Facilitating a Polis Conversation: Notes from the Field
(Begrudgingly co-authored from my messy personal knowledge-base notes using ChatGPT. But pretty unapologetic about that, because my ADHD brain means final outputs languish in my notes and personal conversations for far too long without using AI for the last mile…!)
Over the years of helping people host conversations in Polis, I’ve found that the art of facilitation is less about control and more about composition. A good Polis conversation isn’t designed like a survey. It’s chalked like a playing field.
1. Chalking the lines on the pitch
Think of seed statements as chalk lines on the field.
If you want people to feel permission to speak informally, or to be a bit silly, or even troll a little, then you need to seed those corners of the field too. The boundaries you draw are the invitations you make.
You don’t need to represent every possible position in your seed set — in fact, it’s often better to leave holes that beg to be filled. People who submit one statement tend to submit more. They just need that first spark of permission.
2. Start from “I feel…”
When writing or curating seed statements, try starting with “I feel,” “I believe,” or “I think that…”.
This isn’t fluff. It primes participants to understand that emotions and subjectivities are legitimate input. People can ground their responses in their own sensibilities rather than in presumed expertise.
“Sometimes people know clearly how they feel, but their collectivist brain freezes when asked what we should do. ‘I feel…’ gives permission to shoot from the hip.”
That permission lowers the barrier to entry and increases participation. It’s one of the simplest, most powerful facilitator hacks I know.
3. Moderation without sterilization
Moderating too much can make a Polis conversation sterile. A bit of off-topic energy keeps minds fresh.
If everything is “on-topic,” the session starts to feel like a policy document rather than a public exploration. Allowing some weirdness — playful or tangential statements — helps participants maintain curiosity and agency.
The facilitator’s job isn’t to prune but to tend.
4. The crowd finds alignment; the facilitator divides
There’s a useful tension between what the crowd and the facilitator are doing.
The facilitator is trying to divide and articulate; the crowd is trying to find alignment.
Those are not opposites. They’re phases of the same process. Facilitators expand the space of articulation — decomposing and clarifying — while the crowd, through voting, explores where alignment unexpectedly appears.
Sometimes polarization is just unarticulated nuance.
5. Acceptance, seen through fractal boundaries
I once tried to build facilitation around the idea of “acceptance,” and realized it’s too compound to stand alone.
Acceptance is what you get and what you give. It’s what your in-group wins — your non-negotiables — and what you can let go of.
The danger isn’t that we fail to compromise. It’s that we fail to even see the full resolution of the fractal boundaries — the boundaries over which compromise is necessary — and that we instead assume all out-groups to be simplistic, uncompromising circular fortresses.
“Acceptance,” then, isn’t a virtue in itself. It’s a process unfolding in a landscape. You can accept something and simply label it as unjust, or you can accept it with full awareness of what’s been gained and lost. The latter form is deeply empowering. The former is borderline apathy.
(Thoughts on “acceptance” articulated in conversation with Stefan Roch of Bertelsmann Stiftung.)
6. Navigating between common ground and concreteness
Abstract statements can be great for finding conceptual overlap, but the real work of surprise often happens at the concrete level.
There’s a tension worth holding between:
Common ground: abstractions that help people recognize shared feeling.
Concreteness: specificity that allows testing of real-world scenarios.
Surprise often emerges right at the boundary — when a concrete scenario reveals a shared value no one expected to agree on.
7. The “Pass” button and the right to abstain
Polis’s “Pass/Unsure” button is one of the most misunderstood features.
It’s not indecision — it’s an escape hatch for what can’t be resolved to a clear signal. Sometimes a statement is a mix of good and bad, or rests on a premise you reject, or uses language too strong for your sensibilities. In those cases, “Pass” is a legitimate and important choice. It allows you to react in such a way that you neither walk toward the idea, nor away from it, but instead observe and remain intentionally neutral to it, like a unjudged thought in meditation. (These metaphors of attraction/repulsion/neutrality are grounded in the fundamental mathematical operations of the underlying system.)
Following a “Pass”, you then have a job: to find the better words. By doing the work to find what you can “full heart agree” or “full heart disagree” with, you put it into the system for others to react to. This is what makes the simple agree/disagree/pass format such a generative constraint — it drives the nuance out of simplest part of the system, and into the versatile language part. When the reductive “agree/disagree” options fail, we bring the task into the complexity of language itself. You can then write your own corrected version, one you can take a firm position on, or simply break the complex statement into a few simpler ones. This is how the collective statement pool becomes more nuanced — how the map of our shared emotional landscape gains resolution. Things that hold true for many with percolate up into the collective map, for all to see as common knowledge.
(Reflections on the nature of the “pass” button came from Colin Megill and a discussion on Hacker News, where I wrote about the many valid meanings of “Pass.”)
8. Truth, consensus, and direction
People often ask: Does Polis find the truth?
No. But it can find enough agreement to move together.
Consensus here isn’t epistemic; it’s pragmatic. The value isn’t in uncovering what’s objectively true, but in surfacing shared sentiment sturdy enough to stand on — to take a step forward.
Truth can shift under our feet. What matters is the collective commitment to keep moving.
“Unlike the hyper-division fostered by social media, Polis helps build stable common ground to stand on.”
That stability makes longer-form deliberation possible.
9. Trolls, edges, and intuition
Trolls don’t disappear in Polis, but they become legible.
In a room, we instinctively know who’s trolling by body language — who people look toward or away from, who gets ignored. In Polis, the visualization recreates that intuition spatially. Outlier clusters appear far from the main conversation, and the interface doesn’t invite further engagement with them.
They don’t disappear — they just lose gravity.
10. Polis as sacred infrastructure
There’s a real risk that tools like Polis could be misused — to harvest sentiment, shape media campaigns, or manipulate populations.
That risk is exactly why these systems should be governed as public infrastructure. Run openly by trusted stewards — governments, journalists, cooperatives. Protected by cryptographic and data-trust frameworks.
These conversations aren’t just another civic app. They’re experiments in self-legibility — ways for a society to see itself clearly.
“The process of running these processes must eventually be comparably sacred to those of running elections or parliaments.”
If democracy is the operating system, Polis is one of the debuggers. It deserves the same care we give our constitutions.
Closing
Facilitating Polis is both technical and emotional work. You’re tending an ecosystem where feelings become data and data becomes empathy.
Chalk the field generously.
Leave space for surprise.
Protect the process like it matters — because it does.
I particularly liked the part you say about people writing “I feel”. That’s really useful. I agree that in an ideal world tools like Polis need to be run by something like the UN, but until such time, my experience- though I wish it was not so - is that in order to deploy tech at scale, capital is needed. To get capital, sadly such an organisation needs to play the tech game. If there’s another way that helps such altruistic tech to spread, I’m unaware of it and would gratefully stand corrected. Good stufff.