An Essay on Human Nature and the Art of Living

Text from GitHub, as it stood on 2026-11-04. Source: https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/essay-on-human-nature.md

Updated October 2026. Licensed under CC BY 4.0.

Note on method

This essay began as two manuscripts Simon Kreis wrote in his early twenties, while at university. He was not studying philosophy; he was interested in it, ambitiously and without much method. The manuscripts argued for a single chain of ideas: human behavior is driven by affect, freedom does not survive that fact, ethics must rest on sensibility alone, and happiness recedes when pursued directly. He sent them to a few publishers. Nothing came of it, and the manuscripts stayed in a folder for about ten years. The ideas held together; the writing did not.

Claude rewrote them: it traced each idea to the thinkers who had developed it properly, set it against its strongest objections, corrected what the original got wrong, and kept only the practices that established philosophical, spiritual, and religious traditions actually recommend.

This is the kind of project that was never worth finishing by hand. As Jensen Huang has put it in substance, AI makes worth doing many things that were not worth doing before. The result is neither the young author’s book nor a work of professional philosophy. It is a synthesis, and it is published as one.

---

Contents

• Introduction
• Part I. Human Nature: The Reign of the Affects
  • 1. What the word “passion” means
  • 2. The final end
  • 3. Instinct
  • 4. The substrate
  • 5. When the symbol takes over
• Part II. The Condition of Freedom
  • 6. Reason in the service of the passions
  • 7. The collapse of choice
  • 8. Merit and blame
  • 9. The compatibilist objection
  • 10. Four ways of inhabiting necessity
• Part III. The Ethical Transposition
  • 11. The problem
  • 12. Sensibility as the only ground
  • 13. Where moral sentiment comes from
  • 14. The good life: fulfillment or state?
  • 15. The paradox of hedonism
• Part IV. The Ways of Emancipation
  • 16. Shifting attention
  • 17. Other people
  • 18. The lever of judgment
  • 19. What depends on us
  • 20. The economy of desire
  • 21. The loosening of the ego
  • 22. Time, death, the world
  • 23. Discipline
  • 24. Suspending judgment
  • 25. The radical pole: renunciation
  • Where philosophy stops
• Conclusion

---

Introduction

The question

A commonplace observation serves as the starting point: nothing a human being does can be explained without bringing in, somewhere, what that person feels. A gesture can be described by its mechanics, a decision by its reasons, an institution by its rules. But as soon as one asks why someone got up, insisted, or gave up, the answer always ends by mentioning something felt.

This observation looks harmless. It is not. If the affects really are what sets us in motion, several ordinary certainties become hard to hold.

The first concerns reason. It is usually taken to be what lifts us above our impulses and decides with sovereign authority. But if reason cannot by itself produce any desire, its role changes. It calculates, compares and illuminates; it does not decide what is worth pursuing.

The second concerns freedom. If decisions follow the slope of the affects, and if the affects are not chosen, then choice looks less like an origin than like a result. This at once calls into question merit, fault, and the whole apparatus of blame.

The third concerns morality. If no external authority gives values their standing, and if reason can argue for almost any position, what can a moral judgment still rest on? The nihilist answer (nothing, and so everything is of equal worth) is coherent. The refusal to stop there needs to be explained.

The fourth concerns happiness itself. On analysis, it appears as the term at which every chain of justification stops. Yet it has a strange property: pursuing it directly seems to make it recede. Moral philosophy calls this the paradox of hedonism, and everything else is organized around it.

The route

The argument moves in four stages, each inheriting the problem the previous one leaves open.

Part I. Human nature. How does a being function when it acts under the impulse of its affects? Three levels of explanation fit together: the mechanism of learning by consequences, its evolutionary cause, and its implementation in the brain. To these is added a distinctively human feature, the capacity to transfer reward into the order of the symbol, to the point of dying for an idea.

Part II. The condition of freedom. If the affects steer even the reason that claims to govern them, what remains of choice? The determinist thesis is examined in its tightest version, together with its consequences for merit and responsibility. The compatibilist objection, the strongest that can be raised against it, is taken seriously and not dismissed. The part ends with a reversal: the traditions that thought most deeply about necessity did not see it as a prison.

Part III. The ethical transposition. What becomes of the good if nothing outside us grounds it? The anti-realist argument is carried through to its end, including the nihilist impasse. A way out is proposed (sensibility as the only natural moral fact), and its weaknesses are set out as carefully as its merits. The part closes on the paradox that governs the next one.

Part IV. The ways of emancipation. What do the philosophical traditions offer once the paradox is granted? Their answers fall into two families. The first turns attention outward: engagement in an activity pursued for its own sake, and attachment to other people. The second works from within: on judgment, the dichotomy of control, the economy of desire, the loosening of the ego, the relation to time and death, discipline, and the suspension of judgment. It goes as far as radical renunciation, which abolishes the problem by abolishing willing. These ways do not agree with one another. None is crowned.

The route leaves aside the collective conditions of the good life: institutions, law, the organization of work. This division is a choice, not a given, and the conclusion states what it costs.

The approach

There are two ways to build a moral philosophy, and the choice between them governs everything else.

The first starts from a principle (the dignity of the person, the categorical imperative, an original contract) and deduces what ought to be. Its defect is structural: it produces systems that are coherent and unverifiable. On a priori premises one can build a flawless architecture that never meets a fact capable of refuting it. It nonetheless remains the most serious rival to the second, and chapter 11 gives it its best chance.

The second starts from what can be observed (how organisms learn, what moves them, what ethology documents, what neurophysiology measures) and works back toward what might have value. That is the path followed here.

It has its own defect, and the defect is famous. From what is, one cannot deduce what ought to be. This is Hume’s law, taken up by Moore under the name of the naturalistic fallacy. Observing that evolution produced empathy does not establish that empathy is good. The pages that follow do not resolve this difficulty. They locate it in chapter 12, the exact point where the argument is most vulnerable, and treat it there for what it is: a deliberate choice rather than a deduction.

Chapter 12 will also argue that this defect is not peculiar to naturalism.

How the argument proceeds

One last clarification, about manner.

The theses examined here are strong, and they are stated as such: a watered-down thesis is not worth discussing. But each is systematically confronted with its best objections. When an objection holds, the text says so, and does not treat it as a reason to hurry past.

This standard produces a result unusual in a work of this kind: several chapters end without a clear conclusion. Compatibilism is not refuted. The passage from fact to value is not grounded. Whether desire is an engine or a servitude remains an open question, and Part IV refrains from choosing between ways that contradict one another.

This is not rhetorical caution. It follows from a finding the argument itself reaches in chapter 24: when two reasons balance, honesty consists in suspending judgment rather than manufacturing a preference. The ancient skeptics made this a method, and it happens to be the method that suits this subject.

The quotations come from works in the public domain. Contemporary contributions (the psychology of decision, the neurophysiology of motivation, cognitive therapies) are introduced as such. They shed light on mechanisms; they validate no philosophical thesis.

This precaution is not a formality. Part of the experimental psychology of recent decades has gone through a replication crisis. Results taken as established have proved fragile, and some of those invoked here belong to the affected fields. The most debated cases are flagged along the way. But the general rule matters more than the cases: none of the philosophical theses defended in these pages depends on any particular experimental result. The data serve to make a mechanism intelligible, never to establish a conclusion. If one of them were to fall, the argument it illustrated would remain exactly as strong, or as weak, as it was before.

---

Part I. Human Nature: The Reign of the Affects

How do we actually work?

---

1. What the word “passion” means

The vocabulary used here is old, and its everyday sense has drifted. Its meaning has to be fixed first.

Passion here means anything that can be felt: emotions and sensations, pleasant and unpleasant alike. This is the classical usage, that of Descartes’s Passions of the Soul and Spinoza’s Ethics. It does not mean consuming love or spectacular anger. It means the whole range of what is felt, including its dullest forms. A pain in the knee is a passion. The boredom of a Sunday afternoon is another.

The word also keeps its root: pati, to undergo. A passion is what one undergoes, as opposed to what one does. Part II will show that this distinction is less sharp than it looks.

Happiness and suffering will mean simply the pleasant and the unpleasant, with no distinction between body and mind. This identification puts on the same level experiences that ordinary usage ranks. It is deliberate, and it is contestable. Mill held, on the contrary, that there are higher and lower pleasures, and that a competent judge who has known both will always prefer the former: “better to be Socrates dissatisfied than a fool satisfied.” The objection is serious, and chapter 14 takes it up. One point can be noted at once, however. The hierarchy Mill defends is itself an evaluation, and it cannot be read off the experience of the one who feels. From the standpoint of the one feeling, an intellectual pleasure and a sensory pleasure belong to the same register, whatever one thinks of them otherwise.

A first point follows from this definition. It is almost tautological, but the rest depends on it. Happiness and suffering are not passions among others: they are the two ends of the scale on which all the others are ranged. Every passion carries a valence (it is pleasant or unpleasant, to some degree), and only in that capacity does it count in conduct.

Two reservations about this image of a scale are in order.

The first is that it suggests a single magnitude, as though all experiences could be converted into a common currency. Nothing guarantees this. It is not obvious that a humiliation and a toothache are commensurable, or that a grief and a satisfaction can be added to yield a balance. Bentham believed he could construct a calculus of pleasures (intensity, duration, certainty, propinquity, fecundity, purity, extent). The difficulty he ran into, interpersonal comparison, has never been removed. Welfare economics still stumbles on it.

The second reservation goes deeper: the image presupposes an observer who reads the dial. Yet it is not certain that anyone is behind it. Chapter 21 will show that this inner spectator is itself a construction.

One last distinction, a matter of usage rather than principle, will serve until the end.

The affects can be sorted by what triggers them. Some arise from sensation: hunger, a burn, fatigue, the pleasure of a meal. They are good or bad in themselves, with no judgment needed; a burn does not wait for anyone’s opinion before it hurts. Others arise from judgment, that is, from the way a situation is interpreted: humiliation, hope, regret, worry. The same event (a refusal, a silence, a promotion) can produce opposite affects in two people, because they judge it differently.

What this distinction is worth should be said at once. It does not describe two separate systems. Pain is almost always accompanied by an interpretation that worsens or eases it, and the most abstract judgment is felt in the body; the affective sciences draw no sharp line between the two. The distinction therefore has no theoretical value here. Its value is practical. It comes down to an observation the traditions made long before anyone measured it: the second source, not the first, is where some grip is possible. Part IV draws the consequences.

2. The final end

The thesis

Here is the thesis in its strong form: happiness, in the sense just fixed, is the end of every action; everything else is only a means to it.

In moral philosophy it has a name: psychological hedonism. It should be kept carefully apart from a neighboring and quite different thesis, ethical hedonism, which holds that pleasure is what ought to be pursued. The first is a description of how human beings function, the second a prescription. They are independent: one can hold either without the other. Chapter 12 will show that moving from the first to the second requires an argument nobody has supplied.

The argument for psychological hedonism is a regress. Take any action and ask for its reason, then for the reason behind that reason. One studies in order to work, works in order to earn a living, and earns a living in order to acquire goods. One acquires those goods for the states they produce: security, comfort, status, pleasure. At each step, the link is valued for the next one. Followed far enough, the chain always stops at the same place: an affective state sought for its own sake.

Aristotle states the principle in the first book of the Nicomachean Ethics: if there were no final end, desired for itself and not for the sake of something else, desire would be empty and would lose itself in an infinite regress. Epicurus names this end without hesitation: “we must meditate on the causes that can produce happiness, since when it is ours we have everything, and when it is lacking we do everything to have it.”¹ This is exactly the structure of the regress: what, once obtained, leaves nothing more to desire.

This conception is not exotic. It is the implicit background of ordinary practical discourse. Economics builds its rational agent on the maximization of utility. The psychology of learning describes conduct as a management of pleasure and pain. Common sense finds it “logical” to want to be happy and “irrational” to inflict suffering on oneself with nothing in return. The classic experiment on delayed gratification (a child forgoes an immediate treat in order to get two later) is read as a feat of rationality. It is read, that is, as good management of the timing of pleasure.

It is precisely this universality that should arouse suspicion. A self-evidence that is never stated is rarely a discovery; it is often a grammar.

Butler’s objection

The decisive objection to psychological hedonism was formulated by Joseph Butler in his Sermons of 1726, and it has never received a fully satisfactory answer.

Butler observes that the pleasure taken in a thing presupposes a desire for that thing, and so cannot explain it. Eating gives pleasure only to someone who is hungry; it is the appetite for food that makes the meal pleasant, not the other way round. If one desired only pleasure, one could never obtain any, for lack of an appetite for anything in particular. The passions aim first at external objects; satisfaction is their consequence, not their target.

Applied to altruism, the objection is devastating. To say that a mother rescues her child “for the satisfaction she gets from it” inverts the real order: she feels that satisfaction only because she cares about the child. Remove the care for the child, and the satisfaction disappears with it. Affection therefore comes first, and pleasure derives from it.

The psychological hedonist can reply that the mechanism that selects conduct does operate on affective valence; the theory of reinforcement, in the next chapter, describes exactly this. But the hedonist must then concede the essential point. The objects of desire are not reducible to the pleasure they bring, and a being that wanted only pleasure would want nothing.

The experience machine

A second, more recent objection targets ethical hedonism, and it has become the standard test.

Robert Nozick invites the reader to imagine a machine that can produce any desired experience, with a quality indistinguishable from reality. Plugged into it, one would believe one was writing a great book, making friends, living a successful life. One would not know one was plugged in. The question is simple: should one plug in for life?

Most people answer no. Yet if only the affective state counted, the answer should be an unhesitating yes. The refusal suggests three things. We want to do certain things and not merely to experience doing them. We want to be a certain way and not merely to believe we are. And contact with an independent reality matters to us.

The objection is not unanswerable: one may suspect that the refusal reflects a status quo bias, or distrust of a technical device. It nonetheless establishes something solid. If happiness is indeed the end of action, it is not so in the sense of a pleasant state of consciousness that could be obtained by any means whatever. Something else is being asked for, which chapter 14 will try to name.

What remains of the thesis

One difficulty runs through this whole part, and it is better named at once.

Suppose every action, whatever it is, can be reconverted into a subjective benefit (the satisfaction of duty done, the avoidance of remorse, the preservation of inner coherence). Then the thesis no longer excludes anything. What behavior could refute it? None, by construction: it will always be enough to postulate a sufficient internal affect. A proposition that no observation can refute is not an empirical discovery.

So what it is needs to be stated. It is not a law of human nature but a grammar: a coherent and often illuminating way of redescribing conduct, which gains in scope what it loses in risk. It accounts well for the force of social norms, for the workings of remorse, and for the failure of coercion that is sometimes observed. It does not establish that nothing else exists.

Compare “all behavior is explained by the pursuit of affect” with “all behavior can be narrated in terms of affect.” The distance between them is slight to the ear and considerable in logic. The weak version is the one retained here. It is enough for everything that follows.

3. Instinct

The proximate mechanism

Seeking pleasure and avoiding pain are not peculiar to human beings. They extend to everything that feels, and psychology has given them a precise description: operant conditioning.

Skinner distinguishes respondent behaviors, reflexes triggered by a stimulus, from operant behaviors, those by which an organism acts on its environment. An operant is not mechanically triggered: it is emitted, then selected or eliminated according to its consequences. The central concept is contingency, a stable link between a behavior and what follows it. If an action is regularly followed by a certain kind of consequence, the organism learns this structure. It does so not by representing the structure but by changing the probabilities of its actions. It does not need to understand the rule in order to obey it.

Skinner himself makes the kinship with natural selection explicit: in both cases a repertoire of variations is sorted by its effects, with no intention involved. Behavior is selected by its consequences as organisms are selected by their environment.

