Manual of Practical Rational Action — Version 1.0
Prepared by Dave Kelly with Claude
STATUS
This document is a standalone corpus document. It is supplementary explanatory material and is not a source of governing propositions for framework runs. Governing propositions are drawn exclusively from the primary corpus: Core Stoicism, SLE v3.1, Nine Excerpts, Sterling’s Decision Framework v3, and Sterling Activation v4.
SOURCES
Grant C. Sterling, “My Action Is My Choice" (email to the ISF) — primary source for the definition of action, the three requirements, and the reservation doctrine. Keith Seddon, Stoic Serenity (2006), pg. 30 and Seddon’s Glossary — source for the interests/projects vocabulary, the manner/content distinction, and the askēsis entry (§10). The synthesis, analysis, and manual architecture are Dave Kelly’s independent contribution. The governing framework is Sterling’s.
Part One: What an Action Is
1.1 The Stoic Definition
The first thing this manual must establish is what an action is on the Stoic account, because it is not what common usage assumes. Sterling states it directly: on the Stoic view, an “action” is the agent’s choice, not anything the agent physically does.
This is not a terminological nicety. It is the architecturally central claim on which everything else in this manual rests. When Sterling agreed to go to lunch with a colleague, walked to the restaurant, ate, and returned, he did not perform one action. He performed a sequence of distinct choices: the choice to agree to go, the choice to walk out the door, the choice to continue toward the restaurant, the choice to converse and say particular things, the choice to order a specific item. Each of those choices was, at the instant it was made, either appropriate or inappropriate. Each was a separate moral event.
The physical events — the walking, the eating, the returning — are not the actions. They are the externals that follow from actions, or that were aimed at by actions, or that accompanied actions. They are outside purview in the sense that Providence can interrupt them at any point: a car could have struck Sterling before he reached the restaurant; the restaurant could have been closed when he arrived; the colleague could have changed his mind. None of that would have altered the appropriateness of the choices already made. The choices were complete at the instant of making.
Sterling’s formulation: “My choice to agree to go was based on several considerations — I needed to eat some food, the walk would give me exercise, the weather was nice, the restaurant has good food that is not too expensive, the other professor is a colleague in my department so the conversation was likely to be both enjoyable and productive, etc. Given these considerations, I think it was correct…rational…appropriate of me to agree to accompany him when he asked me to go.”
Notice what governs the determination of appropriateness: the considerations available to the agent at the moment of choice, assessed rationally. Not the outcome. Not what Providence subsequently arranges. The action is appropriate or inappropriate at the instant the choice is made.
1.2 What This Changes
If physical outcomes determined the appropriateness of actions, then the appropriateness of any given choice would be unknown until the outcomes were in — and since outcomes depend on Providence as well as on the agent, appropriateness would be partly outside purview. That would be incoherent. The agent cannot be held morally responsible for what he cannot control.
The Stoic account removes this incoherence. Appropriateness is determined entirely by what the agent controls: the rationality of his assessment of the situation, the correctness of the goal he identifies, the rationality of the means he selects, and the reservation with which he holds the whole. If he has done all of these correctly, the action is appropriate. Full stop. What happens next is Providence’s domain, not his.
Sterling’s example makes this vivid: “A choice to unnecessarily walk along ice and dangerous sidewalks is inappropriate, even if we manage to safely negotiate the dangers unharmed.” The inappropriate choice does not become appropriate because the outcome happened to be fine. By the same token, an appropriate choice to take a route that turned out to be blocked does not become inappropriate because the destination was not reached. The choice and the outcome are morally separate events.
Part Two: The Three Requirements of Rational Action
Sterling identifies three requirements that every rational action must satisfy. These are not a procedure to be worked through consciously at each moment for simple decisions, but they are the structure that governs whether any given choice is appropriate or not. The agent who has internalized them applies them as a matter of formed judgment.
2.1 Requirement One: Identify a Rational Goal
The agent must be aiming at something that is rational to aim at. Not everything an agent might want to pursue is a rational goal. Sterling specifies two conditions under which a goal would fail this requirement: if pursuing it would be immoral (the restaurant’s proceeds sponsor terrorist attacks) or if it would be irrational (the restaurant is prohibitively expensive, or is known to serve spoiled food).
