Ministry of Education V3.0 · Mechanisms of Education Systems · Vol. 019
People do not learn subjects merely to possess information. They learn so that understanding can become capability: the ability to recognise a problem, choose an approach, perform a task, adapt under changing conditions and explain the result.
The 50-second route
Read sections 1–10 for the difference between content, competence, performance and transfer. Read sections 11–20 for the Capability Graph and its nodes, edges, prerequisites and evidence. Read sections 21–30 for how schools, workplaces, families, libraries and digital systems contribute to capability development. Read sections 31–40 for curriculum alignment, assessment, feedback, credentials and lifelong learning. Read sections 41–50 for operating the graph inside a Live Ministry.
The central proposition is: education becomes coherent when it can show how knowledge, practice, judgement and transfer connect to capabilities that people can actually use.
1. Content is not capability
A learner can recite a definition and still be unable to use the underlying idea. V3.0 separates exposure, understanding, practice, performance and transfer.
2. Competence and performance
Competence is potential under stated conditions. Performance is what actually happens in a particular context. Education systems need both views.
3. Transfer is the real bridge
Transfer occurs when a learner recognises that a capability built in one context can be used meaningfully in another. V3.0 maps transfer intentionally.
4. Prerequisites are hidden infrastructure
Every capability rests on earlier capabilities. The graph makes prerequisites explicit for diagnosis and coherent sequencing.
5. Conceptual knowledge
Conceptual knowledge concerns relationships, categories, mechanisms and explanatory structures. It lets a learner answer why and how.
6. Procedural knowledge
Procedural knowledge concerns how to perform. It develops through explanation, modelling, deliberate practice and feedback.
7. Dispositional and judgement knowledge
Capabilities also include habits of attention: checking units, seeking counterevidence, recognising when to ask for help and noticing safety conditions.
8. Practice changes what is learned
Practice changes retrieval strength, fluency, error detection and adaptability. Identical practice can create pattern speed without structural understanding.
9. Capability is contextual
Capabilities operate under conditions. A person may be strong with calculations but struggle when data are incomplete, time is short or constraints conflict.
10. The receiver’s task completes learning
The final test is what the learner can now do independently: choose, perform, explain, adapt and recover.
11. The Capability Graph
The Capability Graph maps knowledge, skills, practices, dispositions and outcomes as nodes, with prerequisite, support, transfer, application and feedback relationships as edges.
12. Nodes, edges and pathways
Relationships tell learners what to learn next and institutions what support is required. Edges can be labelled by function.
13. Prerequisite chains
Some knowledge is truly foundational; some can be learned in parallel; some is useful but not required. Making this distinction explicit improves diagnosis.
14. Evidence nodes
Evidence nodes connect capabilities to demonstrations, assessments, workplace artefacts and observations. Different capabilities require different evidence.
15. Misconception edges
The graph can store misconception patterns and the experiences that expose them, helping teachers diagnose before reteaching.
16. Transfer edges
Transfer edges show where a capability should travel, but transfer claims require evidence rather than assumed similarity.
17. Practice intensity
High-stakes, safety-critical and foundational capabilities can require more varied practice than optional knowledge. Practice time is finite.
18. Feedback loops
Feedback changes the next attempt and tells the system which part of the route may be weak. Repeated error patterns can trigger mechanism review.
19. Recovery and re-entry
Learners interrupt routes for many reasons. Resilient systems provide re-entry points rather than forcing unnecessary restarts.
20. Graph maintenance
Capabilities change as tools, occupations, evidence and social demands change. The graph therefore requires versioning, stewardship and retirement.
Part III — Where capability is built
21. Schools as capability engines
Schools are not simply places where lessons happen. They are capability engines because they combine protected time, sequence, professional guidance, peers, practice, assessment and recovery. The engine fails when these parts pull in different directions.
A timetable can be full while capability growth is weak. A curriculum can be accurate while prerequisites are missing. An assessment can be reliable while it measures a narrower skill than the public purpose requires. V3.0 therefore maps each major school function to the capability nodes it is supposed to activate.
