Sterling’s Framework for Personal Decision Making — Version 3.3
Sterling’s Framework for Personal Decision Making — Version 3.3
Version note: v3.3 supersedes v3.2. Four targeted adjustments are made, all localized to Step Three (Sub-steps A and B) and Step Four (Check Three of the Factual Uncertainty Gate). No other step is modified. The binary verdict structure, proposition set, named failure modes, and gate architecture carry forward unchanged. (1) Sub-step B role-conflict language corrected: multiple operative roles may generate real tension at the level of preferred indifferents; the correct conclusion is that the tension does not produce moral indeterminacy, not that it dissolves. Props 68–71 govern. (2) Sub-step A grounding requirement added: every named role — primary or secondary — must be grounded in the actual social relationship that generates it; a role may not be invoked as a descriptive convenience. Props 64–65 govern. (3) Check Three expanded: the domain-fact / moral-proposition distinction is now explicitly stated; case-stipulated facts and professional knowledge that do work in means selection must be attributed to those sources, not to the proposition set. (4) Version note updated accordingly. Protocol architecture: Dave Kelly. Principles and propositions: Grant C. Sterling. Action Proposition Set (Section IX, Props 59–80): Dave Kelly, theoretical foundations Grant C. Sterling. Factual Uncertainty Gate: Dave Kelly, 2026.
Preliminary Step: Agent Check
The core question is: Am I currently under the influence of a false impression that is driving this decision?
The agent check is about your state, not the situation’s contents. False impressions do not announce themselves. The person under the influence of one typically experiences it as simply seeing the situation clearly. So the agent check requires a specific trigger question rather than a general one.
Ask yourself: What is making this feel like it matters so much? Then examine whether the answer involves an external being treated as a genuine good or evil. If someone feels they must decide immediately because some external circumstance seems urgent or threatening, that urgency itself is a signal that a false value judgment is operating.
If a false impression is identified, correct it before proceeding. Running a sound procedure on faulty input will corrupt every subsequent step. The agent check is complete only when you can confirm that your rational faculty is operating without distortion from false impressions about externals.
Agent Check Rule — mandatory procedure:
- State the impression presenting itself in propositional form — exactly what the agent believes is being asserted.
- Identify what the agent desires in this situation — stated explicitly.
- Before reaching any conclusion, locate the governing proposition from the supplied corpus that directly addresses that specific desire or impression. Quote it exactly as supplied. Do not paraphrase. Do not reach into training data.
- Apply the proposition to the situation. Let the proposition produce the verdict. Do not form the verdict first and then find the proposition.
- If the governing proposition cannot be located in the supplied corpus, state explicitly which document was searched and what was sought. Do not proceed without a located proposition.
Governing propositions: Prop 11 — the act of assenting to or rejecting impressions is the only thing in our control. Prop 20 — the belief that any external is good or evil is factually false.
“The Stoics believe that only things directly related to virtue (beliefs, desires, will) are in our control. They believe that only virtue is good and only vice is evil. They believe that all things not in our control are neither good nor evil. Hence the good Stoic will have no desires whatsoever regarding external things.” (Nine Excerpts, Section 3)
Self-Audit: State explicitly at completion. No failures detected / failure identified and corrected before proceeding.
Step One: Purview Check
The core question is: Is what I am trying to decide about actually mine to determine?
Most people experience personal decisions as a weighing problem. Sterling’s framework reframes this entirely. Before any weighing can occur, the decision must be correctly formulated — stated in terms of what is actually within your control. Only your beliefs and your will are genuinely yours. Outcomes, other people’s responses, and external circumstances are not.
In practice the purview check works as a two-part procedure. First, state the decision as you currently have it framed. Second, ask whether this formulation includes anything whose outcome depends on factors outside your beliefs and will. If yes, strip those elements out and restate. Keep restating until what remains is formulated purely in terms of your own beliefs and will. That restated version is the actual decision you are facing.
This step often reveals that the decision is much simpler than it appeared. Or it reveals there is no decision at all — only a situation to be accepted and responded to virtuously. If the remaining decision feels trivial or anticlimactic after the restatement, that is a signal the purview check worked correctly. The false weight has been identified and removed.
Governing propositions: Prop 11 — the act of assenting to or rejecting impressions is the only thing in our control. Prop 16 — only things directly related to virtue are in our control.