In Sterling’s framework, rationality of a goal is not separable from its moral status, because the framework’s ethics are rationalist: what is genuinely good is determined by correct reason, not by desire. A goal generated by false value judgment — by treating a preferred indifferent as a genuine good, or a dispreferred indifferent as a genuine evil — is not a rational goal in the framework’s sense, even if it is a goal the agent earnestly wants to pursue.
Seddon’s interests/projects vocabulary provides the practical entry point for this requirement. When the agent asks whether his goal is rational, the first question is whether that goal concerns a genuine good or a preferred indifferent. Interests — earning income, maintaining health, raising children, gaining education — concern preferred indifferents. They are rationally appropriate objects of aim. They are not genuine goods whose attainment would constitute real benefit, nor genuine evils whose frustration would constitute real harm. The singular exception is the interest in perfecting one’s character. That is not a preferred indifferent. Virtue is the only genuine good. Its success is entirely within purview at every moment.
The practical implication for identifying a rational goal: the agent who aims at a preferred indifferent — a meal, a productive conversation, a completed task, a maintained relationship — is aiming at something rational. The agent who aims at the same external object but holds it as a genuine good whose attainment is necessary for his wellbeing is not, in the framework’s strict sense, identifying a rational goal. He is identifying a desired outcome and treating it as something it is not.
2.2 Requirement Two: Select Rational Means
Having identified a rational goal, the agent must select means that are themselves rational — that is, means that are genuinely designed to realize the goal, that are not themselves immoral, and that are proportionate to the circumstances.
Sterling’s example of irrational means: the restaurant is 50 miles away and the agent plans to walk there during a lunch hour. The means fail not because they are immoral but because they are not genuinely designed to realize the goal given the constraints. The agent who selects irrational means has made an inappropriate choice even if his goal was entirely rational.
Sterling adds a qualification in the postscript that this manual must preserve: on rare occasions it will be rational to perform a task in an inferior way. The agent may have good reason to shower but equally good reason to spend no more than five minutes doing so. He cannot shower thoroughly in five minutes, so he showers somewhat sloppily. This is not an inappropriate choice. It is a rational trade-off between competing rational goals given real constraints. The agent who demands perfect execution of every means regardless of circumstances is not being virtuous — he is failing to assess the whole situation rationally.
Seddon’s distinction between the content of a project and the way it is carried out speaks directly to Requirement Two. The way the agent carries out the project — the means he selects and how he executes them — is entirely within purview and is where the moral character of the action is located. Two agents pursuing the same goal by the same general means can still differ in whether their pursuit is virtuous, depending on the specific manner of execution: whether it is honest, whether it respects the role-duties the situation generates, whether it treats others as their roles require.
2.3 Requirement Three: Act with Reservation
This is the requirement that most sharply distinguishes rational action from ordinary goal-directed behavior, and it is the one Seddon’s vocabulary alone does not supply. Sterling formulates it precisely: the agent must make all his choices “with the reservation that these outcomes are never really under my control, and so if the all-wise gods will otherwise — not my will but theirs be done.”
Sterling’s illustration is luminous. Having made all correct choices on the way to the restaurant, the agent arrives to find it closed. “I am not in the least upset, because all along I was not aiming to produce the outcome of eating at that restaurant, but rather aiming at the outcome of eating at that restaurant if possible. Now I recognize that it was not possible — the gods did not will it. Nevertheless all my choices were correct at the time, and so I am content.”
This is not a subtle point. It is the hinge on which the entire account of rational action turns. The agent who aims at eating at the restaurant simpliciter has made the outcome a condition of his contentment. The agent who aims at eating at the restaurant if Providence allows has made his contentment depend entirely on his own correct choosing — which is always within purview. The first agent’s happiness is hostage to the closing hours of a restaurant. The second agent’s is not.
