The Five Primitives
The conditions that make reliable coordination possible.
Organizations run on coordination.
Someone promises something. Someone else depends on it. Evidence establishes whether it happened. Authority determines what can happen next. Outcomes affect future decisions. And the organization either learns from what happened—or eventually repeats it.
Most of this occurs thousands of times every day without anyone describing it as governance.
But these ordinary interactions determine whether an enterprise can reliably carry its intentions through to completion.
Enterprise Governability Lab uses five primitives to describe the conditions underneath that behavior:
Commitment. Evidence. Consequence. Permission. Memory.
Together, they provide a simple language for examining why coordination works, why it fails, and what must be true for increasingly complex organizations to remain governable.
The Five Primitives
Commitment
What has been promised?
The basic unit of coordination is not the task.
It is the commitment.
A task describes an activity. A commitment connects that activity to someone or something that depends on its completion. A commitment establishes:
Who or what is responsible?
What outcome is expected?
By when?
Under what conditions?
What will count as completion?
This can be as ordinary as delivering a shipment, approving an invoice, resolving an exception, completing an inspection, producing a forecast, or responding to a customer.
The principle is the same. One part of the organization needs to know what it can reliably expect from another.
When commitments are unclear, people fill the gaps with assumptions. Responsibility becomes ambiguous. Handoffs become fragile. Follow-up increases. The organization begins spending energy simply determining whether something that was supposed to happen actually will.
Commitment turns intention into an accountable expectation.
Evidence
What establishes what actually happened?
A commitment tells us what was supposed to happen. Evidence tells us what did.
Evidence may be a delivery receipt, approval, inspection result, payment record, authenticated system event, completed work product, acceptance record, or other verifiable indication of an outcome.
Its importance is easy to underestimate.
When reliable evidence exists, the next person, system, or agent can act from a shared understanding of reality. When it does not, the organization must reconstruct reality.
People search email.
They compare systems.
They make phone calls.
They reconcile records.
They dispute invoices.
They ask someone what happened.
This is why evidence is not simply documentation. It is an operating condition.
Evidence turns claims about reality into something the enterprise can rely on.
Consequence
What changes because of what happened?
A governing system must respond to outcomes. When commitments are reliably fulfilled, confidence can increase. When commitments repeatedly fail, something should change.
That change might affect approval requirements, authority, escalation, counterparty treatment, workflow design, future permissions, or the level of evidence required before proceeding.
Consequence does not mean punishment. It means that what actually happens matters to what the organization does next. Without consequence, the enterprise can observe the same failure repeatedly while continuing to operate as though nothing has changed.
Policies remain intact.
Dashboards record the problem.
Meetings discuss it.
But future behavior is unaffected.
A governable organization closes that loop.
Consequence allows experience to change future operating conditions.
Permission
Given what is now known, what is allowed to happen next?
Organizations cannot coordinate reliably if every action depends on someone improvising whether it should proceed. Permission connects verified conditions to authorized action.
Has the required evidence been received?
Has the previous obligation been completed?
Is this person, system, or agent authorized to act?
Has a threshold been exceeded?
Does this exception require escalation?
Permission determines when the conditions are sufficient for the next action to occur.
This becomes especially important as AI agents participate more directly in enterprise activity.
The question cannot simply be whether an agent can take an action. The enterprise must establish the conditions under which it may take that action.
Permission therefore does something fundamental: It converts organizational authority into bounded operating action.
Memory
What should the organization remember?
Organizations store enormous amounts of information. That doesn't mean they possess useful institutional memory.
Governability requires more than retaining documents. The enterprise needs to preserve meaningful operating history:
What was committed?
What actually happened?
What evidence established it?
What exception occurred?
What consequence followed?
What changed afterward?
Without that memory, organizations repeatedly rediscover what they already learned — and its future looks like history.
The same supplier fails in the same way.
The same exception requires the same workaround.
The same project encounters the same dependency.
The same operating lesson disappears when the person who remembers it leaves.
AI makes this distinction even more important. If intelligent systems are going to learn from enterprise history, the organization must be able to distinguish verified operating experience from assumption, interpretation, and noise.
Memory allows past behavior to inform future coordination.
The Five Conditions Work Together
The primitives are not five steps in a process. They're interacting conditions.
Commitment establishes what should happen.