A technical point deserves mention, because it is almost always missed. Negative reinforcement is not punishment. It refers to the removal of an unpleasant stimulus, a removal that increases the frequency of the behavior. Putting on a coat stops the cold: the gesture is negatively reinforced. Turning off an alarm, or paying a bill to stop the reminders, has the same structure. “Negative” refers to a subtraction, not to a value judgment.

This distinction matters for what follows. A considerable share of human conduct aims not at obtaining a good but at ending a discomfort, and these two logics produce very different behaviors. Avoidance, in particular, has a formidable property: it sustains itself without ever being tested. Someone who flees a dreaded situation never discovers that it was bearable, and the avoidance is reinforced by its own apparent success. A large part of chronic anxiety works this way.

The distal cause

Behaviorism describes a mechanism without saying why the mechanism exists. The answer is evolutionary.

In the perspective Dawkins popularized, an organism is not the hero of its own story. It is a temporary vehicle, built by genes that secure their own continuity. The formulation is provocative, and its mechanics are useful. A gene that favors, even marginally, the survival or reproduction of its carrier has a better chance of being passed on. Generation after generation, the genes that increase the number of their copies become dominant. Selection never needed intentions.

The circuits that make rewards pleasant and sanctions aversive were therefore not shaped at random, nor because they are good. They were retained because they worked. Fleeing an alarming noise, seeking food, attachment to the group, cautious exploration of a territory: all are legacies that allowed lineages to persist. Reinforcement operates at the level of the individual, selection at the level of the gene. The two layers interlock, and what looks like conduct charged with meaning is often the visible tip of a blind engineering.

The correction

The cynical reading of nature sometimes drawn from this is a misreading, and it needs careful correction.

Darwin himself did not stop at competition. In The Descent of Man, he derives the social instincts (sympathy, mutual help, a nascent moral sense) from the evolution of social species, and he sees in them the origin of conscience. Kropotkin opposed the social Darwinism that retained only the struggle for existence. In Mutual Aid he shows that cooperation is a factor of evolution at least as powerful as rivalry: groups whose members help one another survive better.

Twentieth-century biology gave this intuition precise mechanisms. Kin selection explains why a behavior costly to the individual can be retained if it benefits relatives carrying the same genes; Hamilton’s calculation gives it its form. Reciprocal altruism explains cooperation between non-relatives when interactions are repeated: helping those who will help you in return pays off in the long run. These models do not contradict the “selfish gene”; they show that genetic selfishness produces, at the level of organisms, genuinely cooperative conduct.

Ethology supplied the observational counterpart. The work of Frans de Waal documents, in chimpanzees and bonobos, consolation after conflict, sharing, reciprocity, and the soothing of relations: what he calls the building blocks of morality. These behaviors stem from no reasoning. They stem from social affects that secured the cohesion of groups. His target is explicit: veneer theory, the idea that morality is a thin civilized varnish over a fundamentally self-interested animal.

Both terms have to be held together. Cooperation is natural; so is rivalry. This is not a contradiction. What the double reading forbids is only the easy conclusion: evolution proves that man is a wolf. It proves no such thing. It describes a social animal whose affects are ambivalent.

The survival objection

Here is the serious objection to the thesis of chapter 2, and it comes from biology itself.

If the affects were selected because they favored survival and reproduction, then happiness is not the end of life but its instrument. The true end would be the perpetuation of life, and pleasure a mere guidance device: a signal, not a goal.

The objection is not refuted. It dissolves once one separates two levels that it confuses.

At the level of evolutionary explanation, it is right: survival explains why these mechanisms exist. But explaining the origin of a device does not fix the end pursued by whoever is equipped with it. An organism does not pursue the transmission of its genes. It pursues what does it good, and its genes are transmitted because it pursues that. The proof lies in every case where the two levels come apart: contraception, adoption, voluntary celibacy, the use of substances that short-circuit the reward circuit. A being that aimed at survival would not behave this way. A being that aims at an affective state would.

Survival is the cause; affect is the motive. The objection nonetheless leaves a trace that it would be dishonest to erase. If pleasure is only a signal, organizing one’s life around it amounts to confusing the indicator with what it indicates. Chapter 15 will show that this confusion has a price.

The conatus

One last piece, older and more radical.

Spinoza calls conatus the striving by which each thing endeavors to persevere in its being and to increase its power of acting. “Desire is the very essence of man.” This is not a property we have but what we are. Ethics then comes down to understanding this movement. What increases the power of acting is experienced as good, what diminishes it as bad, and the affects (joy, sadness) are its internal indicators.

The reversal Spinoza performs is crucial and often misquoted: we do not desire a thing because we judge it good; we judge it good because we desire it. Value follows appetite; it does not precede it.

Three centuries before behaviorism and evolutionary biology documented it, the Ethics thus lays down the general framework. The good is not imposed from outside; it can be read in the mechanics of body and mind. Spinoza will also draw from this the hardest determinism there is, and an unexpected conception of freedom: the subject of the next part.

4. The substrate

This chapter introduces contemporary results. They shed light on the mechanisms described so far; they validate none of the philosophical theses at stake.

Prediction error

Where behaviorism described reinforcement from the outside, neurophysiology shows how it is implemented.

The pivot is dopamine, and the popular image of the “pleasure neurotransmitter” is misleading. What this system mainly encodes is a reward prediction error signal: the gap between the reward expected and the reward obtained². Receiving what one expected produces almost nothing; receiving more than expected produces a spike; receiving less produces a dip. The organism does not register pleasure; it registers surprise.

This is exactly the variable that a reinforcement learning algorithm optimizes. The convergence between the two fields is not a literary analogy: the formal models are the same. The organism adjusts its future actions according to this gap, without having to represent the rule.

One consequence deserves notice. Work on the distinction between wanting and liking shows that two sets of circuits are partly dissociable³. Those that produce the attraction to a thing differ from those that produce the pleasure of obtaining it. One can intensely want what gives no satisfaction. This is the pattern of addictive behavior, and also, in a milder form, an ordinary experience. Desire is therefore not a good predictor of pleasure. Part IV returns to what this dissociation makes possible.

Symmetrically, aversive circuits treat punishment, fear and relief as internal signals. The decisive conceptual step is this: affect is not a passive inner commentary; it is a reinforcer. An emotion assigns a valence to a situation and shifts the probability of the next action.

Adaptation

The most important contribution for what follows comes from elsewhere: from work on what has been called the hedonic treadmill⁴.

The finding is simple and devastating. Satisfying a desire produces an effect that fades quickly, as the individual returns to a baseline level. Two mechanisms combine. Habituation: every pleasure becomes the norm, and so stops being felt as pleasure. And escalating expectations: once used to a level of satisfaction, one needs more to feel the same thing.

Desire thus becomes a self-referential engine. The more one obtains, the more one must desire to feel anything at all.

Consider what this result does to the thesis of chapter 2. Suppose happiness is indeed the end of every action, and satisfying desires does not durably change the level of happiness. Then we methodically pursue an end our means do not reach. This is no longer an external objection: it is a contradiction produced by the system itself. It has a name, the paradox of hedonism, and chapter 15 is devoted to it.

The bare formula is false, however, and needs an immediate qualification. Adaptation is real, but it is not universal. Some conditions are not forgotten: chronic pain, noise, bereavement, lasting humiliation, prolonged loneliness. Nor are some satisfactions, especially those that come from an activity rather than an acquisition. This unevenness in the speed of adaptation is probably the most useful practical finding in this whole chapter: it indicates where to put one’s effort.

Two measures of well-being

One last result complicates the very notion of happiness. It belongs here because it will weigh on Parts III and IV.

Work on the retrospective evaluation of experiences shows that the memory one keeps of an episode does not match the sum of what was felt while it lasted⁵. What one remembers depends mainly on the moments of peak intensity and on how the episode ended; its duration hardly counts. Take a longer painful episode whose end is less painful, and a shorter one that ends at its worst moment. The first will be judged preferable in retrospect, and it is the first that people choose to repeat.

The result is a difficulty that no philosophy of happiness has really absorbed: there are at least two things that can be called “being happy.” One is experiencing well-being moment by moment. The other is judging one’s life satisfying when one evaluates it. These two measures regularly diverge, and one cannot maximize both.

When a tradition recommends some conduct “in order to be happy,” it is therefore legitimate to ask which of the two it means. The question is not rhetorical: several of the ways set out in Part IV clearly aim at the second at the expense of the first.

5. When the symbol takes over

The change of scale

The mechanics of reinforcement do not disappear as complexity rises. They change scale and dress themselves in symbols.

In human beings, consequences stop being merely physical. Punishment is no longer a shock but a contemptuous silence, an exclusion, a damaged reputation. Reward is no longer a treat but a status, a recognition, a belonging. Stigma acts by anticipation: not through material pain but through the place it assigns in the eyes of others.

The division of labor is clear. Biology supplies the architecture (seek what brings pleasure, flee what produces loss), and culture supplies the contents, turning approval into pleasure and stigma into pain. Social norms are not arbitrary rules stuck onto the individual. They are regularities stabilized by millions of micro-reinforcements. These signals weigh so heavily because, on the evolutionary scale, exclusion from the group meant vulnerability, often death. Research on social rejection shows, moreover, that it partly engages the same circuits as physical pain: the metaphor of the wound may not be a metaphor.

What this framework does not explain should be noted. It accounts well for the force of norms, much less for their content. Why does one society prize warrior honor and another humility? The model says how a norm becomes established and persists, not why it is this norm rather than another. It sheds light on a mechanism; it does not replace a history.

The limiting case

There remains a distinctively human property, and it is disconcerting. Human beings can reconfigure their own mechanisms of reinforcement. They can do so to the point of producing conduct that flatly contradicts the most fundamental instincts, up to and including survival.

No animal crosses this line. None gives up its survival for a reward it will never perceive. Abstraction has no direct hold on its emotional circuits: the reward must remain perceptible.

In human beings, the barrier falls. A narrative, a cause perceived as sacred, a collective identity can produce affects as intense as physical threats. The martyr does not sacrifice his life by ignoring his biological signals. A symbolic reward (salvation, fulfillment, meaning) has acquired a higher valence. The prisoner who resists torture shows the mechanism in reverse. Coercion fails precisely when the symbolic valence, dignity or the refusal to betray, outweighs the bodily valence of pain.

Language is what makes this transfer possible. It allows one to learn without having experienced, and to be conditioned by a description rather than an experience. It allows one to place a reward in a future no animal can represent, including a posthumous one. This is the same architecture as social reinforcement, pushed to its limit. Reinforcement moves into the space of meanings, and this move permits acts that nonhuman living beings cannot produce.

This plasticity makes human beings extraordinarily inventive. It also makes them extraordinarily vulnerable to any narrative that can turn an idea into an imperative stronger than survival. It is probably the most important fact in this chapter, and it is not reassuring.

The loop

It remains to describe the general form of the mechanism, because it has a curious property.

A piece of conduct arises from an affect (an impulse, a tension, a lack, an anger) that acts as a trigger. And that same conduct aims to produce another affect, which will in turn serve as reinforcement. The impulse and the reward belong to the same family of phenomena: cause and end are of the same order.

What is aimed at, moreover, is never the object itself but the affective transformation that accompanies obtaining it. Conduct works as a bridge between two internal states. It is a regulatory mechanism: triggered by affect, acting toward affect, and reorganizing itself according to the affect obtained.

This circularity can be seen as a weakness: the system would explain everything by itself. Yet it correctly describes a fact. In an organism with a nervous system, the end of a piece of conduct is never external to the subject; it is the internal state the conduct manages to produce.

But what this elegance costs must be acknowledged. It is the same circularity that made the thesis of chapter 2 irrefutable, and it returns here in another form. A framework that absorbs every observation is comfortable and suspect. It will be kept because it organizes well what is observed, provided organizing is not mistaken for demonstrating.

---

Where we stand

The animal described in this part is as follows. Its conduct is shaped by its consequences. Its affective architecture was selected because it worked. Its brain learns by computing the gap between what was expected and what was obtained. Its culture has extended the play of rewards into the domain of symbols, to the point where it can die for an idea. And satisfying its desires does not durably change its level of contentment.

Two problems follow.

The first is immediate. If all this is accurate, where is choice to be located? We believe we deliberate; perhaps we only execute. This is the subject of Part II.

The second is deferred but already visible. We pursue a state that the pursuit itself seems to put out of reach. This is the subject of chapter 15, and it is the real question of this essay.

---

---

Part II. The Condition of Freedom

If the affects decide, what is left of choice?

---

6. Reason in the service of the passions

The reversal

Reason is ordinarily taken to be what distinguishes humans from animals: a faculty that rises above impulses, weighs them, and decides. Reason would be what makes us free, in the precise sense that we could act otherwise than where the affects push.

Hume reverses this picture, and the reversal governs the whole of this part. On his account, reason is an instrument, not an engine. It establishes relations of ideas and matters of fact; by itself it can produce no desire. A flawless chain of reasoning has never set anyone in motion unless something at its end was wanted. Hence the most famous formula of the Treatise: “Reason is, and ought only to be the slave of the passions.”⁶

The formula is often quoted without its argument, which is more interesting than the formula. Hume observes that a belief alone motivates nothing. Knowing that an object has some property leads to action only if that property already matters. Reason can tell someone that a path leads to a place; it cannot make them want to go there. It therefore intervenes at two points, and only two: in the choice of means, and in the correction of the mistaken beliefs on which a desire rests.

Hence the provocative consequence that Hume accepts: there is nothing contrary to reason in preferring the destruction of the whole world to a scratch on one’s finger. A preference is neither true nor false; only beliefs are. The position is extreme, and it has the merit of showing clearly what is at stake.

What deliberation does

The immediate objection is empirical: we deliberate, and sometimes we act against what we want. We would rather stay in bed, and we get up. We want to answer cruelly, and we hold our tongue. Reason does seem to win.

Consider what happens in these cases. Reason does not win against all passion; it arbitrates between passions that work on different time scales. We hold our tongue not because a disembodied faculty has so decided, but because the prospect of a falling-out weighs more than the immediate relief of the retort. The anger is real, and so is the fear of the damage. What is called self-control is most often a distant passion prevailing over a near one.

Two configurations cover the essentials: either the action occurs under the sway of an affect, or it occurs for the sake of an affect. In the first case, impulse settles what seems most pleasant in the moment. In the second, calculation settles what will seem most advantageous later. The criterion stays the same; only the time horizon changes.

The opposite phenomenon is instructive, and the Greeks gave it a name: akrasia, weakness of will. One knows what is better and does something else. Socrates held the phenomenon to be impossible, since no one does wrong knowingly. Aristotle, more attentive, saw in it the victory of a vivid representation over general knowledge. Contemporary psychology describes the same thing as temporal discounting: a reward loses value as it recedes in time, and it loses value much faster than a consistent calculation would allow. Hence the classic reversal of preferences: one chooses long-term health, and the cigarette now.

Akrasia is therefore not a mysterious failure of reason. It is the predictable result of a system in which two evaluations, one immediate and one deferred, are not made at the same rate.

There is more. Reason serves the affects not only in its ends but in its operation. We do not stop to reflect on just anything: attention is selective, and what holds it is what affects us. No one deliberates about questions that leave them cold, not out of laziness but because nothing directs the effort. Affectivity carves out the domain of the thinkable before reasoning begins.

The exact scope of the thesis

This thesis is regularly pushed beyond its scope, so its scope should be stated precisely.

To say that reason serves the affects means neither that it is useless nor that it changes nothing. A sound calculation and a faulty one do not lead to the same place, even when both are motivated. Reason is an instrument, and the quality of an instrument matters. Someone who understands the real consequences of their acts will behave differently from someone who does not. This is not because they are freer. It is because their map of the world is better, and their affects apply to a more accurate map.

