Forced anonymity perhaps? Suppose you're a Striper and I'm a Dotter (from the deadly enemy Stripe and Dot tribes), but we both like to play guitar. Well if you show up on my guitar forum I'll be 'who let that filthy Striper in here?!' and will interpret even innocent participation as attempts to stripe up the place, regardless of whether I have been trying to take it over for the Dots or not.
But if we're just Guitarfan_1172 and Guitarfan_3654 then we can chat about guitars and build bonds that transcend the decor war, as long as neither of us goes off topic and starts discussing patterns. Of course the problem is that the positivity is latent and confined to the interior of the guitar forum, unless we meet in battle and then become aware of our mutual guitar friendship which will cause us to re-evaluate our whole worldview.
Think of it as a poker-like game where you have public and private cards. Your public dot or stripe cards are your loyalty to your design tribe, and the # of stripes or dots across all your cards indicate your degree of group identification. But you also have hidden cards reflecting various private interests.
Let's say when you battle another player you're required to reveal one set of hidden cards if any other have the same or more points (or some other ratio) than your public loyalty cards. In this example you really love guitar and have to reveal your guitar cards. If the other player has the same cards they must also reveal theirs, or perhaps the revelation depends on the strength of their affiliation. Of course, by revealing your similarity of interest maybe you lose points among your own (stripe/dot) peer group because you have something in common with an enemy player...
You see where I'm going with this. Of course, any modeling of relationships and disruption built on this framework assumes that tribal affiliation is fundamentally arbitrary, eg as an accident of birth. If it turns out that the Stripers really and truly want to slice up their enemies or that the Dotters really and truly want to punch holes in them then mutual interest in guitars/crochet/pokemon may never be sufficient to overcome their differences. Thus a good deal depends on whether the enemies are opposed to each other by default or have made a conscious choice to fight each other, in which case that's likely to predominate over all other factors.
"Forced" anonymity exists in some online communities today. If most participants maintain and assume good faith, and are tolerant of some efforts by bad actors to troll, then it works remarkably well. Problems tend to emerge once these assumptions no longer hold true.
Anonymity allows the content to be judged by every reader on the content's own merit, and colored solely by the reader's perception.
However, it's also rather depersonalizing, as every participant will often interpret the others as manifestations of a generic individual, based on their own mental model of what generic individuals look, act, and sound like. It therefore creates the illusion of homogeneity from every point of view, where in truth the real membership might be quite diverse.
The resulting interaction means that over time, the community will develop a sort of institutional identity, complete with etiquette, mannerisms, and tropes they are expected to adhere to to be welcome in the in-group. Some people will don the institutional identity despite it being radically different from their own, just to feel part of a community.
I remember internet forums as a teenager. Back when people didn't use real names and profile pics were of random characters or celebrities not our faces.
It honestly never crossed my mind that the people I was talking to could be anything other than white and mostly guys. I come from a country where races basically don't exist so that worldview made sense kinda.
Sometimes I wonder how many racist jokes I made to people who were black or asian and I had no idea. In my defense, those were just jokes/opinions I picked up on the internet, completely learned behavior as I was trying to learn how this "American Society" thing works because most of the internet seemed to follow american society rules.
Anonymity is great, but it has its drawbacks. Had I known the people I'm talking to are black/asian, I would've asked about their cultures and how they perceive this American Society thing I was exploring.
There's an interesting contrast between completely-anonymous communities (like the chans) and pseudonymous communities (like forums where people have usernames). The issue is that people tend to want to not be constantly on guard in all their interactions, and if they chat about subject A they will often want to chat with the same people about subject B. This is why forums so often have an "off-topic" section that tends toward flamewars.
One issue with the sort of pseudonymous communities, where there is a common identity that persists across discussions, is that once someone is "outed," they can never be un-outed. Systems that rely on forced anonymity assume a single communication channel and no leaks. Also, if a group of dedicated people wants to subvert a community, they can easily do so in a group where their opponents are "playing fair" under completely individual/atomized notions that "we're all just guitar fans!" If a collection of Stripers is able to know that they are Stripers beforehand, or tend to have certain ways of talking about things that secretly reveal their Striperness to each other but not to Dotters, then they can coordinate and increase their own influence to drive Dotters out of the community.
At the end of the day, these systems require some sense of common civility where enough people on both sides are wanting to have the system actually work. If you have to design the system to be completely trustless, that's a hard thing.
Hmm that's interesting. I could imagine some social network which asks you for basic but controversial details about yourself and treats those as your hidden cards, say religious affiliation, political beliefs, etc. and non-controversial details about yourself. It would match you up with potential friends based on the "seen cards" (non-controversial facts) and would schedule the release of hidden card information to the match over time. In that sense it would be almost like forcing the progression of friendship, sometimes your hidden cards match and you're both happy, sometimes they don't and you have to work through it.
Even when staying on-topic, shibboleths will still leak through. The discrimination rate will remain above zero, especially for motivated participants.
Part of me wonders whether separating parts of identities through multiple accounts could be a solution for doing this on future social media sites.
You'd have one login, but the ability to set a different persona for certain audiences, like on specific subreddits, Twitter hashtags, etc. That way, the people who want to talk about gaming or TV or movies won't also be able to go through your history and tell that you also post about say, politics or personal fetishes or whatever else.
And bots couldn't be developed to auto ban people for 'wrongthink' based on the other parts of the network they post on.
But if we're just Guitarfan_1172 and Guitarfan_3654 then we can chat about guitars and build bonds that transcend the decor war, as long as neither of us goes off topic and starts discussing patterns. Of course the problem is that the positivity is latent and confined to the interior of the guitar forum, unless we meet in battle and then become aware of our mutual guitar friendship which will cause us to re-evaluate our whole worldview.
Think of it as a poker-like game where you have public and private cards. Your public dot or stripe cards are your loyalty to your design tribe, and the # of stripes or dots across all your cards indicate your degree of group identification. But you also have hidden cards reflecting various private interests.
Let's say when you battle another player you're required to reveal one set of hidden cards if any other have the same or more points (or some other ratio) than your public loyalty cards. In this example you really love guitar and have to reveal your guitar cards. If the other player has the same cards they must also reveal theirs, or perhaps the revelation depends on the strength of their affiliation. Of course, by revealing your similarity of interest maybe you lose points among your own (stripe/dot) peer group because you have something in common with an enemy player...
You see where I'm going with this. Of course, any modeling of relationships and disruption built on this framework assumes that tribal affiliation is fundamentally arbitrary, eg as an accident of birth. If it turns out that the Stripers really and truly want to slice up their enemies or that the Dotters really and truly want to punch holes in them then mutual interest in guitars/crochet/pokemon may never be sufficient to overcome their differences. Thus a good deal depends on whether the enemies are opposed to each other by default or have made a conscious choice to fight each other, in which case that's likely to predominate over all other factors.