In regard to a debate on Twitter
“What I was really trying to teach them is that we have very few laws that do this type of protection—and we might not even want them—but that there have to be some type of protections for privacy,” said Kate Klonick in an interview with NPR on March 9.[i] Klonick is an assistant professor of law at St. John’s University, where she teaches about the intersection between law and privacy. The interview concerned a “creepy” extra-credit assignment she’d assigned to her class over spring break.
The homework was simple: observe and listen to people in a public space and try to de-anonymize them by using Google’s search engine. The students seemed surprised by how, with just a few details about a someone, they could watch the person become transparent before their very eyes.
In one case, a student was able to de-anonymize someone using the person’s first name paired with the college insignia printed on their sweatshirt. Klonick claims the lesson was about privacy—that is, about how little of it we have left. In fact, she implies public obscurity might be a thing of the past.
NPR’s coverage was the third promotion of the experiment. Klonick published an op-ed in the New York Times on March 8,[ii] which was presumably in response to the interest her experiment garnered initially by route of Twitter on March 5. Over the course of the ensuing days, according to the laws of social media, a small skirmish took place: a number of privacy advocates questioned whether the experiment was ethical.
Some wondered aloud if the experiment would meet Institutional Review Board subject consent standards.[iii] Others noted it put vulnerable populations at risk,[iv] emphasizing this was no way to teach privileged second-year law students empathy.[v] Klonick’s response seemed to imply that the lesson was more important than the severity of the violations involved.[vi]
In other words, the room was clearly divided. Some critics denounced Klonick as a bit detached[vii] herself. But many socially aloof professors were excited to get busy trying it in their own classrooms.[viii] Needless to say, it all made for a few days of priceless back and forth on the Twitter scene.
This might seem like a “big to do” about nothing for the casual observer. But a vast portion of Klonick’s work concerns looking for meaningful solutions to Internet shaming, harassment, cyberbullying, and hate speech. To those whose lives have been devastated by the Internet’s many forms toxic behavior, the experiment could have looked like an object lesson in surveillance,[ix] rather than an empirical field study on privacy.
But more worrisome is the fact that the experiment and its promotion were an exercise in elevating toxic social norms. And perhaps more troubling still, the experiment and its consequent promotion can be viewed cynically as the voice of a more sinister party by way of proxy.
Garnering takeaways from Klonick’s research
To better understand what’s at stake, it would be instructive to summarize some of the key takeaways from Klonick’s two major articles on the subject: “Re-Shaming the Debate” and “The New Governors.”
Re-shaming the Debate
In “Re-Shaming the Debate,” Klonick attempts to show that normative solutions, legal remedies, and private actions are to some extent viable answers to online shaming. Whereas, because of the potential for state restrictions to conflict with other values, like free speech, right-to-be-forgotten laws and similar ideas are not tenable solutions.
Klonick sets the stage with an exploration of social norms, telling us “the internet is a normative place” and that “norms are the primary social control mechanism of the Internet,” (“Re-Shaming the Debate,” 1044). Through an analysis of online social norm enforcement, Klonick reveals the meanings, effects, and the precision of normative solutions have changed in the wake of the internet. Social norm enforcement is the process by which communities instill group norms in their members through legal and extrajudicial means.
Often, in cases of extrajudicial social norm enforcement, shaming is the vehicle that drives sanctions against violators. Klonick conveys to her readers the taxonomic difference between cyberbullying or harassment and online shaming is that online shaming involves an attempt to punish a violation of group norms, (ibid, 1034). For instances of cyberbullying, harassment, or hate speech, that needn’t be the case. Each of these phenomena may not necessarily be initiated by a violation of social norms.
Regardless, the online environment distorts the effects of social norm enforcement, making it unlike its offline kindred. Klonick shows the social meaning behind shaming a violator may be indeterminate, the degree of punishment uncalibrated, and the accuracy of who or what punishes suspect, (ibid,1052-1059).