The point matters for what follows. All of Part IV rests on the idea that judgments can be worked on. If reason were an ornament without effect, that work would be pointless. The Humean thesis says that reason does not set the goals; it does not say that reason leaves trajectories unchanged.

7. The collapse of choice

The argument

If instinct follows affect, and if reason, which was supposed to free us from it, follows it as well, where is freedom to be found?

The idea of freedom rests on the idea of choice: the conviction that at a given moment several options were really open and that one of them was taken. Consider how a choice is made. One takes what, all things considered, seems most advantageous in the broad sense: the most pleasant or the least painful, now or later. Decisions vary because tastes, skills, representations and opportunities vary. If one and the same path invariably seemed most advantageous to everyone, everyone would take it.

It follows that, strictly speaking, the agent is not what chooses. The configuration of the agent’s affects, information and circumstances is what makes one option win. Spinoza put it better than anyone: “men believe themselves free for this sole reason, that they are conscious of their actions and ignorant of the causes by which they are determined.”⁷ We see the decision; we do not see what produces it. From this blindness comes the feeling of being its source.

The image he gives elsewhere is more striking still: a thrown stone that was conscious of its motion would believe itself free to fly⁸.

The limit cases

Some actions seem to contradict this model: suicide, the crime of passion, sacrifice, resistance under coercion. No one would call them advantageous.

The answer is that they need not be advantageous objectively. They need only have seemed so at the moment they were committed. To the person who resolves on it, suicide presents itself as relief. Violence seems less unbearable than contained rage; resisting seems less unbearable than betraying. These are disastrous evaluations, and they are real evaluations.

This answer is correct, and too convenient. It has exactly the form of the problem flagged in chapter 2: whatever the conduct observed, one can always posit the internal evaluation that makes it coherent. The model survives everything because it reconfigures itself after the fact. This is the second occurrence of that structure, and it will not be the last.

The empirical support and its fragility

Experimental psychology has supplied determinism with arguments the seventeenth century lacked. Several lines of results converge on one intuition: consciousness comments on and justifies action more than it initiates it.

Libet’s experiment showed that a preparatory brain signal precedes the conscious impression of deciding⁹. A large share of behavior (walking, driving, grasping) unfolds without conscious supervision, which steps in only when something breaks down. In patients whose corpus callosum has been severed, one hemisphere acts and the other, when questioned, invents a coherent explanation: the brain fills the gap rather than admit its ignorance¹⁰. Some pathologies produce actions without conscious intention. Broader work on the sense of agency shows that it can be manufactured experimentally, by giving someone the impression of having caused an event they did not cause¹¹.

The overall picture suggests an organism that acts first and assigns itself an intention afterward, with language serving as an instrument of retrospective rationalization.

This enthusiasm needs cooling, however, because it has produced many abuses. The interpretation of Libet is seriously disputed. The readiness potential may reflect an accumulation of random fluctuations rather than a decision already made. And the experiment concerns only arbitrary gestures with nothing at stake, not the deliberations that matter, those that stretch over months. Confabulation proves that consciousness can be wrong about its causes; it does not prove that consciousness is always out of play.

There is a body of converging evidence. There is no proof.

8. Merit and blame

What collapses

Here is the consequence, and its implications are the heaviest of the whole argument.

If acts flow from affective configurations that were not chosen, the notions of merit and responsibility lose their footing. A person born or raised with a benevolent temperament deserves no more than a cruel one: each follows the slope of their affects. One finds satisfaction in helping, the other in harming. Were their dispositions swapped, each would take up the other’s life where it stood.

Whether one does a kindness or commits a crime, the structure is the same: one does what seems best in the moment. No one, then, is guilty or virtuous in the strong sense of the term.

Plato already had Socrates say that “no one is unjust voluntarily, but all who do wrong do so against their will.”¹² This is not naivety. It is the exact consequence of the idea that each person pursues what they take to be their good. Whoever does harm does not aim at evil as such. They aim at a good for themselves, of which the harm to others is a by-product.

Moral luck

A separate argument leads to the same place without even assuming determinism, and it is harder to dismiss.

A considerable part of what people are held responsible for depends on factors they do not control. Two drivers commit the same recklessness; a child crosses in front of one and not the other. The first will be tried for homicide; the second will get a ticket. Their conduct is identical. The only difference lies in what depended on neither of them.

The phenomenon has been named moral luck, and it takes several forms. There is resultant luck, luck in outcomes, as in the example of the drivers. There is circumstantial luck: no one knows how they would have behaved under a regime that rewarded informing on others. Those lucky enough never to have been put to the test are judged more favorably than those who failed it. Finally, there is constitutive luck: temperament, intelligence, the capacity for empathy, none of which anyone authored.

Yet morality was supposed to be precisely the domain exempt from luck. The finding is not that we judge badly. It is that the notion of merit, pursued consistently, dissolves in one’s hands.

The genealogy of blame

Nietzsche asks a different question: where does the very idea of responsibility come from?

His answer, in Twilight of the Idols, is brutal. Wherever responsibility is sought, what is generally at work is the instinct to punish and to judge. On this view, the doctrine of the will was invented essentially to make punishment legitimate: men were declared free so that they could be found guilty¹³.

The analysis is powerful, and it holds as a hypothesis rather than as a fact. It explains the social function of assigning responsibility very well, and its universality less well. It nonetheless sheds light on a real connection. In the West, the insistence on free will has a legible theological origin. Faced with the problem of evil, the accepted solution is to say that God left man free, and therefore capable of doing wrong. This answer ranks freedom above happiness: a trade-off that many have inherited without examining it.

What survives

If no one is guilty, must everything be permitted? No, and two distinct answers are available.

The first is consequentialist. That punishment is not deserved does not imply that it is useless. Everything above establishes that conduct is shaped by its consequences. Sanctions are among those consequences, and they therefore modify future behavior. What collapses is not the sanction but its retributive justification: the idea that inflicting suffering is good in itself because it is owed. A justice consistent with this framework would be preventive and protective. It would prevent reoffending, deter, repair, and incapacitate. It would never take as its task to make people suffer in proportion to their fault.

The second answer is subtler, and it has profoundly changed the debate. It observes that the question is badly posed.

Resentment, gratitude, indignation, forgiveness (what have been called the reactive attitudes) are not conclusions drawn from a metaphysical theory. They are immediate reactions to the good or ill will shown toward us, and they are constitutive of relations between persons. No one gets angry at a friend after first checking that the friend has libertarian free will.

The decisive observation is that we already know how to suspend these attitudes, and that we do so without ever invoking determinism. Faced with a very young child, a person in a psychotic crisis, or someone who acted under coercion, we adopt what may be called an objective stance. We stop blaming, and we seek to understand, to treat, to protect. Universal determinism would require adopting this stance toward everyone, forever.

That is where the difficulty lies: this generalization is probably beyond our reach. No one decides to stop being hurt by a betrayal. And even if it were within reach, it is not clear that it would be desirable. The same suspension that removes resentment also removes gratitude, esteem, love, and everything that involves treating another as someone rather than something.

This answer does not refute determinism. It holds that blame and gratitude did not depend on what anyone believed, and that a metaphysical truth has no power to abolish them.

Two worries

Two difficulties remain, and it would be dishonest to pass over them.

Abandoning merit cuts both ways. If there is no fault, there is no achievement either: no one deserves what they have built, nor the recognition they draw from it. People readily accept this consequence for criminals, and much less readily for themselves.

Pure prevention has no internal brake. If sanctions are proportioned no longer to what was done but to what one wants to prevent, nothing limits their severity. Nothing protects an innocent person judged dangerous. Retribution is philosophically ill-founded, yet it serves as a safeguard: it caps the penalty. No satisfactory answer is available here, and the difficulty remains open.

9. The compatibilist objection

This is the strongest objection that can be raised against everything above, and it deserves more than a mention.

The dilemma

Start with the tightest argument against free will, to measure what the objection has to face.

The brain is a material system in a universe governed by laws. Whether those laws are deterministic or probabilistic, either reading closes the door on free will understood as an autonomous origin of action.

If the act is determined by the prior state of the system (history, learning, biological state, situation), it follows from a causal chain and is not a creation out of nothing. If it is due to chance (the opening some believe they find in quantum indeterminacy), it is no freer for that. Introducing randomness into a system does not create a will, and an act drawn by lot is not a willed act. Chance is the absence of control, not freedom.

There is no third term. A decision is either caused or not. If caused, it traces back to antecedents; if uncaused, it is arbitrary. Positing an immaterial entity that decides freely only moves the problem, since that entity must in turn be either caused or random.

The argument admits an even more radical formulation. To be truly responsible for one’s acts, one would have to be responsible for the way one is: for one’s character, values and preferences. But to be responsible for the way one is, one would have had to choose to be that way. That choice presupposes earlier preferences, for which one would in turn have to be responsible. The regress is infinite and nothing stops it¹⁴. No one has made themselves.

The reply

The compatibilist grants all of this, and replies that none of it was ever needed.

The compatibilist’s thesis is that the problem was manufactured by demanding that the word “freedom” name an exception to causality, a demand no one makes in ordinary life. To say that someone acted freely is to say that they acted on their own reasons and desires, without external constraint, understanding what they were doing. It is also to say that they would have acted otherwise had their reasons been different. Nothing in this requires indeterminism.

A finer version identifies what distinguishes a free agent. What characterizes a person is not having desires (an animal has those) but having desires about one’s desires. Take an addict who wishes he did not want his dose, and another who endorses his own desire. Both have exactly the same impulse. They differ in that the second recognizes himself in what he wants. Freedom, on this reading, is that agreement between what one wants and what one would want to want. It requires no gap in causality. And it explains why inner alienation feels like a loss of freedom even when nothing external constrains.

A third argument attacks a principle that seemed obvious: that responsibility requires having been able to do otherwise. Imagine someone who deliberates and decides alone. A device, however, would have forced them to do the same thing had they been about to decide otherwise. The device never intervenes. The agent could not have done otherwise, and yet intuition says they are responsible, since it was their own deliberation that produced the act. What counts, then, would be the actual origin of the action, not the existence of alternatives.

This kind of freedom, moreover, is enough for the work the notion is asked to do. The law has long understood this. It exempts no one on the grounds that the universe is determined. It does exempt the very young child, the person whose judgment is impaired, and the person who acted under irresistible coercion. What the law requires is not metaphysical freedom but a functional capacity for self-determination: enough room to say that the act truly belongs to its author. The distinction works, it sorts real cases, and it presupposes no magic.

The state of the question

Compatibilism is not refuted. What can be set against it is more modest: it changes the subject.

It is right about how the word is used. It is right that this kind of freedom exists and that it is enough to organize a society. But the question this essay pursues is not how the term is used. It is whether, when someone bitterly reproaches themselves for an old decision, the reproach targets something real. The compatibilist answers that the person acted on their reasons. The question asked whether, those reasons being what they were, the person could have done anything else. The compatibilist does not answer it, but explains that the question is badly posed.

Perhaps it is. One can only note that it keeps arising, and that the relief being sought (discussed at length in chapter 19) depends on the answer.

The objection stands. It is the most contested point in everything so far, and it has not been settled.

10. Four ways of inhabiting necessity

With this chapter the perspective reverses; it marks the first turn of the argument.

So far, determinism has appeared as a loss: freedom disappears, merit collapses. Yet the traditions that thought most deeply about necessity did not see it as a prison. Each, in its own way, saw a door.

The Stoics: consenting

For the Stoics, the universe is a rational and necessary order, governed by the logos, the cosmic reason that runs through all things. Nothing happens by chance: every event expresses a causal chain inscribed in the rationality of the world. Fate is not blind fatality but an intelligible web.

Chrysippus defended a causal sequence so complete that it left no room for any real indeterminacy. He nonetheless upheld responsibility, by means of a famous distinction: the causes that set a movement off are not those that determine its form. A cylinder that is pushed rolls. The push sets it in motion, but its own shape makes it roll rather than slide. Likewise, external impressions solicit, but it is character that responds, and character is ours, even though it was itself produced.

The human being is therefore included in the causal flow, and what belongs to us must be sought within necessity. That is the prohairesis: the faculty of judgment, of giving or withholding assent. It is an inner freedom that consists not in changing the course of things but in bringing the right judgment to them. Stoic determinism does not crush responsibility; it redefines it. Human beings answer for their acts not because they are their first source, an idea foreign to this tradition, but because they are able to fit their representations to reality.

Spinoza: understanding

Spinoza pushes necessity further still, and draws from it a remarkable practical consequence: “each person has the power to understand himself and his affects clearly and distinctly, if not absolutely, then at least in part, and consequently to bring it about that he suffers less from them.”¹⁵

This is the heart of his proposal. An affect undergone governs; the same affect understood governs less. Understanding why anger is there (what representation, what disappointed expectation, what history) does not make it vanish, but it loosens its hold. A passive affect becomes active as soon as one forms a clear idea of it.

Freedom, on this reading, is not the absence of determination: it is the understanding of necessity. It admits of degrees, it can be cultivated, and it requires no exception to the laws of nature.

This idea will have a considerable lineage, and chapters 18 and 19 return to it. It is exactly the structure of what cognitive therapies would formalize three centuries later.

Buddhism: dissolving

Buddhism invokes no cosmic reason. It posits dependent origination, pratītyasamutpāda: every phenomenon exists in dependence on other phenomena, and nothing happens without a cause. Chance is excluded, but the order is neither rational nor directed toward a good.

Liberation therefore comes not through consent to an order but through the dissolution of an illusion. Suffering, dukkha, comes from attachment, that is, from the belief that something can be lastingly possessed or controlled. Understanding impermanence undoes that belief.

The divergence from Stoicism is real and often blurred: the Stoics accept a rational order, while Buddhists undo the idea that there is a substantial order. The same apparent result, peace, is reached by two metaphysically opposed paths. Treating them as two dialects of a single wisdom is an easy shortcut, and one to avoid.

Taoism: not forcing

The Tao Te Ching gives the name Dao to the spontaneous order of the world, prior and ungraspable: “the Dao that can be named is not the eternal Dao.” It is not understood; one attunes oneself to it. Hence wu wei, non-action, which is not inaction but action without forcing, action that follows the course of things. The model is water: yielding, it flows around obstacles and in the end wears down rock. And the movement of the Dao is return: what is hard breaks, what is supple endures.

The Taoist differs from the Stoic on one important point. For the Stoic, peace comes from intellectual effort, from discipline, from conscious consent. For the Taoist, it comes from the relaxing of the will. Disturbance arises from rigid willing and excessive control, and peace comes when one becomes supple.

The objection is immediate: “try not to try” is a paradoxical instruction, and wu wei cannot easily be commanded. The tradition answers that it is an acquired skill rather than an act of will, comparable to the mastery of a craftsman who no longer has to think about his movements. This is probably the best answer available, and it moves the difficulty without removing it.

The common reversal

Four different metaphysics, at least two of them incompatible. And yet one common structure:

Peace is obtained by giving up control.

The mechanism is identical in all four cases. Suffering comes not from circumstances but from resistance to circumstances: from the gap between what is and what one demands. Remove the demand and the gap disappears, though nothing in the world has changed.

This convergence provides no argument for truth. That four traditions separated by continents arrive at a similar instruction shows that they met the same human problem. It does not show that their solution is the right one. These are two distinct things, and they are often confused.

Finally, note what none of these ways does: abolish pain. All of them work on judgment; none works on sensation. This is a structural limit, and it will matter again in Part IV.

A fifth answer

There is another answer, which neither accepts, nor understands, nor dissolves, nor relaxes.

Nietzsche denies neither forces, nor instincts, nor conditioning. He refuses to set freedom against necessity. The world is made of forces; these forces can be interpreted; interpretation creates values; these values determine a form of life. Emancipation consists not in escaping determinism but in becoming the author of one’s own meaning. That is the sense of amor fati: not to endure necessity but to love it, to want nothing to be otherwise, not forward, not backward, not in all eternity.

