Codes of conduct - adoption of standards

I am aware of two CoCs that have some degree of adoption as standards. I assume @wolftune has already considered them. For what reasons did we decide not to adopt one of them, ourselves?

tl;dr:

Those examples are from an interconnected family of CoCs. Within that, a different one, the “Citizen Code of Conduct”, is the most acknowledged “standard” and Berlin is a grandchild of that. But since then, a true FLO standard has risen to much more prominance: the Contributor Covenant, and that would be the one we’d adopt now if we went for “standard”.

We never decided not to adopt a standard. I started off seeing the need for a CoC and system around it before any of those existed at all, and we didn’t proactively switch later. Ours started based on a college honor system + KDE CoC and evolved to include reference to ICA principles and some other things from community feedback. And our CoC still has elements the others lack. Aligning with co-op values, our CoC assumes a more Transformative Justice, learn-and-grow, conflict-resolution prioritization, while the other CoCs tend toward more top-down, zero-tolerance approaches.

I am anyway now thoroughly revisiting all this (with others in the team and community) in conjunction with announcing Discourse (along with adapting its relevant tooling). And we’ll be considering all the best practices and options, including the idea of using more standard terms.


The following could be considered an incomplete draft of what should become a blog post on the details, history, and issues around our CoC.

Our initial CoC: based on Randolph College Honor System

Our initial CoC from 2013 actually came from adapting the Randolph College Honor Code after I attended a presentation by psychology researchers showing very strongly that such a traditional honor system at older universities was actually far more effective than the modern style of giant lists of strictly prohibited top-down authority-focused codes at many large public universities.[1]

The initial version of Snowdrift.coop had a stage after sign-up where users would accept an “honor pledge” comparable to Randolph’s (incidentally, all the best research shows that getting people to think about these things in advance and to care about the community etc. sets us up for success more than focusing on it being a punitive issue later). I could (but won’t) go into the long list of elements (beyond the code itself) that make a holistic healthy culture in the case of the Randolph studies. Of course, we can’t directly translate a university culture to an online platform co-op. And now we have to fit our updated workflow and tools.

In the same direction, I’m today just learning about the concept of “Transformative Justice” which evolved from “Restorative Justice” which is all similar to the traditional college honor systems (as opposed to the hard-line top-down modern codes). I remain convinced that these are extremely important concepts missing from the way most CoCs are treated these days.

Geek Feminism, Citizen CoC family evolution

The Geek Feminism “anti-harassment policy” emphasizes that one issue (harassment) and focuses primarily on a long list of what behavior counts as harassment. It doesn’t cover many other things a good CoC should also address. It also emphasizes a blunt banning-style enforcement instead of restorative justice and conflict-resolution.

Those issues are among the many reasons Stumpton Syndicate made a much better, more-standard http://citizencodeofconduct.org (using elements from Django CoC as well). And that Citizen CoC itself has seen a number of revisions and is the basis for PDXRuby CoC which the Berlin one then used and adapted.

I’m not sure if there’s a connection (there’s no attribution stating such), but the true FLO standard CoC today is actually https://www.contributor-covenant.org (and I think it well deserves that status as being better overall than any other recognized competitor in the space; and it’s been translated into a huge number of languages.

Of course, there’s tons of others, like the widely-used and adapted https://www.kde.org/code-of-conduct/ (which itself was adapted by Ubuntu and others). We consulted that one when drafting our initial CoC.

Co-op connection

We recently added a reference to the ICA values & principles which is the most standard document for co-ops internationally. By deferring to that, we can simply say that we do not tolerate people who make statements or take action in opposition to those principles. Anyway, those are the key standard principles we prioritize while we otherwise figure out specifics as needed to fit our particular case.

The social.coop CoC (which I’ve been involved in working out) references the ICA principles specifically as well.

Reporting/enforcement/resolution process

One of the key ICA values is “self-help” which relates to enabling and encouraging members to do conflict resolution and not to focus primarily on engaging authorities (moderators) etc. to take judicial action.