Reservation is not passivity or indifference. The agent who acts with reservation still selects the most rational means to the best rational goal he can identify. He still acts with full attention and genuine effort. What he does not do is treat the outcome as a condition of the action’s appropriateness, or treat its frustration as something genuinely bad. He holds the aim correctly — as a preferred indifferent pursued with reservation, not as a desired outcome held as a genuine good.
Part Three: The Grain of Daily Life
3.1 Actions Are Granular
Sterling’s account of the lunch example reveals a fact about rational action that practical manuals often obscure: action is granular. What appears from the outside as a single activity — going to lunch — is from the inside a continuous sequence of distinct choices, each of which is appropriate or inappropriate in its own right.
The choice to agree was one action. The choice of route was another. The choice of what to order was another. Each conversation turn was an action. Each decision about how to respond to the colleague’s statements was an action. The choice of how to handle the bill was an action. None of these is trivial. Each is a site where the agent either chooses correctly or does not.
Seddon’s interests/projects structure maps onto this granularity. An interest (maintaining collegial relationships) may generate a project (regular shared meals). A project generates a sequence of specific actions. Each action in the sequence is subject to all three requirements: rational goal, rational means, reservation. The interest and the project describe the general shape of what the agent is doing. The action is what the agent does at each specific moment within that shape.
This matters because it means that no moment in daily life is too small to be the site of correct or incorrect action. The manner in which the agent speaks to a colleague during a meal is an action. The manner in which he handles an unexpected obstacle — the closed restaurant — is an action. The manner in which he responds to frustration is an action. These are not morally neutral background events. They are choices, made at instants, each appropriate or inappropriate.
3.2 The Full Scope of Seddon’s Distinction
Seddon distinguishes between interests and projects on the one hand and the way of carrying out projects on the other. The distinction is between what the agent is doing and how he is doing it. Seddon identifies the “how” as what is good or bad and what is of supreme importance.
Sterling’s account of the lunch example shows that the “how” is not a single judgment made once about the general manner of the project. It is a series of specific choices, each of which constitutes the “how” of that particular moment. The way of carrying out the project is not a background feature — it is the accumulation of the agent’s choices at each moment of the project’s execution.
This has a liberating practical implication. The agent does not need to make a single global commitment to pursue an interest virtuously and then execute mechanically. He makes the commitment anew at each moment of choice. Each moment is within purview. Each moment can be correct regardless of what preceded it. If earlier moments in the same project were handled incorrectly, the agent cannot go back and correct them — they are complete. But the present moment is always available for a correct choice.
3.3 Interests as the Map; Actions as the Territory
The practical relationship between Seddon’s interests/projects structure and Sterling’s action account can be stated precisely. Interests describe the standing concerns that organize the agent’s daily life at the level of general orientation. Projects describe the organized activities that serve those interests. Actions are what the agent actually does, moment by moment, in the course of executing those projects.
The interests/projects structure is the map. The action is the territory. The map is useful for understanding the general shape of what the agent is doing and why. But the moral work happens in the territory — at the level of specific choices made at specific moments under specific circumstances.
Two agents can have identical maps — identical interests, identical projects — and differ entirely in the territory, because one makes correct choices at the moments of action and the other does not. And two agents can have different maps while both making consistently correct choices in the territory. The map describes the content. The territory is where virtue lives.
Part Four: The Three Practical Operations
The three requirements of rational action — rational goal, rational means, reservation — translate into three practical operations the agent applies at the moment of choice. These are not the six steps of the Decision Framework, which govern deliberate analysis of significant decisions. They are the moment-by-moment operations that constitute correct action in ordinary daily life.
4.1 The Goal Check
Is what I am about to pursue a rational goal?
The agent identifies, before acting, what he is actually aiming at. In many cases this is simple and obvious: he is aiming at a meal, a conversation, a completed task, a maintained relationship. In other cases it requires more careful attention, because what appears to be a rational goal may on inspection be an object of desire held as a genuine good.
The markers of desire masquerading as rational aim: the agent notices that he needs this specific outcome rather than merely preferring it; that a failure to achieve it would feel like something bad happening to his rather than merely something rationally dispreferred; that his assessment of his own situation is tied to whether this outcome is achieved. These are markers of a goal held as a genuine good rather than a preferred indifferent. The Goal Check names this and corrects it before the action proceeds.
