Framework for Moderation
emi do PublicSeen by 257It’s been exciting to see many new tooters on social.coop and though each of our timelines may look a little different, we are looking for ways to ensure that the social.coop corner of the fediverse continues to reflect our values and ethos for being a cooperative and collaborative space. The CWG Ops team has been hard at work at our moderation duties through this large growth phase and we have been grappling with some moderation questions.
We would like to add clarification to encouraged and discouraged behaviours outlined in our CoC and Reporting Guidelines.
In an effort to encourage and model positive and constructive modes of discourse that demonstrate a mode of online communication that can prove counter to the norms on other platforms,
we’d like to introduce nuance to moderation based on the guidelines outlined on this blog: https://wiki.xxiivv.com/site/discourse.html
In this thread we will discuss different moderation challenges and based on those discussions, will propose amendments to the CoC for approval by the membership.
Item removed
I think that there should be something like this in there but I don't think this text is the right text. Posting references to unknown or unreliable sources to point out that they were unreliable or to ask others what they thought about the reliability of it is fine. It is the deliberate use of such sources to mislead that is the problem. Possibly shorten to just "Disinformation, meaning false information which is intended to mislead."?
simple, elegant, strengthens the CoC
Item removed
And who’s to say what an “unreliable or unknown” source is? I’m sympathetic, to a degree, but my Nature may well be your Joe Rogan, and I’m not sure we want to get into the infinite regress of defining what constitutes valid authority or criteria for credibility.
This would place an unreasonable burden on the CWG to have to be the arbiters of what is true and what is false. People who want to spread disinformation could just endlessly debate the merits of their claims, exhausting our CWG and putting our community at risk of losing our invaluable CWG members.
This is why I suggest broad prohibitions that allow the CWG to act quickly and decisively using their judgment, while we have the CWG election process to resolve any issues with said judgment
Except in a small number of well-known cases, I don't think it's realistic for moderators to judge whether or not posts meet this definition of disinformation, and individual users are going to have different perspectives on what is disinformation or not based on what they have or haven't seen and read elsewhere.
I think we should substitute intention with the potential for harm. Something like "Disinformation, meaning false information that is misleading and has a high risk of causing harm to an individual, group or wider society."
Holocaust denial has a high risk of causing harm to both a group and to wider society. But claiming that the moon landings were faked does not, as far as I'm aware, pose a high risk.
Cooperation with efforts to moderate seems like a reasonable thing to expect of users.
Item removed
I strongly object to including anything about disinformation in our CoC.
1) Who determines intent to mislead?
2) It requires way too much original research, and even expertise, to discover actual facts these days. This is too high an expectation for moderators, or other working groups.
3) People should be able to honestly and respectfully discuss divisive issues. The big platforms systematically thwart free speech, it is a really important service to provide here.
I think this is reasonable language that will make the moderation teams job easier, while still leveraging their discretion and judgement. Agree.
I agree with @lynnfoster and strongly object.
I think the blog post linked to at the very top of this thread (https://wiki.xxiivv.com/site/discourse.html) is very thought-provoking, but if I understand it correctly, it's about how people engage in discourse, not about the content of that discourse.
I feel generally okay about establishing behavior norms for how to engage, e.g. no harassment, not trying to force people into conversation after they've indicated they want to disengage, no "sea-lioning."
I have a harder time thinking of clear policies with regard to the information content. Certainly there are kinds of ideas that feel inappropriate to push, but I have a much harder time figuring out how you can formulate general rules about what kinds of content is or isn't acceptable, and so I'm generally reluctant to "go there."
Repost for reference:
I'll post this again:
https://realityteam.org/resources/credible-sources/
I'd say if someone argues that the World Heath Organization is wrong, and they provide a link to "Bob" on facebook, it's reasonable to go with the WHO.
Not yet sure, I see some issues, but tend to agree in general. I think some things could pass tagged as fictional or not testified, if the sources are weak. It depends on the framework. When there is an appeal procedure, performed by a democratically legitimated supervising body (not the ops-team) transparent by published transcript or minutes, we can delegate the case-to-case discussion there and more or less call it a day.
I think support the intent of the proposition, but i think the language needs to be refined. A helpful exercise may be to think through the the type 1 and type 2 errors (false positives and false negatives). Also, do you really mean not to target political disinformation?
While disagreeing with the disinfo proposal as is, I support having "disinformation" as a topic subjected to moderation. The process I imagine: CWG moderators are explicitly granted leeway to hide or delete (giving limitations of tech) posts they judge to be disinformation; then, they invite the poster to show them if they've made a mistake, that the post has some minimum level of apparent merit and should be allowed; and the CoC somehow tells people that this process is a thing here.
