• Devin Nunes and the Power of Keyword Signaling | WIRED
    https://www.wired.com/story/devin-nunes-and-the-dark-power-of-keyword-signaling

    We increasingly turn to search engines to seek out information. Since Google’s earliest days, marketers have relied on “search engine optimization” to try to maximize the likelihood that Google returns content that highlights their cause or company. In today’s media landscape, organizations and individuals also use these tactics to manipulate the algorithms behind Facebook/Instagram and Twitter feeds. The problem is, whether or not we’re aware, the key words we search are coded with political biases. My research demonstrates that it’s possible to position ideological searches to maximize the exposure of their content.

    When there is limited or no content available on a topic, it’s possible to game search engines to guarantee that certain keywords will be directed to content that includes these terms or is tagged accordingly. This is why conspiracy theorists were able to capitalize on the concept of a “crisis actor.” By producing a plethora of insidious content rife with the term and maximizing SEO, conspiracy theorists filled what Microsoft’s Michael Golebiewski and danah boyd referred to as a “data void.” Searches for “crisis actor” got conspiratorial results until other sources filled the void with more legitimate content debunking the theory.
    screenshot of a google search
    Courtesy of Francesca Tripodi

    To demonstrate how this works in politics, I Googled a few key phrases used in both of Nunes’ speeches. The results demonstrate how politicians and pundits can exploit data voids to create ideological information silos. During each hearing, Nunes describes “the Russia collusion hoax.” When you search for “collusion hoax,” the links returned support the position that investigations into the president are bogus. The top links are from a story in The New York Post published just last week that Dems are trying to block Barr’s probe into the “Russian collusion hoax” and a link to Amazon to purchase a book titled The Russia Hoax: The Illicit Scheme to Clear Hillary Clinton and Frame Donald Trump, by Fox News legal analyst Gregg Jarrett.

    Strategic signaling also drew attention to what the Mueller report did not focus on. On June 12, Nunes noted that the report had not procured any “useful information on figures who played key roles in the investigation such as Joseph Mifsud,” a Maltese academic and figure in the George Papadopoulos case, “or the Democrat paid operative, former spy Christopher Steele,” the British intelligence officer behind the now notorious pee tape allegations. In the days following Nunes’ remarks, the search returns were primarily conservative content published anywhere between two weeks to 12 minutes before Nunes’ speech. In addition to traditional conservative sources like Fox News, Washington Examiner, and National Review, there are also digital-first sources like the Daily Caller and the Daily Wire, as well as stories posted from more dubious publications like the Epoch Times.

    These findings reveal that existing studies on algorithms, filter bubbles, and misinformation online are missing a crucial component regarding the problem of political polarization, specifically data focused on how we access news and information. Epistemological frameworks can lead us into algorithmic rabbit holes. Understanding keyword signaling is an essential part of studying political polarization. While most focus on how output (e.g., search results or social media newsfeeds) keeps us in filter bubbles, more research is needed on how inputs are manipulated for political gain. This level of sophistication highlights how conservative groups systematically work to optimize their content for search and social media. Not unlike the tactics of Republican strategist Frank Luntz, political players and members of the right-wing media ecosystem are able to fill data voids with their own ideas and stories.

    #Politique_numérique #Moteurs_recherche #Faschosphère #Mots_clés #Fake_news