An open forum
Information Governance is the shared stewardship of the conditions under which public knowledge is produced, verified, preserved, made accessible, connected and corrected.
This forum does three things and no more. It publishes a framework of five functions. It hosts named contributions and replies to that framework, including the ones that reject it. And it holds a citable, versioned reference that any institution can adopt, test or discard.
Five arcs, one interrupted circle, with the gap at twelve o'clock where the ring would otherwise close. The opening is the argument. An information environment is never complete, no single institution closes it, and governance of it is a continuing practice rather than a finished state.
The colour deepens and then lifts as the functions move outward. The first three happen inside a publication. The fourth exists only between publications. The fifth spans the environment as a whole, and belongs to no one in particular, which is the difficulty.
Functions, not values. A function can fail observably, and the failure can be described, dated and disputed. A value cannot be tested, only professed.
First-hand knowledge enters public view: someone goes, asks, records and publishes what was not previously on the record. Nothing downstream can compensate for its absence. A commentary layer can be rich while the reporting layer beneath it is thin, and the thinness is only visible if the two are counted separately.
Inside one publication
Claims are tested before they circulate, and the test leaves a trace a reader can inspect: a named source, a document, a method, a correction when the test failed. Verification that cannot be examined is indistinguishable from assertion.
Inside one publication
A fact is placed among the facts that qualify it: the scale it sits at, the trend it belongs to, the disputes it enters. Accurate statements assembled without context can produce a false account, which is why accuracy alone is too weak a standard for a governance framework.
Inside one publication
No publication holds every relevant viewpoint, and none should claim to. The question is whether partial views remain within reach of one another: whether a reader following one account can find the others, and whether enough evidence is shared for the disagreement between them to stay intelligible.
Between publications
Separate accounts are drawn together into something a citizen can hold: the state of a question rather than the state of the coverage. This is the function no single publisher can perform alone, and the one least often anyone's responsibility.
Across the environmentThe framework identifies functions, not approved opinions. It asks whether the conditions for judgement hold, never what a citizen should conclude.
One reading for each function. Each takes a single arc and asks what it would mean to observe it: what evidence would show the function holding, what would show it failing, and what part of it cannot be measured at all. Addendums extend or correct the readings and are versioned alongside them.
A framework that cannot be quoted precisely cannot be tested, and one that cannot be tested has no claim on anyone. Contributions are signed, dated and answerable. Disagreement is published beside the argument it disputes, never filed away from it.
There are no view counts here, no recommendation rail and nothing that sorts an argument by how many people liked it. A forum about the conditions for judgement cannot use devices designed to bypass it.