That emphasis on the co-op community, building a culture of respect, resolution, healthy communication etc. has been guiding this whole process. It’s the thing missing from most CoCs that either (A) take a weak “be excellent” sort of form that does little to protect anyone or (B) take a very hard-line zero-tolerance top-down-enforcement approach.

We want a lot more energy put in doing the hard work of maintaining healthy community, avoiding tragic public shaming or other bad patterns common to online conflicts, etc. but still have enough structure in place to protect everyone from harassment, bad actors, etc. and to assure maximum participation, especially from diverse, marginalized people.

More than wording of the CoC itself, the main issue is how it’s enforced. I recently added https://wiki.snowdrift.coop/community/conduct-enforcement, and we need to improve on that as we build a functioning co-op organizational structure.

People need both systems and people worthy of trust to assure that participation at Snowdrift.coop will be healthy and safe for all. We’ve done a lot in this direction, but we’re not done. Regardless of the exact text of a CoC, where the rubber hits the road is how we deal with it. And that’s where we get into our organizational structure, how our mod tools and flagging work, our Bylaws etc.

Future plans

We’re certainly in the space where solving the issues of online conflicts and harassment etc. is far from figured out. We’re aiming to use the resources that others have developed and have no interest in wheel-reinvention, only in getting the policies that are best for the long-term health of our co-op.

We certainly invite anyone to mention any specific concerns about the current CoC.

I’ve heard some already. For example, “Assume Good Faith” itself was added to our CoC after feedback from others. That itself is pretty standard (one of the key principles at Wikipedia), but there are concerns about it. We don’t really want assumptions at all (certainly not to let bad-faith actors get away with careful wordings that let them hide behind an appeal to AGF). Sure, we want innocent-until-proven-guilty overall, but not to the extent that we allow real harm to our community. So, that should probably change to more the direction of “don’t assume bad faith”.

I’m currently making my way through all the more recent resources to verify that our policies cover everything worthwhile. Sometime soon, I’ll be proposing some updates. Whether or not I’ll propose adapting to a foundation based on another CoC, I’m not sure at this time.


  1. https://randolphcollege.blogspot.com/2013/08/psychology-professors-present-honor-code-research.html references the presentation I saw, but they discussed a lot of other details, and I discussed several things with them in person. They emphasized the deep ownership and care that students have for the honor system and how central it is to the community. Students even take tests on their own schedule with no supervision. A student jury handles conflict and violation resolutions, usually with a restorative goal rather than a zero-tolerance policy, etc. While the original wiki stuff is buried in different locations in git or old databases, we started with basically Snowdrift-adapted use of their honor system, CoC stuff, enforcement process, etc. and there were footnotes explaining more details about it. These insights certainly formed our initial approach and my thinking about this whole topic. ↩︎

1 Like

For some updates on the points above:

The following article describes the concerns about Assume Good Faith that I mentioned in passing above. It’s insightful in many respects, though makes some unfounded assertions in a few cases.

Acknowledging that AGF can be weaponized, the position of the article is strong beyond reason. It says that people just will take AGF in a CoC to mean that moderators will minimize concerns that are reported and so on.

Ironically, the best way to read the article is by assuming good faith! With that perspective, we can see real issues that need to be addressed to assure that AGF (if we keep it) will not have the unintended consequences. The article does a great job of describing the experience of more marginalized people.

But if one reads the article itself with an assumption of bad faith, then statements like:

It tells marginalized people that you don’t see codes of conduct as tools to address systemic discrimination, but as tools to manage personal conflicts without taking power differences into account.

can be interpreted as the author demanding “badge beliefs” and mounting a political attack on anyone who doesn’t say all the prescribed things (as part of the trend in today’s politics toward condescending, dismissal of anyone who questions any aspects of a certain set of left-wing dogmas).

Clearly, AGF matters here. It makes the difference between reading an article like this and taking real lessons from it or using it to further divide people and create additional tensions. Even read charitably, the article is far too dismissive of the importance of norms like AGF.

