Partitioning the Public Sphere: search engines, echo chambers and technological/linguistic balkanization

Since the election of Donald Trump1 to the White House at the beginning of November, there has been a renewed attention to the mechanisms of information dissemination in the 21st century and how this might affect electoral outcomes.

Manipulation of information for political ends is not new. The use of propaganda and the capture of key broadcasting and print media outlets has been a strategy for the power hungry since such technologies existed. Journalists have long been targeted by despotic regimes of all flavours, especially those who refuse to tow particular lines and/or self-censor.

We are reminded in South Africa by the ongoing debacle regarding the SABC 8 and the problematic appointment of Hlaudi Motsoeneng as COO of the public broadcaster that rather crude manipulations of state media outlets is not confined to historical accounts of Goebbels, the Kremlin or the Khmer Rouge. This latest video of local musicians singing the praise of Motsoeneng for placing quotas on local content for TV and radio is a good example of how old-fashioned techniques are still in use.

However, there is a far greater threat to what Charles Taylor refers to as the Public Sphere: online media dissemination.

Taylor, in his essay “Modern Social Imaginaries”, outlines the way that printing technology enabled the creation of the Public Sphere which in turn had a profound effect on the modern social imaginary. The Public Sphere was that open space in which dialogues, discourses, information and debate occurred between people who were not co-located and were not in direct discussion with each other. It describes the social space popularly referred to as ‘current affairs’, ‘popular opinion’ and ‘news’: those stories, issues, arguments that bounce back and forth between strangers across TV and radio waves, through newspapers and magazines. Critically, the Public Sphere enabled a means of information sharing and contestation that has become one of the bed rocks of modern democracies.

Prior to the Gutenberg moment, such a common social information space didn’t exist: current information was limited to the gossip of your local town, from people you knew. Now, with the arrival of the printing press, information could be sent out far and wide, with communities who were unaware of each others’ issues suddenly engaging in ways never seen before. Mimeographs, newspapers and columns prompted dinner table discussions previously unimaginable. Disconnected groups of people could still envision themselves as a piece of a greater whole, beyond the limitations of direct sensory perception regarding who was directly seen and heard.

Then came the wireless, and then television, and information moved even faster. The boundaries of the social imagination–about who we shared our world with and how we fitted in–grew even further as information about distant happenings was teleported into living rooms.

While broadcasts and printed materials could be tailored and tweaked to specific audiences, the control of dissemination at the individual level was relatively weak. Even if radio and TV were localized and ‘personalized’ in their content, anyone with the correct receiving apparatus living in the same area would receive the same content. As for printed media: once in hard copy and in circulation, there was no controlling who sent copies of what to whom. It was out there, and, like feathers in the wind, almost impossible to retrieve. Anyone with the correct production means (a radio transmitter, a printer) could inject information into the Public Sphere. One man and a mimeograph could disrupt the whole news cycle.

The strength of the Public Sphere is not to be under-estimated. It is akin to the centre 8 blocks of a chess board: the player who controls those blocks holds the dominant position. That media outlets are often the first to be targeted and/or captured by those suffering a crisis of legitimacy is testimony to the importance of the Public Sphere to influence opinions and political moods2.

But now there is a new technology to disseminate information, one that is not easily controlled by governments (although China is doing its best): social media and online news. Never before have those with internet access had so much information of such a broad variety at their fingertips. In countries with deep connectivity saturation, such as the United States and Western European countries, manipulating TV shows and newspapers is liking farting in the face of the internet wind.

However, while the World Wide Web promised a robust Public Sphere with access to information previously never seen, there was a down side. The consumption of information via the internet is almost always an individual activity. Unlike watching the TV or listening to the radio, people do not read news at their laptops together: they each have their own screen and keyboard, with external sensory occlusion often completed by the use of headphones. This has opened up the opportunity for individualization of information consumption like never before.

