Show where important claims come from
A source has to support the claim being made. Finding a source is not the same as reading it, and a citation is not proof by itself.
Standards
What readers should be able to expect from this work.
The Institute publishes these standards so a reader can use them to question the work, not just take the Institute's word for it.
Say what kind of claim it is
A reader should be able to tell whether something is a verified fact, an interpretation, a proposal, an uncertainty, or a question that belongs to another authority.
Correct material errors visibly
When a public document changes in an important way, use a new version and leave a record of the change rather than silently replacing the history.
Make necessary complexity easier to carry
Public writing should preserve the complexity the subject requires while removing avoidable burden in wording, structure, navigation, and emphasis.
The Audit standard
The Institute can hold a release when something important is unresolved.
The maintained standard checks whether a release has support for its claims, stays inside its authority, protects private or governed information, considers foreseeable harms, remains usable, and leaves a path for correction.
It can narrow, hold, refer, or send work back for repair. It cannot create expertise, consent, authority, or evidence that does not exist.
Christopher Knorr’s scholarly paper explains the argument behind this standard but remains a separate authored work with its own publication and correction history.
Tool adoption boundary
A file becomes an Institute tool only when the Institute agrees to maintain it.
The Institute has not adopted an operational edition of SIFT v3 or the existing Long Return field sheets. A link or shared history does not by itself make something an Institute tool.
SIFT v3
Christopher Knorr’s authored media-sorting framework. The Long Meanwhile is scheduled to open September 25, 2026 as its publication home; until then, the current reading route remains on The Long Return.
Existing field sheets
Context-bound appendices to Long Return reporting and community-meeting practice. Their current editions, routes, safety limits, and correction context remain there.
Before adoption, a generalized Institute tool needs a clear scope, defined users, a maintainer, testing and safety review, a public correction path, release history, and a way to retire or hand it off.
Building projects
Institute projects follow a documented development process.
The public manual explains how projects are scoped, built, reviewed, released, maintained, corrected, and, when appropriate, stopped or handed off.
AI assistance
AI may assist the work. A person remains responsible for the release.
AI materially assists some research, drafting, comparison, formatting, indexing, coding, and quality-control tasks. It may also assist review when used inside a human-governed process.
AI cannot create evidence, consent, cultural permission, professional expertise, or authority. Questions that require a clinician, lawyer, government, community, cultural authority, or other qualified person still have to go there.
Why say so. Fluent writing can make an unsupported claim look finished. Naming AI assistance is one part of letting readers judge how the work was made.
QR codes and stable routes
A QR code should make a route easier to reach, not hide where it goes.
Use a stable Institute address
Codes point to short addresses such as institute.thelongreturn.org/care. The Institute can update the page behind that address without reprinting the code.
Print the address beside the code
A person should still know where to go if they cannot or do not want to scan a QR code.
Keep private and protected information out
QR routes do not encode patient information, family case details, protected cultural knowledge, or other confidential material.