The status of this answer should be stated plainly. It is more a stance than a solution, and it is not clear that it meets the argument of chapter 9, since interpretation is itself caused. Yet it names something the other four leave aside. All of them propose attuning oneself to reality; none proposes affirming anything in it. The conclusion returns to this.

---

Where we stand

Choice, examined closely, looks less like an origin than like a result. Merit and blame lose their foundation, while punishment keeps its usefulness and the reactive attitudes do not disappear. The compatibilist objection still stands and has not been overturned. And the traditions that took necessity most seriously drew from it not despair but a method.

One question remains, which this part has sidestepped. If nothing is good or bad in itself, if no one is guilty, if reason can argue for almost any position, what can a moral judgment still rest on?

That is the problem of Part III, and it is the hardest.

---

---

Part III. The Ethical Transposition

What becomes of the good if nothing outside us grounds it?

---

11. The problem

A forgotten distinction

Two kinds of proposition need to be separated first, because ordinary conversation mixes them.

There are facts: what can be established by observation, measured, corroborated, refuted. Water boils at a hundred degrees at sea level; this person lied; unemployment has risen. One can be wrong about them, and that is precisely what makes them facts: a procedure exists for settling the matter.

There are value judgments: this is good, that is unjust, this act is contemptible. From a standpoint that invokes no transcendent authority, these propositions do not have the same status. No observable fact makes them true. Nothing is good or bad in itself; things become so because they are judged so.

Two arguments support this position, and they deserve to be kept apart.

The argument from relativity. Moral conceptions vary considerably from one era and one society to another. This variation is better explained by commitment to different forms of life than by uneven discovery of a single objective moral order. Scientific disagreements eventually get resolved; moral disagreements persist with remarkable regularity. The classic objection (the disagreements concern facts, not values) holds for some cases, not for all. People can agree on every consequence of a practice and still differ on whether it is acceptable.

The argument from queerness. If objective values existed, they would be entities of an unprecedented kind: properties that, besides existing, would by themselves demand that one act in a certain way. Nothing else in the inventory of the world behaves like this. No fact prescribes; a desire must always be added for an action to follow. And a special faculty would also be needed to perceive such values, a faculty no one has ever described convincingly.

From this follows a position sometimes called error theory. Our moral statements do purport to describe objective facts (they are not simply our preferences in disguise), but they are systematically false, because there is nothing for them to describe.

This variability is easy to forget when everyone around shares the same values. An uncontradicted consensus comes to pass for the obvious, and the obvious for the true.

Reason does not decide

Then comes the question that governs everything else: if morality is only judgment, what makes one moral judgment better than another?

The spontaneous answer is reason. Reason would be the criterion of objectivity in non-factual matters.

There are good grounds for doubt. Reason manages to justify almost any moral position with arguments of comparable force. Take any lasting divide: individual rights against collective rights, equality against merit, protection against liberty. Each side produces a coherent, well-articulated case resting on defensible premises.

The claim is not that all arguments are equal: some are badly built, and that can be shown. The claim is that when two competent minds have conducted a debate properly, reason alone is generally not enough to decide between them. The disagreement lies in the premises of value, and such premises are not the kind of thing one demonstrates.

The Greek skeptics gave the formula: “to every argument, an argument of equal force can be opposed.”¹⁶ They drew from it a consequence that chapter 24 examines.

Reason comes afterward

There is something more serious, and here contemporary psychology has vindicated the ancients rather spectacularly.

We do not build our moral positions the way we think we do. The usual story runs as follows: one gathers information, examines the arguments, weighs them, concludes. The real sequence is almost always the reverse. A judgment arrives first, fast and affective; the argument comes afterward to justify it. Reason works here less as a judge than as a lawyer, charged not with establishing the truth but with defending a position already taken¹⁷.

The studies on this point are telling. People are placed in situations designed to offend an intuition without causing any harm. They keep condemning the act after all their arguments have been refuted. They do not change their minds; they simply declare that they cannot explain why it is wrong, but that it is. The judgment survives the collapse of its reasons, which shows fairly clearly that those reasons were not holding it up.

Two mechanisms make the bias especially visible. The first is the appeal to feeling: a symbol that can evoke pity, indignation or anger tips an opinion without any argument being exchanged. The second is interest: we are more moved by a stake that concerns us, and more favorable to what benefits us.

Schopenhauer put it bluntly. The intellect remains so far excluded from the real resolutions of the will that it often learns of them only afterward, as one learns those of a stranger, by spying on him¹⁸.

One precaution is needed, however. That moral judgments are affective first does not show that they are false. It is a fact about their genesis, not about their worth. To confuse the two would be to commit exactly the error Nietzsche denounced: believing that exposing the origin of a belief is enough to refute it.

Kant’s objection

Before following the consequence, it has to face its most serious adversary, because this whole chapter assumes what that adversary disputes.

Kant grants Hume much of his starting point. From inclinations one can derive only hypothetical imperatives: if one wants a given end, then one must take a given means. These imperatives bind only those who have the corresponding desire, and no morality can rest on them, since one would need only to stop wanting the end to be released. On this ground, reason is indeed the instrument of the passions.

But Kant holds that there is another use of reason, one that borrows nothing from desire: a pure practical use. It supplies no content; it supplies a form. “Act only according to that maxim through which you can at the same time will that it become a universal law.”¹⁹ The test requires no premise of value. It asks only whether the rule one gives oneself could be willed by everyone without destroying itself. Someone who makes a promise with no intention of keeping it cannot pass the test. A false promise, universalized, abolishes the very practice of promising, which it needs in order to deceive. The contradiction is logical, not affective.

A second formulation gives the same principle a face: treat humanity, in one’s own person as in that of another, always at the same time as an end and never merely as a means²⁰. Kant adds that this law binds categorically (independently of what one desires). That is exactly the claim error theory recognized in moral statements. Where that theory concluded the claim is false, Kant holds that it is well founded, because reason itself makes it.

If this holds, everything above collapses: reason decides, and it decides without the help of any affect.

Three difficulties nevertheless prevent the matter from resting there.

Formalism. The universalization test decides only according to how the maxim is described. A hateful rule, stated precisely enough, universalizes without contradiction; a harmless rule, stated clumsily, fails. Hegel made this his central reproach: form alone is empty, and it delivers verdicts only once it is fed prior judgments about what counts as a relevant description. The premises of value the test claimed to exclude come back in through this door.

The motive. Kant must explain how a purely rational law sets anyone in motion. His answer is respect: a feeling of a unique kind, produced not by an object but by the representation of the law. He thereby concedes Hume’s point at the very moment he fights it. Even for Kant, what moves is something felt. Calling this feeling rational does not change its place in the explanation.

The starting point. Kant calls consciousness of the moral law a fact of reason: it is observed, not demonstrated. It is an axiom, and the next chapter argues that every ethics posits one.

Something of Kant nonetheless remains that none of these objections reaches, and the rest of this essay will need it. The demand for impartiality (making no exception for oneself that one would refuse to others) cannot be drawn from any spontaneous affect. Chapter 13 shows that natural moral sentiments are partial, and that their extension to all sentient beings is the work of an argument. In other words, if reason does not ground morality, it may be what extends it. Kant’s position is not refuted. It is brought down to the rank of its rivals, which all posit an axiom, and it leaves behind a demand that naturalism cannot produce on its own.

The impasse

Set aside what Kant salvaged, and follow the consequence through all the same.

Suppose morality is subjective; suppose reason can defend almost any position with equivalent force; suppose nothing makes an opinion that conforms to reason objectively better than another. Then no moral judgment has objective value, and one must admit that everything is equally valid.

This is nihilism in the strict sense: the absence of values. It is not the rebellious pose attributed to adolescence. It is the conclusion of an honestly conducted argument.

Yet there is one thing this argument has left out.

12. Sensibility as the only ground

The one moral fact

The proposed way out fits in a sentence: there exists one natural moral fact, and only one: suffering matters because it is felt.

The reasoning of the previous chapter shows that constructed values (those people argue for, defend and pass on) have no objective foundation. It does not show that there is nothing at all. Something remains that was not constructed and that no argument undoes: some states are lived as desirable and others as unbearable, by us and by everything that feels.

This fact is not an opinion. A burn is not unpleasant because someone judged it so, and no effort of reasoning will make torture pleasant. This is therefore the only serious candidate for the title of natural value. It is also already, beneath the cover of moral reasoning, the principle that actually directs conduct.

Bentham stated it once and for all: “Nature has placed mankind under the governance of two sovereign masters, pain and pleasure. It is for them alone to point out what we ought to do, as well as to determine what we shall do.”²¹ The sentence states a prescription and a description at once, and this double reach is exactly what is at stake.

He draws from it a consequence whose radicalism should be measured. What brings a being into moral consideration is neither its intelligence nor its language: “The question is not, Can they reason? nor, Can they talk? but, Can they suffer?”²² The criterion is sensibility, and nothing else. It crosses the boundaries of species, as it crosses those of sex, age and social condition, which was not obvious when he wrote it.

An argument from convergence can be added. Nearly all moral schools, all cultures and all religions hold that what causes suffering is bad. They diverge radically on who counts, on priorities, on exceptions; very few diverge on whether the suffering of the beings they consider is an evil. This is not proof, but it is a sign that this foundation is not cultural in origin.

A minimal definition follows: morality concerns only what can affect the happiness or suffering of sentient beings. Everything else (ritual purity, symbolic order, conformity for its own sake) belongs to something else.

The gap

This is where the argument is most vulnerable, and it is better to attack it than to dodge.

The objection is old. Hume observes that treatises on morality proceed for a while by propositions of fact, then pass without warning to propositions containing an ought. This passage would need an explanation that no one ever supplies. Moore restates the difficulty through what is called the open-question argument. Whatever natural property one uses to define the good (pleasure, the desirable, what accords with nature), the question “but is that good?” still makes sense. If the two terms were synonymous, it would not.

Observing that suffering is universally avoided therefore does not establish that it ought to be avoided. Three remarks follow, in place of a refutation.

The first is that the objection applies to everyone. No ethics crosses this gap by demonstration. Kant starts from a fact of reason, as the previous chapter showed; contractualism starts from a hypothetical situation, religious ethics from a revelation. Each posits an axiom. Naturalism merely has the drawback of making its own visible.

The second is that sensibility has a singular status among facts. Most facts are neutral: the mass of an object, the temperature of a room. An affective state is not neutral for the one who undergoes it: it is constitutively good or bad for that being. This may be the only place in the world where a fact arrives already charged with a valence. That does not close the logical gap. It indicates that if a bridge exists anywhere, it is probably here.

The third is an observation. The choice to ground ethics on sensibility is not demonstrated; it is assumed. It is assumed because no better candidate presents itself once transcendent foundations are set aside, and because an ethics resting on anything other than what is felt would be indifferent to what matters. That is a reason; it is not a proof.

Whose sensibility?

A second difficulty is less conspicuous and more serious.

Part I argued that each individual acts only for the sake of their own affective state. If morality rests on sensibility, whose sensibility is meant: one’s own, or that of all sentient beings?

The two answers do not lead to the same place, and passing from one to the other is not legitimate. That each person is motivated by their own states is a thesis of psychology: psychological egoism. That each person ought to consider only their own states is another thesis: ethical egoism. The first does not entail the second. If anything, it tends to make the second superfluous: one does not prescribe what is inevitable.

And ethical egoism defends itself poorly on its own ground. If what makes suffering bad is the fact that it is felt, then the suffering of others is bad for the same reason: the foundation does not distinguish between those who bear it. To claim that only one’s own suffering counts forces one to hold that one is a special case, without being able to say how. Sidgwick called the alternative the point of view of the universe: once one adopts an impartial perspective, one’s own situation has nothing privileged about it.

A gap therefore remains between what moves people and what the foundation designates. This gap is not closed here. Closing it by suppressing one of the two terms would be the lazy solution.

13. Where moral sentiment comes from

If morality has natural roots, it should be possible to show them.

Sympathy

Hume, who, as chapter 6 recalled, makes reason the instrument of the passions, draws the moral consequence. Ethics needs no external foundation: it arises from human constitution. Its central axis is sympathy: the capacity to be affected by what another feels.

Sympathy is not benevolence by decree; it is a mechanism. Another’s visible distress produces an unpleasant state in the observer; that person’s relief produces the opposite. From there, approval and indignation become intelligible without any appeal to a moral law. Virtue, Hume writes, is whatever produces, when contemplated, a feeling of approval in an informed and disinterested spectator.

Adam Smith refines the scheme by introducing the impartial spectator. We judge our own acts by imagining the gaze of a well-informed observer with no interest in the matter, and this internalized figure produces conscience. The construction has a considerable merit: it explains how a partial sensibility can generate judgments that claim impartiality.

The ethological evidence

Hume did not speak of natural selection, but his intuition anticipates what ethology would document.

The work mentioned in Part I carries its full weight here: consolation after a conflict, sharing, reciprocity, refusal of conditions perceived as unfair. Repeated experiments show primates rejecting a reward they normally like when a companion receives something better for the same task. These behaviors proceed from no moral reasoning. They proceed from social affects that secured the cohesion of groups.

Moral sentiment therefore has a biological prehistory. It is not a veneer laid over a fundamentally self-interested animal.

Three limits

The first. Showing that morality has natural roots does not establish that it reduces to them. Between a chimpanzee’s consolation and a system of universal rights lies a leap that animal observation cannot make on its own. The continuity of affects does not explain the invention of institutions.

The second. These roots ground nothing. A genealogy explains; it does not legitimate. This brings back exactly the difficulty of the previous chapter.

The third is the most awkward. Evolution did not select a general benevolence. It selected a partial sensibility: stronger toward kin, toward those alike, toward members of the group. Spontaneous moral affects are exactly as narrow as the groups in which they formed. Whoever looks to nature for a universal morality will not find one there. They will find an animal capable of remarkable devotion to its own and remarkable indifference to others.

The natural foundation of morality and its universal extension therefore do not come from the same place. The history of the last three centuries can be read as the gradual widening of the circle of beings whose suffering counts: beyond the tribe, then the nation, then the species. Spontaneous affect did not carry this widening, since it did not push that way. An argument carried it, the very one Bentham formulated: if sensibility is what makes suffering bad, no boundary is relevant.

In other words, universality is a rational construction applied to a partial affective base. On this precise point, that gives back to Kant what chapter 11 took from him elsewhere. The position is unstable: reason lacks the strength to motivate on its own, and affect does not reach as far as reason does. This can be seen as the central difficulty of any naturalist ethics.

14. The good life: fulfillment or state?

A duty to oneself?

One question deserves to be asked before the paradox: does naturalist ethics generate a demand on oneself?

The appealing formula, being happy is a duty, does not hold as it stands. A duty presupposes that one can fail in it. If the pursuit of happiness is what everyone does necessarily, by construction, then it is not a duty but a description. One does not prescribe what is inevitable, and presenting a necessity as an obligation lends a fact the prestige of a value.

The underlying intuition nonetheless has a serious lineage, in three distinct forms.

For Aristotle, happiness, eudaimonia, is not a subjective state but the fulfillment of what a being is made to be. Every being tends toward its telos; for a human being, this fulfillment runs through the activity of reason and the balance of the affects. The virtues are the stable dispositions, acquired by habit, that realize this function. Happiness is not a pleasant moment but the activity of a whole life in accordance with excellence. And he adds, in a point often forgotten, that one cannot call a life happy before it is complete.

For Epicurus, ataraxia is the natural end of human life, and failing to aim at it means living in needless turmoil. He makes it less a duty than a debt to oneself: freeing oneself from groundless fears and vain desires is a form of care, almost a therapy. Philosophy, he writes, is empty if it heals no suffering of the soul.

