A colleague of mine usually tells me she “lives under a rock” when I’m telling her about whatever it is that has set me off in this news this week. There’s times when I speak with her or other people in my life that I feel like my views are out of synch with theirs. Somewhere, some machine has slipped a gear. There are many possible explanations for why each of us sees the world a little differently, and the nature of those misperceptions are part of what Samuel Woolley explores in Manufacturing Consensus: Understanding Propaganda in the Era of Automation and Anonymity. You may have heard the phrase manufactured consent and so you’re thinking this book is a new update on that idea, and you’d be correct. Woolley argues that the nature of what used to be known as mass media has changed significantly in the days since Herman and Chomsky gave a name to the idea of a small number of broadcasters being able to create wide support for certain ideas through one-way messaging. What’s different now is that we have devices with worldwide reach in our pockets and we can use them to remain anonymous while sharing our perspectives with each other directly. More than that, we can hide our perspective behind that of a bot that interacts with other users according to our programming instructions. As a result, Woolley contends that instead of manufacturing consent, we are instead involved in a process of manufacturing consensus; it may seem that the majority of people share the same viewpoints, even when that is not the case.
The book’s cover communicates this idea beautifully. There are 28 dots on the cover that comprise shades from blue to red. Only the bottom three dots are completely red, and they also have shoulders attached to them that make them appear as generic social media avatars. What it seems to be saying is there is a diversity of voices available, but only the three voices that are fully red are taken as representing the whole group. It’s a biased sample. Our perception may be that three out of every three people agree with a certain perspective, but those three are from a population of 28 and they are in the minority. It’s easy to be confused in a context where those three voices dominate, even if you occasionally have evidence that the other 25 voices exist. You might feel that you live under a rock if one of those 25 voices speak to you, regardless of whether you agree with the perspective espoused by the three red voices.
Woolley provides an operational definition for propaganda early in the text, writing that it describes “the use of politically biased information in considered attempts to manipulate or influence the opinions and actions of individuals and, more broadly, society” (p. 4). He goes on to state that the purpose of propaganda may not be to effect concrete changes in its targets’ behavior. It can be enough to seed certain emotions (anger, apathy) that result in a target taking no concrete actions or behavioral changes. Think about people who have convinced themselves that not voting is some kind of message. It’s not. Non-voters are telling candidates that their opinions can be safely ignored. Getting enough people to feel cynical about the process of elections or other social participation is the goal of some propaganda. It seems easier than ever to get people to sit home and rot these days, which is just what the propagandists want.
Later in the first chapter, Woolley is explaining how powerful propaganda can be in the world of social media. He uses the case of Martha Coakley’s failed 2010 run for U.S. Senate in Massachusetts as an example of how social media users influenced the result of a race. Widely seen as an easy win for Coakley, the race went to her Republican opponent. Woolley explains how a small number of social media accounts were able to successfully pounce on Coakley’s gaffes and missteps, which helped to blunt her momentum and cost her the seat. More generally, Woolley uses a section of this chapter to explore the truism that “if you don’t know what the product is, the product is you” with regard to social media (p. 15). I was a little surprised to not see Shoshana Zuboff’s revision of this statement cited instead. She claims that its not “you,” but “your behavior” that is the product on social media. Her analysis dovetails nicely with Woolley’s because what is important is not just the social media user, but how propagandists can manipulate their behaviors through anonymity and automation that is important. So, what is key is not just the users on social media, but the behavior of the people that a small number of social media users manipulated to prevent Coakley from coasting to victory in her 2010 race.
Further into the text, Woolley defines more clearly three levels of manufacturing consensus and how they interact. He describes them as “a kind of ouroboros of manipulative information” because of how they rely on each other (p. 55). Each level is based on different types of users. There are political-bot-, sockpuppet-, and partisan nanoinfluencer-based propagandists. Then there are social-media-algorithm-, recommendation-, and trend-based propagandists. Finally, there are news-media-based propagandists. Each level relies on and interacts with the others. The first type (bots and sockpuppets) give the illusion of wide support for marginal ideas. The second type (algorithms and trends) are taken as genuine public opinion polling, even though they can be manipulated or gamed. The third type (news media) gives an institutional sheen to the propaganda created by the other two and broadcasts it to wider audiences as they “reproduce, recreate, and further launder content” (p. 55). One of Woolley’s interview participants grasps these levels of manufacturing consensus with his rhetorical question, “Why would I focus on trying to change someone's mind with a bot barely capable of communication… when I could get the trending algorithm on a site to reprioritize and reshare the content I'm pushing with five thousand bots?” (p. 121). It’s clear that Woolley’s levels are grounded in the reality of social media users. This particular user (among his other interview participants) is able to articulate exactly how to move from the first to the second level of manufacturing consensus, and there is reason to believe that he could easily understand why getting the news media to cover his propaganda campaign would lend it further legitimacy.
At this point, you yourself may be feeling there is no sense is fighting back against these interlocking systems because there is no way to undo all the damage propagandists have wrought. Woolley tells us later that “there are, sadly, no easy fixes” (p. 84). One that he proposes is based on the work of Joan Donovan and danah boyd, who argue for “strategic silence” from news media (p. 138). The attention the news media give to trends on social media amplifies, launders, and legitimizes propagandists’ campaigns, as Woolley has explained. (He uses 2016’s Pizzagate incident as an illustrative example of how this process unfolds.) So, part of the solution does rest on the shoulders of mass media. How and whether the local or national news covers certain trends has a large influence on whether it is seen as an idea with consensus.
Woolley doesn’t let us down in the end with a despairing conclusion. One immediate change that would be helpful would be to replace the U.S.’s Section 230 with “legislation that takes account of the massive rise of social media” (p. 176). He also calls for interaction between the Federal Elections Commission, Federal Trade Commission, and Federal Communications Commission to “prevent manipulation via digital tools” (p. 176). That he cites a thirty-year-old law and three large governmental bodies that need to work together to solve this problem gives a sense of how immense the problem is. Both of these changes in regulation sound like excellent ideas.
Normally, about here is where I’d sign off with a comment like “well Trump won again after this book was published, so there’s no hope.” However, this past weekend’s comically mismanaged Great American State Fair shows just how little actual power these folks have. The consensus they appeared to wield is has been manufactured and their low level of support is undeniable. They’re outnumbered and we will win.
YOU MAY ALSO ENJOY THESE REVIEWS:
More Than Words: How to Think about Writing in the Age of AI by John Warner
Waiting for Robots: The Hired Hands of Automation by Antonio A. Casilli
Paved Paradise: How Parking Explains the World by Henry Grabar
No comments:
Post a Comment