An institution can get better at what it does while leaving people less able to change what it does. That is the pattern this series is about. It is not a story about bad intentions. It is a story about how ordinary success can reorganize the conditions under which future choices are made.
The central distinction is between functional improvement and structural progress. Functional improvement makes an existing arrangement work better: faster, cheaper, smoother, more reliable. Structural progress changes the distribution of capabilities, authority, and viable alternatives through which people can determine what happens next. The two can reinforce each other. They are not the same thing.
An institution’s success should be judged not only by what it enables people to accomplish, but also by what it enables them to question, govern, revise, and eventually do without.
Six patterns
Each pattern names a different way an apparent gain can coexist with a limit on human agency. The professional case is Success Is Not Independence.
| Pattern | What conventional success can overlook |
|---|---|
| Improvement vs. progress | A better-functioning system can leave its underlying constraints intact. |
| Assistance vs. authority | Receiving benefits does not establish power over their terms. |
| Transparency vs. correction | Knowing what is wrong does not establish the ability to change it. |
| Innovation vs. dependency | A useful solution can gradually become a compulsory intermediary. |
| Formal vs. effective accountability | The existence of rights does not guarantee the resources to exercise them. |
| Shallow vs. recursive corrigibility | Correcting decisions can preserve the rules and power structures producing them. |
1. Improvement vs. progress
A system can become easier to navigate while remaining just as difficult to govern. The left side of the diagram shows the ordinary case: one route, optimized, with the alternatives already foreclosed. The right side shows what structural progress would add: routes to appeal, to redesign the arrangement, and to leave it.
View source
%% title: Improvement vs. progress
%% caption: A smoother path is not the same as more paths. The dashed branches on the left are the alternatives the arrangement forecloses.
flowchart LR
subgraph Improvement["Improvement"]
P1((person)) --> I1[institution]
I1 -.-> F1[foreclosed]
I1 -.-> F2[foreclosed]
I1 -.-> F3[foreclosed]
end
subgraph Progress["Progress"]
P2((person)) --> I2[institution]
P2 --> A[appeal]
P2 ==> R[redesign]
P2 --> E[exit]
R ==> I2
end
classDef ghost stroke-dasharray:4 5,opacity:0.35;
class F1,F2,F3 ghost
The difference is not the quality of the corridor. It is what remains possible after the person has walked through it.
2. Assistance vs. authority
A provider can deliver real benefits and still leave the people receiving them with no say over the terms. The left side shows a one-way service relationship with a nominal return path. The right side shows people not only receiving benefits but sharing in setting, revising, and governing the system.
View source
%% title: Assistance vs. authority
%% caption: The left side delivers benefits with a nominal return path. The right side gives people governing power over the terms.
flowchart TB
subgraph Assistance["Assistance"]
I1[institution] --> P1((person))
I1 --> P2((person))
I1 --> P3((person))
P1 -.-> I1
end
subgraph Authority["Authority"]
I2[institution]
P4((person)) <--> P5((person))
P4 ==> I2
P5 ==> I2
end
Being served is not self-government. A service can be excellent and still be something that happens to people rather than something they govern.
3. Transparency vs. correction
Visibility is not control. An institution can open its machinery to inspection and still leave no lever connected to the machinery. The left side shows a glass box: the gears are visible, the people can see them, and the path from seeing to changing is blocked. The right side connects what people see to what they can change.
View source
%% title: Transparency vs. correction
%% caption: Visibility without a lever is not accountability. The right side connects what people see to what they can change.
flowchart TB
subgraph Visible["Transparency without correction"]
P1((people)) -.-> B1[glass box]
B1 --- G1[gears]
B1 -.- X[no lever]
end
subgraph Corrective["Transparency with correction"]
P2((people)) --> C[control panel]
C ==> B2[glass box]
B2 --- G2[gears]
end
A complaints office, an audit, or an apology can absorb friction without touching the rule that produced it. The test is whether the challenge can reach the category, the budget, or the authority behind the decision.
4. Useful innovation vs. dependency
A solution can improve lives while making alternatives harder to build. The left side shows a helpful service becoming the hub for more and more activities, with the surrounding nodes returning to it because there is nowhere else to go. The right side shows a useful solution that stays modular and interoperable, keeping multiple paths viable.
View source
%% title: Useful innovation vs. dependency
%% caption: A helpful solution can become the hub for more and more activity, or it can stay modular and interoperable.
flowchart LR
subgraph Lockin["Innovation → lock-in"]
H[hub]
H --> A1[people]
H --> A2[data]
H --> A3[workflow]
H --> A4[standards]
A1 -.-> H
A2 -.-> H
end
subgraph Options["Innovation → option space"]
N1[module] --- N2[module]
N2 --- N3[module]
N1 --- N3
end
The helpful bridge that becomes the only bridge is still a bridge. The question is whether the next generation can build another one.
5. Formal vs. effective accountability
The routes can exist and still be unusable. The left side shows the formal channels — complain, appeal, vote, exit — each meeting a practical barrier: time, cost, expertise, risk. The right side shows the same routes with the support that makes them traversable: legal aid, organization, data access, an enforceable remedy.
