Are Algorithms Enabling Autonomous Audiences?

Part four in an ongoing series

When considering algorithms as media, or algorithmic media, it is important to factor in how this media connects with its audience. What kind of audience is created, enabled, and served by algorithmic media? What sort of power does algorithmic media have in relation to its audience? How can we study and understand these audiences?

The study of audiences has progressed from regarding them first as passive objects (subjects granted audience to the ruler or their representative), to understanding audiences as active, engaged, and in possession of considerable agency and autonomy (Sullivan, 2012).

In the Internet era audiences are often not even addressed as such, as the phrase still evokes a passivity, instead we have friends, followers, and fans, who we regularly prod for feedback and responses.

Audiences, however, still exist, but with steadily growing agency and autonomy that makes it even harder to study and understand them. One way to do so is to use the theoretical lens of automodernity to examine how audiences are changing.

Do Audiences Have Autonomy?

The automodern audience speaks to the contradictory power that individuals and groups have when it comes to the media they choose to consume or create. We live in an era where seemingly anyone can publish, to a potentially global audience, who have considerable choice, and agency when it comes to what they choose to consume.

This is articulated in the description of automodernity, developed by Robert Samuels (2008), and elaborated in his book “New Media, Cultural Studies, and Critical Theory After Postmodernism” (2010), as being the result of “contemporary individuals turning to automation in order to express their autonomy.” (3)

Adina Baya (2013) was the first audience researcher to link the concept of automodernity with the study of media audiences. She successfully links the evolution of audience research with the evolution of cultural theory from modernity, to post-modernity, to this emerging concept of automodernity.

She argues that the contradictory mixture of automation and autonomy offers a better portrayal of audiences and how to understand their agency and context. Separating the political and aesthetic aspects of new media consumption allows us to look at the range of motivations and effects that new media foster (Baya, 2013).

For example, just because people use Facebook as a primary source of news does not mean that they are aware of how Facebook chooses what is newsworthy. With newspapers and cable television news it was possible to choose a perspective or even ideology based on the institution that produced that news. You could look at a publication and understand what its editorial agenda might be.

However, with Facebook the desire to consume information via the newsfeed is more of a choice about efficiency, what works on your smartphone. It is not a reflection of Facebook’s political agenda, which remains opaque, and a mystery to most users.

Therefore, we shouldn’t make conclusions about an audience’s political positions based on the new media they choose to consume. That audience wants autonomy to consume what they want and decide what they believe regardless of the role Facebook and other algorithmic media play.

There is of course a contradiction to this, as their desire for autonomy results in a giving up of social control via automation. People don’t understand why Facebook chooses what appears in their news feed, however it makes them feel as if they are in control and are able to decide what to believe or not.

Samuels argues that “automodernity reacts to the postmodern stress on social and cultural conflict by celebrating the autonomous individual’s ability to exploit unregulated, automated social systems.” (3) Specifically Samuels describes how “the automation of the public sphere helps to produce a heightened sense of individual control and freedom.” (7)

However, this does not mean that people actually have more individual control and freedom. While we may live in an era where anyone can publish, we also live in an era of near total surveillance, where everything you post online can and will be used against you. This surveillance society also lends itself to all sorts of controls that can and are being used against people (Schneier, 2015). Thus the “heightened sense” is really just that. Only a sense, and not a reality. It’s not as if we can examine or verify the surveillance that is conducted upon us, let alone verify that the automated services we depend upon have “editorial integrity” (Pasquale, 2010).

Algorithms enable anyone to create their own public, to create their own online community, however the contradiction exists around the ownership of said public. While the user may have the perception of control, they legally and technically have no control, as it is the algorithm, and by extension the platform, that actually enable and ultimately control the public that was created.

Orwellian Algorithms

In this context, automodernity could be viewed as the institutionalization of Orwellian doublespeak, the use of language without connection to actual meaning. Taking advantage of the subjectivity of digital media, that we each see our own screen, and applied to meaning, so that we each perceive our own meaning, all with the illusion that we are autonomous thanks to automation.