This is not an either/or situation. We want a CoC that both deals with personal conflicts (even between those with no power or privilege differential) and addresses systemic discrimination patterns. We need to have norms of healthy interactions while working to not let them be gamed in ways that cause harms.

When there is a power or privilege differential between people, we should emphasize deference for the less-privileged side, and some of that can live in the CoC, though much of it is in the enforcement mechanisms.

As the article describes, there are bad ways to enforce a CoC (e.g. the unfairness in making victims apologize for having emotional reactions), and we certainly don’t want to do those things.

Of course, these issues are complex, such as noted in this perspective brought up well by someone else in the social.coop CoC working group who linked to this series of tweets:

And besides that, expectations become self-fulfilling prophecies.[1] So, actually expecting good faith or not has dramatic positive or negative effects on everyone. But while AGF remains a valuable concept, enforcing such assumptions/expectations may not be the right approach.


  1. This Invisibilia video on expectations introduces the topic well. And, as a side-note: expectations driving behavior is another reason that economic models positing humans as utility maximizing machines are so damaging, especially when taught to future business leaders. ↩︎

Here’s a very important point on the original topic here. In the social.coop context,[1] one of the participants (who, incidentally, supports the Geek-Feminism type direction) said:

all the value of a code of conduct comes from the process of generating it – identify your values, discussing them, codifying them, and then creating the organizational capacity to act based on it, and enforce it – the words matter … but you can get away with a pretty lousy one as long as you have a strong sense of unity around your values and a competent enforcement and reporting mechanism …

The now-common practice of adopting a boilerplate code of conduct for a project is really an antipattern — it just undermines confidence in codes of conduct

So their point is that a CoC and everything around it is inherently specific to each community and all the others are resources to draw upon.

And for our CoC going forward, I’ve learned a lot and will propose important updates. And in the end, along with really announcing this forum and launching the site fully, I want to see a real sense of buy-in on the end result from the core team and extended community such that everyone feels it is truly an appropriate reflection of our values and will accomplish what we mean it to.

(side note: on the enforcement side, tooling definitely matters, and Discourse remains so much better than most of what I see anywhere else)

[update]: I found a thorough and amazingly excellent article by our friend and Stumptown Syndicate / Open Source Bridge co-founder Christie Koehler which references the multiple CoCs mentioned in our topic here: https://authenticengine.com/2016/adopting-a-code-of-conduct-is-an-adaptive-challenge-not-a-technical-one/ was posted by someone with the tag-line:

Adopting a code of conduct is an adaptive challenge not a technical one

Christie’s post even describes the tensions around wheel-reinvention and more. Furthermore, it emphasizes how important conflict resolution is. She describes that, unlike the expected use of Citizen CoC to block serious abuse, almost all the reports that came up at conferences were complex, nuanced, subtle interpersonal issues that needed conflict resolution.

I think I’ll reference her article when I eventually blog about the CoC situation that we end up with (which is still in progress as of this edit to this post).


  1. Incidentally, Social.coop had some unfortunate tensions arise that were themselves instructive about some of the relevant issues. They involved all-too-common challenges of bringing together people of different views and experiences online. Ironically, some of the tensions involved some people demanding a zero-tolerance abuse-focused CoC, complaining aggressively about CoC drafts that had conflict-resolution emphasis. And yet, all the problems were themselves the sort of disagreements and misunderstandings that needed better conflict-resolution rather than problems with any abuse. Among many other insights, it highlighted for me how important it is to have a good CoC in place early. Their CoC now is an interesting compromise but needs further work and the governance and tooling remains particularly lacking. ↩︎

Thanks for your responses! This was, in fact, the informative exposition I was hoping to get.

My takeaway is the following. Like all documents, a CoC needs a beginning, middle, and end that invite the reader to follow along with the author’s thoughts. The beginning should straight up say things like, “We want this document to address interpersonal conflict, mindful of power differentials, as well as address systemic discrimination patterns.” Too many of these sorts of documents cut straight to the chase, giving detractors easy targets for misinterpretation of motive. Stating our motives will give people the power to discuss the document’s effectiveness.

2 Likes