As a final point against its lone viability as a solution, Klonick shows online social norm enforcement can create feedback loops where those pointing fingers at violators can be counter shamed as violators for breaking other norms in the process, (ibid, 1057-1059). Klonick concludes normative solutions can still be viable though, if self-aware users make it a point to shame the violation instead of the violator, as exemplified by the campaign against manspreading, (ibid,1055-1056).
Having shown the true degree to which normative solutions are viable, Klonick moves on to explore the viability of legal remedies for providing relief to shaming victims. It turns out, unfortunately, they’re also of limited worth.
As defamation torts are contingent upon statements about plaintiffs being false, their span is only applicable to the narrow range of cases where statements about the accused are indeed inaccurate, (ibid, 1060). Klonick identifies a second possible tort for “intentional or reckless infliction of emotional distress with ‘extreme and outrageous conduct,’” (ibid, 1060). But she quickly points out that legal remedies are costly, time consuming, and may in the meantime exacerbate the problem while the claimant awaits resolution.
A small analysis of Eugene Volokh’s scholarship gives Klonick all the reason she needs to dismiss the Right to Be Forgotten as a workable solution. Because right-to-be-forgotten laws butt up against the competing value of free speech, the amount to which they can be seen as workable seems questionable to Klonick, (ibid, 1061).
But Klonick doesn’t completely abandon the idea of third party intervention. She claims adding a component, such as a government or private watchdog agency, to perform a type dispute resolution, or simply to issue official letters of apology from the accused, would provide relief by allowing shaming victims “to change and inform social meaning” surrounding the underlying issue, (ibid, 1065).
This leaves private actions to be examined as a viable solution. Although private actions taken by users can obscure search results, Klonick understands such solutions are limited, (ibid, 1062). The success of private action is largely dependent upon an individual’s power as a user (the strength of their profile), ability to profitably leverage technology to their advantage, or financial capacity to retain a reputation management service for the purpose.
The final takeaway from Klonick’s “Re-Shaming the Debate” is that the viability of each of these solutions is largely contingent upon the individual circumstances encompassing the issue. What works as a remedy in one case, may not work at all in another. While one-size-fits-all solutions, like the Right to Be Forgotten, run into trouble because they trample on competing values, answers specifically tailored to individual cases may provide relief, even though none of them may be a panacea.
The New Governors
“The New Governors” examines the self-regulatory regimes of social media platforms. Klonick first details the justification and legal grounds for content moderation, arguing CDA 230’s “Good Samaritan” provision provides a reason for both, (“The New Governors,” 1602).
She next explores the platforms’ motivations for self-regulation, claiming their motivations are driven by “free speech norms, corporate responsibility, and the economic necessity of creating an environment that reflects the expectations of their users,” (ibid, 1602). The third section of “The New Governors” argues the platforms have established models of governance that are not unlike other legal or governance systems, (ibid, 1602).
The important takeaways from this opening salvo of sections are two: One, Klonick believes the platforms’ content moderation systems embody democratic values—namely free speech norms. And two, the platforms, being dependent upon meeting user expectations, reflect and in turn perpetuate group norms.
As we needn’t debate these points, we’ll eschew their coverage here and instead concentrate on what Klonick identifies as problematic: The loss of fair opportunity to participate and the lack of accountability platforms have to their users, (ibid, 1603). Although absence of transparency within the content moderation regimes leads to concerns about unjust censorship, Klonick decides not discuss these matters at any great length, saving extended criticism for another day.
In a multitude of ways, social media platforms treat powerful users differently. Because of their stronger profiles and newsworthiness, powerful users have elevated platforms that give them an unfair advantage and extended reach, (ibid, 1665). On the flipside of the coin, powerful users can also be unfairly treated because of their elevated status and newsworthiness. Both leave powerful users at the center of a profound quandary: How do we ensure fair participation on platforms when some users, for better or worse, are given preference?
Klonick doesn’t know the answer, but she suggests once more that a one-size-fits-all solution shouldn’t be implemented. Instead, if such legislation should be found necessary, it should be carefully crafted so that it considers the self-regulatory structures already in place, (ibid,1666).
