Sterling on Control — The Primary-Source Record (v1.0)
Sterling on Control — The Primary-Source Record (v1.0)
Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Analysis and synthesis: Dave Kelly. Prose rendering: Claude (Anthropic). 2026.
Context
Three ratified documents rest on two dated passages that have never been filed as primary source. Use and Correct Use of Impressions — The Genus and the Qualified Case §II quotes Sterling’s definition of control (30 August 2019) and §III quotes the yield/resist sentence to Michael Edelstein (3 September 2019); The Decision Point and the Therapy Thesis — Defense Four §V quotes the yield/resist sentence again; the CPA Run — Michael R. Edelstein records at P5 the claim Sterling was answering. The corpus files Sterling’s dated words when they carry that weight. These were not filed.
This document supplies the record. It reproduces the control passages of August and September 2019 in full, states the question each answers, gives the interlocutor’s side of the exchange, and identifies what the corpus has drawn from them. A fourth passage is added that the corpus has almost entirely left unquoted: Sterling’s 26 August account of how he decides what to do.
Every passage below was retrieved by full-thread pull during this run and read from message text. Line wrapping introduced by the list mailer has been closed up; wording, punctuation, underscore emphasis, and typographical errors are as found. No passage is quoted from a search preview.
I. The Question: How Do You Know It Is the Right Choice? (17–26 August 2019)
On 15 August, in the thread Making correct use of impressions, Sterling had written that it is virtually never the case that it is difficult to decide what is the right choice. George Richards took hold of the sentence and would not let go of it for three weeks. His first challenge, 17 August, was one line:
“‘Actually, I find that it is virtually never the case that it is difficult to decide what is the right choice.’
How to you _know_ it’s the right choice?”
Sterling’s reply, 19 August, was equally short:
“G: Because I have kataleptic impressions of the principles on which the judgments are based.”
Richards judged his own question too brief and put it again on 24 August, this time as a demand for procedure rather than for a warrant:
“Grant, consider any day-to-day situation that might crop up in your life that requires your attention. Suppose you have several options in front of you concerning what to do, including the option to do nothing. From amongst these available options, you _choose_ one of them. How do you _know_ this choice is the _right_ choice? How do you _measure_ whether this choice is the _right_ choice or not.”
Adrian answered first, on 25 August, before Sterling did — and his answer is worth recording because Sterling’s differs from it at every point. Adrian held that in most of life there is no particular right choice, only a choice that feels appropriate at that moment in time, arrived at by examining one’s impressions and judgements in full awareness of one’s oneness with God, and choosing what will be of greatest benefit to the play of life as a whole.
Sterling replied on 26 August. This is the passage the corpus has not had:
“G: I look at the likely effects of each action, and consider which are preferred and which are dispreferred indifferents. I also consider whether I have any role-duties that are relevant to the situation. Deciding the right thing to do is virtually always simple. (And my simplistic earlier answer plugs in here—I have kataleptic impressions of the truth of propositions about preferredness and duties.)
Of course sometimes there is the _mechanical_ question of which actions will be most likely to produce the best results. But those aren’t moral problems, or moral dilemmas. (I.e., I see that it is my duty to try to explain the ideas of Parmenides to my students in a way that will allow them to understand his thought. But I may be in some doubt as to whether delivering a standard lecture, assigning them to small groups for discussion, or having them read a secondary source is the most effective way to achieve that end. But this isn’t a moral dilemma, or anything that expertise in Stoicism would allow me to determine—I know what I ought to do in general, I just have to figure out which specific action is most likely to move towards that goal.)
I can’t actually remember the last time that I faced a moral dilemma. I have faced many pseudo-dilemmas that result from my not in fact living up to the standards of rationality. (To give an example not from my life but from the life of a colleague...a student offers money in return for a higher grade. The professor’s car is broken, and the money would be quite useful to him. So the professor begins rationalizing to himself—after all, no-one would really be hurt by giving this student a higher grade, and the student’s life would go better, and the student’s parents are rich, so I need the money more than they do.... After a while, it seems like we have a dilemma—it seems like we have _two_ reasonable options. But, of course, we don’t. Taking the money is wrong, it’s a violation of the teacher’s duty to grade students fairly. And the professor _knows_, deep down, that it’s wrong. But he really _desires_ the money....)
