attention-focus

Dunbar Layer Awareness

Also known as:

Understand that human intimacy operates in natural layers—5 intimate, 15 close, 50 friends, 150 acquaintances—and invest accordingly.

Understand that human intimacy operates in natural layers—5 intimate, 15 close, 50 friends, 150 acquaintances—and invest your attention and resources according to these layers, not against them.

[!NOTE] Confidence Rating: ★★★ (Established) This pattern draws on Robin Dunbar’s research into human cognitive limits and social group sizes.

Section 1: Context

Most commons-building efforts collapse not from lack of intention but from attention distributed wrongly across scale. A tech startup scales its community to 500 members and wonders why decision-making fractures. An activist cell grows from 8 to 60 people and loses the trust that held it together. A municipal government tries to govern 150,000 citizens as though they were a single, intimate community, then blames “disengagement” when people stop showing up. The system isn’t broken—the container is misaligned with human relational capacity.

Dunbar’s research reveals that human beings maintain stable relationships in discrete, nested tiers. Each tier requires different communication frequency, emotional investment, and governance attention. Most organisations ignore this structure entirely, treating all relationships as fungible. The result: energy leaks upward into administration, downward into fragmentation, sideways into parallel power structures that actually do honor these layers.

This pattern surfaces when you recognise that the system’s vitality depends not on flat, uniform connection but on nested, honest acknowledgment of where genuine relational capacity actually lives.

Section 2: Problem

The core conflict is Dunbar vs. Awareness.

Dunbar’s limits are cognitive—the human brain can maintain stable relationships with roughly 150 people, period. This number is biologically grounded; it correlates with neocortex size. Attempting to scale intimate decision-making beyond this creates cognitive overload, where the system must choose between authentic relationship or authentic participation.

Awareness presses in the opposite direction: we want to know everyone. We want to be known. We want the commons to feel inclusive, horizontal, boundless. We resist the idea that some relationships must be thinner to keep others real. We equate layers with hierarchy and recoil.

When Dunbar is ignored, the pattern breaks: governance becomes opaque (only a hidden 5–15 can actually coordinate), decision cycles lengthen (trying to include 150 in every choice), and trust erodes (members sense they’re not truly seen). The system performs the rituals of commons-building but lives in fragmentation.

When Awareness is ignored, the pattern calcifies: tight inner circles hoard power, outer layers feel excluded by design, and the commons becomes a pyramid disguised as a circle. Vitality dies from the outside in.

The tension is not resolvable through will alone. You cannot decide to maintain intimate relationship with 500 people. But you can design structures that honour both Dunbar’s limits and the human need to matter.

Section 3: Solution

Therefore, map your actual relational layers explicitly and allocate communication, decision-rights, and stewardship investment proportionally to each layer—then make this architecture transparent.

This shift from denial to honesty transforms governance itself. Instead of pretending everyone relates equally (while actually operating in hidden tiers), you make the tiers visible and design them consciously.

Dunbar’s layers are not imposed; they emerge naturally wherever humans form stable groups. What changes is intentionality. The 5–15 intimate core still exists—but instead of being a shadow power structure that operates outside formal governance, it becomes a named, accountable stewardship layer. The 50 friends still cluster—but now they’re explicitly the decision-making cohort for shared direction. The 150 acquaintances still form the boundary—but now they’re designed as the participation layer where newcomers learn, not as failures who didn’t make it “in.”

This pattern works because it stops fighting human relational capacity and starts leveraging it. The mechanism is composability: small, human-scale groups (5, 15, 50) can coordinate with clarity and speed. These groups then nest upward. Five people deeply aligned can steward a 50-person team; a 50-person team can represent and coordinate a 150-person commons. There is no longer a gap between the actual relational architecture and the formal one.

From a living systems view, this is like understanding root depth before designing a garden. You don’t make roots grow deeper through willpower; you plant species suited to the soil and water accordingly.

Section 4: Implementation

In corporate settings: Map your organisation’s actual decision-making tiers. Typically: a 5–7 person executive stewardship layer makes resource and direction choices. This layer meets weekly, with high psychological safety and real disagreement. Below it, 12–20 senior individual contributors form the next tier—they attend monthly strategy sessions and drive quarterly OKRs. Below that, 40–60 core contributors form the working teams where daily collaboration happens. Make this architecture explicit in your governance documents. Name the stewardship layer openly and explain what decisions they own vs. what they surface to the wider group. Stop pretending the org is “flat”—it isn’t, and people suffer from the lie.

