# Ownership

> **Living document**, continuously evolving.
> First version 2025-08-26. Published publicly 2026-05-20. Last update 2026-05-10.
>
> Part of [Thoughtful Social AI](../README.md). License: CC-BY-SA 4.0.
> [View raw markdown](https://raw.githubusercontent.com/joostliebregts/thoughtful-social-ai/main/en/OwnershipPrinciples.md) for copy-paste into your AI.

> **What this document mentions.** As you read, you'll come across a few specific people and projects that are central to the author's practice. The principles and techniques are generally applicable; these concrete references serve as context, not as a requirement.
>
> - **Floor de Ruiter**: facilitator and thinker on bottom-up change ([bottom-upchange.nl](https://bottom-upchange.nl))
> - **the Doesburg engagement**: a bottom-up community engagement around a caring community; these ownership principles were formalized there (autumn 2025)

Formalized in the Doesburg engagement (a bottom-up community engagement around a caring community), autumn 2025.

## Core philosophy

**Ownership comes from within.** It is not something we assign, but something we recognize and respect in the words and actions of participants.

## Design principle: ownership is local or it isn't

Ownership only grows in concrete, close-by situations where people can experience it, not in abstract guiding frameworks. Mastery experiences (Bandura) beat pep talks and vision decks. Locus of control shifts through doing, not through understanding.

What this means for session design and project architecture:

- **Scale down to where it's tangible.** Better three neighbors over one street-level issue than thirty over "the caring neighborhood".
- **Places that are already in motion > new meeting cycles.** A conversation at the coffee machine or in the church hall stands a better chance of felt ownership than a steering group every three weeks.
- **Wait until the body can come along.** Cognitive understanding without felt safety is mastery that doesn't land. Pace is not a soft precondition, it is the condition for the principles to work at all.
- **The facilitator is a matchmaker, not a director.** We connect people who need each other in a local context, and step back once the conversation starts.
- **Don't tell, let them experience.** Floor de Ruiter: "You shouldn't bother people with your principles." Behavior changes first, the brain follows. Values change based on experiences, not based on explanation (Marquet, *Turn The Ship Around*).

This is also the mechanism beneath bottom-up work: self-efficacy is local or it isn't, and invitation is the only form that doesn't produce pseudo-participation. What makes an individual grow is what makes a group grow, the same stuff.

## Origin Story: AI-Assisted Heuristic Development

**How These Principles Were Formalized** (from a conversation about the origins of the work in the Doesburg engagement, October 2025):

> "I was working on this system five, six weeks ago on trying to get sort of this little agent factory going on analyzing meetings, meeting transcripts to look at: Are the feelings of ownership rising? [...] And I don't know, Claude, or I think it was Claude, it came back to me with like this whole markdown file of ownership heuristics. And I looked at it and I'm like, this makes a lot of sense. And it looked at language like... High ownership is like, yes, I will do that. Or, yeah, I'll take a look at this. I'll do this now. And then there's all these in-between stuff."

**The Process**:
1. **Tacit Expertise**: Years of facilitation work developed implicit ability to recognize ownership signals
2. **AI Formalization Request**: Asked Claude to analyze transcripts for ownership patterns
3. **AI Generation**: Claude produced structured markdown of ownership heuristics based on language patterns
4. **Human Validation**: The author verified heuristics against lived facilitation experience ("this makes a lot of sense")
5. **Integration**: Formalized heuristics became explicit scoring methodology

**Key Insight**: Sometimes AI can help articulate tacit expertise that humans struggle to formalize explicitly. The AI didn't create new knowledge, it structured and made explicit what the author already knew implicitly from practice.

**Validation Methodology**:
- **Test against real transcripts**: Do the heuristics match actual ownership you observe?
- **Compare with facilitation intuition**: Does the AI score align with your gut feeling?
- **Iterate with feedback**: Refine heuristics based on cases where AI and human judgment diverge
- **Trust but verify**: AI formalization is useful but requires human expertise validation

**This is NOT**: AI replacing human judgment about ownership
**This IS**: AI helping humans articulate and systematize their existing expertise

**Meta-Learning**: This collaboration model (AI formalizes, human validates) may be valuable for other tacit expertise domains: facilitation skills, pattern recognition, contextual judgment.

## Language principles

### 1. Preserving the authentic voice
- **Use their own words** in summaries and analyses
- **Quote verbatim** where possible to ground ownership
- **Avoid paraphrasing** when the original wording is stronger
- **Respect their framing** of problems and solutions

### 2. Respecting context
- **Consent-based naming**:
  - `consent=true` → use real names for ownership
  - `consent=false` → use pseudonyms, but preserve their language
- **Cultural context**: Respect for local terminology
- **Professional context**: Preserve professional jargon when it strengthens their ownership

## Recognizing ownership

### What DOES Show Ownership

**0.7-1.0: High Ownership**
```
"I'm going to do something about that": direct intention to act
"We need to approach this differently": collective ownership
"I'll try that next week": concrete planning
"I've already reached out to...": already active
```

**Characteristics:**
- Uses "I", "we" for solutions
- Concrete action plans
- Takes responsibility for outcomes
- Sees connections and possibilities