Regards, GCS”
Three things are stated here that the corpus has held on other grounds.
The first sentence is the content of clause (d) in Sterling’s own first-person voice. The corpus states clause (d) as: by attending to preferred and dispreferred outcomes and to the duties attached to one’s roles, one can recognise what would actually be correct in each situation. Sterling here reports doing exactly that — likely effects, preferred and dispreferred indifferents, relevant role-duties — as his actual procedure, not as a summary of doctrine. The clause itself is stated in Excerpt 7.
The second paragraph draws the line the corpus draws between the moral question and the mechanical one. Which action best serves a duty already known is not a moral problem and not a matter on which expertise in Stoicism has anything to say. Clause (d) is therefore not a discovery procedure; it presupposes that recognition is available and directs attention to what is already there, as The Fixed Target states.
The third paragraph is the decision point in miniature, four months before the control exchange sharpens it. The professor is not in doubt about what is right. He knows, deep down, and desires otherwise; the dilemma is manufactured by the desire. Sterling classes this as a pseudo-dilemma arising from his own failure to live up to rationality — not as a deficit of information, conviction, or emotional access. The phenomenon is stated at The Decision Point — We Usually Know What to Do, but Desire to Do Something Else.
II. The Extension: Ordinary Aims Are Not in Our Control (27 August 2019)
Richards next challenged the dichotomy itself, on 27 August, by denying that he could find any instance of the thing it warns against:
“What are these ‘things not in our control’ that we (mistakenly) ‘desire?’ I can’t think of any outcome I’ve pursued in the past that was clearly ‘not in my control.’ Perhaps I’m missing something. Can you, Grant, think any such thing that’s occurred in your lifetime?”
Sterling replied the same day:
“Evidently you are missing something. Note that the examples below are certainly not cases where I was acting as a Sage would act....
I applied for a job. I desired to get it. But it was not in my control whether I got it or not.
I asked a girl out for a date. I desired her to say ‘yes’, but it was not in my control whether she did or didn’t.
I got in my car to drive to work. It was not in my control whether my car started, or whether I made it to work.
I exercised in order to improve my health. It was not in my control whether or not my health improved.
Look at your own life and consider all the cases where the outcome you desired was something other than the act of will itself. In 100% of those cases, you were aiming at something not in your control.
Sincerely, GCS”
The four examples are the material Richards spent the next week attacking, and they are what provoked the definition. Note the scope of the closing sentence: the class of things not in our control is not a short list of remote outcomes but everything aimed at other than the act of will. This is Core Stoicism’s boundary stated in autobiography.
III. The Definition (30 August 2019)
Richards’ reply of 30 August opened by accusing Sterling of leaving the field — at this point Grant appears to have walked away from the conversation, why, I don’t know — and then set out the trichotomy at length, quoting Greg Sadler on Irvine:
“The difficulty lies with Epictetus’ dichotomy of control: ‘we should focus our lives on determining what is in our control and what is not.’ The difficulty is that this dichotomy does not work! In real life there is a trichotomy of control, as Epictetus recognises in many places in the Discourses.”
The Sadler material Richards quoted holds that “some things aren’t up to us” is ambiguous between total absence of control and absence of complete control; that we have some measure of control over our bodies, finances, and employment; that things over which we have some but not complete control ought to be treated as important and valuable rather than with indifference, lest we become withdrawn underachievers; and that the remedy is to internalise our goals and adopt reserve.
Sterling answered the same day. The reply is the origin of the definition:
“G: Sadler, and you, fundamentally misunderstand the entire point. There is no ‘trichotomy of control’. The very term ‘some measure of control’ (or ‘partial control’) is logical nonsense.
Suppose that I want to build a new skyscraper, but the location where I want to put it up is not zoned for such buildings. My friend says ‘go ahead and buy up the property—I control the zoning board.’ So I buy up the property, and my application for a zoning change fails. I accost my (now former-)friend and say ‘I thought you said that you controlled the board!’ He says ‘Oh, I only meant that I _have some measure of control_ over the board. I can create situations in which the board voting my way may occur.’
