Reason Commons
Where Does Our Thinking Go?
Trees of change, forum discussions, and getting Second Renaissance to constantly evolve its vision.
Rufus Pollock & David Joseph · Second Renaissance Research Group
Title
Hold this while the room settles. Say nothing about the software yet.
Visual: ghosted issue-tree behind the type, drawn rather than photographed. If an out-of-focus canopy photo turns up in the image pass it can replace the SVG — keep the opacity at ~0.16 either way, the type has to stay the figure.
Keys → next build · ← back · ↑↓ slide · G overview · N notes · F fullscreen · T timer
“What is the goal of the Second Renaissance, and how would we know we were moving toward it?”
— a research group session, earlier this year
It is still there, in a Zoom recording somewhere.
The question that started it
Three clicks to build the question. Hold. Fourth click drops it to 20% and leaves it there — do not narrate the fade, let the room watch it happen. It comes back at full opacity in the close.
“It was a good question. It got a good hour. People said things I still think about.”
A week later:
nowhere.
- No note.
- No thread.
- The thinking stayed in the room.
And people leave rooms.
A week later
“nowhere” lands first and lands hard. Then the three short lines underneath, one at a time, quietly.
“I don’t think that was unusual. I think it is close to the normal outcome here.”
Where does our thinking go?
Spine line
Cut to black. Let it sit for three full seconds before you speak. This exact design returns once, at the end, answered.
“So that is what this talk is about. We have been trying to answer it in a practical way, and it turned into a project: Reason Commons.”
One scope note, before we start.
- Cohering our thinking — our theory of change. The tree of change, today.
- Coordinating our action — the tree of action. A talk of its own.
A tree of action builds on a tree of change — how you should act follows from how you think you should.
Two halves
Said once, plainly, right after the cold open and before the problem lands. There are two halves to all of this: cohering our thinking, and coordinating our action. This whole talk is the first half. The second, how work gets organised and tracked, is a talk of its own. Today is about the thinking.
The tree of change is itself partly a thesis about what to do, so the second idea is present implicitly — we are just choosing to focus on the tree of change and how it gets updated as people talk. The closing line makes that dependency explicit: the tree of action would build on the tree of change, not stand apart from it.
1
The problem
A movement whose thinking doesn’t accumulate.
A movement for civilizational renewal.
- A living theory of change, meant to get better as more people work on it.
- That only works if the improving adds up.
What the movement is trying to be
One or two sentences on the 2R thesis, no more — this room knows it cold. The load-bearing line is the third: that only works if the improving adds up.
The two-halves scope note (cohering our thinking vs. coordinating our action) already got its own slide, right before this act opened — nothing to say about it here.
- Same conclusion. Twice in three years.
- Both times the room agreed.
- Both times it evaporated.
The research circle
“A research circle meets every week. Twice in three years it has worked its way to the same conclusion. Both times the agreement had nobody’s name on it and nothing downstream had changed.”
On the third line the first two dim. The people in that room are not the problem — say that out loud. Individual understanding compounds; the group’s version doesn’t.
Agreement that survives only in the people who were present is a rumour. the one dark slide in this stretch
The line
Say it once, at normal volume, and move on. It does not need help.
Not a shortage of intelligence. Or goodwill. Or ideas.
A shortage of somewhere to put them.
Not a shortage of
One click. The first sentence leaves and the second arrives in the same place — do not let them coexist.
“We are not short of ideas.”
Six or seven white papers. ~100,000 words.
A claim in the middle of page 40 has no address.
No address
“A claim buried in the middle of a paragraph on page forty has no address. You cannot point at it. If you cannot point at it, you cannot disagree with it precisely — so in the end nobody does.”
The sentence ringed here is the 2R goal statement itself. It is the sentence the tree in the next act is built on, and in the prose it has nowhere to live.
“The most valuable thing a new person brings, a precise disagreement, is the thing a long document is least able to receive.”