In government / community scale: Design your policy bodies according to layers. A neighbourhood council of 12–15 active stewards sets direction and allocates resources. They meet fortnightly. A working group of 40–50 community members (drawn by genuine interest, not appointment) develops specific initiatives and feeds back to the stewards monthly. The wider 150–200 person community participates in quarterly assemblies where decisions are explained and feedback is taken. Crucially: the stewards are not hidden or self-appointed. They’re acknowledged as the layer that can coordinate quickly because they’ve invested the relational time to do so. Transparency replaces the resentment that comes from invisible leadership.

In activist movements: Structure cell networks according to Dunbar layers. A 5–7 person affinity group coordinates strategy and carries institutional memory. They meet weekly. This group is embedded in a 15–20 person local chapter that handles recruitment, training, and adaptation. The chapter connects to a 50-person regional network of chapters where larger strategy is debated. The wider movement (150+ people regionally) participates in action days and annual gatherings. Each layer has explicit decision rights: the affinity group decides how to execute; the chapter decides what to work on locally; the regional network decides which campaigns to prioritise. People are not frustrated by exclusion because the architecture is transparent and the pathways for advancement are clear.

In tech / social layer management: If you’re building platform governance, don’t try to simulate intimate relationship at scale. Instead, design explicit tiers: power users and stewards (5–15) set community norms and guide algorithm tuning; active moderators and contributors (40–60) manage daily conversation and surface issues; regular members (150–500) participate in forums and votes; casual participants (everyone else) consume and occasionally create. Build the tech architecture to make these layers visible. Let the 15-person stewardship layer see conversation heat maps and flag emerging issues. Let the 60-person moderator layer have real-time communication tools. Don’t obscure governance—surface it. Make it clear that the platform itself operates in Dunbar layers, and that this is by design, not accident.

In all contexts, the implementation move is the same: audit your actual relational and decision-making structure; name it; make it legible; then redesign your formal governance to match reality rather than fight it.

Section 5: Consequences

What flourishes:

Clarity around decision velocity. The 5–15 person stewardship layer can move quickly because relational trust is real. Major direction shifts don’t require consensus from 150 people; they require alignment among the 15 who have invested the relational capital to understand the full context. This doesn’t exclude others—it frees the 50 and 150 from paralysis.

Genuine inclusion without false equality. People stop resenting “hidden hierarchies” because the layers are acknowledged. A person in the 50-person layer knows they don’t attend every decision meeting—and they’re okay with it because they understand why the 15 stewards exist. They know the pathways for moving into deeper involvement if they invest more time.

Vitality renewal cycles. Each layer can refresh at its own pace. The 5 might turn over every 18 months. The 15 every 3 years. The 50 every 5 years. The 150 continuously. This prevents the decay that comes from either unchanging leadership or constant churn.

What risks emerge:

Ossification of layers. If the boundaries between tiers become rigid, the system calcifies. The 5 stop inviting new people into relationship; the 15 become a club; the 50 feel permanent exclusion. Watch for this specifically in activist and corporate contexts (both scored 3.0 on autonomy). The pattern only works if movement between layers is genuinely possible—not easy, but possible through demonstrated commitment.

Hidden power consolidation. The legitimisation of inner layers can mask real power hoarding if stewardship is not accountable. Transparency is the antidote: the 5-person layer must regularly report to the 15 on what they decided and why. The 15 must host the 50 quarterly. Without this accountability infrastructure, the pattern becomes a permission structure for invisible control.

AI-driven layer collapse. As algorithmic systems attempt to scale personal messaging and recommendation, they can create the illusion that you have intimate relationship with thousands. This weakens the felt reality of actual tiers and can hollow out the pattern. See Section 7 for detail.

Section 6: Known Uses

Robin Dunbar’s own research, beginning in the 1990s, tracked the relational structures of military units, church congregations, and online communities. He found that the most stable, high-trust military units had 150–180 people, organised internally into squads of 5–8 and platoons of 30–50. These weren’t designed by Dunbar’s theory—they evolved over centuries. When militaries tried to flatten the structure or push units larger, cohesion fractured and casualty rates rose. Dunbar’s insight was that the organisation wasn’t creating the layers; it was recognising and formalising layers that humans naturally maintain.