Evidence establishes what did happen.
Permission determines what may happen next.
Consequence allows outcomes to change future conditions.
Memory preserves what the organization has learned.
Consider what happens when any one is weak.
A commitment without evidence produces uncertainty.
Evidence without consequence produces observation without adaptation.
Permission without clear conditions produces arbitrary authority.
Consequence without memory produces temporary correction.
Memory without reliable evidence preserves stories rather than operating truth.
The strength of the system comes from the relationships among them. Together, these conditions help determine whether an enterprise can continue behaving according to its intentions as complexity increases.
Why We Call Them Primitives
A primitive is a basic building block from which more complicated structures can be constructed.
These five concepts appear repeatedly across very different forms of enterprise activity.
A shipment and a software release look nothing alike. Neither looks much like a financial close, patient-care workflow, insurance claim, manufacturing process, energy transaction, or AI-agent interaction.
But underneath them we can ask the same questions:
What was committed?
What evidence establishes what happened?
What changes because of the outcome?
What is permitted now?
What should be remembered?
That makes the primitives useful across functions, systems, and industries. They give us a common language for describing coordination without requiring every organization to operate the same way.
Why We Call Them Constitutional
This is where the constitutional idea enters.
A constitution does something different from a procedure. A procedure tells an actor what steps to follow. A constitutional structure establishes the conditions within which actors are allowed to operate.
That distinction becomes increasingly important as enterprises contain more people, systems, algorithms, external parties, and autonomous agents.
It's neither practical nor desirable for leadership to prescribe every action in advance. Instead, the enterprise can establish governing conditions:
What constitutes a valid commitment?
What evidence is required?
What changes when obligations are fulfilled or violated?
Who or what has permission to act?
What operating history must persist?
Those conditions shape behavior without requiring central management of every interaction.
Constitutional enterprise design governs the environment in which coordination occurs—not merely the actors operating inside it.
That is why EGL refers to Commitment, Evidence, Permission, Consequence, and Memory as the Five Constitutional Primitives.
Why AI Raises the Stakes
AI increases enterprise capability.
It also changes the speed and scale at which coordination can occur.
An intelligent agent can interpret information, generate recommendations, initiate workflows, communicate with other systems, and increasingly take actions that create consequences elsewhere in the enterprise. Human-speed management cannot supervise every machine-speed interaction.
That creates a design problem. How can organizations allow greater autonomy without surrendering control over organizational behavior?
The Five Primitives provide one way to frame the answer.
An agent should not act merely because it generated an answer. Instead, its action can occur inside an environment where:
Commitments are explicit.
Evidence is required.
Consequences affect future behavior.
Permissions are bounded.
Memory is grounded in what actually occurred.
The objective is to make greater autonomy trustworthy. It is not to constrain intelligence unnecessarily.
Moving From Management by Intervention to Governability by Design
Traditional organizations compensate for weak coordination through human intervention. Managers chase commitments. Employees reconcile conflicting information. Executives resolve escalations. Experienced people remember what the system forgets. High performers learn which procedures can safely be worked around.
These behaviors can keep an organization functioning. They can also hide how dependent the enterprise has become on extraordinary human effort to accomplish ordinary work.
The Five Primitives suggest another possibility.
Instead of relying primarily on people to repair coordination after it breaks, organizations can begin designing the conditions under which reliable coordination occurs in the first place.
That is the deeper purpose of constitutional enterprise design.
Not more rules.
Not more bureaucracy.
Better conditions for reliable action.
And as organizations become more complex, more autonomous, and more intelligent, those conditions may increasingly determine whether capability becomes performance—or simply creates more complexity to manage.
Continue the Research →
For more about the framework, see:
Foundational Library
Leading the Organization of the Future (new book available)
The first comprehensive introduction to Enterprise Governability [Learn More]
Foundational Essays
What Governs Enterprise Performance? (WP-No. 1)
The foundational theoretical paper proposing Enterprise Governability as the missing explanatory layer between operations and financial performance.
The foundational paper proposing that enterprise performance undergoes a phase transition in which governability becomes the dominant determinant of realized performance as organizational complexity increases.
Featured Board Brief
AI, Coordination Debt, and the Case for Constitutional Governability
Read the Manifesto
An invitation to begin seeing today's organizations differently.