I don't think the CoC can possibly have adequate wording that clarifies what counts as disinformation. I think the essential thing is to transparently say to the community effectively: "we do consider the question of whether something seems to be disinformation when moderating", and we can describe something about how moderators are to make such judgments.
What if we simply say something to the effect of:
We don't want disinformation, we see it as harmful
It's extremely hard to nail down what counts as disinfo once it gets fuzzy
Here's a list of topics we have already voted on and decided to count as disinfo
And we develop such a list via actually encountering specific issues. So, we don't curate a perfect list, we basically say, "these came up ever and we came to the conclusion that they will be barred here"
Maybe nobody ever says vaccines include microchips and so it doesn't come up. Maybe someone starts pushing that, so it goes through moderation process, and the co-op confirms "yes, we agree that's disinfo" so it goes on the list. And maybe we require a supermajority vote?
I just think while there's so many pitfalls in this topic, but we shouldn't throw up our hands and deny having disinfo as a valid reason to report a post. So, the minimum could acknowledge disinfo as a type of thing that is reportable even as we also acknowledge that drawing the line is hard.
I also condemn disinformation but I think that kind of rules are either too detailed to cover specific cases (and must be updated every time a new case appears), either too vague and subject too dangerous interpretations.
Maybe CWG members could explain difficulties they encounter moderating disinformation to better the need of such tool in our CoC
I agree this is important to articulate but would suggest a couple of language changes.
We could just say "conspiracy theories," as they can be harmful whether or not they've been debunked.
Its clearer to insert the examples in the first clause: "false information, e.g. [list categories], with the intention to decieve." There are reasons to post false info; to debunk, warn, or even make a joke, and this would help clarify that intention is what makes it DISinformation.
A number of people have suggested inclusion of language about intention to deceive. If we were to go in that direction, how would we distinguish between someone whose intention is to deceive vs. someone who has themselves been deceived and is posting sincerely?
I agree that this is good enough as the whataboutery from other commenters is preventing something that is "good enough for now, and safe enough to try". I'd also point to some work that I've done in this area which references this excellent article with a graphic I found super-helpful.
I support the aim of this, but there's much open to interpretation, espec "references to unreliable or unknown sources".
If we want to avoid disinfo on SC it seems either:
– deep specificity - specific policies on Ukraine, Covid, Gates, homeopathy, 4G etc as/when issues emerge. Maybe set by moderators but debateable/votable by the community somewhere.
– something general - like this proposal but shorter & giving power to the mods (espec if they can be transparent on # of blocks/common topics)
Whilst I support the aim of this, I'm concerned that there is a lot of subjectivity in this. what is "false information", what is an "unreliable or unknown source"?
I agree that this should be included as long as you're sure that you can enforce it fairly and without too much cost in time to the moderators.
Like @Andrew Shead, I disagree with this proposal for reasons similar to those raised by @Lynn Foster.
Item removed
This addition makes the CoC better, and I see no reason not to include it.
I am disappointed to see that, as a community, social.coop is more interested in rules-lawyering "what is disinformation" or "what does cooperating mean" than in trusting the moderation team do their job as moderators. Surely none of you really believes that we can write a set of rules precise and concrete enough that it will cover all possible situations, and remove the need for moderators to make a judgement call.
I agree with the principle but have read many good suggestions on this thread that may improve the wording, and which I'd like to see incorporated, e.g., to better emphasize harmful intent.
Some further thoughts. The discussion has made it clear that there is disinformation and there is disinformation.... we don't all mean the same thing.
Your Ukraine example is a great one @Nic Wistreich . From what I've read, I have understood for a long time what you became aware of 2 weeks ago, as well as the historical context since 2014 and before. My worst fear in all of this is that information to support that kind of view is blocked from social.coop only because the predominate mainstream presentation is the opposite. Or that someone saying that partially since delta and for the most part since omicron, vaccines haven't prevented transmission, with studies to back it up, will be automatically blocked as anti-vax. Not that anti-vax views should be automatically blocked either, if there is science behind the viewpoint and it is presented honestly.
On the other hand, on @Doug Belshaw 's chart above, some of those categories seem like possibly realistic ways to identify actual intentional disinformation (as opposed to the kind that the powers that be have just declared as disinformation for their purposes). Like I don't support allowing pictures of Palestinian children presented as Ukranian children, which has been exposed in mainstream media. I'd be curious to hear from the moderators if that kind of thing has been encountered.
So, while I have some doubts that this can be done, if there are ways we can challenge clear disinformation techniques (various kinds of manipulation of data), without discouraging views honestly held by a minority, and if it *is* really a problem, then that is worth considering. And encouraging adding /sarc or content warnings or whatever to clarify satire seems like a reasonable idea.