“Only things directly related to virtue (beliefs, desires, will) are in our control.” (Nine Excerpts, Section 3)
Self-Audit: State explicitly at completion. No failures detected / failure identified and corrected before proceeding.
Step Two: Value Strip
The core question is: Am I treating anything in this situation as a genuine good or evil that is actually an indifferent?
This step concerns the situation’s contents, not your condition as agent — that was addressed in the preliminary step. Here you take everything that remains after the purview check and classify each element correctly: virtue, vice, or indifferent.
The list of indifferents is long and includes things people routinely treat as genuine goods — health, financial security, relationships, reputation, career advancement, comfort, other people’s approval. All of these belong in the indifferent column. Only virtue is a genuine good. Only vice is a genuine evil. Everything else is indifferent.
In practice this step involves two moves. First, list everything in the situation that feels like it is at stake. Second, ask of each item: is this something whose presence or absence would make me a better or worse person, or just a more or less fortunate one? Only virtue and vice affect the first. Everything else affects only the second. That distinction is the heart of the value strip.
The practical difficulty is that indifferents do not feel indifferent. They feel like they matter enormously. The value strip does not ask you to stop feeling that — it asks you to correctly classify what you are feeling about. The feeling can remain. The false moral weight gets removed.
What gets reassigned in this step is important. The indifferents are not discarded — they are reclassified as the practical context within which the real decision will be made. They inform the shape of the action without determining its moral content.
Note on Theorem 19: Positive feelings that arise in the present moment without desire are not pathological. However desiring to achieve such feelings or desiring them to continue beyond the present involves the judgment that they are good, which is a false value judgment. The feeling arriving is legitimate. The desire for more of it is not.
Governing propositions: Prop 17 — only virtue is genuinely good; only vice is genuinely evil. Prop 20 — the belief that any external is good or evil is factually false. Prop 22 — preferred indifferents are appropriate objects to aim at, though not genuinely good. Props 23–26 — all emotions caused by beliefs about external value are pathological.
“All emotions are caused by beliefs about what is good or evil. Specifically, emotions result from beliefs that externals have genuine value. All beliefs that externals have value are false.” (SLE v4.0, Section IV)
“Such positive feelings are not irrational or inappropriate. Though if we desire to achieve them or desire for them to continue beyond the present, then that would involve the judgment that they are good, and hence that would be irrational.” (Nine Excerpts, Theorem 19)
Self-Audit: State explicitly at completion. No failures detected / failure identified and corrected before proceeding.
Step Three: Virtue Identification
The core question is: Which preferred indifferent is the appropriate object of aim in this situation?
The previous steps have cleared the ground. The agent is operating without distortion, the decision is correctly framed within purview, and false value has been stripped from the situation’s contents. What remains is a clarified situation whose appropriate object of aim is now visible without obstruction.
Virtue consists of the pursuit of appropriate objects of aim, not the pursuit of the external objects of our desires. Appropriate objects of aim are preferred indifferents. They are not genuine goods. They are what reason identifies as rationally correct to pursue given the situation, with the reservation that their achievement is not in the agent’s control and is not the agent’s good.
Identifying the appropriate object of aim in a specific situation requires three propositionally governed sub-steps, now explicit:
Sub-step A — Role identification. Identify all roles the agent currently occupies that are operative in this situation (Prop 64). Roles are identified by actual social relationships, not by preferences or desires (Prop 65). When the situation makes a role operative, that role’s duties govern (Prop 66). Each role named — primary or secondary — must be grounded in the actual social relationship that generates it. A role may not be invoked as a descriptive convenience or as a repository for manner considerations that have not been independently established. If a secondary role is named, state explicitly what actual social relationship generates it and what specific duties it creates in this situation.
Sub-step B — Role conflict resolution. Where multiple roles are present, identify which role is most directly operative (Prop 70). Apply the determination rule: all other things being equal, maximize preferred indifferents across all operative roles simultaneously (Prop 69). This is a necessary moral truth known by reason, not a calculated outcome. Multiple operative roles may generate real tension at the level of preferred indifferents even when they do not generate moral indeterminacy. The correct conclusion of this sub-step is not that the conflict dissolves but that it does not produce moral indeterminacy: reason, working from Props 68–71, is competent to determine which preferred indifferent is the appropriate object of aim. The tension is real; the determination problem is soluble.