For Spinoza, joy is not a pleasant extra but the very increase of the power of acting, and so the expression of virtue. Not to seek rational joy would run against one’s own essence.

Three formulations share one intuition: there is an ethics turned toward oneself in which concern for one’s own fulfillment is not a whim.

The decisive distinction

What eudaimonism actually says differs from what the common reading takes from it, and this difference is the pivot of the whole essay.

Eudaimonia is not a pleasant affective state. It is an activity, a fulfilled life, the exercise of what is best in oneself. Aristotle does not say: try to feel good. He says: be fully what you can be, and contentment will follow as an accompaniment, not as a target.

The difference is considerable. The conventional logic of happiness (that of chapter 2, of economics and of common sense) makes it a state to be obtained, a balance to be maximized. Eudaimonism makes it the by-product of an activity pursued for its own sake.

Contemporary psychology has rediscovered this distinction in its own terms. States of absorbed concentration in a demanding task one has mastered, what has been called flow, are marked by the disappearance of self-consciousness and of attention to time. They are not experienced as pleasant while they last, and yet they rank among the episodes people report as the best of their lives²³. This suggests that a large part of what we call happiness is not a felt state but a mode of engagement.

A third measure: meaning

Chapter 4 noted that the word “happiness” covers at least two distinct things: experiencing well-being moment by moment, and judging one’s life satisfying when one evaluates it. There is a third, which eudaimonism points to without naming. It must be isolated here, because it governs what follows.

Studies that measure separately the pleasure experienced in an activity and the meaning attributed to it find a regular dissociation, and sometimes an outright opposition²⁴. Raising young children, caring for a sick relative, and seeing a difficult piece of work through rank among the activities richest in meaning and among the least pleasant at the time. Conversely, the pursuits that maximize immediate pleasure score very low on meaning.

The correlates diverge too. Pleasure is associated with satisfied needs, the absence of constraint, health and material ease, and it bears on the present. Meaning is associated with belonging to groups, with giving rather than receiving, with effort and with hardships endured. It presupposes a place in time, a past and a future linked together.

In other words: a life can be pleasant and empty, or painful and full. No theory that confuses the two can account for what people actually choose. They regularly choose, without being mistaken about what they are doing, what lowers their enjoyment and raises their sense that their life matters.

This third measure clarifies what eudaimonism was aiming at. Eudaimonia is neither pleasure nor retrospective satisfaction. It is something closer to meaning in this sense: a mode of existence that is oriented, engaged and extended in time. It also explains why Aristotle adds that a life cannot be judged before it is complete. Pleasure is measured by instants; meaning can be read only along a trajectory.

Finally, it has a practical consequence on which chapter 16 relies entirely. Meaning is not produced by trying to feel good, but by engaging in something that exceeds concern for oneself. If so, the way out of the paradox of hedonism may not be only to desire less. It could be to aim elsewhere.

Two objections

Two objections to the Aristotelian edifice should be noted.

Teleology. Everything rests on the idea that the human being has a proper function inscribed in its nature. Biology has discarded this notion: evolution assigns no purpose to species. One can try to save eudaimonism without the telos, by rereading it as what makes a social organism of this kind flourish durably. That is the reading adopted here, and it is weaker than the original.

Content. Aristotle makes the contemplative life the highest form of fulfillment. It is hard not to see in this a philosopher’s preference for the activity of philosophers. The difficulty is general. As soon as the good life is defined by a determinate content rather than by what the person concerned feels, one imposes a conception that person may not share.

This last difficulty is the price of the opposing position. Hedonism is democratic and empty: each person judges their own pleasure, but then nothing more can be said about the life of someone who declares themselves satisfied in degraded conditions. Eudaimonism says something, and pays for it with a suspicion of paternalism. Neither position is comfortable.

15. The paradox of hedonism

This question is the heart of the essay. Everything before leads to it, and everything after follows from it.

The contradiction

Set the two propositions side by side, each established in its own part.

The first, from chapter 2: happiness is the end of every action. Everything done is done for the sake of an affective state, the one term that points to nothing beyond itself.

The second, from chapter 4: satisfying a desire does not durably change how happy one is. Habituation dulls what is gained, and expectations rise to meet it, always demanding more for the same effect.

Together they describe a being that pursues one end with great method, by means that cannot reach it.

The phenomenon was named and analyzed long before anyone measured it. Butler had already pointed to it, extending his objection from chapter 2: since pleasure presupposes a desire for something other than pleasure, aiming at pleasure directly defeats itself. Sidgwick called this the paradox of hedonism: hedonism may be true as a criterion of the good, but it defeats itself as a method. Whoever aims at pleasure gets less of it than whoever aims at something else.

Mill, who learned it through his own breakdown, put it more simply. Those only are happy, he wrote, who have their minds fixed on some object other than their own happiness; aiming thus at something else, they find happiness by the way²⁵. The conclusion is exact. “Ask yourself whether you are happy, and you cease to be so.”

The mechanism

The mechanism is precise.

When something is judged good or bad, and is therefore coveted or dreaded, the judgment takes the form of desire or fear. This is the main source of the suffering tied to judgment, and for a structural reason: desire contains lack. To desire is to register that one does not have.

Three outcomes are possible, and none keeps its promise. If one fails, the stronger the desire was, the sharper the frustration. If one succeeds, disappointment often waits: the image was idealized, and the object turns out to be ordinary. Chapter 4 gave the mechanism for this, the dissociation between the circuits of wanting and those of liking. And if the object lives up to expectations, the satisfaction is brief. The elation subsides, the object loses its affective value and keeps only its usefulness, and a new desire arises in another form.

Seneca describes this mechanism with almost clinical accuracy. New agitations follow the first ones, one hope awakens another, ambition calls forth ambition; people do not seek an end to their miseries, they change their object²⁶.

One further spring is often overlooked: the worry of loss. What has been obtained must be defended. Possession does not deliver from agitation; it moves it elsewhere.

The metaphysical formulation

Schopenhauer gave this structure its most systematic version. Behind the world of phenomena, on his view, acts a single force, blind and aimless: the Will. We are expressions of it, a ceaseless willing. And to will is to lack. Satisfied, desire gives way to boredom; unsatisfied, to frustration. Hence the image: life swings like a pendulum between suffering and boredom²⁷.

He adds a disturbing consequence: the higher consciousness rises, the more suffering grows. Plants have no sensibility, and so no pain. The lower animals have a beginning of it. Only in the vertebrates does it grow markedly, and it grows in step with intelligence. It culminates in human beings, and all the more the more lucid they are²⁸.

The idea, in a less somber form, was already present in chapter 5. Reason, which allows happiness to be calculated better, is also what allows it to be missed. It widens the range of desires, projects into a future it does not control, and makes suffering by anticipation possible. The animal is content to satisfy its needs; the human being desires what does not yet exist and dreads what may never happen.

Diversion

Pascal described the same structure two centuries earlier, but from its reverse side: not what desire makes one suffer, but what it protects against.

His starting point is an observation he presents as the key to everything else: “all the unhappiness of men arises from one single fact, that they cannot stay quietly in their own chamber.”²⁹ Rest, which should be the goal of all agitation, becomes unbearable as soon as it is obtained. It leaves the mind alone with its condition: its fragility, its mortality, the absence of anything that could fill it. Hence the need to keep busy: gambling, hunting, conversation, business, office.

The example of the hunt contains the whole analysis. People spend the day chasing a hare they would not want if it were offered to them. The object is not what they seek. Men do not know, he writes, “that it is the chase, and not the quarry, which they seek.”³⁰ The object serves as a pretext for agitation, and agitation serves to keep from thinking.

This analysis lights up the paradox from a side Schopenhauer leaves in shadow. Desire is not only a lack in search of its object. It is also a way of occupying attention, which is why its satisfaction is sure to disappoint: it ends the occupation. Pascal draws from this the most exact formula of the paradox of hedonism to be found before Sidgwick: “we never live, but we hope to live; and, as we are always preparing to be happy, it is inevitable we should never be so.”³¹

His own way out is religious, and the framework adopted here cannot follow it. But the diagnosis is independent of it, and it raises a question Part IV cannot avoid. If activity is what keeps us from seeing ourselves, how can engagement that gives life be told apart from occupation that serves as flight? Chapter 16 meets this question at the very moment it recommends activity.

To hope is to fear

One last point prepares the whole of the next part.

What makes desire painful is not only lack; it is uncertainty. Hope causes suffering because it opens the possibility of failure. Mixed with this possibility is responsibility: the fear that the failure may be one’s own fault.

Spinoza draws an equivalence from this: “there is no hope unmingled with fear, and no fear unmingled with hope.”³² The two affects are one and the same, oriented differently. To hope is already to dread not obtaining.

The consequence is instructive. Whoever holds their lot to be unchangeable resigns themselves and stays at peace. It is the one who hopes, who believes they have a hold on things, whom uncertainty torments. The turmoil is proportional to the feeling of power.

This is the point at which Part III connects directly with Part II. If freedom is illusory, then part of the torment rests on an illusion: that of a power which does not exist. Nearly all the ways of emancipation examined next start from there.

Three warnings

Desire is not only a servitude. It is also the engine of action, effort and creation. Part I established this: without lack, nothing starts moving. A life without desire is not only peaceful; it is motionless. Every technique that proposes to reduce desire soothes, and can extinguish. This tension runs through all of Part IV and is not resolved there.

The paradox is not total. Adaptation is real, but it is not absolute. Some conditions are not forgotten; neither are some satisfactions, particularly those that come from an activity rather than an acquisition. This unevenness is probably the most useful practical finding in the whole chapter.

And the paradox is a datum, not an instruction. Observing that the head-on pursuit of happiness is counterproductive does not dictate what to do. One can draw from it renunciation, asceticism, a shift of attention toward activity, acceptance, or affirmation. These are different ways, and sometimes incompatible ones.

They are the subject of the last part, which presents them all.

---

Where we stand

The only moral ground left is sensibility: a thin ground, imperfectly founded, assumed rather than demonstrated. Moral sentiment has natural roots, but they are partial and carry no universality; universality is a rational construction grafted onto an affect that did not grow toward it. The idea of a duty to be happy does not hold. The eudaimonist intuition it distorted survives, however, and will prove decisive: happiness as the by-product of a fulfilled life rather than as a target.

Above all, what is pursued recedes as one advances.

The traditions have formulated answers to this problem. They do not agree with one another.

---

---

Part IV. The Ways of Emancipation

Once the paradox is granted, what do the traditions propose?

---

The status of this part

The first three parts argued. This one sets out.

It presents ways that traditions have worked out, sometimes over centuries, in answer to the problem left open in chapter 15. None is ranked, none is crowned. This is not caution for form’s sake. These ways are not all compatible, and to manufacture a synthesis that does not exist would do violence to every one of them.

The principle that governs them all

The paradox of hedonism is not merely a discouraging finding: it contains an instruction. If aiming at happiness directly drives it away, then the target must be moved. The whole of this part is an inventory of the possible moves.

They fall into two families, and this division sets the order of the chapters.

The ways from without turn attention toward something other than oneself: an activity pursued for its own sake, other people. They reduce nothing; they substitute one object for another. These are chapters 16 and 17.

The ways from within work on how what happens is received: judgment, the division between what is up to us and what is not, the economy of desire, one’s relation to oneself, to time and to death, discipline, the suspension of judgment. These are chapters 18 to 24. Almost all of them proceed by subtraction.

Chapter 25 stands apart: it does not move the target, it abolishes it.

The two families are not rivals, and this is the point on which the whole part turns. The ways from without assume that one can act; the ways from within matter precisely when one cannot. Commitment to a work is no consolation for an illness that rules it out. Attachment to others is not available to someone who has lost them. That is where the work of judgment takes over. Conversely, a life spent entirely reforming its perception of what happens, while nothing happens, is an empty life. Each family covers the other’s blind spot.

The primacy of judgment

Seven chapters out of ten work on the inside. The reason for this proportion deserves to be given rather than assumed.

Chapter 1 distinguished two sources of the affects: sensation, good or bad in itself, and judgment, which qualifies what happens. The traditions that worked out an art of living converge on one point. The second source weighs most in a life, and it is the only one we can get a grip on.

The Stoics make it the principle of everything else. Things, writes Marcus Aurelius, do not touch the soul; they stay outside, unmoving, and disturbance comes only from the opinion formed of them within³³. Buddhism opens its most widely read collection of sayings with the same claim: mental states have mind as their forerunner, mind as their master, and are made by mind³⁴. It also gives the most precise image of this that we have, the image of the two arrows. Someone struck by a painful sensation grieves, laments, struggles. He thus feels two pains, one of the body and one of the mind, like a man hit by a first arrow and at once by a second in the same spot. The instructed disciple receives the first and not the second³⁵. Sextus’s skeptic, as chapter 24 shows, says the same thing: he is cold and thirsty like everyone else, but he suffers only once.

Three reasons make this primacy plausible without requiring any of these metaphysics.

Duration. Most sensations are brief: a meal, a passing pain, a discomfort. The affects that come from judgment last. A mood, a grudge, a worry can persist for weeks or years, and they follow the person everywhere, at work as at rest. What fills the greater part of an affective life is not what the body feels but what the mind thinks of it.

The added share. Even where sensation dominates, judgment adds to it. Anticipation, fear and interpretation modulate the intensity of a pain without removing it. Someone who tells himself this will never end does not suffer as someone does who feels the same thing without that commentary. Epicurus noted this in his own way: extreme pain does not last, and pain that lasts can be borne³⁶. This is the second arrow, and it is the target of the exercises.

Purchase. No one decides not to be hungry. With training, one can stop treating a refusal as a catastrophe. Judgment is not free, as Part II showed at length. It is only modifiable, slowly, through repetition, which sensation almost never is.

This convergence is not a proof, and Part II rules out any talk of mastery. It does explain why traditions that started from incompatible premises ended up in the same workshop: changing neither the world nor the body, but the way what happens is received. That is what gives the ways from within their clinical descendants, examined in chapter 18. It is also what sets their limit. Where suffering lies in the body rather than in interpretation, no exercise of judgment is enough. The final section of this part is devoted to that limit.

How to read what follows

In antiquity, philosophy was not first of all a body of doctrine. It was a way of life, and its texts served as exercises: repeated practices meant to transform a way of being rather than to establish propositions³⁷. The maxims that follow are therefore not summarized theses but training instructions. They are to be judged less by their truth than by their effect, and they presuppose repetition. Read as theoretical claims, most of them seem flat; that is because they are not made to be read once.

One last point applies to all of them. None of these techniques acts on pain itself, only on what judgment adds to it. No exercise makes a raging toothache a matter of indifference.

What is known of their effectiveness

The pages that follow often claim that an exercise “can be learned,” that a disposition “can be trained,” that a skill “can be acquired.” What supports such claims, and what does not, should be stated.

What is established. Some of these principles have been built into structured protocols (cognitive therapies, attention-based programs, acceptance therapies). There they have been evaluated in comparative trials, and they produce measurable effects: on anxiety, on depressive relapse, on how persistent pain is lived. These protocols share certain features. They last a few weeks or a few months, they target a defined problem, they require daily practice, and they are supervised. Their effects are real and moderate, rarely spectacular.

What is less so. In people with no particular disorder, isolated exercises (keeping a gratitude journal, practicing a few minutes a day) produce on average weak effects, which fade when the practice stops. Several results in this field have held up poorly under attempts at replication. Programs directly inspired by the ancient schools have received only limited evaluation, often without a comparison group.

What is not established at all. That a maxim, read even often, durably transforms anyone. No data show it, and the ancients did not claim it. Their exercises presupposed a school, a teacher, a community, and daily practice kept up for years.

The gap between these three levels governs how what follows should be read. The lineage linking these practices to evaluated therapies gives them part of their credit. It also forces the admission that this credit attaches to supervised, repeated and targeted practices, not to reading a book. What produces the effect, as far as is known, is not the text: it is the exercise, and its duration.