### What Shows LIMITED Ownership

**0.4-0.6: Mixed Ownership**
```
"It should but...": external dependency
"If there were budget then...": conditional action
"It's actually not my job, but...": ownership despite the system
"I do try, but the system...": effort despite obstacles
```

**Characteristics:**
- Frustration with the system, but still looking for solutions
- Conditional willingness to act
- Acknowledges influence, but feels limited
- Tries despite obstacles

### What Shows LOW Ownership

**0.0-0.3: Victim Position**
```
"There's nothing I can do about that": powerlessness
"The system decides...": external control
"They have to solve that": responsibility elsewhere
"It is what it is": resignation
```

**Characteristics:**
- Uses "they", "the system", "the organization" for solutions
- Feels no influence over outcomes
- Passive stance toward problems
- External locus of control

## Scoring methodology

### 1. Quote-Based Evidence
For every ownership score:

**Required:**
- At least 2 supporting quotes
- Exact timestamp or `inference:true`
- Original wording preserved

**Template:**
```markdown
## Ownership (ownership pulse)
**Score:** 0.X

**Why:** [rationale in their words]

**Supporting signals:**
> "Quote 1 that shows ownership" ([NAME], timestamp)
> "Quote 2 that demonstrates agency/action" ([NAME], timestamp)

**Counter-signals:**
> "Quote that shows limitation/frustration" ([NAME], timestamp)
> [Observation about external dependencies]

**Balance:** [Why this score despite counter-signals]
```

### 2. Recording Counter-Evidence
**Always record:**
- Quotes that suggest lower ownership
- System constraints that limit agency
- External dependencies
- Moments of resignation or powerlessness

**Why:** A complete picture of the situation, not just positive signals

### 3. Evolution Over Time
Track ownership evolution through:
- **Longitudinal quotes**: Compare statements over time
- **Action follow-up**: What did they do after the previous conversation?
- **Network effects**: How does collaboration influence ownership?
- **System changes**: Impact of external changes

## In Practice: Examples

### Good: Authentic Ownership Recognition
```markdown
**Anna's Ownership (0.8):**
"I'm going to do something about that, because this can't go on like this," shows her direct
intention to act despite bureaucratic obstacles. She consistently uses
"I'm going to" and "I'll arrange," which demonstrates strong personal ownership.

Supporting: "I've already reached out to the municipality" (00:23:15)
Counter-signal: "But the system just won't cooperate" (00:25:30)
```

### Wrong: External Interpretation
```markdown
❌ "Anna shows high ownership through her proactive attitude"
✅ "Anna's statements 'I'm going to do something about that' show direct ownership"
```

### Good: Nuanced Scoring
```markdown
**Barbara's Ownership (0.6):**
Mixed pattern: strong intention to care but feels limited by the system.
"I really want to help but the rules..." shows ownership despite
obstacles, not alongside them. Her "I do try" attitude gives a 0.6
instead of a lower score.
```

## Ownership in Experiments

### Experiment Design Principles
1. **Within their sphere of influence**: What can THEY influence?
2. **Their own words**: Experiment title in their terminology
3. **Their motivation**: Why would THEY want to try this?
4. **Their success metrics**: How do THEY define success?

### Example:
```json
{
  "hypothesis": "Anna's 'neighborhood care café' idea reduces loneliness",
  "small_bet": "Anna starts a pilot as she proposes: a weekly coffee hour",
  "owner": "Anna (community nurse)",
  "success_metric": "Anna's own definition: 'When people start recognizing each other'"
}
```

## Privacy & Ownership

### When Consent=True
- **Use the real name** for ownership attribution
- **Preserve their specificity**: "Anna's approach" vs "The community nurse's approach"
- **Direct quotes**: Verbatim citations as ownership evidence

### When Consent=False
- **Use a pseudonym**: But preserve their language fully
- **Role-specificity**: "(community nurse)" for context
- **Their own words**: Preserve their wording despite the pseudonym

## Validation & Feedback

### For Participants
**Recognition test:** "Do you recognize yourself in this description of your ownership?"
- Can they find their own words back?
- Does the ownership score feel accurate?
- Are they missing important aspects?

### For the Team
**Consistency check:**
- Similar situations → similar scores?
- Own biases → are they influencing interpretation?
- System attributions → at the expense of personal agency?

---

*"It's not about what we think they can do, but about what they say they're going to do."*

## Changelog

- **2025-08-26**: first version formalized from the Doesburg engagement.
- **2026-05-10**: design principle "ownership is local or it isn't" added, linking to bottom-up work (Bandura, Marquet, Floor de Ruiter).
- **2026-05-20**: publicly shareable version in the Thoughtful Social AI bundle.

---

## Further reading in this bundle

- [Principles](./SOCIAL-AI-PRINCIPES.md): 15 principles for AI with group work
- [Prompts](./PROMPT-BEST-PRACTICES.md): concrete translation into prompt design
- [Bottom-up](./BottomUpPrinciples.md): how change emerges bottom-up
- [Reading guide](../README.md): overview and how these docs fit together

---

*Part of [Thoughtful Social AI](../README.md). CC-BY-SA 4.0.*