The second, and possibly more problematic aspect of content moderation, is the fact that platforms are more beholden to their bottom line than users. In other words, accountability to users is driven more by financial incentives than “corporate values” or a desire to meet “free speech norms,” (ibid, 1666). This leaves the balance of power between platform and users skewed, as users are only given indirect sway over the system.
As these systems of governance are “indirectly democratically responsive to users’ norms,” they’re “disincentivized to allow antinormative content” and “incentivized to filter content to taste,” (ibid, 1666-1667). The first unfortunate result being, of course, antinormative content could be filtered out, leading to less diverse content, (ibid, 1667). The second unfortunate result being filter bubbles, which lead to “nondeliberative polarization,” (ibid, 1667).
Again, Klonick sees the available solutions to these problems to problematical in themselves, such as for the platforms to adopt shareholder models or “for platforms to voluntarily take up a commitment to a notion of ‘technological due process,’” (ibid, 1668). For instance, Klonick believes a shareholder model wouldn’t work “because shareholder values of maximizing company profits are perhaps not well matched with user concerns over equal access and democratic accountability,” (ibid, 1668). Obviously, the problem with a voluntary commitment by the platforms is that they’d never do it voluntarily.
The final takeaway from “The New Governors” is the social media platforms have taken up a model of governance that is normatively and economically motivated to reflect the value of free speech and the social norms of its users. The platforms are the result of an unregulated internet and mirror democratic culture. However, they can only be “as democratic as the democratic culture and democratic participation reflected in them,” (ibid, 1670).
Toward building the new norm
Returning to the Twitter debate with the knowledge Klonick has gifted us in hand, we hope now to detail the problems we believe inherent in the experiment and the underlying consequences of its promotion. The best place to start is with the premise the entirety of Klonick’s teachings are dependent upon: That is, the internet is a normative place and that “norms are the primary social control mechanism of the Internet,” (“Re-Shaming the Debate,” 1044).
It seems fair to say that Klonick sees herself among the realists, (“The New Governors,” 1614). Each of her pieces strive to describe phenomena rather than proscribe actions in regard to them. In “Re-Shaming the Debate” and “The New Governors,” Klonick’s scant usage of modal verbs is impressive. Most often, when Klonick uses one, it’s there to convey possibility—not necessity.
Nowhere within Klonick’s Twitter threads, nor in the corresponding news pieces that covered it, does she drop an “ought” or a “should.” In other words, Klonick never properly formulates any of these propositions:
- “In general, people should observe one another in public and attempt to de-anonymize each other.”
- “In particular, professors teaching about privacy should try this experiment in their classrooms.”
- “As a regular course of action, people should Google everyone they meet.”
Although by no means is Klonick a celebrity, she’s certainly an authority on online shaming, privacy, and content moderation. Klonick’s authority has garnered a following of over six thousand users on Twitter—a number of those followers being important figures within the wider privacy debate. Some of these important figures have, no doubt, quite a bit of influence themselves as faculty members at prominent universities.
As Klonick has taught us, social media platforms don’t treat users fairly. A powerful user has far more clout with algorithms and far more reach with their microphone. Compared to Miley Cyrus (who has more than 41 million followers), Klonick seems like a regular Joe. However, within her niche sphere of influence, Klonick’s voice soars and reechoes throughout the echo chamber.
Albeit there may have never been a singular “ought” or “should” uttered in regard to googling people in general or the experiment in particular, each and every vote of approval potentially reinforced these practices as group norms. Presumably if norm development follows the path of classical accounts of social conditioning, then, for individuals, complying with norms means receiving signals of approval, which reinforce the behavior through social and neurochemical rewards. And over time the beliefs underlying these practices, from reiteration and neurochemical reinforcement, become less deliberative, value-oriented, and plastic, hardening into routine, as though established by rote paired with operant conditioning.
If we can extend this logic to groups, so the same would presumably be true for collections of individuals as well: Each yeasayer added to the pile (keeping in mind some yeasayers are more powerful influencers than others) would reinforce the belief that norm is correct, rewarding the collection of consenters individually with social approval and dopamine, as well as training them to perform the exercise again. It would not, perhaps, necessarily compel them to regroup per se, so much as it would compel them individually to coalesce once more around the idea.
