Incoherence is mismatch, not incompetence
Large successful enterprises already achieve substantial coherence across enormous portions of their activity. Incoherence does not mean the organization lacks expertise, ownership, process, coordination, or sophistication.
It appears when the enterprise’s operative treatment of a consequential reality is materially weaker than the reality requires. Something that can change the enterprise answer is omitted, distorted, separated from a relationship that matters, represented at the wrong boundary or resolution, mistranslated, allowed to become stale, or prevented from changing action when it should.
The relevant question is not whether the enterprise is well run in general. It is whether this movement preserves enough of the reality that governs its economics.
Local rationality can still produce enterprise incoherence
Large enterprises must divide work, knowledge, authority, budgets, systems, objectives, and accountability. Those divisions make scale possible. They also mean that no local representation necessarily contains the whole enterprise logic.
Finance, architecture, engineering, operations, procurement, risk, and business units can each reason correctly inside their own objects while the combined enterprise treatment remains incomplete. Local optimization can shift cost, consume a shared constraint, foreclose an option, or weaken another part of the system without any participant behaving irrationally.
Ownership does not establish coherence. Coordination does not establish coherence. Capability does not establish coherence. The economically relevant whole still has to be reconstructed well enough for the action at hand.
Where incoherence enters
Incoherence can enter throughout enterprise movement.
Reality formation can fail when the wrong object is chosen, the boundary is too narrow, the current state is represented poorly, or superior possibilities never enter the operative alternative set.
Composition can fail when individually accurate representations do not reconnect the relationships, objectives, constraints, configurations, and portfolio effects that determine enterprise value.
Commitment and transformation can fail when the logic that made a move rational degrades as it becomes budgets, architecture, contracts, requirements, workstreams, systems, workflows, and operating practice.
Observation and adaptation can fail when new evidence is measured but does not reconnect to the assumptions or authority required to revise the move.
These are broad causal zones, not an exhaustive taxonomy.
Materiality is the threshold
Not every disagreement, inconsistency, missing fact, or imperfect handoff is incoherence in the economically useful sense.
The inadequacy matters when it can change the judgment, value, configuration, realization requirements, future state, or reasonably attainable opportunity set. This materiality threshold keeps the concept tied to enterprise economics rather than to administrative neatness.
A move can also be incoherent and still produce a favorable result. Luck does not repair the treatment. Conversely, a coherent move can encounter an adverse external state. The concept describes the quality of the enterprise’s relationship to reality, not the outcome alone.
Relationship to coherence and value leakage
Systemicity describes the condition that must be navigated. Coherence describes the positive quality of the enterprise’s relationship to and movement through that condition. Incoherence describes the economically material inadequacy in that relationship.
Value leakage is one economic consequence. When incoherence changes what the enterprise sees, chooses, configures, realizes, learns, or makes possible later, reasonably attainable value can fail to become realized.
Bibliography
Adjacent intellectual terrain includes bounded rationality, distributed knowledge, organizational decomposition and specialization, local optimization, coordination and integration failure, information loss and distortion, sensemaking, socio-technical systems, cybernetics and feedback, organizational learning, and systems-engineering treatments of interfaces and handoffs.