Hello,

We noticed you're browsing in private or incognito mode.

To continue reading this article, please exit incognito mode or log in.

Not an Insider? Subscribe now for unlimited access to online articles.

  • Graphika
  • Connectivity

    This is what filter bubbles actually look like

    Maps of Twitter activity show how political polarization manifests online and why divides are so hard to bridge.

    American public life has become increasingly ideologically segregated as newspapers have given way to screens. But societies have experienced extremism and fragmentation without the assistance of Silicon Valley for centuries. And the polarization in the US began long ago, with the rise of 24-hour cable news. So just how responsible is the internet for today’s divisions? And are they really as bad as they seem?

    In this Twitter map (below, and in 3-D form above) of the US political landscape, accounts that follow one another are clustered together, and they are color-coded by the kinds of content they commonly share. At first glance, it might seem reassuring: although there are clear echo chambers, there is also an intertwined network of elected officials, the press, and political and policy professionals. There are extremes, but they are mediated through a robust middle.

    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Each circle is a Twitter account; its size represents the number of followers. Color reflects the kind of content.

    There are multiple left/right groups, with the political parties closer to the center and anti-/pro-Trump groups at the extremes.
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Resist, anti-Trump
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Progressive movement
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Mainstream, left journos
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Democratic Party
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Local news, politics
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Party politics
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Trump support
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Conservative media
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Republican Party
    Illustration of colorful, bean-shaped map made up of small circles
    Slideshow: The US political landscape

    Other
    This story is part of our September/October 2018 Issue
    See the rest of the issue
    Subscribe

    However, as the following diagrams will show, the middle is a lot weaker than it looks, and this makes public discourse vulnerable both to extremists at home and to manipulation by outside actors such as Russia.

    Image of colorful scatterplot
    Slideshow:

    Noisy, partisan bots
    The center of the political universe is far quieter than the polarized wings. This plot of average daily tweets (vertical axis) from the network seen in the charts above shows that the extreme partisans on both sides are screaming while the center whispers. It also shows divisions being amplified by bots on both sides: we see clearly automated activity, with accounts churning out a hundred tweets a day or more on a common schedule. Hundreds of accounts (especially on the left) have identical daily tweet counts, further evidence that they are bots.
    Image of colorful scatterplot
    Slideshow:

    The silence of the center
    The polarization looks even more extreme when the accounts are plotted according to their “valence,” a measure of how politically homogeneous their connections are. A valence of 0 means an account follows or is followed only by progressive accounts, while 1 means it’s connected only to conservative accounts.

    A press divided

    If social media were merely a distraction from an otherwise normal diet of professional news, its impact on democracy would be limited. But this map (directly below) from scholars at Harvard’s Berkman Klein Center and MIT’s Media Lab, based on “co-citations” (i.e., who links to whom), shows that the media world is bifurcated too.

    Plot of data showing names of media outlets, represented, in comparison, by different sizes and color related to political affiliation
    Image of graph with media outlets marked as data points
    Image of graph with red and blue data points
    (left) This plot shows which news sources are cited the most by the Twitter accounts in the US political landscape map shown at top of story. Traditional mainstream journalistic sources are cited primarily by the left, while the right is served by sources such as Fox, Breitbart, and True Pundit.

    (right) If we analyze the same data by looking at citations of individual articles rather than news sources, the divide is even more stark. The articles that get the most tweets represent the most partisan views on both left and right.

    How Russian trolls exploit division

    The polarization seen in the diagrams above is fertile ground for misinformation operations such as the one Russia conducted to influence the 2016 US election.

    Instead of trying to force their messages into the mainstream, these adversaries target polarized communities and “embed” fake accounts within them. The false personas engage with real people in those communities to build credibility. Once their influence has been established, they can introduce new viewpoints and amplify divisive and inflammatory narratives that are already circulating. It’s the digital equivalent of moving to an isolated and tight-knit community, using its own language quirks and catering to its obsessions, running for mayor, and then using that position to influence national politics.

    The first of the two maps in the GIF image below shows the US political spectrum on the eve of the 2016 election. The second map highlights the followers of a 30-something American woman called Jenna Abrams, a following gained with her viral tweets about slavery, segregation, Donald Trump, and Kim Kardashian. Her far-right views endeared her to conservatives, and her entertaining shock tactics won her attention from several mainstream media outlets and got her into public spats with prominent people on Twitter, including a former US ambassador to Russia. Her following in the right-wing Twittersphere enabled her to influence the broader political conversation. In reality, she was one of many fake personas created by the infamous St. Petersburg troll farm known as the Internet Research Agency.

    The Iranian blogosphere

    The echo-chamber effect on the internet is nothing new. This 2008 map depicts the blogosphere in Iran, clustering together blogs that link to each other and coloring them by their content. Before a sustained government crackdown on online speech, supporters (lower right) and detractors (lower left) of the clerical regime each enjoyed substantial followings.