View source
%% title: Accountability on paper vs. in practice
%% caption: The routes are the same. The left side meets barriers of time, cost, expertise, and risk; the right side has the support that makes the routes traversable.
flowchart LR
subgraph Paper["Formal accountability"]
P1((person)) --> C1[complain] --> B1[time]
P1 --> C2[appeal] --> B2[cost]
P1 --> C3[vote] --> B3[expertise]
P1 --> C4[exit] --> B4[risk]
end
subgraph Practice["Effective accountability"]
P2((person)) --> D1[complain] --> S1[legal aid]
P2 --> D2[appeal] --> S2[organization]
P2 --> D3[vote] --> S3[data access]
P2 --> D4[exit] --> S4[enforceable remedy]
end
A map full of roads is not a map of passable roads. The difference between a right and an effective right is the cost of using it.
6. Shallow vs. recursive corrigibility
A system can appear corrigible while protecting the structure that decides what correction can mean. The left side can challenge a decision but not the rule, the governing arrangement, the membership, or the material conditions. The right side keeps a path open to every level, including the institution’s own necessity.
View source
%% title: Shallow vs. recursive corrigibility
%% caption: The left side can correct a decision but not the rule. The right side keeps a path open to every level.
flowchart BT
subgraph Shallow["Shallow corrigibility"]
S0[Decisions]:::open --> S1[Rules]:::closed
S1 --> S2[Governance]:::closed
S2 --> S3[Membership]:::closed
S3 --> S4[Material conditions]:::closed
end
subgraph Recursive["Recursive corrigibility"]
R0[Decisions]:::open --> R1[Rules]:::open
R1 --> R2[Governance]:::open
R2 --> R3[Membership]:::open
R3 --> R4[Material conditions]:::open
end
classDef open stroke-width:2px;
classDef closed stroke-dasharray:4 4,opacity:0.5;
A building with open doors on the ground floor and locked stairwells above is not a building you can leave. The correction path has to reach the floors that decide what the ground floor is for.
7. What it makes possible, necessary, and leaves possible
The six patterns converge on three questions. What does the institution make possible? What does it make necessary? And what does it leave possible?
View source
%% title: What an institution makes possible, necessary, and leaves possible %% caption: The third column is the one conventional evaluations of success omit. The first two can both improve while the third narrows. flowchart LR I[institution] --> P[what it makes possible] I --> N[what it makes necessary] I --> L[what it leaves possible] P --> B[benefits, knowledge, infrastructure, capabilities] N --> D[dependencies, obligations, constraints] L --> A[independent capacities, resources, authority] classDef third stroke-width:2.4px,font-weight:bold; class L,A third
An institution can improve the first column and the second while narrowing the third. That is the pattern the six panels have been circling. The political principle follows:
No institution’s present contribution to society should, by itself, entitle it to permanent authority over the possibilities of society’s future development.
Three dimensions
The six patterns can be organized into three dimensions.
Capability: what can people accomplish? Services, knowledge, technical capacity, resources, coordination, and material benefits. This is the conventional emphasis of innovation and institutional performance.
Authority: what can people determine? Control over purposes, rules, decisions, resource allocation, and the conditions of participation. This is the distinction between being served and exercising self-government.
Transformability: what can people change or outgrow? Effective correction, institutional replacement, independent alternatives, and the capacity to revise the structures of authority themselves. This is the distinction between adapting successfully and expanding future freedom.
The third dimension connects the concerns about innovation, institutional dependence, democratic authority, and progress across time. An institution may improve capabilities without distributing authority. It may distribute some authority without enabling structural transformation. It may allow formal revision without preserving the resources needed to make alternatives viable. None of the dimensions can stand in for the others.
Four relations, not one quantity
The series makes four distinctions legible:
- Access: can people use or benefit from the system?
- Influence: can they affect its operation?
- Authority: can they make or revise binding decisions about it?
- Transformation: can they alter, replace, or move beyond the arrangement itself?
These are different relations, not simply increasing quantities of one thing. A public institution may legitimately limit individual exit while providing strong collective control. A competitive marketplace may provide exit without providing much influence over shared conditions. Keeping the four apart is what stops the framework from equating decentralization with freedom or centralized coordination with domination.
The unifying thesis
The strongest synthesis is not that every institution should maximize options or become replaceable as quickly as possible. It is this:
Institutions should be evaluated by how their ordinary operation changes people’s effective capacity to determine and revise the conditions under which they live — including the institution’s own purposes, authority, and continued necessity.
The formulation preserves the value of expertise, durable infrastructure, specialization, interdependence, and institutional continuity. What it rejects is the conversion of those achievements into an unchallengeable claim of authority.
What the series is mapping
The underlying subject is the distribution of effective possibility. The arrows, nodes, gates, barriers, and branches do not merely depict the movement of resources or information. They depict how institutional arrangements organize possibilities for action.
A concentrated hub might enable efficient coordination and make independent action prohibitively difficult. A visible pathway might represent a legal right whose practical availability depends on resources and protection from retaliation. A branching structure might represent alternatives that are meaningful only if people can actually develop and pursue them.
The recurring comparison is not centralized versus decentralized, or closed versus open. It is:
Success that entrenches: more activity becomes dependent on an increasingly indispensable center.
Success that enables: useful coordination develops independent capacities and feasible paths for future change.
The important thing about the second picture is not that it has more connections. It is that its participants can develop and exercise meaningful authority over the network’s future configuration.
The invitation
The identity behind the series is ordinary enough to place: a systems designer and independent researcher studying how technology and institutions shape human agency, distribute power, and determine what kinds of change remain possible.
The invitation is simpler:
How can we build systems that improve people’s lives without making people less able to shape what comes next?
Progress is not only the expansion of what a society can do. It is also the development of its capacity to decide, and subsequently revise, what it will do, how it will do it, and which institutions it will continue to need.