The evil supervillain captures the hero, and informs him of his evil scheme (as supervillains are required to do, right before they fail to kill the hero). ‘I have invented a device that controls the weather!’ The hero looks on in horror as he imagines the ways in which the supervillain might use his terrible power. SV continues: ‘By throwing this switch, I can change the chance of rain from 30% to 52%!’ Oh, wait. It turns out that he doesn’t control the weather at all.
‘Control’ means ‘I can make this do what I want it to do, when I want it to do it’. If the thing doesn’t obey me on some occasions, then I don’t control it. I may ‘influence’ it, but I don’t ‘control’ it.”
He then gives the reason the definition matters, which the corpus has not previously carried:
“The reason that this is vitally important for Stoicism is that unlike (for example) Aristotle the Stoics proposed that we can achieve eudaimonia. For Aristotle, we can work towards eudaimonia, but we need luck to get there. If I am hideously ugly, or friendless, or my children die horrible deaths or turn into traitors, then my eudaimonia is marred. If I am enslaved, or my city is sacked by barbarians, I will not have eudaimonia.
If my goal is something only (forgive me for adopting the bastardization) ‘partially under my control’, then it is possible that I will fail to achieve it, and will fail to achieve eudaimonia _even if I am acting as a Sage_. I can be internally perfect, and fail to achieve eudaimonia. That’s a fundamental violation of Stoic thought.
What Sadler (et al) is trying here (and what others have tried before) is to allow us to have a goal that we are ‘concerned with’, but not be upset when we fail to achieve it. But that’s psychologically impossible. If I am genuinely concerned about a goal, then I will be disappointed when I don’t get it. If I am concerned about goals that are not in my control, then over the course of my life I am doomed not to achieve eudaimonia—sometimes I am going to fail to get those goals. Conversely, if I try to do something and I am not in the least bothered when I fail to achieve it, then I wasn’t really concerned with getting it, was I?”
The same message answers the underachiever charge directly, and in doing so states clause (d) a second time — this time in the operational language of preferred indifferents, role-duties, and reserve:
“On the other hand, Sadler (et al) are right that we should not become ‘underachievers’. If I pass someone who has been hit by a car and is dying, I should not say ‘Ah, their life is not in my control’ and step over their prostrate body on my way to the restaurant. Life is a preferred indifferent—I should recognize that it would be rational for me to try to aid the victim. (Whether that aid should come in the form of calling for an ambulance, yelling for help from others nearby, trying to bind the wounds, or something else depends on the circumstances.) Or, if you prefer the ‘role-duties’ approach, I have a duty as a citizen to aid the victim. And this is where ‘reserve’ comes into play. I appraise my situation and try to figure out what action of mine will be most likely to preserve the victim’s life (and health). My goal is to will rationally...but the content of my act of will is finding ways to save the victim’s life. If I do in fact will that which is most rational towards that end, then I am doing something entirely and totally within my control, and I will not be distressed _at all_ if the victim in fact dies—I will be ciontent to have done the right thing...performed the correct act of will.”
And it closes with the corruption argument:
“There is no need for a ‘trichotomy of control’ in order to understand Epictetus or Stoicism in general. But there is a very, very great danger in talking that way. Because once you break away from the radical ‘my body and my possessions and my job (etc.) are not in my control’ and start talking about how we should be concerned about things we don’t totally control, then what people hear is ‘it’s ok for me to desire external things to happen, as long as they’re things that I have influence over’. And once you’ve told them that, they’re doomed. They’re going to turn Stoicism away from ‘I can guarantee myself joy in life, but I need to radically change the way I look at the world’ and into ‘I should try to reduce how upset I am when things don’t go my way’. And then we can throw out Epictetus, and give them Aristotle or Epicurus or ‘I’m ok, you’re ok’ or whatever.
Regards, GCS”
Richards was unmoved. On 31 August he replied that the skyscraper case was nonsensical and offered a simpler one — agreeing on the phone to meet a friend for coffee the next day, where the meeting is the targeted outcome and getting to the shop on time is his part — concluding that this does not alter the fact that whether they meet tomorrow or not is partially in his control. He dismissed the rest as metaphorical hand-waving.
The definition was thus given under direct pressure, restated after rejection, and never withdrawn.
