QUESTION-064: Does an anchor need endorsement as well as addressability, stable identity and discoverability: are entries the record has not accepted built on less than entries it has?

Why it is a question

QUESTION-038 asks what turns a structure that persists into one later structure can be built on, and names three candidates: addressability, stable identity and discoverability. Clark and Chalmers’ Otto, whose notebook stands in for his memory, is the closest case the record holds of an external resource doing a mind’s remembering. NOTE-180 records the features that make that work: “constancy, direct availability, automatic endorsement, and the admittedly ‘arguable’ past conscious endorsement”. The first two have counterparts in QUESTION-038’s list. Endorsement does not.

The record has an endorsement condition of its own. A claim is Proposed or Active, a reading is Read or Skimmed, a LIT is Deferred or Active, and the owner’s acceptance moves entries between them. If endorsement matters to anchoring, Proposed and Deferred entries should be built on less than Active ones, other things equal.

What would count as an answer:

• Yes: entries that are not yet accepted draw fewer rests_on and grounds edges than accepted ones of the same age and line, beyond what the record’s inactive-ok discipline explains.
• No: they are built on as readily, so addressability without endorsement is enough.

NOTE-180 records that the criteria were read off one case, and that the fourth is flagged “arguable”. The active-memory objection, which Clark answers by conceding that the notebook is static (NOTE-363), cuts the other way here: for a collaborator who keeps no memory, static storage is the virtue (CLAIM-180).