22. Families and independent learning
Families support capability through conversation, routines, reading, modelling and encouragement. They also provide evidence about what learners can and cannot yet do outside school. But the system must not confuse family resources with learner potential.
Independent learning becomes stronger when learners have clear next steps, accessible explanations, practice tools and ways to check understanding. The graph can provide these routes without requiring the learner to understand institutional taxonomy.
23. Workplaces and apprenticeships
Workplaces expose capabilities to real constraints: incomplete information, deadlines, customers, safety, collaboration and consequences. Apprenticeship connects conceptual knowledge to situated performance through demonstration, guided practice and feedback.
V3.0 maps workplace tasks back to formal capabilities. This allows education providers to see whether a course prepares people for the kinds of decisions they will actually face.
24. Libraries and public learning
Libraries extend capability routes beyond enrolment. They support independent inquiry, access to evidence and assisted navigation. Public learning also occurs through museums, community programmes, professional associations and digital resources.
The Capability Graph makes these routes visible as alternate paths to the same capability, while preserving differences in quality, evidence and assessment.
25. Assessment and credentials
Assessment should sample the capabilities it claims to represent. Credentials then communicate a bounded claim about what has been demonstrated, under what conditions and at what time.
V3.0 avoids treating one score as the entire capability graph. A learner can show different strengths across concepts, procedures, judgement and transfer. Credential design should communicate enough structure for receivers to understand the claim.
26. Digital systems and AI tutors
Digital tools can personalise practice, surface misconceptions and help learners retrieve explanations. AI tutors can also generate plausible but wrong answers. The graph therefore keeps canonical knowledge and evidence separate from conversational projection.
High-consequence outputs need grounding, provenance, verification and escalation. A tutor that cannot say where a claim came from should not silently become the learner’s final authority.
27. Curriculum coherence
Curriculum coherence means more than avoiding duplicated topics. It means the learner’s route has usable foundations, deliberate transitions and opportunities for transfer. A graph shows whether a later demand assumes a capability that earlier learning never securely developed.
Cross-subject connections should be explicit where they are real. False connections create overloaded curricula. The evidence for a transfer edge should be proportional to the claim being made.
28. Time and attention allocation
Learning time is finite. Every additional capability, assessment, platform or enrichment activity consumes attention that could have been used for something else. V3.0 therefore connects the Capability Graph to the Attention Budget.
The system asks which nodes are foundational, which need maintenance, which unlock many downstream capabilities and which can remain optional. This is not a ranking of people; it is allocation of instructional time.
29. Equity of capability routes
A capability is not meaningfully public when access depends on private tutoring, specialist devices or institutional insider knowledge. V3.0 examines whether multiple legitimate routes exist and where the cost of remediation falls.
Equity is checked at entry, practice, feedback, assessment and transfer. A learner may access content but lack time to practise, receive feedback or use the skill in a real setting. The graph makes these missing edges visible.
30. Lifelong learning
Capabilities decay, change and become newly relevant. Lifelong learning therefore needs re-entry, recognition of prior capability and pathways that do not force every adult through a child-sized curriculum.
The graph can preserve what a person already demonstrates, identify changed prerequisites and generate a targeted route for new capability. This reduces wasted learning while preserving standards.
Part IV — Operating the graph
31. The capability registry
The registry gives each capability a canonical identity, description, evidence type, prerequisites, typical contexts, transfer examples, steward and review state. It is the stable spine beneath many different learning resources.
Without a registry, the same capability can be named differently across schools, courses and agencies, making pathways difficult to align. With one, projections can remain locally useful while sharing a common underlying claim.
32. The pathway planner
The pathway planner starts from a desired outcome and works backward through prerequisites, practice, evidence and transfer. It can also start from a diagnostic and work forward toward the smallest useful next route.
This enables just-in-time repair. Instead of repeating a whole course, a learner can secure the missing nodes and rejoin productive learning.
33. The evidence map
The evidence map connects every high-consequence capability to appropriate demonstrations. It distinguishes formative signals from summative evidence and recognises that some capabilities require longitudinal observation.
Evidence is not only a score. It can include explanations, performances, portfolios, projects, simulations, decisions and verified workplace tasks.