Asset: a real screenshot of a dense 2R white-paper page can drop straight in behind the ring. The drawn version reads correctly from the back of a room, which a screenshot of body text may not.
2
The idea
The tree of change on screen. Every claim has an address. The links are claims too.
What if what we share isn’t a stack of documents, but a structure.
The idea
Two clicks. The stack recedes rather than vanishing — the documents still exist, they just stop being the thing held in common.
“What is on the screen is not just the conclusions. It is the reasoning.”
- A top claim. The sub-claims it depends on. All the way down.
- Every claim has an address.
- The links are claims too: this one is necessary for that one.
Every claim has an address
The tree draws itself rank by rank as you speak. Click 5 is the one that matters: the address tag. Click 6 glosses a link and click 7 says what that buys.
“The lines between claims are themselves claims: this one is necessary for that one.”
Demo A · pre-baked, narrated live
The Second Renaissance tree of change
Demo A — the real tree
Two things and stop. Ninety seconds. Do not tour the tree.
Click 2 — the deep leaf under condition 4, “pockets of coherent new culture”. Its path to the goal lights, everything else recedes. Say: this claim is not floating in a document, it is at an address, and the address tells you what it is holding up.
Click 4 — one link. “The links are claims too, so you can disagree with a link without touching the boxes.” Click 5 shows the assumption that link hangs on: shared action requires a shared worldview. Note it and move on — somebody in the thread breaks it in Act 4, and it lands much harder if the room has already seen it.
“I did not build it for the talk. It is generated from the actual written corpus.” The eight conditions are the goal tree’s own critical success factors; the claims under condition 4 are the tree as it currently stands, which is to say provisional.
Live alternative: screen-share the tool and make the same two clicks. This drawing is the fallback, and it is also the safer choice from the back of a large room.
- A thesis about what has to be true.
- An object you can improve, one claim at a time.
Two things at once
Click 4 turns one leaf dashed. Dashed is the whole grammar in one mark: offered, not yet relied upon. The tree did not have to be reopened to receive it.
Ancestors of the Ideas
Eliyahu Goldratt’s Theory of Constraints diagnoses an organisation with literal cause-and-effect trees. Barbara Minto’s Pyramid Principle structures an argument the same way — one governing thought, branching into what supports it. Dettmer’s Logical Thinking Process, thirty years later, develops and summarises Goldratt’s own trees. Nothing here depends on knowing any of it.
They both ended up drawing trees.
Goldratt
Goldratt & Cox · 1984
Minto · 1987
Dettmer · 1997
Ancestors
One word for a title. The joke lands quietly under the paragraph rather than as the headline — “they both ended up drawing trees” is the wink, not the point.
“There is a separate talk that does it justice. Nothing here depends on knowing any of it.” If someone in the room lights up at Theory of Constraints, take it in the break, not now.
Goldratt’s photo is real (Wikipedia, low-res press photo, non-free/fair-use on that article rather than freely-licensed — same “internal talk” allowance as the book covers). Minto has no photo by request — her book stands in. Four images, one size, so the set reads as one family: two people, two of their books, one further book (Dettmer) that merges the two.
3
The complication
Structure creates friction. One dissolvable assumption is what turns that into a dilemma.
So why don’t we already work like this?
Why don’t we already do this?
A beat slide. Ask it, and wait long enough that somebody nearly answers.
The dilemma
Left, then right, then the clash. Give both sides their due — people are right to be wary of anything that interrupts a good exchange to make it tidy.
This is the Evaporating Cloud, and you are not going to say so.
One assumption holds the dilemma together
The hidden assumption
The two columns collapse inward and hang off this one sentence. Read it slowly. Do not editorialise yet — the next slide is the pivot and it needs the room to have taken the assumption seriously first.
The structure has to be put there by a person, at the moment of contributing.
Not any more.
AI can do the filing now — and that is what makes a Reason Commons possible.
The turn
Cut to black on entry. The assumption is struck through, then it goes, then “Not any more.” Let it sit before you say the small line: the assumption held for decades — AI is what changed it, and that is what makes a Reason Commons possible.