Context... I come from a left anti-capitalist perspective, which does inform my trust or lack thereof of certain sources, for what it's worth.
And... I'm a white older female, for whatever that's worth. The old part might be worth something. :)
Instead of trying to formulate an all-encompassing prohibition, we could treat disinformation case-by-case as we do for the defederation of instances. There is a wiki page that lists Suspended and Muted instances, when and why: https://wiki-dev.social.coop/Defederation_of_instances: Similarly, we could maintain a wiki page that lists community-agreed disinformation with links to the relevant discussions on Loomio and final community vote. As a community, we decide what constitutes disinformation with a traceable record of the decision making process. Doing it this way allows us to revoke a particular decision when new information changes our collective mind. This process should give us maximum flexibility of engagement with disinformation, and assist the CWG in its work.
Is that list of defederated instances updated regularly?
It looks like no-one from the CWG has weighed in on the disinfo policy vote - could y'all tell us if this would be helpful to you? I'd gladly change my vote if so
@Matt Noyes @emi do @Mica Fisher @Sam Whited @Eduardo Mercovich
@adamgreenfield I don't know. If not then it should be kept current. Also, that link is to the proposed new wiki, currently in development.
[ETA4: comment edited to remove no-longer-relevant references to deleted content]
Quite a few other Fedi instances have no problem dealing with misinformation without needing to have a novel-length Loomio discussion or a precise policy outlining every possible edge case.
This is why I like Ana's responses, they usually cut to the heart of the matter.
I love this post so much.
Item removed
I've had some pretty strong feelings about this conversation. Like others, I think @Lynn Foster has been hitting the nail on the head. I feel like I could throw my support behind any moderation policy she supports.
I also have a question. When I mentioned this thread to my spouse, he asked whether something had happened to prompt this conversation on social.coop. And I realized I have no idea. Is there a specific reason why the lack of a disinformation policy is being brought up at this time?
I was a moderator on another Mastodon instance. I resigned, and switched to social.coop, because the site admin and most of the other moderators thought the idea of discussing moderation policies was laughable. I wanted to be on a Mastodon instance that actually took democratic discussion of policy seriously, and that's what social.coop appears to be doing.
@Dynamic as I understand it, the originating point for the current discussion was this post: https://www.loomio.com/d/zT7jiGxI/moderation-of-fake-news
@nicw For what it's worth, Nic, I liked your posts and I was referring to liking Ana's comment that we may not need a gigantic novel to discuss all this, which my proposal has caused us to write.
Item removed
@Nic Wistreich I can see how my original comment was more strongly worded than I had intended. I apologize for that. This ongoing neverending discussion about "framework for moderation" has me more than a little bit hurt/angry and disappointed, and I clumsily pointed those feelings at your comment. Sorry.
@Ana Ulin thanks for saying that - and sorry too for complicating things with the delete, and now making this thread even longer!
Thanks for the context. I never would have guessed, although I guess (knowing how riled up people get about vaccines) that I probably should have.
Based on the description provided, this is actually exactly the kind of content that I *don't* want to see banned from our forum. I support vaccines and I think they are great, but I also think it's really important for people to be able to have these conversations. I have no idea what the quality of the vax-skeptic posts were, so my opinion is based on how they were described in the thread you link to. My impression from what was presented is that author was not forcing their views on people nor intentionally trying to convince people of things that they themselves did not believe (i.e. I am seeing no evidence of intention to deceive).
Assuming that those are the case, I strongly want there to continue to be mixed spaces where people can hash out these kinds of fraught issues without it turning into an abusive free-for-all. I would feel good about social.coop being (or continuing to be?) a place like that.
Hmm... reference to this thread going on forever did prompt me to scroll way up to where @Will Murphy dissected some of the writings from the person in question (the irony is not lost on me), and I'm now less inclined to believe that this Jem Bendell person is acting in good faith.
I could see myself supporting a policy based on the "gotcha" argument structures. I feel much less comfortable with a policy on posts supporting particular beliefs, even wrong ones.
This is as much a test as a proposal. I think as a group, we really don't want to have even the appearance of authoritarian-style censorship. Disinformation is not usually about whatever topic it pretends to have an opinion about, it's about calling into question every verifiable source of information. I posted in mastodon about this, but it bears repeating.
To believe the moon landing was faked, for example, can seem cute, but in order to believe that, you have to stop trusting the government, medical and academic authorities, and generally recognized news sources. Specifically, antivax claims have actually been used to incite violence.
At the same time, how do we disallow fraud, quack remedies, and political disinformation but leave things like spirituality alone. If someone posts about Jesus or astral projection or tarot reading, they should probably slip through. It's not clear that a political comment should inherently be against the rules, either.