The Occasion Router — Closing the Three Declared Gaps v1.3
The Occasion Router — Closing the Three Declared Gaps v1.3
Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Analysis and synthesis: Dave Kelly. Prose rendering: Claude (Anthropic). 2026.
Companion to The Corpus as a Thinking System v1.4 and The Subordination Test v1.0.
I. The Beginning
Everyone wants happiness. Incomplete happiness is not worth accepting if complete happiness is available. Complete happiness must therefore lie wholly within our control. Only assent lies wholly within our control. Happiness is therefore correct assent — correct when it matches the truth about value: only virtue is genuinely good, only vice genuinely evil, everything else indifferent.
The Subordination Test derived from this beginning what may be added to the system and on what terms: material that is outside the field of assent may be brought in as preparation for it, provided it never becomes a second measure of success. Three such omissions were named there and left open — deliberation among indifferents, correction of factual error, and institutional or collective analysis. This document closes them.
It closes them by citation. Every route below names a document already ratified in this corpus, states what it is for, and states the guard that keeps it subordinate. Nothing here is new architecture. Where a router would have to invent a mechanism to make a citation fit, that is the sign the citation is wrong, not a license to build the mechanism.
II. What This Document Is and Is Not
It is a routing table: for each of the three occasions, which existing instrument applies, and under what restriction.
It is not an engine, a dispatcher, or a layered architecture. Those words describe mechanisms that parse, classify, and act — verbs that belong to the practitioner, not to a document. A router does not run automatically; the practitioner recognizes the occasion and reaches for the instrument, the way he reaches for the SDF now, when the SDF applies now. Nothing here changes that. What this document supplies is a stated place to reach, so that reaching for it is recognized as legitimate rather than as inventing on the spot.
It does not touch Clause (e) or the status of joy as an immediate consequence of correct action. That question is not a missing citation; it is a standing correction already on record in Joy as Theorem, Not Premise, and any document asserting joy as following automatically from duty maintained must satisfy that essay’s argument or drop the assertion. This router does neither, because routing is not the tool for it.
III. Route One — Deliberation Among Indifferents
Occasion. An impulse or impression presents a choice among externals — competing preferred indifferents — and the fitting action is not immediately obvious to the practitioner.
Trigger condition, stated precisely. The Decision Point — We Usually Know What to Do, but Desire to Do Something Else establishes that the ordinary case is not uncertainty about the right action but resistance to a right action already seen. This route is for the genuine remainder: occasions on which the practitioner does not yet see what fits. Reaching for deliberative machinery when the action is already known and merely disliked is not this route; it is evasion of clause (b), dressed as diligence.
Instrument. The Sterling Decision Framework, v3.3, applied in full: its six steps, its Factual Uncertainty Gate, its Action Sets, its own six named failure modes. Nothing here compresses, relabels, or substitutes for that structure. A four-step paraphrase that shares the SDF’s name without its Gate is not this route.
Subordination guard. The SDF’s output is a candidate action proposition. It carries no authority until it is formulated as a true action proposition, brought consciously to mind, and assented to under clause (d) — with the reservation Excerpt 10 attaches to every choice, since the outcome is never under the practitioner’s control. The SDF prepares the material of that clause; it does not perform it. An unachieved outcome, or a defect in the SDF’s own factual premises, is not a moral failure; it is a preparation that turned out imperfect, which the framework’s own Factual Uncertainty Gate already anticipates.
IV. Route Two — Correction of Factual Error
Occasion. An impression embeds an unverified or contestable claim about a state of affairs, distinct from any claim about its value.
Instrument, in two parts. First, the Logic of Assent Instrument’s sort of propositions into value-neutral, value-bearing, and action-bearing: the practitioner’s first move is separating what the impression asserts about the world from what it asserts about worth, so that the two are corrected by different standards. Second, the standing caution supplied by two runs of the Thinking System Audit: the Cartesian-method run, showing that a method built entirely on withheld and granted assent can still leave value untouched; and the Bayesian-epistemics-as-practice run, whose central finding — Operator Thinned, the decisive act replaced by a computed magnitude — is the specific failure this route exists to avoid.
Subordination guard. A probability, a calibration, a piece of verified evidence is material for clause (d), and becomes an object of assent only when it is incorporated into a true action proposition there — not before, and never as a standalone assent to a confidence level. This is the terminus the route was missing in v1.0 and v1.1: the corrected fact does not float between preparation and act, it enters clause (d) as part of what the practitioner recognizes it would be correct for him to do. Tagging an impression with a confidence level is preparation. Treating the tag as the verdict is the exact defect the Bayesian run named.