    Illustration of colorful, bean-shaped map made up of small circles

    Turkish Twitter

    This Twitter map of the political landscape in Turkey, analogous to the US map at the beginning of this story, shows multidimensional polarization, with a dense sphere of influence around Erdogan supporters, on the far right side of the map, and two different poles of opposition on the other. These “amplification cores” of highly connected accounts have a disproportionate influence on the conversation and can rapidly boost polarizing messages.

    Illustration of colorful, bean-shaped map made up of small circles

    Russia: The same but different

    This Twitter map of the Russian political landscape shows polarization in another context. There are clear pro- and anti-Putin clusters, but they’re knit together by a broad set of mainly pro-government news and discussion-oriented accounts. A halo of apparently automated “personal” and marketing accounts surround the Putin fans.

    Illustration of colorful, bean-shaped map made up of small circles

    John Kelly is CEO and Camille François is research director of Graphika, a social network analytics company. 

    AI is here. Will you lead or follow?
    Join us at EmTech Digital 2019.

    Register now
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Each circle is a Twitter account; its size represents the number of followers. Color reflects the kind of content.

    There are multiple left/right groups, with the political parties closer to the center and anti-/pro-Trump groups at the extremes.
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Resist, anti-Trump
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Progressive movement
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Mainstream, left journos
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Democratic Party
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Local news, politics
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Party politics
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Trump support
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Conservative media
    Illustration of colorful, bean-shaped map made up of small circles
    The US political landscape

    Republican Party
    Image of colorful scatterplot


    Noisy, partisan bots
    The center of the political universe is far quieter than the polarized wings. This plot of average daily tweets (vertical axis) from the network seen in the charts above shows that the extreme partisans on both sides are screaming while the center whispers. It also shows divisions being amplified by bots on both sides: we see clearly automated activity, with accounts churning out a hundred tweets a day or more on a common schedule. Hundreds of accounts (especially on the left) have identical daily tweet counts, further evidence that they are bots.
    Image of colorful scatterplot


    The silence of the center
    The polarization looks even more extreme when the accounts are plotted according to their “valence,” a measure of how politically homogeneous their connections are. A valence of 0 means an account follows or is followed only by progressive accounts, while 1 means it’s connected only to conservative accounts.
    Plot of data showing names of media outlets, represented, in comparison, by different sizes and color related to political affiliation
    Image of graph with media outlets marked as data points
    (left) This plot shows which news sources are cited the most by the Twitter accounts in the US political landscape map shown at top of story. Traditional mainstream journalistic sources are cited primarily by the left, while the right is served by sources such as Fox, Breitbart, and True Pundit.

    (right) If we analyze the same data by looking at citations of individual articles rather than news sources, the divide is even more stark. The articles that get the most tweets represent the most partisan views on both left and right.
    Image of graph with red and blue data points
    (left) This plot shows which news sources are cited the most by the Twitter accounts in the US political landscape map shown at top of story. Traditional mainstream journalistic sources are cited primarily by the left, while the right is served by sources such as Fox, Breitbart, and True Pundit.

    (right) If we analyze the same data by looking at citations of individual articles rather than news sources, the divide is even more stark. The articles that get the most tweets represent the most partisan views on both left and right.
    Illustration of colorful, bean-shaped map made up of small circles
    Illustration of colorful, bean-shaped map made up of small circles
    Illustration of colorful, bean-shaped map made up of small circles
    More from Connectivity

    What it means to be constantly connected with each other and vast sources of information.

    Want more award-winning journalism? Subscribe and become an Insider.
    • Insider Plus {! insider.prices.plus !}* Best Value

      {! insider.display.menuOptionsLabel !}

      Everything included in Insider Basic, plus the digital magazine, extensive archive, ad-free web experience, and discounts to partner offerings and MIT Technology Review events.

      See details+

      Print + Digital Magazine (6 bi-monthly issues)

      Unlimited online access including all articles, multimedia, and more

      The Download newsletter with top tech stories delivered daily to your inbox

      Technology Review PDF magazine archive, including articles, images, and covers dating back to 1899

      10% Discount to MIT Technology Review events and MIT Press

      Ad-free website experience

    • Insider Basic {! insider.prices.basic !}*

      {! insider.display.menuOptionsLabel !}

      Six issues of our award winning print magazine, unlimited online access plus The Download with the top tech stories delivered daily to your inbox.

      See details+

      Print Magazine (6 bi-monthly issues)

      Unlimited online access including all articles, multimedia, and more

      The Download newsletter with top tech stories delivered daily to your inbox

    • Insider Online Only {! insider.prices.online !}*

      {! insider.display.menuOptionsLabel !}

      Unlimited online access including articles and video, plus The Download with the top tech stories delivered daily to your inbox.

      See details+

      Unlimited online access including all articles, multimedia, and more

      The Download newsletter with top tech stories delivered daily to your inbox

    /3
    You've read of three free articles this month. for unlimited online access. You've read of three free articles this month. for unlimited online access. This is your last free article this month. for unlimited online access. You've read all your free articles this month. for unlimited online access. You've read of three free articles this month. for more, or for unlimited online access. for two more free articles, or for unlimited online access.