Hume’s Guillotine doesn’t come crashing down on Klonick. She never derives an ought from an is, strictly speaking. But, per the logic of social norm enforcement, the is behind norm development is more important than the mere affirmation of a practice. It doesn’t just say, “this is the case” over and over again. It retraces the route of an exercise that is met with approval and neurochemical reinforcement at the end of the day—a reward on the individual level for compliance.
At the same time, this group of yeasayers, not only has their collectivity around the idea reinforced with each member added, but they also stand collectively in opposition to stances that deny it, ready as it were to punish or deflect those who would oppose it. And so, this meager is pronounces a should to fresh entrants to the conversation or the undecided every time its validated by the addition of another member to the group. To avoid the pain associated with social dissonance and reap the rewards of compliance, newcomers are compelled toward the in-group, absconding from antinormative behavior to forego the displeasure of being penalized or pushed outside the margins of the community.
An assistant professor of emergent media pointed out to Klonick on Twitter that her experiment actually reinforced negative behavior by rewarding the students who participated with extra credit.[x] This statement taken in conjunction with the other tweets we’ve noted that stood in opposition to the experiment testify to an antinormative position that contests the practice in general and the experiment in particular.
Far from being democratic, the supposed consensus is one part the result of a filter bubble and one part the result of Klonick and her colleagues’ elevated platforms, (“The New Governors,” 1665). Compounding the issue is the fact that Klonick spread the word about experiment on Twitter, the New York Times, and NPR, finishing her round of promotions by going on CNN to be interviewed by Michael Smerconish on his eponymous program.[xi]
The newsworthiness of the experiment being such lent the message legs, giving it far greater strides than if had merely been circulated on Twitter. While the opposition’s stance has been relegated to the remnants of skirmish left on a few desolate Twitter threads and a fantastic but modestly circulated article[xii] published by one of her opponents in the debate, Professor Chris Gilliard, Klonick’s message has been disseminated widely, thus decreasing the chance the opposition’s position will be given fair play and increasing the likelihood it will normalize the practice and the experiment.
As these “democratic” platforms can only be “as the democratic as the participation reflected in them,” the argument could be made that quite simply the number of votes needed to quell the promotion of the experiment never arrived, (“The New Governors,” 1670). However, these systems of governance do more than reflect norms, they perpetuate them. Rewarding yeasayers for agreement and compliance, they are anything but neutral parties. Making the lack of neutrality more problematic is the fact these platforms have a stake in the issue.
Don’t shoot the messenger
As Klonick relates to us, social norm enforcement is upheld by rewards and punishments, shame being a major tool for enforcing compliance, (“Re-Shaming the Debate,” 1035). Those who attacked Klonick’s experiment were for the most part agreeable to Klonick herself, applying the lesson learned in “Re-Shaming the Debate” by shaming the violation, and not the violator, (“Re-Shaming the Debate,”1055-1056).
As regrettable as it may be, there are times when an action is directly tied to the actor in such a way that if we cut the twine that binds them, we miss an important thread that deserves scrutiny. It is with this conviction in our heart that we now examine the actor. In other words, our goal isn’t to shame Klonick, but instead to shed light on why Klonick’s spread of these practices is deeply troubling.
As Klonick is a New America Foundation Fellow, perhaps, the most cynical of us would point a finger at them, questioning the degree to which normalizing such practices favors their benefactors, in particular Google.[xiii] In 2017, following the EU fining Google 2.7 billion for antitrust violations, Barry C. Lynn was fired from New America Foundation for praising the move.[xiv] At the time of Lynn’s dismissal, former Alphabet CEO Eric Schmidt, along with his foundation and Google, had donated over 20 million dollars to New America.[xv]
This left reasonable spectators to wonder if that wasn’t the behavior of a shill. Anne-Marie Slaughter reassured onlookers, stating:
[New America’s] norm can’t be that. We’re an organization that develops relationships with funders. And you know, these are not just black boxes; they’re people. Google is a person, the Ford Foundation—these are people . . . And particularly when they give you money, which is really a nice thing . . . basic courtesy I think requires—if you know something really bad, you say, ‘Here’s a heads-up.’ (“Has New America Foundation Lost Its Way?”, Washintonian, 6/24/2018)
“These are people.” So they are, but that just doesn’t change the facts of the matter. One being New America fired Lynn for putting their funding in jeopardy. Two being New America fired Lynn for praising what he believed to be the appropriate policy. Both facts point to a funder intervening in a policy decision.