34. The transfer lab
The transfer lab deliberately changes surface conditions while preserving the underlying structure. Learners practise recognising when a familiar principle applies to an unfamiliar problem.
Teachers can use the lab diagnostically. If transfer fails but near transfer succeeds, the missing edge may be abstraction. If near transfer fails, the underlying concept or procedure may not yet be secure.
35. The skills-gap detector
The skills-gap detector compares desired capability with observed evidence. It distinguishes a missing prerequisite from insufficient practice, weak transfer, poor conditions or unclear instruction.
At system level, repeated gaps can identify curriculum weaknesses. At learner level, the same mechanism can produce a focused recovery route.
36. The practice scheduler
The practice scheduler allocates spaced retrieval, mixed examples, deliberate practice and transfer tasks according to the nature of the capability. It avoids endless repetition when a new edge needs to be built.
Practice plans remain adaptable. Forgetting, workload and changing goals affect the schedule. The aim is durable capability, not visible activity.
37. The feedback engine
The feedback engine connects observed errors to the relevant graph node and likely misconception. This makes feedback actionable rather than generic.
At scale, repeated feedback patterns can become system signals. If many learners fail after the same explanation, the system investigates the representation rather than assuming all receivers need more effort.
38. The credential mapper
The credential mapper shows which capabilities a qualification actually evidences, which remain outside its scope and which were demonstrated under controlled versus real-world conditions.
This helps employers, universities and learners interpret credentials without inflating them. Portability improves when the claim is precise.
39. The capability-health dashboard
A capability-health dashboard monitors activation, practice, evidence, transfer, decay and recovery across a population while protecting privacy. It should show where the route is strong and where the mechanism is failing.
Aggregate metrics are useful but not sufficient. Distribution, long-tail failure and local context reveal whether a capability is genuinely accessible.
40. The Learning Commons connection
The Capability Graph sits between the Knowledge Commons and the learner. The Knowledge Commons preserves validated public knowledge. The Capability Graph organises the path by which a person can turn that knowledge into action.
The two systems are coupled. A new validated insight changes the graph. Repeated learner failure can reveal that a knowledge representation needs improvement. Workplace evidence can add a transfer edge. The result is a closed educational learning loop.
Part V — Capability clinics
41. Clinic: reading a difficult science text
A fictional Secondary learner can recite definitions in a chapter on chemical equilibrium but cannot explain what changes when conditions shift. The Capability Graph reveals that the problem is not simply “weak chemistry.” The learner has vocabulary, partial concepts and some procedural fluency, but the causal relationship between disturbance and system response is not securely connected.
The recovery route begins with a diagnostic explanation, then a small set of contrasting examples. The learner predicts before seeing the answer, explains the direction of change and then tests the reasoning against a new context. Evidence is stored against the conceptual node and transfer edge rather than merely recording another completed worksheet.
42. Clinic: writing that works outside the classroom
A learner can produce a five-paragraph composition in exam conditions but struggles to write a clear email explaining a problem to an adult. The form has been mastered; the transferable capability has not.
The graph separates genre knowledge from audience analysis, purpose, tone, sequencing and revision. Practice then changes the receiver, purpose and medium while preserving the underlying communication task. Feedback focuses on decisions that generalise across contexts.
43. Clinic: mathematics for ordinary life
A fictional learner solves textbook percentage questions but cannot compare two mobile plans with different fees, discounts and usage limits. The gap is transfer from symbolic practice to a real decision with distracting information.
The pathway planner identifies the relevant arithmetic, ratio and comparison nodes. The learner builds a representation of each plan, checks units and tests sensitivity to changes in usage. The final evidence is a justified decision, not simply a correct numerical answer.
44. Clinic: a practical laboratory skill
A learner can describe laboratory safety rules but repeatedly reaches for equipment before checking the setup. The knowledge node exists; the disposition and procedural sequencing are weak.
V3.0 uses a simulated routine in which the learner must identify hazards, prepare materials, verbalise the check and then act. A later scenario changes the context so the learner must recognise the same safety structure without memorising the original sequence.
45. Clinic: language vocabulary that never becomes usable
A learner recognises two hundred vocabulary words in multiple-choice tests but rarely uses them in speech or writing. The lexical nodes have recognition evidence but weak retrieval, collocation and transfer edges.