This is the pivot of the talk. If the room is going to lean in anywhere, it is here.
4
The possibility
Contribute normally. A machine recovers the structure afterwards. AI is what makes that affordable.
- The structure doesn’t have to be put there by a person, at the moment of contributing.
- So let the conversation be a conversation.
The possibility
“You post your paragraph on the forum the way you were going to anyway. Nothing interrupts you.”
These are the actual posts. Same thread we are about to read.
Afterwards, separately, a machine reads it against the tree.
Structure recovered afterwards
The lines are dashed and accent-coloured, and that is not decoration: nothing has been connected yet. These are proposals about where things might go.
Demo B · pre-baked, narrated live
A real forum thread — Problems and Possible Responses Within Our Influence
Demo B — watch it read a thread
“Nobody in that thread did anything differently. The structure was recovered afterwards, by something whose whole job is to do the filing.”
Card 1 (click 2) — the easy one. Asimong’s small-tents sentence supports the leaf we looked at in Demo A. Point out that the target is 4.2.1, the same address from ninety seconds ago.
Card 2 (click 3) — the good one. He is not challenging a box, he is challenging a link, and specifically the assumption hanging off it. Watch the assumption strike through on the right as the card lands. “You can disagree with a link without touching either box” stops being an abstraction here.
Card 3 (click 4) — Robert challenges the goal itself. Note that it is not struck through: it is a live disagreement, not a settled defeat, and the record keeps it that way.
Card 4 (click 5) — the honest one. The tree cannot place glennr86’s post at all. Click 6 says why that matters: it is the best evidence in the thread that the diagnosis is right. Twenty posts of good thinking produced a thread the twentieth person could not enter.
Click 7 — the count. Four proposals, nothing merged. Say “nothing merged” out loud; the next slide is about who does that.
Consent: check with Asimong, Robert Bunge and glennr86 before the talk. Posts 6, 7, 8 and 19 are public, but being quoted on a slide is different, and the whole argument depends on people’s contributions being handled well.
- A person, or a group of people, with AI help if needed, decide what merges.
- Not one person. Not the machine alone.
It’s pull requests for discourse.
Who decides
“Not one person, and not the machine on its own, because one party quietly deciding what the map says is how you get a map nobody trusts.”
The model is drawn in the structure hue and has none of the reasoning colours. That is enforced, not stylistic — the machine and the reasoning are different objects, and an AI can help a team think without getting to decide what the team believes.
If someone asks about capture or governance, this is the slide to stay on.
- Cost of contributing: one sentence, where you were already saying it.
- The filing is someone’s job now. Not the contributor’s.
The cost of contributing
Six things on the tall bar, and every one of them is a real reason somebody has already not contributed to something.
Be careful not to overclaim: the filing did not vanish. It moved.
AI is what makes this possible.
Two years ago this ended at “too much hand work.”
Why now
This is a load-bearing slide, not an aside — say the title plainly and let it land. AI is what makes this possible is the whole reason the assumption from the turn no longer holds.
The right-hand note is the honest half: what got cheap is the filing. Judgment did not.
5
You try it
The room writes. One or two taken through fit → gap or disagreement → so what. Then the reverse view.
Your turn · ~1,000 characters
Something important the Second Renaissance may be missing, underestimating, or getting wrong.
Or something we could act on.
Your words. Don’t tidy them.
4:00press T to start · T again to pause
Demo C — write
Leave this up for three or four minutes of genuine quiet. Press T to start the timer. Do not fill the silence.
Step 2, before anything is shown: “What you are about to watch is a prototype. The trees are provisional and wrong in places, the reader is early, and the rules for what merges are about three sentences long and have never been tested by anyone genuinely annoyed. That is what today is for.” The caveat lands here, not in the close.