Sub-step C — Candidate selection. Where multiple preferred indifferents present themselves as candidates, select the one whose pursuit maximizes preferred indifferents accessible across all operative roles (Prop 71). A preferred indifferent that a role makes appropriate cannot be displaced by personal desire for a different preferred indifferent (Prop 72).
The output of this step is a specific preferred indifferent identified as the appropriate object of aim — stated plainly — with the operative role or roles named.
Governing propositions: Prop 22 — preferred indifferents are appropriate objects to aim at, though not genuinely good. Prop 60 — a rational goal is a preferred indifferent held as an appropriate object of aim, not a desired outcome held as a genuine good. Props 64–66 — role identification: roles are determined by actual social relationships; operative roles govern action. Props 68–72 — resolution of multiple roles and competing preferred indifferents; determination rule; role conflict resolution.
“Virtue consists of the pursuit of appropriate objects of aim, not the pursuit of the objects of our desires. Such virtuous acts will give us good feelings, and since we have no desires regarding the actual outcome, they will never produce unhappiness for us.” (Nine Excerpts, Theorem 29)
Self-Audit: State explicitly at completion. No failures detected / failure identified and corrected before proceeding.
Step Four: Action Determination
The core question is: What does pursuing the appropriate object of aim require of me in these specific circumstances?
This is importantly different from asking what will produce the best outcome. The action is not determined by calculating results — it is determined by what the appropriate object of aim demands given the indifferents as practical context. Virtue is not the achievement of the aim. Virtue is the pursuit of the appropriate aim with reservation — acknowledging that the outcome is outside purview and is not the agent’s good.
Factual Uncertainty Gate
Run before Move One. Mandatory. The gate does not suspend action determination. The moral structure established by Steps 0–3 — the false value judgment identified, the appropriate object of aim determined, the operative roles named — is not affected by factual uncertainty and is not revisited here. The gate addresses a prior question: whether the facts the agent actually has access to provide sufficient basis for means selection.
Gate question: Do the facts I actually have access to provide sufficient basis for means selection, or does this action ruling depend on facts that are unavailable, uncertain, or beyond the framework’s domain?
Check One — Facts in hand. State explicitly what facts the agent has direct access to: the situation as it actually presents itself, the roles that are operative, the constraints that are known. Do not import assumed facts. Do not import probable facts as though they were certain. The Purview Check (Step 1) established what is within the agent’s control; this check establishes what is within the agent’s knowledge.
Check Two — Dependence test. Ask: does the means selection about to be made depend on a fact that is not in hand? Three outcomes are possible:
- Fact is known: proceed to Move One without modification.
- Fact is uncertain but the action does not depend on it: note the uncertainty explicitly; proceed to Move One. The uncertainty is carried into the reservation at Move One (Prop 62): the agent aims at the goal if the control dichotomy allows, which includes if the uncertain fact resolves in a way that makes the chosen means rational.
- Fact is unknown and the action materially depends on it: state explicitly which fact is missing and why means selection cannot be completed without it. Do not proceed to fabricate a means on the assumption the missing fact will resolve favorably. The rational means becomes: acquire the missing fact if it is acquirable within the agent’s purview, then return to Move One. If the fact is unacquirable, the rational means is the best available action given acknowledged ignorance — stated as such, with reservation explicitly broadened to cover factual uncertainty as well as outcome uncertainty (Prop 62).
Check Three — Domain boundary. Ask: does this situation require technical expertise, institutional knowledge, or domain-specific judgment that lies outside the framework’s scope? The framework determines the moral structure of action — the appropriate object of aim, the operative roles, the required manner. It does not generate medical diagnoses, legal advice, engineering assessments, or organizational strategy. Where domain expertise is required for means selection, declare this explicitly: the moral structure is determined; the means selection requires domain knowledge the framework does not supply. The agent is directed to acquire that knowledge before proceeding, or to act with explicit acknowledgment that the means are selected under domain uncertainty, carried into the reservation at Prop 62.