To the millions of people suffering from online shaming, the Internet seems like an experiment gone horribly wrong.[xvi] People of color (in particular black women), women, the LGBT+ community, and the indigent have been disproportionately affected by this bit of mad science, and the consequences have been anything but amusing.
And here, perhaps quite innocently, a person affiliated with a think-tank that has been publicly documented doing something questionable involving the party in question, Google, tells us an experiment has primacy over the ethical violations involved. Quite simply, it makes one wince.
The eye of the beholder
In regard to the promotion of the experiment and the practice overall, the truth of the matter is we shouldn’t be eager to assign intention to Klonick’s actions, for only she knows to what degree her aloofness was pure, unfettered by the reins of a sponsor once removed. The power of these platforms to distort the effects of social norm enforcement being what they are, we must remain forever watchful against any form of shaming, (“Re-shaming the Debate,”1052-1059). However, regardless of Klonick’s intent, the possible consequences of her actions were well within the range of her purview, as her work documenting these phenomena testify to.
The platforms’ accountability to users, not withstanding Klonick’s own scrutiny, should be understood as self-interested and driven by the bottom line, (“The New Governors,” 1666). A long way from embodying “democratic norms,” these unneutral entities perpetuate as well as reflect group norms by shepherding like-minded users together into norm-enforcing herds and by turning the powerful speakers who lead them into virtual Pied Pipers. As they owe allegiance to no one but their shareholders, their reverence for free speech seems only a gimmick to keep the money rolling in.
The depiction of these platforms as democratic in “The New Governors” is a patent absurdity. Beside those descriptions in which the words “democratic,” “democratic culture,” “democratic values,” and “democratic norms” are bandied about so freely are the narratives of the shamed, which Klonick chronicles at some great length in “Re-Shaming the Debate.” They, nor we, have representation within these platforms’ systems of governance. And if our votes aren’t of equal worth, albeit this is an all-too-American norm, then how can they be said to be democratic at all?
Earlier we spoke of Klonick’s scant usage of modal verbs within “Re-Shaming the Debate” and “New Governors.” We observed that Klonick’s usage of modal verbs was done more often to convey possibility rather than necessity. There are only two normative statements within “The New Governors” that convey necessity rather than possibility: Both involve protecting these platforms—not people, not democracy.
When we Google a stranger on the internet, inevitably there will be one of three results:
- A neutral outcome: perhaps we might find out the person’s address, where they went to school, listings of old phone numbers, etc.
- A positive result: perhaps we might find the person’s awards, accolades, impressive accomplishments, etc.
- A negative result: perhaps we might find the person has blemished record.
So who cares if we make judgements about a complete and total stranger? We do. Because every click of a link is a vote for that content to be elevated, to be potentially linked to that person indefinitely. As an astute observer and pretty funny guy once noted on Twitter:
“Underneath every LinkedIn profile, there is a person who once wanted to be free.” “Linked in” and “internet links,” as in link this cuff to your ankle and see how far you can drag this dead horse.”[xvii]
Every time we do this experiment in the classroom, we run the risk of normalizing a practice that none of us should truly want. Every time we do this practice in general, we could potentially add a little misery to someone’s life—maybe even complete stranger or someone we don’t know really well. As a representative of what seems to be the antinormative position in this debate, I say these are results we should not want, and so these practices we should not do.
Klonick, K. (2015). Re-Shaming the Debate: Social Norms, Shame, and Regulation in an Internet Age. SSRN Electronic Journal, 75(4), 1029-1065. Retrieved March, 25, 2019, from https://digitalcommons.law.umaryland.edu/cgi/viewcontent.cgi?referer=&httpsredir=1&article=3720&context=mlr Online.pdf