The QR code and short link were cut — both were placeholders (a blank code, and “rc.li/2r”, which points nowhere) and read as confusing rather than useful. Kit still needed before the day: shared form, a real short link, QR, screen operator. Add them back here once the form exists.
Always in this order
- 1 Where does it fit?
- 2 Gap or disagreement?
- 3 So what, and who owns it?
Question one is the easy part, and it is the only part a machine does well. Two and three are the group’s work.
The three questions
This slide stays up while you walk one or two contributions through. Each number lights as you reach it.
1 — where does it fit? The tree proposes a location. Easy part.
2 — gap or disagreement? Is it under a condition nobody is holding, or in tension with a link? Now it is doing work on the group’s reasoning.
3 — so what? If the room broadly agrees: what is the next action, and who might own it?
If the placement is weak, say so, place it by hand, and move on. A wrong first guess that a person corrects in five seconds is still the system working.
The same map tells everyone which condition nobody is currently holding.
The reverse view
“The same structure that tells one person where their idea goes tells everyone which part of the goal nobody is holding.”
Condition 8 is marked by absence as much as by colour: the seven around it fill up with ticks and it does not. Say the number out loud.
Then click 4, which is the important half: this is not an accusation. The redesign brief calls for competence at this moment, not alarm — and the boundary of the alarm is drawn in words and in a count.
Replace the counts with the real figures before the talk. They carry a date on the slide on purpose: a number that arrives without its origin is the exact failure mode this whole architecture exists to refuse.
- A paragraph of ordinary thinking → a specific claim, at a place, with a path to an action.
- A long way past “thanks for your feedback.”
What just happened
Build it one box at a time as you narrate. Click 8 is the loop, and the loop is the argument: without the return arrow this is a filing system, and with it it is a theory of change.
“It is also, more or less, that forty-year-old method from earlier, run without any of its vocabulary.”
One more thing this buys you — not built yet
- So far: the forum tells the tree what’s happening.
- It can go the other way. The tree can tell the forum what already happened.
- A new post touching a claim gets a quiet tag: “related — see the 2023 and 2026 threads on this.”
- The research circle wouldn’t have needed a third pass.
It works the other way, too
Say plainly that this is roadmap, not shipped — same honesty as the Demo C prototype caveat. The diagram is a schematic in the deck’s own drawing grammar, not a mock screenshot, so it can’t be mistaken for a built feature.
The payoff line is a direct callback to “The research circle” (Act 1): the same conclusion reached twice in three years because nobody knew it had already happened. This is the fix for that specific failure, not a generic AI-magic line — hence dating the two prior posts 2023 and 2026, matching that story exactly.
If the diagram doesn’t earn its place once it’s on screen, cut it back to text-only (steps 1–4) — it was optional from the start.
6
Close
Back to the goal question. Where the thinking goes.
“How would we know if we were moving toward what we’re trying to do?”
You can only ask that of something you can point at.
Back to the question
The question comes back at full opacity, in the same face it evaporated in. Do not remark on that. The room will feel it.
“A tree of change, you can point at. A hundred thousand words, you can’t.”
“Goal.” “Progress.” “Moving toward.”
Loaded words here, and that’s fine.
- Some of you think a renewal is something that emerges — not something you plan toward.
- We’d agree. But skilful gardeners plan too, and use reason to tend what grows.
Loaded words
“I know some of you think a renewal is not the kind of thing you plan toward — that it emerges. We’d agree.” Name the discomfort; do not route around it.
“But skilful gardeners and permaculturists still plan, and still use reason to improve what they’re doing. That is what we are doing here: cultivating a commons for reason, a place for better sense-making to emerge.”
This replaces the old Robert-Bunge-challenge visual, which argued a different point (a specific disagreement stands unresolved with a name on it) that no longer matches this beat now the script leads with the emergence/planning tension instead. A visual for the gardening reframe is still open — type alone carries it for now.
Where does our thinking go?
Into the tree. At an address. Using AI.
Spine line, answered
Same slide as the introduction, answered. Hold the question alone for a beat before the second line arrives.