This distinction must be stated plainly in the gate declaration wherever it applies: the propositions govern the moral form of the action; domain-specific facts stipulated by the situation or supplied by the agent’s professional competence govern the content of the means. These are separate inputs. When case-stipulated facts or professional knowledge are doing work in means selection, that work must be attributed to those sources, not to the proposition set. The propositions do not generate factual claims about medicine, law, engineering, or any other domain. They govern what is done with the facts once known.
What the gate does not do. The gate does not introduce “inconclusive” as a verdict. The gate does not address moral underdetermination — the framework holds (Prop 68) that reason is competent to determine the right action from the propositions alone. The gate addresses factual underdetermination only: where the propositions are clear but the facts needed to select specific means are not in hand.
Governing proposition: Prop 73 — the agent selects the means most genuinely designed to realize the goal given the actual constraints of the situation. Factual uncertainty is an actual constraint. Means selected without acknowledging factual uncertainty are not genuinely designed to realize the goal — they are designed to realize the goal if the assumed facts are correct. That conditional must be made explicit and carried into the reservation at Prop 62.
Mandatory Gate Declaration (must appear in output before Move One): Facts in hand: [list]. Uncertain facts: [list or “none identified”]. Action dependence on uncertain facts: [none / noted and carried into reservation / requires acquisition before proceeding]. Domain knowledge required: [none / (domain) required — declared]. If this declaration is absent, the gate has not been run. Absence of the gate declaration is a named failure mode (FACTUAL UNCERTAINTY GATE BYPASS — see Named Failure Modes).
Action determination proceeds through four propositionally governed moves:
Move One — Means identification. Take the appropriate object of aim from Step Three and identify means genuinely designed to realize it given the actual constraints of the situation: time, resources, the requirements of all operative roles, and the rational goals simultaneously in play (Props 61, 73). There is no requirement of perfect means when good means are available and perfect means are not. Where the gate has declared factual uncertainty, carry that uncertainty into the means selection explicitly: the means chosen is the most rational means given what is actually known, held with reservation broadened accordingly (Prop 62).
Move Two — Manner check. The manner of execution is role-constrained and entirely within purview (Prop 67). The same means executed in different manners constitute different actions. The manner must be honest, role-appropriate, and genuinely attentive (Prop 74). Selecting rational means but executing them in a manner that violates role-duty or honesty is an inappropriate action regardless of the rationality of the selection.
Move Three — Appearance check. Confirm the action is chosen because it is the rational means to the rational goal — not because it appears virtuous to others (Prop 75). The external appearance of virtue is an indifferent. Action chosen for appearance is desire for an external outcome dressed as a rational goal.
Move Four — Verification test. Ask: would this action still be chosen if the emotional charge present in this situation were removed entirely (Prop 76)? If yes, proceed. If no, the action is grounded in desire from false value judgment — return to Step Two. The presence of pathos does not automatically disqualify an action if it can be identified as directed at a rational goal by rational means, but requires the verification test be applied with particular care (Prop 77).
The output of this step is a specific action — something concrete that can actually be done, expressed in terms of the agent’s own beliefs and will, with the operative role named, the gate declaration stated, and the governing Section IX propositions cited.
Governing propositions: Prop 59 — every rational action has three components within purview: goal, means, reservation. Props 61, 73 — rational means: genuinely designed to realize the goal, proportionate, not immoral, sensitive to competing rational goals. Prop 62 — reservation is constitutive of every rational act of will. Props 67, 74 — manner is role-constrained and entirely within purview. Prop 75 — action chosen for appearance is not a rational action. Props 76–77 — verification test.
“A rational act of will involves: (a) Identifying rational goals to pursue (preferred indifferents); (b) Selecting rational means designed to help realize these goals; (c) Making these choices with reservation — acknowledging that outcomes are in the hands of Providence.” (SLE v4.0, Section V, Prop 35)
Self-Audit: State explicitly at completion. No failures detected / failure identified and corrected before proceeding.
Step Five: Outcome Acceptance
The core question is: Can I release what follows from the action I have taken?
The action having been taken, the moral work is complete. What follows belongs to circumstances outside purview. Outcomes are indifferent. Attaching moral weight to results at this stage would reintroduce exactly the false value that the value strip removed — it would treat the outcome as though it were a genuine good or evil rather than an indifferent.