V. Route Three — Institutional and Collective Structure
Occasion. An impression involves an institution, a multi-agent arrangement, or a structure with feedback loops, delays, and leverage points not reducible to a single choice.
Instrument. The descriptive tools examined in the Thinking System Audit’s run on systems thinking — mapping feedback loops, identifying delays, locating leverage points, and distinguishing what lies within the practitioner’s own jurisdiction from what belongs to another agent’s prohairesis. That run found the method itself Inverted at the system level, fixing the condition of the world as its end; nothing here adopts that end. What is borrowed is the mapping vocabulary, not the goal it was built to serve.
Subordination guard. Stated exactly as the Subordination Test states it: add the subject, refuse the criterion. The health, survival, or improvement of the institution is an external indifferent. Mapping where the leverage lies is material for clause (d) — for recognizing the duty connected with one’s own role and station — and is never itself the standard by which the practitioner is measured.
Where the mapping belongs. Clause (c) instructs the practitioner to formulate true propositions in advance as far as possible. Standing institutional facts — jurisdiction, delegation, who decides what — are of that kind: they can be established before any occasion arises, and belong there. Mapping them under load, mid-occasion, is the degraded case, available when the advance work was not done but not the ordinary form of this route.
VI. The Relation Between Route Two and Route One
When an occasion carries an unverified fact, Route Two runs first. It separates the fact from any value attached to it and marks it as unverified. Then one question decides what follows: does the fitting action depend on whether the fact is true, or only on the fact’s being unverified?
Where the duty follows from the unverified status itself, no deliberation is needed. The action proposition goes directly to clause (d). A reviewer who lacks evidence of safety owes a request for that evidence, whatever the truth turns out to be.
Where the choice among actions materially depends on the fact’s truth, Route One runs, and the SDF’s Factual Uncertainty Gate — its Dependence Assessment — governs how to act under the uncertainty.
The error this section guards against runs in both directions: invoking Route One’s machinery when the unverified status already settles the duty, which is deliberation used to delay a known act; and skipping Route One when the right act really does depend on a fact not yet established.
VII. What Remains Untouched
Clause (e) and the status of joy as a following consequence. Left exactly where Joy as Theorem, Not Premise left it, and not routed around.
Any packaging of this router as an “engine,” a “dispatcher,” or an autonomous classifier. If such packaging is wanted later, it is Dave Kelly’s own synthesis, openly labeled as such — the way the TSA, DTI, and CSI are — built on these three routes rather than replacing them, and it would still have to name and match what it cites.
VIII. Declared Gaps
Route One assumes the SDF v3.3 text as the corpus states it; this document does not reproduce that text and a reader must consult it directly.
Route Two’s citation of the two TSA runs is to their findings, not to their full texts; a reader wanting the Cartesian method’s or the Bayesian practice’s complete architecture should consult those runs.
Route Three’s citation of the systems-thinking run is to its descriptive apparatus only; the run’s own Contrary and Inverted findings are not imported as endorsements of anything beyond the mapping tools named.
No claim in this document asserts a mechanism not already stated in the cited sources. Where that could not be achieved without inventing one, the gap is left open here rather than closed by invention.
Version Note
Version 1.3 supersedes Version 1.2. Section VI is added, and the later sections renumbered. It resolves an overlap between Route One and Route Two surfaced by the first sealed test under the Impression Test Generator, in which seal and run each took a defensible reading of an ambiguity the Router left open: an unverified fact bearing on a decision fitted both routes. The two now run in sequence, and the Dependence Assessment decides whether Route One is needed at all. The sequence rests on the Dependence Assessment as described in the SDF Scenario Generator v2.1; SDF v3.3’s own statement of the Gate was not opened for this amendment and governs where it differs.
Version 1.2 superseded Versions 1.0 and 1.1 and reversed the v1.1 correction: v1.0’s citation of clause (d) was right, and Excerpt 7 opened in full confirmed it. Route One’s and Route Two’s guards terminate in clause (d), and Route Three places standing institutional facts under clause (c)’s advance work.
Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Analysis and synthesis: Dave Kelly. Prose rendering: Claude (Anthropic). 2026.
0 Comments:
Post a Comment
<< Home