Common categories where false goals appear dressed as rational ones: pursuing income in a way that treats a specific level of income as necessary rather than rationally preferred; pursuing health in a way that treats a specific outcome as a condition of wellbeing; pursuing approval or reputation as though these were genuine goods whose absence constitutes genuine harm; pursuing the success of a project in a way that makes the project’s completion a condition of contentment rather than a preferred indifferent held with reservation.
None of these interests are irrational. All of them are appropriate objects of aim. The error is in how they are held, not in whether they are pursued.
4.2 The Means Check
Are the means I am selecting rational, and is the manner of my execution honest, role-appropriate, and proportionate?
The Means Check covers two distinct questions that Seddon’s distinction between content and manner helps separate.
The first question is whether the means are rational in Sterling’s sense: genuinely designed to realize the goal, not immoral, not wildly disproportionate. This is largely a question of practical reasoning. The agent identifies the most rational path to the goal given the actual constraints of the situation, including time, resources, the interests of others, and the requirements of his various roles.
The second question is whether the manner of execution is virtuous. The same means, executed in different manners, constitute different actions. A difficult conversation can be conducted honestly or dishonestly. A task can be performed with genuine attention or with performed attention directed at the appearance of effort. A role-duty can be discharged with genuine care or with compliance that satisfies the external requirement while internally treating the object of the duty as an obstacle. The manner is entirely within purview, and it is what is good or bad — not the task performed or the obligation discharged.
Seddon’s §46 specifies that the manner must be role-appropriate: the agent discharges what his social roles require. This is a real constraint on the Means Check. It is not enough for the means to be rational in the abstract. They must also honor what the agent’s roles require of him. The parent’s means must be appropriate to a parent. The colleague’s means must be appropriate to a colleague. The role specifies part of what rationality requires in the specific situation.
4.3 The Reservation Check
Am I holding this aim with reservation — aiming at it if Providence allows — or am I making my contentment conditional on its achievement?
This is the check that most directly expresses Sterling’s account and that Seddon’s vocabulary does not by itself supply. It is applied at the moment of action but its effect is prospective: it determines how the agent will relate to outcomes when they arrive.
The practical marker of reservation correctly held: the agent can complete the formulation “I am aiming at X if Providence allows” honestly. He is not merely saying this as a formula while internally treating X as a condition of his contentment. He genuinely aims at X as a preferred indifferent, acknowledging that its achievement is not within his complete control, and that its non-achievement would be rationally dispreferred but not genuinely bad.
The practical marker of reservation incorrectly held: the agent discovers, when the outcome is frustrated, that he is upset in a sense that goes beyond rational acknowledgment of a dispreferred result. He is upset because something he was treating as a genuine good has failed to materialize. This is the signal that the Reservation Check was passed nominally but not actually. The work going forward is to return to the correct relationship to the goal: aim at it, if pursuing it again is still rational, but hold the aim correctly this time.
Sterling’s Gethsemane comparison is worth dwelling on. Jesus’s prayer — “if you are willing, take this cup from me; yet not my will, but yours be done” — expresses the structure exactly. The agent identifies what seems best given what he can assess. He pursues it. He acknowledges that Providence may will otherwise, and that if it does, that is not a genuine evil. This is not resignation or indifference. It is the only rational relationship to outcomes that are not fully within the agent’s control.
Part Five: Virtue, Appropriateness, and the Honest Scope of the Manual
5.1 The Distinction Between Appropriate Action and Virtuous Action
Sterling marks a distinction in the lunch email that this manual must preserve. Appropriate action and virtuous action are not identical. To go beyond making appropriate choices and achieve virtue, the agent must make appropriate choices and those choices must be connected together in a settled disposition to rationally evaluate all information that comes to him.
On the classical Stoic account, one cannot perform a single virtuous action. Virtuous actions come when the agent has reached the stage where his inner rational development has been perfected — which, strictly speaking, only the sage achieves. Sterling is willing, as a practical matter, to call some actions virtuous. But the honest scope of this manual is the domain of appropriate action: choices that satisfy the three requirements, made by an agent who has not yet achieved the settled rational disposition of the sage.