Couple this individualization of information consumption with advanced research in psychology and big online companies now have a terrifyingly powerful apparatus. This apparatus works using two key mechanisms: the ability to change already published content, and the ability to individually funnel readers to specific pages covertly without them being aware of it.

Changing already published content

Unlike hard copy or broadcast media, consumers of online information never really have their ‘own copy’: the information resides on a remote server under someone else’s control, who can update, redact or tweak the information at will3. Only the most observant will notice the removal of a sentence in an article from one day to the next, or the subtle change in wording. The days of declaring errata are over: publishing corporations need never admit their errors or changes unless called to account by the most anal retentive of readers who keep multiple local copies of information and actually bother to compare versions of the same material to look for changes.

This also means that the Public Sphere is now vulnerable to pirating, the  (il)legitimate manipulation by technologically advanced ‘third parties’ of hackers, whose intentions may range from the noble to the anarchic (e.g. Anonymous).

Funneling readers into partitions

But by far the more insidious mechanism has to be the learning algorithm of search engines: in the West this takes the form of Google and Facebook. By tracking people’s preferences and choices over time and using advanced models of psychological trends, readers now have the information on their screens ‘customised’ to show them what Google and Facebook think they want to see. Combined with ever more cunning strategies to entice clicking (known as click-baiting), the very lucrative click-advertising revenue stream is optimized by presenting users of online content with information that affirms their own world views.

The result is a partitioning of the Public Sphere into echo chambers: people are now balkanized into online spaces of shared opinion, rather than into the physical spaces of shared neighbourhoods that might’ve previously demarcated information distribution. It is the technological equivalent of apartheid era spatial geography in the Public Sphere, and it is extremely dangerous. Some are even hypothesizing that partitioning has the capacity to influence critical elections like that between Clinton and Trump by spreading ‘fake news‘.

This has become such a phenomenon that the entire era has been referred to as ‘post-truth’. It is a phenomenon that Donald Trump is keenly aware of, hence his courting of big tech business CEOs, as he attempts to align the interests of tech giants who now control the American Public Sphere with the interests of his administration.

Not only technology can partition

Coming back to the South African case, and the relevance of the threat to the Public Sphere to schooling and education, the current focus in the United States on technological partitioning should ring some bells closer to home. For not only did the apartheid government physically balkanize South Africans, but it linguistically balkanized them too.

While South Africa is not yet as vulnerable to online manipulation due to the high levels of variation in connectivity (as evidenced by the efforts still exerted to capture the public broadcaster), the funnelling in SA is done by language. This is easily seen when comparing the SABC news in English broadcast at 6:30pm to the news in isiZulu broadcast at 7pm. Radio shows are mostly monolingual, or at least conform to language clustering that mirrors that of the population (English and Afrikaans together; indigenous languages together). For the most part, only the official state broadcaster is distributing information in indigenous languages, bar the occasional independent outlet such as Isolezwe, the isiZulu newspaper from Independent Media. The ‘liberal’ independent media and its champions would do well to remember that, by broadcasting and printing in English mainly (and occasionally Afrikaans), their reach is fundamentally limited to the echo chamber of middle class South Africans.

The dismantling of the linguistic partitions in the South African Public Sphere requires the teaching, learning and use of multiple languages not only in the name of social cohesion, but as a fundamental prerequisite for genuine democratic participation: nothing less than the reconstitution of a genuine Public Sphere through which an engaged citizenry can hold the executive–and each other–to account. This, along with critical consumption of online information, must be a key priority in our schools for any teacher who believes in emancipatory education.

Lack of attention to the information consumed by the Other, along with the mechanisms by which our public information is ‘personalized’ and hence used to partition us, will only yield more Trump moments.

  1. Quaintly referred to by some in the US as He-Who-Must-Not-Be-Named.
  2. The Arab Spring of 2011, popularly attributed to Twitter, and the subsequent crack down on social media in Egypt, is a good example.
  3. I have made several minor edits and corrections of typos in this article since publishing it. Have you noticed?