IV. Restated Under Pressure (2 September 2019)
Sterling returned to it on 2 September, replying to Richards’ analysis of the job application into a targeted outcome plus a course of action:
“a) ‘Partial control’ is nonsense. I mean that literally. The very concept of ‘control’ is the concept of being able to make something do what I want it to do. If I can’t make it do what I want it to do, I don’t control it.”
The reply distinguishes emotional investment in the outcome from investment in the means, and concludes:
“b2) The course of action, understood as physical actions, is not in our control—it can go wrong, and so becoming emotionally involved with it hampers one from achieving eudaimonia.
Now if we understand the ‘course of action’ only as _the inner act of will_, then of course now we arrive at something which is in our control, which can never go wrong, and which is therefore the proper source of attachment. But then we’re back at the point that I started at—virtually 100% of people think that an external which is not in their control (the aimed-for outcome, usually) is good or evil. This is a mistake. Stoicism tells us to redirect our attention from _all_ externals, which are _out of our control_ (‘partially’ or wholly doesn’t matter), and instead concentrate on getting the internal act of will right.”
V. What Control Confers, and What It Does Not (3 September 2019)
Kevin’s question, 2 September, opened the second thread:
“Do you think that ‘control’ itself is something special meaning it by itself has something about it which makes something good or bad? Or maybe better to say something is good or bad to me because it is in my control or not? I have always interpreted this as a decision making model that Epictetus used for his students.”
Sterling’s reply, 3 September, denies that control is what confers value while retaining it as a necessary condition:
“2) Nevertheless, I do accept the idea that since my eudaimonia is _mine_, it should involve things that benefit or harm _me_. The real ‘me’ consists of the things that I do, the things that I control. So I find the Stoic idea that my eudaimonia can only be affected by things that are in my control plausible. Theoretically there could be things in my control that do not affect my eudaimonia, so I wouldn’t say that ‘being in my control makes something good or bad’, but being in my control is a _requirement_ for something being good or bad.”
He then distinguishes the practical argument from the grounding one:
“3) Epictetus gives compelling reason to think that judging things not in our control to be good or bad leads to lack of eudaimonia, since it leads to psychological distress (which is incompatible with eudaimonia). That would be sufficient as a _practical_ argument: you want eud.; you can’t get it if you desire things not in your control; so stop desiring things not in your control. Your desires are caused by beliefs about what’s good and bad, so stop thinking of externals as good or bad and you might be able to achieve eud. ... But by associating our true nature with our Will, we can get to the conclusion that this reform of our beliefs about goodness is _right_.”
And states why the practical argument alone will not serve:
“If all we have from the Stoics is the idea that we need to stop believing that externals are good or evil in order to have a chance at eud., that would leave open the possibility that eud. might be built on a (practically beneficial) lie. Associating our true self with our Will is a bridge that allows the Stoics to think that it is the _truth_ about the value of externals that also sets us free from suffering: there is no divide between believing what is true and believing what is beneficial.”
This is the correspondence commitment operating inside the control question. The dichotomy is not a therapeutic device that works because believing it helps; it is asserted as true, and its benefit follows from its truth. Sterling’s standing position on that distinction is Stoicism Is Not Therapy, But Training.
VI. Yielding as an Exercise of Control (3 September 2019)
Michael Edelstein entered the thread the same day with one sentence:
“Re the trichotomy of control: we have some control but not total control over our thoughts. Why does this not add a 3rd type of control?”
Steve Marquis answered before Sterling did, holding that we have no control over which impression is presented and complete control over whether to assent, and that he saw no partial control anywhere in the Stoic model. Edelstein pressed with his clinical case:
“In my practice as an addiction psychologist I perceive clients progressing from little control to complete (often, not always, complete) control over their addictive behaviors.
My impression is you’re saying if I’m addicted to overeating and I have the thought ‘I must have another serving’ I have complete control—not partial control—over whether I act on this thought. What am I missing?”
Sterling’s reply is the passage Use and Correct Use of Impressions §III and Defense Four §V rest on. It is given here in full:
“Michael:
Addicted-Michael has the thought ‘I must have another serving’. There are two possibilities:
1) Addiction is so powerful that it is impossible for you to resist this idea. (Perhaps you do not even consider the possibility of not assenting to the impulse.) In that case, you do not have partial control, you have no control at all. The Stoics talk little about this possibility, although they do consider the possibility of people with insanity who are such that they cannot act rationally. But this is not an exception to the control dichotomy—it is a case of not having control.
