The canonical concepts.

The cognitive enterprise is the successor to the industrial, information, and digital enterprise: an organization whose defining capability is not what it produces or what it knows, but how well it reasons, remembers, and improves its judgment over time.

It is defined not by adopting any particular technology but by treating reasoning — evidence, judgment, decisions, and learning — as a first-class asset worth capturing and compounding.

Organizational cognition treats reasoning as a property of the institution rather than only of the individuals within it. It asks how an organization forms views, tests them, records them, and revises them — and how those capacities can be deliberately built.

It is to judgment what knowledge management was to information: a discipline concerned with a specific organizational asset, but one aimed at the reasoning that uses knowledge rather than the knowledge itself.

Institutional intelligence is the property of a firm that is smart because of how it remembers and reasons, not merely because of who happens to work there now. It outlives the people who currently carry it.

It is the outcome the discipline aims at: an institution whose reasoning improves decade over decade instead of resetting with each generation of leadership.

Institutional memory is not documentation. Documents record conclusions; memory preserves the ability to re-enter the reasoning that produced them — the evidence, the alternatives considered, and the judgment applied.

Because it is the default of organizations to forget, memory is an infrastructure problem: it must be built into how decisions are captured, not left to the people who happen to remember.

Decision architecture treats decisions not as isolated events but as a connected, cumulative system: each decision records its reasoning and evidence, each outcome feeds back to revise judgment, and each new decision starts from the accumulated reasoning of the ones before.

It is the design discipline that turns a sequence of one-off analyses into an organization that learns.

Evidence is the first point at which an organization becomes accountable to reality: a claim connected to a traceable basis. In a cognitive enterprise, evidence is not disposable input to a decision but durable infrastructure — retained, versioned, and reusable.

Without retained evidence, decisions cannot be audited and judgment cannot be improved.

Judgment is where expertise actually lives — and the layer no system of record captures today. It is the reasoned commitment to a view when the evidence is partial, which is to say, always.

Judgment, not knowledge, is the asset that compounds: captured and revised against outcomes, each period’s judgment becomes the principal on which the next is built.

Confidence makes reasoning honest. By attaching a stated degree of belief to evidence and judgment, an organization can distinguish a settled conclusion from a working hypothesis, and can revise appropriately when new evidence arrives.

Confidence is what allows judgment to be tested: a claim held at stated confidence can be shown right or wrong, and the reasoning updated accordingly.

A knowledge fabric is not a repository but a fabric: a connective structure in which documents, data, evidence, and judgment are linked so that reasoning can move across them. It is what lets one part of an organization’s reasoning inform another’s.

Each domain — industrial, government, healthcare, and others — has its own fabric, sharing a common architecture but populated with domain-specific reasoning objects.

The Industrial Graph is the domain-specific instantiation of a knowledge fabric for industrial systems: assets, suppliers, processes, and their relationships, linked to the evidence and judgment that concern them.

It is the substrate over which domain reasoning — about supply chains, risk, and operations — is performed and remembered.

A cognitive enterprise treats relationships as first-class: which supplier depends on which, which decision rested on which evidence, which judgment a person actually holds. The systems of record store entities well and relationships poorly, and it is in the relationships that reasoning lives.

Capturing relationships is what turns a collection of records into a fabric that can be reasoned over.

Trust in a cognitive enterprise is not a sentiment but a set of guarantees: that a claim can be traced to its evidence, that reasoning is explainable, that oversight is possible, and that data sovereignty is preserved. It is the precondition for anyone acting on the organization’s judgment.

Trust is what allows reasoning to be reused across the organization without each user having to re-verify it from scratch.

Progressive connectivity describes how a knowledge fabric compounds: each new piece of evidence, judgment, and decision links to those already present, so the value of the whole grows faster than the sum of its parts.

It is the mechanism by which institutional intelligence accumulates rather than merely accretes.