This limitation is not discouraging. The domain of appropriate action is exactly where the agent who is making progress in Stoic practice lives and works. The manual’s three operations — Goal Check, Means Check, Reservation Check — describe what appropriate action requires. They do not claim to describe what full virtue requires, which is something that emerges only from the long sustained work of character development.
5.2 The Long Work
Seddon’s §10 identifies askēsis — practice, exercise, training — as the sustained work by which the agent builds the capacity for correct action. The long work is the accumulation of appropriate choices over time, in the direction of an increasingly settled rational disposition.
The accumulation is itself held with reservation. The agent does not pursue character development as a genuine good whose achievement would constitute a real benefit. He pursues it as the work that correct rational activity requires of him, acknowledging that Providence governs whether and when the settled disposition emerges. What he controls is the next choice. That choice is always within purview.
Two supporting practices serve the long work. Retrospective review examines past choices not to punish but to identify where the three requirements were failed — where the goal was held as a genuine good, where the means were irrational or manner was distorted, where reservation was nominal rather than real. This examination is itself an action, made at a moment of choice, and is held with reservation.
Prospective preparation formulates correct propositions before entering situations where the three checks are likely to be difficult. The form is established in Nine Excerpts Section 7: the external object at stake is not in my control; its attainment or frustration is neither good nor evil in itself; my capacity for correct action is intact regardless of outcome. Assenting to these propositions before the situation begins means that the moment of action is not the first time the agent has engaged the correct value judgment.
Part Six: Summary
The Definition
An action is a choice, not a physical event. Each choice is appropriate or inappropriate at the instant it is made. Outcomes do not determine appropriateness. Appropriateness is determined entirely by what the agent controls: the rationality of his goal, the rationality of his means, and the reservation with which he holds the whole.
The Three Requirements
Rational Goal: The agent aims at something that is rational to aim at. On Sterling’s framework, rational goals are preferred indifferents held as appropriate objects of aim, not objects of desire held as genuine goods. The singular exception — the interest in perfecting one’s character — is not a preferred indifferent but the only genuine good, and its success is entirely within purview at every moment.
Rational Means: The agent selects means genuinely designed to realize the goal, that are not immoral, and that are proportionate to the circumstances. He may on occasion perform a task in an inferior way when competing rational considerations require it. The manner of execution — honest, role-appropriate, attentive — is itself within purview and is where virtue is located at the level of concrete activity.
Reservation: The agent aims at the goal if Providence allows, not unconditionally. He does not make his contentment conditional on the outcome. When the restaurant is closed, he is content, because all along he was aiming at eating there if possible — not aiming at eating there as a condition of his wellbeing. He now makes new choices about what to do.
The Three Practical Operations
Goal Check: Is what I am about to pursue a rational goal — a preferred indifferent held as an appropriate object of aim — or a desired outcome held as a genuine good?
Means Check: Are my means rational and proportionate? Is the manner of my execution honest, role-appropriate, and genuinely aimed at the goal?
Reservation Check: Am I holding this aim with reservation — aiming at it if Providence allows — or am I making my contentment conditional on its achievement?
The Governing Formulation
We should always be pursuing our interests as appropriate objects of aim — that is, as preferred indifferents — and not as objects of desire. The appropriateness of the pursuit is determined not by its external shape and not by its outcome, but by the value judgment operative in the agent at the moment of action. That moment is always within purview. That judgment is always the agent’s to make.
ATTRIBUTION NOTE
The Sterling email is Grant C. Sterling’s work. The Seddon passages are from Keith Seddon, Stoic Serenity (2006), pg. 30 and Seddon’s Glossary. The synthesis, analysis, and manual architecture are Dave Kelly’s independent contribution. The governing framework is Sterling’s. The attribution standard of the Project applies throughout: Sterling’s theoretical contributions and Dave Kelly’s practical contributions are not conflated.
Manual of Practical Rational Action — Version 1.0. Dave Kelly. Standalone corpus document.
No comments:
Post a Comment