The phrase automodernity follows modernity, which represents science, democracy, reason, equality, individualism, alienation, colonization, and conformity, and then postmodernity, which represents diverse social movements working to increase social justice and equality (8). Samuels suggests that automodernity is the logical successor as “automodern literacies based on television, advertising, movies, and the Internet do not undermine people’s belief in modern universal reason and unified subjectivity; instead, automodern technologies help to provide a greater sense of technological neutrality, universalized information, and individual power.” (15)

Samuels points to automobiles, personal computers, word processors, mobile devices, blogs, remote controlled TV, and first person shooter video games as examples of automodern media. He argues that “these technological objects share a common emphasis on combining together a high level of mechanical automation with a heightened sense of personal autonomy.” (15)

The automobile is a great example as it empowers the individual on the surface, but actually enslaves them to a broader automotive culture that created suburbs, structured work life, and in many cities now consumes many hours of each day resulting in stress, obesity, and reduced lifespans (McCormack et al, 2014). The illusion of autonomy continues to drive the desire for automation, as now self-driving cars are regarded as the solution to the existing problems of traffic congestion and automotive culture.

In the view of postmodernity, automation represent a loss of personal control, and yet in automodernity automation leads to individual autonomy. Automodernity reflects an evolving popular culture that assumes all technological progress is good, and that rather than be threatened by automation we should embrace it as path to an increased quality of living.

Certainly this view matches evolving understanding of audiences, as consumers of media exercise considerable power with regard to what media they choose to consume or use. Or so they believe.

Without knowledge of how algorithmic media works, how can people actually choose what they consume? Rather the trending topics and viral media that infest their feeds have a power that most people ascribe to magic but is actually embedded into the logic of these platforms as a whole. There is also an insular logic, as the Facebook newsfeed facilitates echo chambers, homogenous perspectives among “friends” that enable the spread of misinformation and conspiracy theories (Vicario et al, 2015; DiResta, 2016).

The aesthetic of scrolling through a feed, or choosing who to follow, or even surfing the web as a whole, gives us the illusion of autonomy and control, yet how much of what we see is actually fed to us by algorithms that sort, rank, and serve up what they deem newsworthy or fit for our consumption?

A primary feature of automodernity is contradiction. Individuals may feel or even exercise autonomy, but at what cost? Our desire to be free and autonomous becomes an opportunity for automation to deliver the illusion without the reality. What if the social controls currently in place are even stronger and more effective than any known previously in history? Yet because they are invisible we do not question them and instead assume that we are free.

This is best embodied by the way algorithms enable search engines to work. They effortlessly make the complicated task of finding something online seemingly easy (Samuels, 2010, 19). Let us not for a moment forget that finding a needle in a haystack is hard, but finding information in a pool of billions of web pages and database entries is of a significantly greater magnitude of mathematical difficulty. And yet we forget this and are conditioned to think of how easy it is to just “Google” something.

Another contradiction within our automodern existence is the construction of self, and how selfie-taking social media users “do not feel that they exist unless someone else hears about their current presence. Here, autonomy is shown to be dependent on the recognition of others.” (21)

Explicitly how are we autonomous if we require the recognition of others to exist? The compulsions of social media complicate how we consume media and provide another context in which we feel autonomous but instead are subject to significant social pressures.

Samuels also talks about the work of Henry Jenkins and the attempt to push back against the idea that audiences are manipulated, but instead recognize that they make meaning in all sorts of ways (27). However, their ability to make meaning does not preclude manipulation.

For example, look at how much YouTube content is a remix, commentary, or copy of pop culture. The algorithms encourage and reward emulation of existing models rather than rewarding original or unique content. While many can play the YouTube game, a tiny few actually succeed in achieving the fame and wealth they desire.

Automodernity suggests that rather than be carved away by digital media, audiences, and by extension publics are thriving and growing as automation makes them easier to create and larger in reach. However, questions of their control, autonomy, and power, especially in relation to algorithmic media still need to be addressed.

Therefore, automodernity also offers us insights as to the contradiction that frames our current era. On the one hand users perceive themselves as having more autonomy, and seem to possess more power to create media than any other period in history. However, on the other hand, they are now dependent upon, and arguably controlled, by the algorithms that enable their perceived autonomy.

Continued in Part 5