---

16. Shifting attention

The instruction drawn from the paradox

Chapter 15 established that the direct pursuit of happiness defeats itself, and chapter 14 supplied the reason. What we call happiness is not only a state; it is also a mode of engagement. Meaning (a third measure, distinct from pleasure and from retrospective satisfaction) is produced in action rather than in introspection.

The consequence follows directly. If happiness is not obtained by aiming at it, perhaps it is obtained by aiming at something else, and aiming hard enough to stop monitoring oneself.

Mill is the most useful witness here, because he arrived at this answer through a breakdown. At about twenty, he put a question to himself: if everything he was working for came to pass, would it give him great joy? His inner answer was no, and that answer swept away his reasons for acting. What he drew from this crisis was not renunciation but a shift. Those only are happy, he writes, who have their minds fixed on some object other than their own happiness: on the good of others, on the progress of an art or a pursuit, followed not as means but as ends. Aiming at something else, they find happiness along the way³⁸. And he adds the only instruction the paradox allows: do not ask whether you are happy, for asking is enough to cease to be so.

Activity as an end

Aristotle had given this structure its technical formulation, and it is more precise than the usual paraphrases of it.

Pleasure, for him, is not what activity aims at. It is added to the completed activity as a supplement, “as the bloom of youth to the young.”³⁹ The image is exact. The beauty of a young person is not obtained by seeking it; it accompanies a state. In the same way, pleasure accompanies an activity done well and for its own sake, and it vanishes if one tries to aim at it separately.

Two consequences follow.

The first is that the activity must be desired for its own sake. Aristotle distinguishes activities performed for the sake of something else from those that are sufficient in themselves, and he places the second kind on the side of fulfillment. Work done solely for what it brings in does not produce this kind of pleasure, however well it is paid.

The second is that the activity must be exercised with excellence. This is not a moral ornament. Pleasure grows with the quality of the exercise, which is why learning a trade or an art transforms one’s relation to it. What was painful because one did it badly becomes absorbing once one does it well.

What experience confirms

Contemporary psychology has rediscovered this structure and described it precisely.

States of complete absorption in a task (what is called optimal experience, or flow) arise under identifiable conditions: a clear goal, immediate feedback on what one is doing, and above all a balance between the difficulty of the task and the skill available. Too easy, and the result is boredom; too hard, anxiety; well matched, absorption⁴⁰.

Three features of these states matter for the paradox.

First, self-consciousness disappears. This is the shift of attention carried to its end: during absorption, no one is left to check whether he is happy. Mill’s question literally cannot arise.

Second, they are not pleasant while they last. A climber in the middle of an ascent, a musician in the middle of a difficult passage, report no enjoyment; they report an absence of distance. The pleasure is noticed afterward, or shows in the fact that they go back. This is exactly Aristotle’s structure of a pleasure that is added.

Finally, and this is the decisive point, these states escape adaptation, for a mechanical reason.

Why this way resists the treadmill

Chapter 4 established that the satisfaction of a desire fades quickly: habituation devalues what is gained, and escalating expectations always demand more. This mechanism applies to acquisitions: an object, a status, a sum of money. It applies much less to activities, and the difference is not mysterious.

An acquisition is static. Once obtained, it no longer changes, and that is precisely what makes it something one gets used to: the nervous system stops registering what does not vary. Chapter 4 put this in the vocabulary of reward prediction error: what is perfectly predicted no longer produces any signal.

An activity exercised with excellence is the reverse. Keeping it going requires difficulty to keep pace with skill. As soon as one improves, the task becomes too easy and boredom sets in, which forces one to raise the demand. Absorption is thus paid for in renewed effort, and this obligation is what protects it from habituation. The hedonic treadmill keeps one running without moving forward. Activity moves one forward, which is the only known way not to run in vain.

A second mechanism, drawn from chapter 4, points the same way. The dissociation between wanting and liking (the circuits that produce attraction are not those that produce satisfaction) explains why wanting more does not deliver more. Engagement in an activity, by contrast, produces satisfaction without going through the inflation of wanting: there is nothing to obtain, so nothing whose anticipation can swell. This is probably the deepest reason this way works where accumulation fails.

The limits, which call for the following chapters

There are four limits. Each deserves to be taken seriously, because each marks exactly where another way must take over.

It requires conditions. Absorption presupposes time, adequate health, a minimum of material security, and access to an activity that admits of mastery. These conditions are not equally distributed, and recommending engagement to those who lack them is a form of indifference. Chapter 19 returns to this problem, which runs through the whole part.

The restriction is broader than it looks, and it bears on what fills most of the day: work. The conditions of absorption (a clear goal, immediate feedback, a difficulty that keeps pace with skill) are not natural properties of tasks. They depend on how work is organized. A fragmented task, whose result is never seen, whose pace is imposed from outside and which leaves no room for progress, does not meet them, whatever the disposition of the person who performs it. Adam Smith, who praised the division of labor, had himself seen its price. The man whose life is spent repeating a few simple operations, he writes, never has occasion to exercise his understanding, and loses even the habit of doing so⁴¹.

The picture is not uniform, however. Studies of optimal experience found an unexpected result. People report more of it at work than during their leisure, much of which is passive, and yet they say they prefer their leisure⁴². Work can therefore offer the conditions of absorption, and it often offers them better than free time does; it does not offer them to everyone, or everywhere.

Recording this inequality is a matter of description, and it belongs in this book. Saying how to remedy it belongs to another work. But a doctrine of engagement that kept silent about it would recommend something that many lack the means to practice where they spend their days.

It can become an escape. Absorption that makes self-consciousness disappear can be sought for that very disappearance. This is exactly Pascal’s diversion, met in chapter 15: the hunt rather than the catch, keeping busy so as not to see oneself. Endless work, compulsive activity and permanent busyness are proven ways of not meeting one’s own life. The difference between engaging and fleeing does not show in behavior. It shows in what happens when the activity stops. The person who is engaged can stop without feeling threatened. The person who is diverting himself cannot, and the stopping is what reveals it.

It is morally neutral. The conditions of absorption are formal: a goal, feedback, a matched difficulty. A well-run swindle meets them as fully as a work of art does. Optimal experience says nothing about the value of what one is absorbed in, and to claim otherwise would confuse intensity with the good. This is the point of chapter 14: the activity must be in accordance with excellence, which presupposes a judgment that absorption does not supply.

And it fails when action is impossible. Illness that immobilizes, age, constraint, the definitive failure of a project: there are situations in which there is nothing to be done. That is exactly the domain of chapters 18 to 24. A doctrine that offered only engagement would leave without resources those who need resources most.

---

17. Other people

What the text established without drawing on it

Three results from the earlier parts converge on this chapter, and none of them has yet been followed to its consequences.

Chapter 5 noted that social exclusion partly recruits the circuits of physical pain, and that this fact has an evolutionary reason: in the groups where our affective architecture took shape, being rejected often amounted to dying. Chapter 13 grounded moral sentiment in sympathy (the capacity to be affected by what another feels) and documented it even in other primates. Finally, chapter 2 reported that most people refuse to plug into a machine that would produce any experience they desire. Asked what they do not want to lose, they overwhelmingly name real relationships with real people.

A being so constituted cannot be correctly described as an individual managing its internal states in a sealed vessel. Yet that is what an art of living would look like if it spoke only of judgment, desire and death.

Friendship as the first of goods

Epicurus, whose classification of desires chapter 20 sets out, does not place friendship among the pleasant extras one could do without. He makes it a condition of the good life.

He goes further. Of all the things wisdom provides for the happiness of a whole life, he writes, by far the greatest is the possession of friendship⁴³. Coming from a doctrine that elsewhere advises reducing one’s dependencies, the claim is surprising, and that is exactly what makes it interesting.

His justification is twofold. Friendship first provides security. What makes life anxious is uncertainty, and nothing remedies it better than confidence that help will come, even more than the help itself⁴⁴. But Epicurus adds a movement that calculation does not exhaust: friendship begins in need and ceases to be reducible to it. One first becomes attached out of interest, and then the attachment becomes its own end. Every friendship, he writes, is desirable for its own sake, though it began from utility⁴⁵.

This is exactly the structure of the previous chapter, applied to persons. What is first a means becomes an end, and it is by ceasing to be a means that it produces what was expected of it. A friendship kept up for its usefulness does not provide the security hoped for, because whoever calculates knows that the other calculates too.

The three friendships

Aristotle produced the most complete analysis. Two entire books of the Nicomachean Ethics are devoted to it, a proportion that says enough about the rank he gives it. His opening thesis is plain: no one would choose to live without friends, even possessing all other goods⁴⁶.

He distinguishes three kinds, according to what they rest on.

Friendship of utility rests on services rendered. It is real, it is common, and it dissolves as soon as the usefulness ends: business partners, neighbors, working relationships.

Friendship of pleasure rests on the enjoyment found in another’s company. It is livelier and no more lasting: it lasts as long as the taste for it, and tastes change, especially early in life.

Friendship based on virtue rests on what the other is, not on what the other provides. It is the only one that lasts, because its object does not vary with circumstances. It is rare, it takes time, and it presupposes a certain likeness: one admires durably only what one understands.

The formula that sums up this last kind (the friend is “another self”) is not sentimental. It means that the boundary of concern has moved: the other’s good has become one’s own good, without any calculation entering in. The Magna Moralia, attributed to Aristotle, adds a remark that chapter 21 can take up. We see ourselves poorly, and we come to know ourselves in a friend as we see our face in a mirror⁴⁷.

Widening the circles

Stoicism, readily taken for a doctrine of withdrawal, says on this point the opposite of its reputation.

Its central notion is oikeiosis: appropriation, the movement by which a being recognizes something as its own. Hierocles gave it its famous image. Each of us stands at the center of a series of concentric circles. The first contains one’s own body; then come close family, wider kin, neighbors, fellow citizens, and finally all of humanity. The moral task consists in drawing the outer circles toward the center: treating cousins as brothers, fellow citizens as relatives, strangers as fellow citizens. Hierocles makes it an exercise down to the words one uses: calling one’s cousins brothers, and one’s uncles and aunts fathers and mothers⁴⁸.

This device resolves, better than any other in these pages, the difficulty that chapter 13 left open. Spontaneous moral affect is partial: it bears strongly on those close to us and weakly on those far away, and universality cannot be deduced from it. Oikeiosis does not claim that affect is naturally universal. It proposes an exercise to extend it. Universality is recognized for what it is (a construction) and treated as a practice rather than a fact.

Marcus Aurelius draws from it his most repeated formula. Rational beings are made for one another, and withdrawing from the human community makes no more sense than a limb detaching itself from the body⁴⁹.

Why this bond escapes adaptation

The question arises, since chapter 4 showed that almost everything loses value through habit: why would relationships resist?

The answer is that of chapter 16. A relationship is not an acquisition; it is an activity. One does not possess a friend, one tends the friendship, and what is not tended simply ceases to exist. A relationship therefore cannot be got used to like an object, because it is never acquired once and for all. This very precariousness, which makes it demanding, is what protects it from wear.

Two features add to this. A relationship is unpredictable: the other is a system one never fully models, which guarantees a constant flow of unanticipated events, the exact opposite of the conditions for habituation. And it is cumulative: unlike a pleasure, a shared past adds up instead of running out. This ties it directly to the third measure of chapter 14, meaning, which is read over a span of time and not moment by moment.

This is probably why studies that follow people over long periods single out the quality of relationships as the most stable factor associated with lasting well-being: ahead of income, professional success and most of the circumstances assumed to be decisive.

The tension with the rest of this part

What this chapter costs now needs stating, because it stands in open conflict with several of the chapters that follow.

To become attached is to become exposed. Chapter 19 will recommend placing one’s identity on the side of what is up to us, and treating the rest as foreign. Yet the life of a loved one is in no way up to us. A consistent doctrine should therefore count other people among the indifferents. That is exactly what Stoicism does, with consequences that have always made its readers uneasy.

Epictetus gives the harshest illustration: when kissing one’s child, to tell oneself that it is mortal, and that what one loves may be taken back tomorrow⁵⁰. The instruction is consistent with the whole system. Taken literally, it is also a way of never being fully present to anyone.

Two readings are available, and this text will not decide between them. Either the exercise makes it possible to love without demanding: to be fully there without claiming to possess, which is perhaps the only form of attachment that does not destroy its object. Or it installs a permanent reserve that deprives attachment of what makes it worth having. The tradition affirms the first reading. The second is what many readers have found in it, and it is not certain they were wrong.

Grief is the price, and there is no discount. Chapter 22 will offer several ways of regarding loss. None removes the fact that attachment mechanically produces suffering in proportion to its depth. Whoever wants to be invulnerable has a simple method: to care for no one. It works, and chapter 25 presents it in systematic form.

And this way lends itself poorly to method. No one decides to have friends or to be loved. The ancients nonetheless left a few instructions, besides Hierocles’s exercise. Seneca advises judging at length before granting one’s friendship, and then trusting entirely the person one has admitted⁵¹. Aristotle recalls that true friendship takes time: one knows someone, as the proverb has it, only after eating a good deal of salt together⁵². The rest depends on the time one gives to these relationships.

---

18. The lever of judgment

The principle

This is the parent technique of the ways from within, the one from which nearly all the others derive. Epictetus states it in one sentence: “Men are disturbed not by the things which happen, but by the opinions about the things.”⁵³

The move separates two elements that are experienced as one. There is the fact: someone has delivered an insult. And there is the added value judgment: this is unjust, this is intolerable. The disturbance comes not from the first but from the second, and the second, unlike the first, depends on the one who makes it.

The exercise is to describe without qualifying. Epictetus gives the example: of someone who drinks a great deal of wine, say that he drinks a great deal, not that he does wrong⁵⁴. The fact is not denied. What is suspended is assent to the verdict, and the verdict alone wounds. The Stoics call this the discipline of assent: between the impression (this is an evil) and the act of assenting to it there is an interval, and this interval is the only domain that is truly one’s own. Marcus Aurelius gives it its driest form: “Take away thy opinion, and then there is taken away the complaint, I have been harmed. Take away the complaint, I have been harmed, and the harm is taken away.”⁵⁵

Three extensions

Distancing the impression. Some representations do not yield to cool reinterpretation: they force themselves on the mind, charged with anxiety or desire. Epictetus then proposes speaking to them. As soon as a harsh idea arises, say to it: “You are an appearance, and in no manner what you appear to be.”⁵⁶ The move is nominal, and it works. It slips an interval between the subject and the representation, which ceases to be reality and becomes once more an object that can be examined.

Laying bare. Marcus Aurelius breaks objects of prestige down into their matter: wine is only grape juice, a fine dish the corpse of an animal, purple cloth wool dyed in the blood of a shellfish⁵⁷. Deflating the legend of a thing loosens its hold, whether the thing is a coveted object or a dreaded humiliation.

Not suffering twice. This is the most useful of the three, and it is the Buddhist second arrow under another name. Seneca: “He suffers more than is necessary, who suffers before it is necessary.”⁵⁸ If the feared evil comes, one will suffer when it comes. Nothing requires suffering in addition, and for a long time, from its mere representation. Once is enough. Seneca draws from this the complementary exercise, the premeditatio malorum: to possess what one loves “as something one cannot possess forever.”⁵⁹ To anticipate a loss is not to summon it; it strips the event of its power to surprise.

Reframing

Once the verdict is suspended, a second operation becomes possible: choosing, among the defensible readings of a situation, the one that does not destroy. This does not mean lying about the facts. It means redirecting attention.

Plutarch gives the clearest image: the sage is like the bee, which draws an exquisite honey from the bitterest plants⁶⁰. The instruction fits in one sentence: look at what remains, not at what is missing. Aristippus, having just lost an estate, answers the friend who pities him that he still has three, while the friend owns only one: it is rather for him to pity the friend⁶¹.

