’s report on Zimbabwe’s internet outage reveals how below-water communication can become above-water communication. An analysis of the Zimbabwe internet story – using either Media Cloud or Polymeme – would likely cite Ethan’s blog post as the source of the story. But, in fact, the original source of information came from a private email sent from Denford Magora to Ndesanjo Macha, which was then forwarded to Ethan and a few others. It was below-water information which would never have been tracked by either Media Cloud or Polymeme because Ethan doesn’t mention either Denford or Ndesanjo as a source. I have a feeling that this is frequently the case in mainstream reporting. Many if not most articles probably begin with a tip from some email, twitter message, or blog post, but there is no way for us to track such influences other than surveying (and trusting) what reporters tell us.


Gladwell’s monotonous voice never did put me to sleep. Instead I arrived at Brussels exhausted, but needing to stay awake so as not to miss my next flight. Finally I arrived to Monrovia feeling and looking like the devil incarnate. I hitched a ride into town with Lawrence Randall of the Liberian Media Center and Philip A. Sandi of the Press Union of Liberia. During the long journey we argued about the ICC’s indictment of Bashir. For the most part Philip and Lawrence included me in their conversation, but occasionally they’d go off into a dialect of which I could only capture certain words and phrases, as a way of having their own separate “below water” conversation.

At one point Philip said he heard people tell him that there might be a civil war soon in the United States. Philip is a social connector, a bridge between Liberian and American societies. I can imagine him spreading this opinion that the US is headed toward civil war and it expanding exponentially like Paul Revere’s midnight ride. Rumor and suspicion have always been central aspects of Liberian society and communication. The majority of Liberians still believe that former president Samuel Doe was drawn into his gruesome death by a CIA conspiracy despite little evidence. But where did that explanation originate and how did it spread so quickly and so effectively? What are the catalysts that transform minor headlines into national and international stories? What are the obstacles that kill a story? And how does a story change as it is passed from hand to hand? What is the epidemiology of news? As Ethan points out, we’ve got a long way to go until we really understand the process, but tools like Open Calais and Google Trends’ API – and their various implementations – are opening up possibilities that didn’t exist just a couple years ago.