The recovery route moves from recognition to retrieval, then collocation, sentence construction, short conversation and purposeful writing. The graph stores use contexts and revisits words after spacing rather than treating a list score as mastery.
46. Clinic: learning after an interruption
A fictional learner moves schools halfway through a mathematics sequence. The receiving teacher has limited evidence about what was already mastered. Restarting the entire course wastes time; assuming full continuity risks hidden gaps.
The graph compares demonstrated nodes with the receiving course prerequisites. Short diagnostics target uncertain edges. The learner re-enters at the first productive point rather than repeating everything or being pushed into content that depends on unsecured foundations.
47. Clinic: a workplace capability with changing tools
A technician learned a maintenance procedure on one generation of equipment. The new system changes the interface but preserves the underlying failure modes. Training that teaches every click from the old system creates fragility.
The graph distinguishes invariant principles from tool-specific procedures. The worker learns to diagnose the underlying state, then maps that reasoning onto the new interface. Transfer becomes the bridge between stable capability and changing technology.
48. Clinic: when assessment becomes the curriculum
A fictional school notices that teachers devote increasing time to likely test formats. Students become faster at familiar item structures while struggling with open-ended tasks.
The Capability Graph identifies an assessment edge that has become too dominant. V3.0 restores balance by mapping the public capability to multiple evidence forms, including explanation, novel application and transfer. Assessment returns to being evidence of learning rather than the entire definition of learning.
49. Clinic: the learner who looks “average” in an aggregate
An aggregate dashboard shows a learner near the middle. A graph view reveals a different picture: strong language capabilities, weak quantitative prerequisites and one severe misconception blocking several later nodes.
This is a systems diagnosis rather than a ranking of the learner. The next action can be precise because the graph shows the bottleneck. Once the prerequisite is repaired, several downstream routes become accessible.
50. Clinic: a skill that matters but has no owner
A ministry identifies that many young adults struggle to understand contracts, bills, subscriptions and basic financial commitments. The need sits between subjects and agencies, so no unit sees itself as the canonical owner.
The Capability Graph creates a capability owner, evidence definition and public entry route. Existing mathematics, language, citizenship and consumer resources are connected without inventing a new standalone subject unnecessarily. The gap becomes visible and governable.
Part VI — The long-term learning architecture
51. Capability decay
Skills can weaken when unused. The system should distinguish forgotten capability from never-acquired capability. A learner returning to algebra after several years may recover quickly with retrieval and a few targeted prerequisites.
V3.0 therefore records maintenance schedules for capabilities where continued fluency matters. These can be lighter than initial acquisition and should focus on high-value retrieval and authentic use.
52. Capability growth
Advanced capability is not simply “more of the same.” It often adds abstraction, judgement, complexity management and independence. The graph represents these as new nodes and edges rather than stretching one beginner node forever.
This supports progression without creating the illusion that every advanced skill is just a harder worksheet.
53. Capability clusters
Some real tasks depend on bundles: scientific investigation, project management, emergency response, persuasive communication or financial planning. A cluster shows how several capabilities must cooperate under a common task.
Cluster evidence should not erase individual components. A learner can succeed in a task through one strong compensating skill while another component remains weak. The graph keeps the underlying nodes visible.
54. Capability bottlenecks
A bottleneck is a node whose weakness blocks many downstream pathways. Bottlenecks deserve attention because small improvement can unlock disproportionate opportunity, but the system must verify the causal relationship rather than assuming every highly connected node is inherently more important.
V3.0 uses evidence from learner routes to identify bottlenecks. It then tests whether repairing the node actually changes downstream performance.
55. Capability debt
Capability debt arises when a person or institution moves ahead while carrying unsecured prerequisites. The cost appears later as repeated relearning, error, dependence on others or inability to handle new contexts.
The graph makes debt visible. A learner can keep progressing while the system marks unresolved prerequisite risk and creates a recovery window before the missing foundation becomes a larger barrier.
56. Capability opportunities
The opposite of debt is latent capacity: a learner possesses several related nodes and is close to an important capability but has not yet encountered the right task. Good education can reveal these openings without forcing every learner into every possible extension.