2) Although you are addicted, and the thought of overeating is extremely tempting, it is not literally overpowering—you could resist it if you forced yourself to do so. But in this case you have control over the decision to assent. Another helping may appear very tempting, but you could refuse.
I understand (really I do!) the tendency for people (both psychologists and laymen) to think of this as ‘partial control’, because someone who is trying to break their addition will sometimes choose to yield to the temptation, and will sometimes resist it. But this obscures the reality of the situation. The fact that in situation X I sometimes choose to yield to a temptation and sometimes choose not to yield does not mean that I have partial control, it means _either_ that on some of those occasions I had no control and on others I had full control, or else (more commonly) on some occasions I used my control to yield, and on some I used my control to resist.
I have never been clinically addicted to anything. But I have certainly had occasions where I knew that it would be better for me in the long run not to do something which at that moment seemed very inviting. And on some of those occasions, I yielded to the temptation and on some I didn’t. The goal of Stoic practice is to get me to be in the position where I will no longer be tempted to yield irrationally to temptation...and, luckily, even if I don’t make it to being a Sage, the percentage of the time that I yield will decrease as I train myself to resist.
The Stoic view is often presented starkly, and so it seems to ignore realities. It is often presented as ‘you just choose to do the rational/right thing, and that’s totally in your control’, and that makes it sound _easy_. But when I have a habit of seeing something as good (whether that habit counts as an ‘addiction’ or not), and I get the appearance that it would be good, it’s _easy_ to give in, and requires mental effort to resist. Not surprisingly, people often yield, even people who are training as Stoics or undergoing therapy. But that doesn’t mean that they have partial control....
Indeed, I once more claim that the notion of partial control doesn’t even make sense. If the addiction is overwhelming so that I cannot resist, then I have no control. If the addiction is not overwhelming, then what could it mean to say that I have partial control? That I will to resist but that some percentage of the time my body takes a bite anyway and part of the time it doesn’t? Is that what you observe in your patients—that they genuinely choose not to indulge but that their bodies sporadically indulge even while they will themselves not to—they will their hands to remain at their sides, but intermittently their hands raise up and shove food into their mouths? I’ve never witnessed such a thing, or experienced it.
Regards, GCS”
Two features of the passage have not been used in the corpus and are recorded here.
The first is the concession about difficulty. Sterling grants that resisting a settled habit of seeing something as good requires mental effort, that giving in is easy, and that people in training and in therapy often yield. The strict dichotomy is not maintained by denying the phenomenology; it is maintained by refusing to let the phenomenology reclassify the faculty. The corpus’s account of what the ordinary correction case actually is stands at The Pathos Already Occurred.
The second is his own first-person case: he has known what it would be better not to do, and has yielded on some occasions and not on others. The man at the decision point is not a construct; Sterling reports himself as that man in the same breath in which he denies that this is partial control.
Later the same day, replying to Marquis, he supplied the definition of correct use in the same vocabulary as the 26 August passage:
“4) What is ‘correct use of will’? using our will in connection with true value beliefs instead of false ones. True ones include the idea that externals are neither good nor evil, that some things are preferred and others dispreferred, that we have role-duties, etc.”
VII. What the Corpus Draws
Use and Correct Use of Impressions §II takes the 30 August definition as the fixing of the term — control admits no degrees; a thing either obeys on every occasion or is not controlled at all. The full passage supports this and adds the ground: the strictness is required because eudaimonia must be achievable by the Sage without luck, which fails if the goal is anything the agent may not obtain.
Use and Correct Use §III and Defense Four §V both take the yield/resist sentence as establishing that yielding is an exercise of the faculty performed at full strength, not a lapse of it. The full passage supports this and rules out the alternative Sterling was offered: the dilemma he forces is exhaustive, and neither horn admits a third category. The wider defence of the same boundary is Three Defenses — Assent, Termination, and the Therapy Thesis.
The CPA Run — Michael R. Edelstein records at P5 the claim that control over thought is partial and grounds a third category. The primary text confirms the attribution: it is Edelstein’s own sentence of 3 September, offered as a question and pressed a second time with the clinical case.