The same move applies to unavoidable evils. Seneca recalls that illness, bereavement and old age are the tribute life exacts: not accidents, but the price of a good already received⁶².

The clinical legacy

This lineage is not a scholar’s curiosity. The founders of the cognitive therapies explicitly claimed the Stoic inheritance: the idea that interpretations, not events, produce distress is their founding principle⁶³. Twenty centuries separate the Enchiridion from the consulting room, and the move is the same: intercept the automatic judgment.

Later developments in this tradition made a shift that brings it still closer to the ancient sources. Rather than disputing the content of a thought in order to replace it with a more accurate one, patients learn to change its status: to notice it, name it, and stop treating it as a description of reality. This is exactly Epictetus’s move of speaking to his impression⁶⁴. Attention practices drawn from contemplative traditions were folded into the same movement.

To date, this is the only place in the whole account where an ancient intuition has been submitted to systematic empirical evaluation and has held up: within supervised protocols, and with the modest scope specified in the introduction to this part.

The limits

The discipline of judgment slides easily into denial. Suspending the verdict “this is an evil” can serve to avoid looking at a situation that needs to be changed: a real injustice, a real danger. Contemporary psychology distinguishes reframing, which modifies the interpretation upstream, from expressive suppression, which stifles the emotion once it has taken hold. The first works; the second is costly and fails.

Reframing has its own pathology: the injunction to look on the bright side, which invalidates legitimate pain and makes those who cannot manage it feel guilty. Some losses need not be converted into lessons, and some anger is just. The technique helps one not to suffer beyond what is necessary; it does harm when it demands not suffering at all.

Finally, an empirical reservation. These techniques assume that one can detach from one’s affects through attention. Chapter 4 qualifies this: a charged representation engages circuits that do not switch off on command. Distancing is learned slowly and fails under acute stress. It is a fallible skill, not a switch.

19. What depends on us

The division

Epictetus opens his Enchiridion with a cut: “Of things some are in our power, and others are not. In our power are opinion, movement towards a thing, desire, aversion […]; not in our power are the body, property, reputation, offices […].”⁶⁵

The whole art lies in placing one’s energy and identity on the right side of the line. To treat as one’s own what is not is to condemn oneself to disturbance. Epictetus’s image is the traveler at sea: choosing the right ship, the right pilot and the right season depends on oneself; once the storm has risen, it is no longer one’s affair.

The exercise Epictetus draws from this is a single question, to be asked of every painful impression: is this among the things that depend on me? If not, have the answer ready: it is nothing to me⁶⁶. Act fully on what depends on oneself; let go of the rest.

Cicero, setting out the Stoic doctrine, gives the image of the archer: it depends on him to do everything to aim true; hitting the target does not, and that is not where his goal lies⁶⁷. What depends on us is our efforts, never the results. The Stoics called this acting “with reservation”: pursuing a goal fully while holding its outcome to be uncertain.

Willing what happens

That leaves the second column. The Stoic answer goes further than acceptance: “Seek not that the things which happen should happen as you wish; but wish the things which happen to be as they are, and you will have a tranquil flow of life.”⁶⁸

The difference lies between enduring and consenting. Seneca illustrates it with the image of the soldier: the order will be carried out in any case. The one who marches willingly spares himself the bitterest part of constraint, which is doing reluctantly what he will do anyway⁶⁹. Marcus Aurelius goes as far as love, addressed to the world: “I love as thou lovest.”⁷⁰

Understanding rather than blaming

This is where the link with Part II is closest.

If acts follow from configurations that were not chosen, reproach, toward oneself as toward others, aims at a target that does not exist. Spinoza draws the practical consequence: “each person has the power to understand himself and his affects clearly and distinctly […] and consequently to bring it about that he suffers less from them.”⁷¹ Understanding which passions conditioned which act does not make the act disappear; it takes away the object of resentment.

The move has two sides. Toward others: reconstruct the motive before condemning. Marcus Aurelius makes it a rule: “When a man has done thee any wrong, immediately consider with what opinion about good or evil he has done wrong.”⁷² Its foundation is the thesis of chapter 8: no one does evil while seeing it as evil. Epictetus pushes toward compassion: we pity the blind and the lame, so why not the wicked, who are wicked in spite of themselves⁷³? He also gives the image of the two handles: everything can be taken by two sides, “he has done wrong,” by which it cannot be carried, or “he is my brother,” by which it can⁷⁴.

And toward oneself, the same movement. Spinoza denies repentance the status of a virtue: whoever repents of what he has done is, he writes, twice wretched or powerless⁷⁵. Understanding the causes of a past act does not excuse it; it spares one from adding to the wrong a suffering that repairs nothing.

This is probably the most valuable counterpart of determinism. What Part II took away from merit, it also takes away from guilt.

The social objection

A serious objection deserves treatment here rather than dismissal, because it targets the core of the method and not only its excesses.

Telling people that their distress comes from their judgment is not a neutral operation. When the cause is real and changeable (a job that destroys, a violent relationship, an avoidable deprivation), the advice amounts to asking them to adapt to what ought to stop. And it weighs more heavily the less power the person has over the situation: someone who can change their circumstances does not need to be taught to reinterpret them.

The objection becomes sharper still once one notices who formulated this doctrine. Epictetus was a slave. The fact is cited as proof that his method works under any condition. It can be read the other way, as the sign of a doctrine born of powerlessness and cut to fit it. Both readings are available.

Two considerations, however, block the conclusion that the doctrine is quietist.

The first is doctrinal. The Stoics never held that circumstances are without value. They distinguished the preferred indifferents (health, ease, bodily integrity) from those to be avoided. These things are not the good, but they are preferable, and the sage pursues them. The doctrine says that happiness does not depend on them; it does not say that one should renounce them.

The second is biographical. The best-known Stoics were not recluses: Cato resisted to the point of suicide, Seneca governed, Marcus Aurelius reigned and waged war. Roman Stoicism is a doctrine of people who acted a great deal.

Yet the doctrine, detached from this context, invites the quietist reading, and the quietist reading is the one in circulation. The honest formulation is therefore this: the dichotomy of control is a tool for what cannot be changed; applied to what can be changed, it is an anesthetic. The sorting has to bear on that question first, and it is not always easy, because it is often more comfortable to file a thing on the wrong side than to set about changing it.

The other limits

The line is less sharp than the binary division suggests. Many things are partly under control: health, a career, a relationship. The border between what depends on us and what does not is thin, and people judge themselves in any case by how they face what they did not control.

Acceptance borders on resignation. Where does lucid consent end, and where does abdication begin? The Stoics answer with the dichotomy: one acts on what depends on us. Camus objects head-on: against acceptance he sets revolt, the lucid refusal to consent. This is a choice of how to exist, not a theorem, and the conclusion returns to it.

Understanding is not excusing. If no one is wicked willingly, what is to be done about real wrongs? Chapter 8 accepted where this leads (a justice that prevents rather than punishes), but compassion for the motive must not erase protection against the act.

20. The economy of desire

This way answers chapter 15 directly. If desire makes happiness unattainable, desire is what has to be acted on.

Sorting

Epicurus offers the sharpest tool of antiquity: “Of our desires, some are natural, others vain; and of the natural, some are necessary, others natural only.”⁷⁶

Three categories. Natural and necessary desires (not to be hungry, not to be thirsty, not to be cold⁷⁷) are few, easy to fulfill and indispensable: they are to be satisfied. Natural but unnecessary desires (the refined meal rather than bread, comfort rather than sufficiency) are harmless in themselves and dispensable: they are to be moderated. Vain desires, neither natural nor necessary (wealth, glory, status), are born of opinion: they are to be pruned.

The strength of the analysis lies in the word limit. Natural desires are measured by the body: a stomach fills up. Vain desires proceed from opinion, which knows no satiety. This is exactly the mechanism of chapter 4: the escalation of expectations has no ceiling because nothing bodily stops it. Epicurus says so himself: “poverty measured by the end of nature is great wealth, and wealth without limit is great poverty.”⁷⁸

Epicureanism adds a thesis rarely credited to it, so far has common usage of the word distorted the doctrine: the pleasure in question is not an accumulation of enjoyments but the absence of disturbance. Pleasure reaches its peak as soon as pain is removed; beyond that point, pleasures can be varied but no longer increased. The regimen Epicurus recommends is frugal, and the chief good he retains is friendship.

The exercise comes from Epicurus himself: ask of every desire what will happen if it is satisfied, and what will happen if it is not⁷⁹.

Wanting is not liking

Chapter 4 introduced a result that illuminates this sorting better than any moral argument, and it can now be put to work.

The circuits that produce the attraction to a thing (wanting) and those that produce the pleasure of obtaining it (liking) are partly dissociable. One can intensely want what yields no satisfaction. This is not a laboratory curiosity. It is the pattern of addictive behavior, in which desire intensifies as enjoyment fades. In a milder form it is also an everyday experience: the long-coveted object that, once obtained, does almost nothing.

Two practical consequences follow, and they make Epicurus’s classification sharper than it was.

The first is that the intensity of a desire is no indication of the satisfaction it will bring. Often it is the reverse: the most insistent desires are frequently those whose valence has been inflated by anticipation. The criterion for sorting therefore cannot be the strength of the wanting, which is what one feels. It has to be an examination of what satisfaction actually produced on previous occasions.

The second is that vain desires are not merely without limit: they are designed to be so. This leads to the next point.

Manufactured desire

Epicurus attributed vain desires to opinion. He lacked the vocabulary to describe what happens when opinion is deliberately produced.

Chapter 3 set out operant conditioning and the notion of contingency. Skinner had established one further detail experimentally. Of all schedules of reinforcement, the most powerful for sustaining a behavior is not systematic reward but unpredictable reward, delivered on a variable ratio, so that the organism can never predict when it will come. A behavior maintained this way is remarkably resistant to extinction: it persists long after the rewards have stopped.

This is the principle of the slot machine. It is also, exactly, the structure of a scrolling news feed: occasionally gratifying content, delivered at unpredictable intervals, through a device that costs nothing to access.

The conceptual apparatus of the earlier parts suffices to describe the result. Variable-ratio reinforcement maximizes wanting. The dissociation described in chapter 4 means that this wanting need not be accompanied by pleasure to sustain itself. Hedonic adaptation devalues each gratification obtained, which restarts the search. And the attention thus captured is precisely the resource that, as chapter 16 showed, conditions absorption in an activity, which is the one thing that escapes the treadmill.

The result deserves to be stated without emphasis. It is possible to manufacture, on an industrial scale, desires that bring no satisfaction, and to do so all the more effectively because they bring none.

This observation is not a moral one and calls for no normative conclusion here. It simply supplements Epicurus’s classification with an observation it could not contain: among vain desires, some are now the object of engineering. The exercise of sorting thereby becomes harder, since it applies to devices designed to resist it, and more necessary, for the same reason.

Stepping off the treadmill

The way off the treadmill is not to run faster. Epictetus is categorical: “Freedom is not acquired by satisfying one’s desires, but by suppressing them.”⁸⁰ Epicurus puts it his own way: nothing is enough for one to whom what is enough seems little⁸¹.

Revaluing what one has

Against the adaptation that devalues what is possessed, three moves.

Active gratitude. Marcus Aurelius: “Think not so much of what thou hast not as of what thou hast: but of the things which thou hast select the best, and then reflect how eagerly they would have been sought, if thou hadst them not.”⁸² The exercise is to recall the value a good had before it was obtained: the effort it cost, the desire one had for it. A harsher variant is to imagine losing it, in order to rediscover it.

Downward comparison. Human beings spontaneously compare upward, and so enjoy poorly what they have. Democritus and Plutarch reverse the direction: thinking of those whose lot is worse recalibrates a perception distorted by habit⁸³.

Enjoying the past, which Epicureanism understood better than Stoicism. The future is uncertain; the past is secured. Past goods are therefore the only guaranteed goods, immune to uncertainty and available at will through memory. Epicurus, dying in extreme pain, writes that he sets against his present suffering the joy of remembering his past conversations⁸⁴. Hence the rule he draws from it: heal loss by grateful memory of what has been, and by the awareness that it is impossible to make what has happened not to have happened⁸⁵.

The limit

It runs through the whole chapter, and it is the most serious in this entire part.

Reducing desire can impoverish a life as much as free it. Ambition, effort and creation are also born of lack. Part I established that desire is the engine of action. Stepping off the treadmill brings calm; it can also extinguish. A doctrine that recommended only the reduction of desire would produce people who are tranquil and inert, and nothing guarantees that this would be progress.

Here chapter 16 supplies the decisive corrective, and it is better stated explicitly. The solution is not to arbitrate between desiring much and desiring little. It is to distinguish desires aimed at acquisitions (which habituate, escalate and can be manufactured) from desires aimed at activities and people, which do not habituate because they must be continually exercised. Epicurean pruning applies to the first. Applied to the second, it produces exactly the tranquil, empty life just described.

The border between “natural but unnecessary” and “vain” is, moreover, more cultural than Epicurus admits: which box does a book, or a journey, belong in? The taxonomy is an instrument of lucidity, not a program of deprivation.

And gratitude has its own drift: complacency (others have it worse, so all is well), or the numbing of a real lack that could be corrected.

21. The loosening of the ego

This way answers what Part II did to the subject. If there is no pilot, what becomes of self-love?

Two ancient theses

Buddhism offers two descriptions of the self which, taken together, dissolve its substance.

The first is the doctrine of the aggregates (khandha). What is called a person breaks down into five sets of phenomena: bodily form, sensations, perceptions, mental formations, consciousness. None of them is a self. One of the oldest sermons proposes an exercise: examine them one by one and say of each, this is not mine, I am not this, this is not my self⁸⁶. The Milindapañha gives the image that made it famous: a chariot is not its wheels, nor its axle, nor its body; the word names their assembly and nothing more, and the same holds for the name of a person⁸⁷.

The second is impermanence: anicca. Nothing in the self is constant: the body renews itself, and mental states follow one another without pause. Heraclitus had already said that one cannot step twice into the same river⁸⁸.

From these two theses follows anattā, non-self: what is called “I” is not a thing but a process. And if the ego is neither stable nor independent, the question becomes: why be wounded by an attack aimed at it?

The contemporary convergence

Neuroscience has gradually dismantled the idea of a unified inner center. The brain appears as a distributed system: regions that cooperate and sometimes contradict one another, and whose impression of unity is a late synthesis. In split-brain patients, each hemisphere handles its tasks autonomously. Hemispherectomy in children shows the complementary reverse: one hemisphere is removed and the other redistributes the functions. A center would be destroyed; a distributed system reconfigures itself. Some lesions erase autobiographical memory while leaving mood intact; others alter personality without touching reasoning⁸⁹.

What is called “self” therefore resembles a narrative more than a substance: a narrative thread drawn, second by second, across a fabric of parallel activities.

This convergence between contemplative introspection and brain imaging is striking, and that is precisely why it calls for suspicion. Two frameworks with opposite aims (to explain on one side, to free oneself on the other) can converge by verbal analogy without saying the same thing. And “no center” does not mean “no coherence”: the system produces a functional continuity that is entirely real. The unity of the self is a construction: useful, not false.

A point from analytic philosophy reaches the same place by another route. Suppose personal identity consists only in continuity of memory, character and intention, with no core persisting underneath. Then the question “Will I still be me in thirty years?” has no clear-cut answer, and the importance one attaches to one’s own survival must be revised downward. This is not a bleak thought: it lessens the fear of death at the same time as it lessens attachment to oneself⁹⁰.

The other reading of the narrative

The same finding licenses an opposite conclusion, and the choice between the two is not a trivial one.