Opportunity maps can support enrichment, elective choice and career exploration while leaving the learner free to decide.
57. Capability under uncertainty
Real work often happens without complete information. A graph that only tests clean textbook tasks creates an unrealistic picture of competence.
V3.0 includes bounded ambiguity where appropriate. Learners practise identifying what is known, what is missing, what assumptions are being made and when additional evidence is worth the cost.
58. Capability and collaboration
Some outcomes are individual; others depend on coordination. A person may possess every component skill but fail a team task because roles, communication or timing break down.
The graph therefore distinguishes individual capability from system capability. Collaborative practice can include shared planning, handoffs, conflict resolution and collective correction.
59. Capability and ethical judgement
Knowing how to do something does not determine whether it should be done. Professional and civic capability includes recognising constraints, stakeholders, consequences and legitimate authority.
V3.0 keeps ethical judgement connected to factual and procedural knowledge while distinguishing value choices from empirical claims. Learners can be asked to justify decisions and identify which evidence supports the factual parts of their reasoning.
60. Capability as a public infrastructure
When enough people share foundational capabilities, society gains coordination capacity. People can interpret public information, use technology, participate in institutions, work safely and learn new skills.
The public education system therefore produces more than individual achievement. It maintains a distributed capability infrastructure that other institutions rely upon.
Part VII — Field tests for V3.0
61. The explain-it-tomorrow test
If the learner forgets the exact wording tomorrow, can the underlying structure still be reconstructed? This tests durable understanding rather than short-term performance.
62. The unfamiliar-example test
Present a new surface form. Does the learner recognise and use the same underlying principle? This tests transfer.
63. The no-prompt test
Remove the cue that usually signals the required procedure. Can the learner select an appropriate strategy? This tests independent recognition.
64. The recovery test
Introduce a predictable mistake. Can the learner notice, diagnose and repair it? Capability includes error recovery.
65. The pressure test
Change time, information or competing constraints within safe bounds. Does performance remain coherent? This tests contextual robustness.
66. The teach-someone test
Ask the learner to explain the mechanism to another person and anticipate a likely misconception. Teaching exposes structural understanding.
67. The transfer-to-life test
Ask where the capability could improve an ordinary decision outside school. The learner should identify a legitimate use without forcing superficial analogies.
68. The cross-subject test
Where genuine transfer exists across disciplines, test it. This prevents siloed learning while avoiding claims of connection unsupported by evidence.
69. The credential-meaning test
Can a receiver of a certificate understand what the holder demonstrated and what remains outside the credential’s scope?
70. The next-generation test
Can the capability be taught to someone new, preserved through resources and improved through feedback? This links individual learning to the Knowledge Commons.
Frequently asked questions
Is the Capability Graph a replacement for a curriculum?
No. A curriculum is a planned learning route. The graph is the underlying relationship model that helps a curriculum remain coherent, diagnosable and transferable.
Does every learner need the same graph?
The shared capability definitions can be common while routes differ. Prior knowledge, goals, contexts and pace can change the path without changing the intended standard.
Can capability be reduced to numbers?
No. Numbers can summarise evidence, but complex capability needs multiple forms of evidence, context and professional interpretation.
How does this connect to assessment?
Assessment supplies evidence attached to capability nodes. It should not become the node itself unless the capability is genuinely the ability to perform that assessment.
How does this connect to AI?
AI can help diagnose, explain, practise and route, but the canonical capability definition, evidence standard and correction process remain governed by accountable educational stewards.
Terminal proposition
The purpose of a learning system is not to produce completed lessons. It is to produce people who can use knowledge well. The Capability Graph makes that purpose inspectable.
Its core loop is: desired capability → prerequisites → explanation → practice → feedback → performance → transfer → real-world use → new evidence → correction → renewed practice.
A Live Ministry can then see where learning routes work, where they break and what infrastructure is needed to repair them. This turns education from a catalogue of offerings into an observable system of capability formation.
Series route: How Education Works · Ministry of Education V3.0 | Live Ministry · Vol. 018 | The Knowledge Commons