Phil offers a sociology of online abuse, separating it into two broad forms: “abuse from above” and “abuse from below”. Abuse from below, he says:
offers [the abuser] a simulation, a simulacra of mattering, of being someone who counts — especially if the target responds or reacts. The unnoticed is suddenly noticed, even if they’re hiding behind an anonymous handle and all they’re doing is shitposting, success in getting a rise is still success.
This is a good analysis, which rings true. But analysis only takes us so far. If our normative position is that abuse is a bad thing which should be removed from the internet, then we need to look deeper into what triggers abuse. After all, not everyone lacking in power is an online abuser.
One way to do this is to move from sociology to social psychology. I am thinking in particular of the groundbreaking The Moral Economy, in which Samuel Bowles analyses in detail how “market signals” assumed to be a core driver for economic efficiency actually often contribute to market corruption and failure because of the way they are interpreted.
Let’s regard social media — and I have in mind twitter and Facebook particularly on account of their “like” and “retweet” functions — as a market.
People on twitter/Facebook labour over messages in the hope that lots of others will “buy” them, in the form of likes, retweets and other engagement, generally measured in followers. Of course this doesn’t generate cash of itself, but particularly successful sellers can build enough profile that, overall, the investment in time sending messages become worth it.
One way to generate likes and retweets is to be abusive. This works, potentially, for both those “above” and those “below”.
But the externalised costs of this kind of abuse can be so great that public policy intervention is deemed necessary, by those responsible (or those who want to assume responsibility). That is, online abuse is a form of environmental pollution which makes life worse for all of us, but especially those nearest the pollutant output point. Online, that’s women, black people, gay people.
Public policy intervention against the online abusers/cost externalisers can take two forms.
The first is for the online operator to ban the abuser. This happens every now and then, when the operator decides that need to retain overall reputation as a market outweighs the costs of losing the online usage associated. Such a decision does not take any account of that abuse moving over to a different, perhaps well-policed, platform, or of the possible though under-researched and still evolving link between online abuse and real world abuse. It is not, after all, the job of the Twitter market to worry about the health of the Snapchat market.
The second, more common public policy intervention is the act of “calling out”, or seeking to shame abusers in public. The online operators outsource this intervention to members of the line community itself, who are paid in the same currency of retweets & followers via a form of general taxation.
In market terms, the abusers are then sent a “signal” that they should change their product, on the basis that consumers do not like it.
The problem is that it is difficult to see how this can ever work.
Some abusers may be quite stupid, but even they can tell that the market signal sent as warning about to change their product is in fact a message that their product is doing well enough to be noticed.
In fact, the signal sent to the abuser fulfills all three of Bowles’ research -grounded reasons  for which those receiving such a signal actually interpret it in such as a way that the very sending of the signal worsens their behaviour by “crowding out” remaining decency.
First, it’s a “bad news” signal, indicating to the abuser that the unwritten rules they thought operated (“I am free to say what I like because it’s on the internet and I’m not actually murdering a gay person”) are not as they thought. This means all previously understood “contracts” are off.
Second, the signal leads to a process of moral disengagement. A message to the effect that the abuser, who may be anonymous online anyway, is now not to be regarded as a reliable seller to the market means that any prior moral duty to comply with social norms can now be disregarded.
Third, and very much in line with Phil’s sociological take, the signal messes with the abuser’s self-conception as a web-worker with a degree of autonomy over their production. Being now told that he must meet minimum quality standards not of his own making is unlikely to motivate them to improved online performance, because it is no longer his responsibility to do so. “Calling out” becomes, by mistake, a form of managerialism.
In the real world, as Bowles sets out, the market does not completely collapse because of these misplaced “crowding-out” signals, because they are mitigated by broader social commitments to civic culture. But on the web, even though people generally bring their real world decency to their i-phone when they type a twitter message, there is no such established civic culture , and it is all too easy to interact with strangers in a way you would never contemplate elsewhere.
In such circumstances, it might be argued, those who take on the role of “calling out” might be adding to the abuse problem, not helping to tackle it.
So is there a better way?
In the real world, Samuel Bowles thinks there is. He argues that current public policy is based on Machiavellian/Hobbesian assumptions (or the assumptions of their later interpreters) about the innate selfishness and corruptibility of people, and that this leads to a public policy based on sanctions. But based on what we now know about humans in society, he goes on, public policy is actually best constructed on an Aristotlean model, in which citizens are motivated to civic virtue by incentives which foster innate values of altruism.
When it comes to stopping abuse on social media, a neo-Aristolean approach might look something like this:
- Stop rewarding abusers by “buying” their online goods under a pretext of exposing them;
- At the same time, create routes towards virtuous online behaviour by rewarding generous social media use through likes, retweets and followers;
- But at the same time, avoid the pitfalls of signal misplacement by developing these rewards to benefit all users of the social media service;
- Working offline to develop civic cultures of generosity and virtue which are then consciously transferred into social media use;
- Most important of all, we need a tax regime linked to an independent regulator/assessor of online services which, while ensuring that Facebook et al. actually pay their taxes, also allows a route to some tax breaks if the regulator assesses that they have contributed to a more virtuous online environment through their own contols and rewards systems.
Some of what is needed is technological (algorithms geared to cooperation, not to who competes best), and could be a matter for public policy engagement with the operators, based on advances in social psychology, and of a type which reflect (but also presage) Ulrich Beck’s vision for a democratization of scientific progress.
But more of what is needed is likely to lie in wider commitment, especially within left-leaning political parties and social movements, to equality of participation in social decisions as a whole, and to the kind of material change that is needed to drive those new/renewed freedoms.
In such circumstances, social media might come to be an engine that drives a neo-Freirian project of the liberation ofthe abuser, rather than a form of self-generating control of the abusers with no power by the abusers without.
 The principal research evidence brought forward by Bowles to support these three reasons for signals going awry are:
“Bad news”: the reaction to a Boston Fire Commissioner’s decision to clamp down on sick days (p.9–10);
Moral disengagement: the ability of Berlin students, under explicit market conditions to let go of their morals over the gassing of mice (p. 91–92);
Loss of autonomy: setting expectations for engagement in and quality of art work reduces children’s engagement in and quality of art work (p.99–99)