The quality of the decision is determined entirely by the quality of the deliberation and the virtue expressed in the action — not by what the action produces (Prop 63). A decision made correctly through this framework is a good decision regardless of its results. A decision made badly does not become good because it happens to produce a favorable outcome.
This step is not passive resignation. It is the correct understanding of where moral responsibility ends. The agent is responsible for the quality of beliefs, will, and the virtue brought to action. What lies outside purview is not the agent’s responsibility. Outcome acceptance is the recognition of that boundary.
Where the outcome is frustrating, prospective preparation formulated before the situation (Prop 78) is what makes release possible without distress. Where past choices failed the three requirements, retrospective review identifies the failure without punishment and orients the next choice correctly (Prop 79). The accumulation of correct choices over time is the work of character development — the only genuine good — pursued through the sequence of individual correct choices (Prop 80).
Governing propositions: Prop 38 — the appropriateness of a choice is determined at the moment of choice, regardless of outcomes. Prop 63 — outcomes do not retroactively alter the appropriateness of a choice. Prop 62 — reservation: the agent aimed at the goal if the control dichotomy allowed; contentment was never dependent on the outcome. Props 78–80 — prospective preparation, retrospective review, accumulation of correct choices.
“The appropriateness or inappropriateness of a choice is determined at the moment of choice, regardless of outcomes.” (SLE v4.0, Section V, Prop 38)
Self-Audit: State explicitly at completion. No failures detected / failure identified and corrected before proceeding.
Named Failure Modes
REASSURANCE BIAS: The instrument has a training-level disposition toward validating the user’s emotional state and producing reassuring conclusions. This disposition is a malfunction in this context. Any conclusion reached before the governing proposition has been quoted exactly from the supplied corpus must be treated as a reassurance bias failure and discarded. Restart the step.
MID-SESSION TEXT REQUEST: The instrument may not ask the user to supply propositions or corpus text during a decision session. The uploaded documents are the complete working corpus. If the instrument cannot locate the governing proposition from those documents it must state explicitly which document it searched and what it was looking for. It may not transfer the retrieval burden to the user under any circumstances.
TRAINING DATA CONTAMINATION: The instrument may not reach into training data on Stoicism, classical virtue ethics, or any other philosophical framework to supplement the supplied corpus. When a step requires grounding not present in the supplied corpus, the instrument must name the gap explicitly rather than filling it from training data. Action audit steps that do not cite specific Section IX proposition numbers by name are presumed to be training data contamination failures until the citations are supplied.
PREFERRED INDIFFERENT AS GENUINE GOOD: The instrument may not classify a preferred indifferent as a genuine good at any step. Health, relationships, knowledge, justice, truth-telling and all other preferred indifferents are appropriate objects of aim. They are not genuine goods. Their presence or absence does not affect the agent’s virtue. Virtue is entirely in the quality of the pursuit. Any output that treats a preferred indifferent as constitutive of the agent’s good rather than as an appropriate object of aim is a correspondence failure and must be restarted.
SECTION IX BYPASS: The instrument may not determine action at Steps 3 or 4 by training-data judgment when Section IX propositions govern the determination. Failure to cite Props 64–72 at Step 3 or Props 59–77 at Step 4 by number is a Section IX Bypass failure. State the failure, identify the uncited propositions, rerun the step with explicit propositional citation.
FACTUAL UNCERTAINTY GATE BYPASS: The instrument may not proceed from Step 3 to Move One of Step 4 without running the Factual Uncertainty Gate and producing the mandatory Gate Declaration. Absence of the declaration — Facts in hand / Uncertain facts / Action dependence / Domain knowledge required — is a named failure. State the failure, run the gate, produce the declaration, then proceed to Move One.
Principles and 58 Unified Stoic Propositions: Grant C. Sterling. Action Proposition Set (Section IX, Props 59–80): Dave Kelly, theoretical foundations Grant C. Sterling. Factual Uncertainty Gate: Dave Kelly. Protocol Architecture and Synthesis: Dave Kelly. Governing instrument: SLE v4.0 (80 Unified Stoic Propositions). Version 3.3 — four targeted adjustments to Sub-steps A and B of Step Three and Check Three of the Factual Uncertainty Gate; binary verdict structure, proposition set, named failure modes, and gate architecture unchanged from v3.2. 2026.
0 Comments:
Post a Comment
<< Home