The Mondragon Cooperatives in the Basque region operate according to Dunbar-aware design, though they didn’t use Dunbar’s language. Each cooperative is capped at roughly 500–800 members. Within each, a 12–15 person general assembly holds core decision rights. Work teams cluster around 30–50 people. The wider membership gathers quarterly. Mondragon has maintained worker ownership, profitability, and low turnover for 70+ years—partly because they never tried to govern 2,000 people as though they were 50. When they did try to grow larger, they split the cooperative into new legal entities rather than scaling the old one. This structure choice has been crucial to their resilience.

Valve Software’s flat organisational structure, often cited as radically non-hierarchical, actually operates in hidden Dunbar layers. The company hovers around 400 people. Within that, small project teams (5–12 people) coordinate directly. The broader company self-organises into affinity networks around technology areas (30–60 people each). Leadership emerges organically, not through title, but through people who invest deeply in relationship and trust-building. When Valve grew too large and tried to flatten further, people reported confusion and fragmentation—not because flatness failed, but because they’d violated their own natural tiers and refused to acknowledge it. Recent restructuring has moved toward more explicit layers.

Section 7: Cognitive Era

Dunbar’s limits are neurobiological and unlikely to shift. But the experience of these layers has changed radically in an AI-mediated world.

Algorithmic systems now simulate intimacy at scale. A social platform can deliver personalised messages from thousands of sources, creating the false sense that you have deep relationship with many people. This is cognitively and emotionally deceptive. Your brain still only has relational bandwidth for 150; the algorithm is lying to both you and your attention.

The risk: if AI allows us to pretend we’re operating at scale without Dunbar constraints, we’ll build commons that ignore the pattern entirely. A platform might claim to enable “community governance” for 50,000 members—with an algorithm managing “participation” at each layer. But humans can’t actually feel represented by or trust a system they don’t relate to at Dunbar scale. The result is governance theatre—the form of participation without the substance of relationship.

The lever: use AI to make the layers more transparent and functional, not to obscure them. AI can help the 15-person stewardship layer identify signals from the 150 by processing volume they couldn’t handle manually. AI can route questions to the right Dunbar layer rather than creating a single feedback channel that drowns. AI can visualise the relational network and show where actual clusters exist, helping you align formal structure with lived relational reality.

The key move: treat Dunbar Layer Awareness as a constraint that AI governance must respect, not a problem that AI can solve. If your platform can’t clarify which decisions are made by which human layers, don’t add AI complexity—simplify the architecture until it’s transparent.

Section 8: Vitality

Signs of life:

  • People can name the decision-making tiers without confusion. A member at any layer can explain: “The 12-person council sets annual direction. The 45-person working group figures out how to deliver it. I’m in the working group.” No ambiguity.

  • Movement between layers happens regularly and is seen as natural. Someone moves from 50-person participant into the 15-person core because they’ve shown up consistently and deepened relationship. It takes 6–18 months, not a secret vote. Others move back out when life circumstances change, without shame.

  • Each layer meets at its natural frequency. The 5–7 meet weekly or bi-weekly. The 15 meet monthly. The 50 meet quarterly. The 150 gather annually or semi-annually. These rhythms are non-negotiable and protected.

  • Turnover and refresh happen predictably at each layer, and new people are explicitly invited and onboarded rather than left to figure it out.

Signs of decay:

  • The formal structure says “flat” or “non-hierarchical,” but people point to a hidden 3–5 person power group that actually decides things. The mismatch between stated and lived structure breeds cynicism.

  • Meetings bloat: the 15-person stewardship layer now has 40 people attending because “everyone should be included.” Decision velocity collapses. The group splits into a hidden 5 that coordinates offline.

  • Layers are sealed: new people join the outer ring but have no clear pathway inward. Tenure becomes the only criterion. The system calcifies and stops renewing.

  • The layers exist but aren’t named or discussed openly. People sense the tiers but experience them as exclusion rather than design. Resentment accumulates.

When to replant:

Restart this practice when you notice the discrepancy between stated and actual relational structure has become visible. This is the signal that you’ve grown past your last architecture and are ready to redesign honestly. The best moment is when you’re still functional but beginning to feel the friction—before collapse.

Redesign is also needed if you’ve been operating the layers for 3–4 years without refresh. The original 5 may be tired. The 15 may have calcified. Explicitly sunset one layer and invite fresh cohorts. Vitality depends on renewal, not permanence.