And the same dynamics can occur with individuals hurling abuse online. Those individuals could experience consequences for their actions (from twitter, valve, potentially the criminal justice system in some cases, etc.) from institutions as well as from society as a whole. Telling a game developer "I will rape you!" should be an action that dramatically curtails the perpetrators career options and results in losing friends, but today it doesn't.
Sure, people hold dark impulses inside them routinely, everyone understands that. That doesn't mean they are uncontrollable. This is a recent phenomenon, which itself is proof enough that it is not due to some uncontrollable universal urge.
For example, when you're interacting in a customer service context on, say, steam, you are not anonymous. Valve can change its TOU to make it so that you forfeit all claims to redress if you use abuse during a customer service interaction, and can further scale that up to include additional consequences (banning you from participation in certain ways or from certain benefits).
Additionally, "in band" consequences don't have to compromise anonymity. On twitter or facebook, for example, you can have abuse reports that are verified result in different levels of account restriction. You can make it so that the account's posts are no longer visible in other people's timelines as replies, for example (e.g. they are only visible to people who specifically follow that user or when specific posts are linked to directly). You can restrict accounts in other ways and even block access to accounts for short periods of time. Or you can ban people forever. All of these techniques have been used for years and years to prevent abusive behavior, the only reason they haven't been used effectively against this particular sort of abuse is because of apathy on the part of most platform owners, but that can change easily.
Road rage is another related symptom. There’s something special about pseudo-anonymity than enables a small percentage of people to act like jerks.
The horrible comments are just a function of anonymity at scale.
Surprisingly, the part of the article that comes close to talking about what you brought up gets very little coverage and doesn't start the article (burying the lede). There too, according to the article, those posts (which you refer to) make up "an even smaller proportion". So those posts which might justifiably concern not only aren't featured prominently in the article, but constitute an indefinably smaller portion of what was already vaguely described as small. One has to wonder why the language you're talking about isn't the basis for the article. Seems a shame to focus on petty namecalling if publishing someone else's address presumably to encourage the public to shame or harass them in person or to "try and find their family" (quoting the article) presumably for some in-person dealing is commonplace. I don't know if that's the case, and neither your reply nor the article help anyone determine how frequent that is.
The article claims "What I’m drawing attention to here is the normalisation of online abuse — both from the perspective of content creators and from the point of view of the audience. It is now, for a significant part of our audience, the way in which they communicate with creators.". That's an evidenceless innumerate claim from the article which very much needs some backing. We simply aren't given information we need to make sense of the claims. What constitutes "normalisation"? How much is "significant"?
But what we are given is propaganda -- "creators" -- a word used to compare the programmer to a deity so as to "elevate authors' moral standing above that of ordinary people" (as https://www.gnu.org/philosophy/words-to-avoid.html#Creator explains), which is preposterous. They're programmers or software developers, working people who sometimes write programs some people don't like and less than 1% of whom (we're told) express themselves very poorly.
You are going well beyond the article with your evidenceless claims about the police, police resources, and everyone's speech rights.
I assume they are a very small proportion of feedback. On the other hand, in my current non-game-development career, customers never threaten to rape or kill me. Never ever! It's never happened!
It's great that people threatening rape or homicide comprise less than 1% of game customers and not 20% or 50% or 80%. But a little goes a long way, when it comes to people threatening to rape or kill you. Many people are going to treat those threats as a significant incentive to stay out of game development.
We are a tribal species; in the past this type of behavior would be socialized out of you or else you would be a recluse. You didn't have the opportunity to sit in your own filth and project your self-loathing onto anyone across the Internet.
As a species we're just not emotionally equipped to deal with the current reality.