Sterling on Kataleptic Impressions: The Primary-Source Record §II quotes the 19 August avowal and the opening paragraph of the 26 August reply, under the kataleptic-impression heading.
What the corpus has not drawn: the eudaimonia argument for strictness; the corruption argument for why trichotomy talk is dangerous; the car-accident case as a worked instance of preferred indifferent, role-duty, and reserve together; the requirement-not-sufficiency answer to Kevin and the truth-over-benefit ground it rests on; Sterling’s concession about difficulty and his own first-person report of yielding; and the whole of the 26 August passage after its first paragraph — the mechanical/moral distinction, the Parmenides example, and the professor-and-bribe pseudo-dilemma.
VIII. Corrections to the Retrieval Map
Stated explicitly rather than silently absorbed.
- The 30 August message is misplaced in the map. The map lists thread 16ce85b0bcdda958 as a single message of 30 August 2019 carrying Sterling’s definition of control. That thread holds one message dated 31 August 2019 15:46:04Z, authored by George Richards, which carries Sterling’s 30 August text only as quoted matter. Sterling’s own message is 16ce306b115a1992, 30 August 2019 14:55:53Z, within thread 16ca05c165a33b10. The corpus citation “Sterling, ISF, 30 August 2019” is correct; the archive location recorded in the map is not.
- Thread 16ca05c165a33b10 is far larger than the map states. The map lists four messages, 17–26 August. The thread holds 37 messages, 17 August – 3 September 2019, and contains the 27 August four-examples reply, the 30 August definition, and the 2 September restatement — three of the six Sterling passages recorded above. The map’s four entries are accurate as far as they go.
- The 26 August passage is not wholly unquoted. Its first paragraph is already quoted in Sterling on Kataleptic Impressions: The Primary-Source Record v1.1 §II, where it serves as the second of two personal avowals of kataleptic impression. The remaining three paragraphs are unquoted anywhere in the corpus. The clause-(d) reading is new; the sentence itself is not.
- Confirmed as the map has it: the 3 September Sterling-to-Kevin content (requirement, not sufficient), Edelstein’s 3 September question at 16cf7fe1b1bf9c04, the yield/resist message, and thread 16cf436dfbb90f6e at 16 messages, 2–4 September.
Sources
All from the International Stoic Forum (Yahoo Groups era), threads titled Making correct use of impressions and control (Making correct use of impressions. Retrieved by full-thread pull, 7 September 2026. Times are as recorded by the archive (UTC); dates are unaffected by conversion to Sterling’s local time.
Thread 16ca05c165a33b10 (37 messages, 17 August – 3 September 2019):
- 16ca05c165a33b10 — 17 August, George Richards.
- 16cab8ae698ffe36 — 19 August, Sterling.
- 16cc39e28cb2ba0c — 24 August, Richards.
- 16cc8f57faff98a9 — 25 August, Adrian.
- 16ccedf59f31ef82 — 26 August, Sterling.
- 16cd3414979be141 — 27 August, Richards.
- 16cd3f4a63c24635 — 27 August, Sterling.
- 16ce295dc163bca8 — 30 August, Richards (quoting Sadler).
- 16ce306b115a1992 — 30 August, Sterling (the definition).
- 16cf383ae5cec467 — 2 September, Sterling.
Thread 16ce85b0bcdda958 (1 message):
- 16ce85b0bcdda958 — 31 August, Richards.
Thread 16cf436dfbb90f6e (16 messages, 2–4 September 2019):
- 16cf436dfbb90f6e — 2 September, Kevin.
- 16cf7b7b8ef35e91 — 3 September, Sterling.
- 16cf7fe1b1bf9c04 — 3 September, Michael Edelstein.
- 16cf807ef756679f — 3 September, Steve Marquis.
- 16cf82a08bd02d93 — 3 September, Edelstein.
- 16cf85cf77b24a87 — 3 September, Sterling (yield/resist).
- 16cf8e7914f0cb21 — 3 September, Sterling.
Preserved by Dave Kelly, 2026.
Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Analysis and synthesis: Dave Kelly. Prose rendering: Claude (Anthropic). 2026.
0 Comments:
Post a Comment
<< Home