To say that the self is a narrative can mean: it is only a narrative, so there is no one, so there is nothing to defend. That is the dissolving reading, the one just set out.

But a narrative is also something one composes, well or badly. A life can have a shape, a direction, a coherence between what was undertaken and what is being pursued; or it can be a mere series of unrelated episodes. The difference is perceptible, it matters to those who live it, and it presupposes no hidden substance. A river has no permanent core either, and yet it has a course.

This second reading connects with three points established earlier. It connects with the third measure of chapter 14: meaning, which is read not in instants but along a trajectory. It connects with Aristotle’s formula that eudaimonia is the activity of a whole life, and with his remark that a life cannot be judged before its end. And it connects with chapter 16: an activity pursued over years, with a deepening competence, is precisely what gives a narrative its continuity.

The two readings do not contradict each other: the absence of a core is compatible with the presence of a form. But they do not lead to the same place, and this fork reproduces exactly the division that runs through all of Part IV. The dissolving reading leads to detachment: less self, so less for things to take hold of. The compositional reading leads to engagement: since the self is a work, there is something to do.

No decisive reason settles the choice. Perhaps the first suits those who suffer from watching themselves too much, and the second those who suffer from drifting.

The practical consequence

Indifference to the gaze of others. For the Stoics, reputation is an “indifferent”: it does not depend on us and adds nothing to what we are. Epictetus: “If you want to make progress, be content to be thought a fool in external things; and if some people think you are somebody, distrust yourself.”⁹¹ The gesture comes down to two refusals: not to justify oneself, and not to depend on approval. Told that someone has spoken ill of you, reply only: “He did not know the rest of my faults, or he would not have mentioned only these.”⁹²

Inner honor. What replaces outside esteem is not a void but a relocated tribunal. For the Stoics, the only good is virtue, that is, the quality of the intention and of the act; the outcome belongs among the indifferents⁹³. Democritus puts it in one sentence: one should learn to feel shame before oneself rather than before others⁹⁴.

The limits

Indifference to the gaze of others can cover a hardness. No longer seeking approval does not license arrogance. Nor is reputation always a pure indifferent: it sometimes carries real obligations (the trust one owes, the example one sets). The Stoic aims at freedom from vanity, not from all regard for others. Chapter 17, moreover, marks the limit of this precept: a well-conducted indifference to public judgment has nothing to do with indifference to the gaze of those who matter, which is not wisdom but a loss.

And a logical difficulty deserves to be flagged rather than sidestepped: if the self has no center, who practices the loosening? All these techniques presuppose an agent able to work on itself, which is exactly what the thesis has just dissolved. The traditions concerned reply that the exercise is itself one process among others, and that no separate agent is needed for a system to modify itself. The reply is coherent; it is not certain that it is satisfying.

22. Time, death, the world

Keeping the end in view

The memento mori is not a taste for the morbid: it reveals value. Epictetus: “Let death and exile and everything that seems terrible be before your eyes every day, but most of all death; and you will never think anything mean, nor desire anything to excess.”⁹⁵

The exercise has two effects. It recalibrates desires: in the face of finitude, boundless ambition and the wounds to self-love lose their weight. And it makes the present precious. Marcus Aurelius urges living each day as if it were the last, “without agitation or torpor,” and leaving life as a ripe olive falls, praising the earth that nourished it⁹⁶. The image is gentle, not gloomy: death appears in it as a completion, not a catastrophe.

The Epicureans add an argument that is genuinely soothing: death is not an evil to be experienced, for when we are, death is not; when death is, we are no more⁹⁷. The fear bears on a state that no one will live through. Lucretius completes it with the symmetry argument: the eternity that preceded birth troubles no one, and nothing distinguishes it from the eternity that will follow death.

Are these arguments convincing? They treat death as a state and show that there is nothing in it to fear. One can object that what is dreaded is not a state but a deprivation: everything that will not take place. The Epicurean argument answers the fear of suffering after death; it answers less well the regret at no longer being.

Montaigne tested this discipline on himself, and his course provides its most precise critique. As a young man, he adopts it under Cicero’s formula (that to philosophize is to learn to die) and advises growing familiar with death by thinking of it constantly: “let us have nothing so often in our minds as death.”⁹⁸ Growing old, he goes back on this advice. He observes that the peasants around him die without ever having given it a thought, and often with more composure than the philosophers. He observes that the preparation troubles us more than the thing prepared for. And he concludes: “We trouble life with the care of death, and death with the care of life.” Then, more bluntly still: “If you do not know how to die, never mind; nature will teach you on the spot, fully and sufficiently.”⁹⁹

This reversal does not refute the memento mori. It marks its internal limit: the exercise meant to free one from fear can become a way of keeping it alive. Montaigne does not propose forgetting death, only not making it the object of training. By another route, this joins the limit stated below for anxious temperaments.

What is passed on

There is a fourth answer to death, the most widespread in human history, and it appears in none of the preceding ones: one survives in part in what one passes on.

Passing on takes ordinary forms: raising a child, training someone in a trade, leaving behind a work, a house, a skill, a way of doing things that others will continue without knowing where it came from. None of these forms is survival in the proper sense; all of them extend something of oneself beyond oneself.

The ancients knew this. In the Symposium, Diotima makes this desire a form of love itself: mortals seek immortality as best they can, some through children, others through the works and laws they leave behind¹⁰⁰. Aristotle asks in earnest whether the fortunes of descendants still affect the happiness of someone who has died, and answers that they affect it a little, without being able to overturn it¹⁰¹. Epicurus, who held that death is nothing to us, nonetheless arranged by will for the future of his school and for the celebration of his friends’ memory¹⁰². Even the doctrine most determined to disarm death did not give up on what would survive it.

This answer has a peculiarity: it does not soothe by removing weight, but by moving value outside oneself. It belongs to the family of the ways from without, applied to the one thing that seemed to belong only to the within. And it gives a name to what chapter 14 left vague: if meaning is tied to what one gives rather than to what one receives, passing on is the most common form of that giving.

It has its difficulties. The most immediate comes from this very chapter. The exercise of the view from above, set out below, recalls that those who remember will in turn be forgotten, and that every trace fades. If passing on is worth something only through its duration, it loses against time. It withstands this objection only if one stops measuring it by its duration and measures it instead by what it does now: by what it changes in the lives of those who receive. The second difficulty is more intimate. One can pass things on in order to prolong oneself, and what is passed on then becomes one more possession, with the anxiety of loss that chapter 15 attached to every possession. What is given in order to outlive oneself is not entirely given.

Flux

A large share of suffering comes from treating as stable what is not. Marcus Aurelius compares substance to a river in perpetual flow, whose activities and causes never stop changing¹⁰³.

From this he draws a valuable reversal: change is not the enemy; it is the condition of life. Can one be nourished, Marcus asks, unless food is transformed, and can anything useful be accomplished without transformation¹⁰⁴? The death of one thing is the birth of another. To dread change is therefore to dread what makes all existence possible. Death is not an anomaly in a world that could have been permanent: it is the exact price of a world in which something can happen.

The present

The practical corollary: time cannot be put off. Seneca makes this the heart of On the Shortness of Life: we do not lack time, we waste it. “Everything else is borrowed; time alone is ours. While we put it off, life goes by.”¹⁰⁵

The diagnosis is severe. We defer endlessly, as if certain of reaching the end, and we reserve for living the only time that is good for nothing. The antidote is not the hedonism of the instant but the fullness of the present: doing the work of the moment, expecting nothing and fleeing nothing. Counting each day as a whole life.

Changing scale

When a worry fills the whole field, there remains the option of rising above it. This is the most characteristic exercise of the Meditations: “Asia and Europe are corners of the world; the whole sea is a drop in the universe; the whole present is but a point in time.”¹⁰⁶

The exercise works along two axes. Along time: the eternity before birth and the one after death, entire reigns of which only names remain, the oblivion that erases even fame¹⁰⁷. Along space: the whole earth, Seneca writes, is only a point, and it is this point that men divide among themselves by fire and sword¹⁰⁸.

The reverse movement is also available, and it is more sober: “People seek retreats for themselves: the countryside, the seashore, the mountains. But you can, whenever you wish, retire into yourself. Nowhere does a person find a calmer retreat than in his own soul.”¹⁰⁹ This retreat is no daydream. Marcus specifies that the point is to find there a few brief and essential principles that suffice to dispel disturbance, and to return to them often. The Buddhist tradition codified a neighboring gesture, attention to the breath, which it counts among the foundations of meditation¹¹⁰.

Looking at what is there

There is a third movement, neither widening nor withdrawal, and it deserves to be set apart because it does not seek to relativize.

It consists in paying attention: not to the universe as a crushing totality, but to what lies in front of one. Marcus Aurelius makes this a recurring exercise: noticing that bread splits as it bakes, that the ripe ear of grain bends, that the foam at a boar’s mouth has a beauty of its own¹¹¹. These things are beautiful neither by intention nor by design; they become beautiful for whoever looks long enough.

This practice corrects a tendency of the whole chapter. The memento mori, the view from above and the inner retreat work by subtraction: they deflate, relativize, remove weight. This one adds. It does not ask that the world be judged less important; it asks that it be looked at better. Along with chapters 16 and 17, it is one of the few places in this part where something increases instead of diminishing.

Contemplation

This attention received its most complete theory from an author already met twice, and the argument so far has lacked that theory.

Schopenhauer supplied chapter 15 with the pendulum of desire, and he will supply chapter 25 with the denial of the will. Between the two, his system contains a third term. There exists, on his view, a state in which the will falls silent without having been destroyed: aesthetic contemplation. Before a work of art, a landscape, any form looked at for its own sake, the individual ceases for a moment to be a desiring being. He no longer asks what the thing can bring him or take from him. He becomes, Schopenhauer says, a “pure subject of knowledge,” freed from willing. And for that time, the suffering that willing produces is suspended along with it. This is, he writes, the rest from the forced labor of willing, the moment when “the wheel of Ixion stands still.”¹¹²

Kant had given this pleasure a definition that marks its originality: it is disinterested. One can find beautiful what one has no desire to possess, and the satisfaction taken in it does not depend on the existence of the object¹¹³. It is the only pleasure that the structure described in chapter 15 does not reach, because it does not pass through lack.

The value of this way, for the argument, lies in its exact position. Chapter 15 says that suffering comes from willing. Chapter 25 proposes to abolish willing, and must acknowledge that the price is exorbitant. Contemplation obtains intermittently, without destroying anything, what renunciation obtains permanently by destroying everything. It is brief, it is free, and it requires no doctrine.

It is also distinct from the absorption of chapter 16, with which it might be confused. In both cases, self-consciousness disappears. But in absorbing activity the will is fully mobilized, strained toward a goal; in contemplation it is at rest. These are two opposite ways of ceasing to watch oneself: one through effort, the other through the suspension of effort.

Its limits are the expected ones. It is brief: Schopenhauer himself acknowledges that the will reasserts its claims as soon as attention slackens; this state cannot be inhabited. It presupposes leisure and training (time, availability, often an education of the eye), which are unequally distributed, like the conditions of chapter 16. And it can become a way of living at a distance: under the name of the aesthetic stage, Kierkegaard described an existence that seeks only intense moments and evades every commitment. He saw in it a form of despair. Contemplation brings relief from life; it cannot take its place.

The limits

The memento mori does not free everyone. For some temperaments, meditating on death feeds anguish rather than gratitude. The exercise presupposes a ground of serenity already in place; applied to acute anxiety, it can make things worse. That is the lesson of the late Montaigne.

Meditating on impermanence can devalue the present. If everything passes, why invest oneself, love, build? The detachment that frees and the engagement that gives meaning pull in opposite directions, and chapter 17 showed that this tension cannot be resolved by argument.

Decentering can tip into nihilism. If nothing has value on a cosmic scale, why act? The Stoic replies that the view from above lightens troubles, not duties: Marcus Aurelius always adds that one must act according to one’s nature. It is a sedative, not a metaphysics.

And “living in the present” runs up against prudence. Putting everything off is one flight; preparing for nothing is another. The ancients aimed at the fullness of each day, not at improvidence: a nuance easily lost.

23. Discipline

All the preceding ways assume that one can practice them. The Stoics turned this practice into an explicit method.

Training

They saw themselves as athletes: virtue, like strength, is built by training. This is askēsis. Seneca: “If the body can be brought by exercise to endure blows, how much more easily could the soul be hardened to take the blows of Fortune without breaking?”¹¹⁴

Two practices follow. The first: turn every annoyance into an exercise. Epictetus advises starting small. Your oil is spilled, your wine is stolen: “this is the price at which calm is sold.”¹¹⁵ Every daily irritation becomes a rehearsal in patience.

The second, more demanding: voluntary discomfort. One deliberately exposes oneself to want (fasting, cold, simplicity) to verify that one can bear it, and so defuses the fear of someday going without. Seneca recommends setting aside three or four days from time to time to live like a poor man, asking oneself: is this what I was afraid of?¹¹⁶

Epictetus adds a rule of prudence that holds beyond its original subject: before undertaking anything, examine “what comes before and what follows,” and know one’s own measure; otherwise one starts out on fire and then gives up in shame¹¹⁷. Training is not recklessness.

The model

A simple technique: keep a figure of excellence before one’s eyes and measure one’s acts against it. Seneca: “Choose a Cato; and if Cato seems too austere, choose a man of gentler nature. Always keep before your eyes a witness to your thoughts, someone whose authority makes even your solitude sacred.”¹¹⁸

The exercise is twofold: an observed model (how would he have acted here?) and an internalized model, an imaginary witness before whom one has nothing to be ashamed of. The point is not slavish imitation but a concrete compass where the abstraction “be virtuous” says nothing.

The limits

The line between discipline and the cult of suffering is thin. Seeking discomfort can slide into valuing suffering for its own sake, or into a harshness toward oneself with nothing liberating left in it. The goal is freedom in the face of want, not the merit of suffering. If the exercise feeds anguish rather than serenity, it has missed its target.

The model can become an idol: guilt at never equaling it, or thought frozen in deference to an authority. The Stoics chose several figures and treated them as reference points, not masters. The good model is one that can be criticized.

24. Suspending judgment

The epochē

Pyrrhonian skepticism offers an unexpected path to tranquility: not finding the truth, but ceasing to decide. Sextus Empiricus turns this into a method. One sets things in opposition to one another, “by means of which we arrive first at epochē, suspension of judgment, and then at ataraxia, freedom from disturbance.”¹¹⁹

The mechanism is precise. For any thesis, one looks for an opposing thesis of equal force: this is isostheneia, the equipollence of reasons. Since the two cancel each other out, the mind cannot prefer either and withholds its assent. From this suspension, almost by surprise, calm arises.

Sextus tells the anecdote that sets the tone of his doctrine. The painter Apelles, despairing of rendering a horse’s foam, threw at the painting the sponge he used to clean his brushes, and the sponge produced the foam. The first skeptics hoped to reach tranquility by finally deciding. When they failed, they stopped, and ataraxia followed epochē “as the shadow follows the body.”

The Pyrrhonist does not assert that nothing is true: that would be one more dogma. He abstains. His diagnosis is exactly that of chapter 15: disturbance comes from holding firmly that something is good or bad by nature. Whoever leaves this undecided “neither flees nor pursues anything too vehemently,” and is therefore free from disturbance.

The doctrine does not claim to abolish pain. Sextus is explicit: the skeptic is cold and thirsty like everyone else. He merely suffers once instead of twice, because he does not add to the discomfort the opinion that it is bad by nature.

The practicable version

Complete skepticism is hard to live. A more modest variant exists, which bears not on truth but on consequences: one can never judge in advance whether an event is good or bad, because one does not know what will follow from it.

This is the tale of the old man of the frontier. His horse runs off to the nomads; people come to commiserate; he asks: who knows whether this is not good fortune? A few months later, the horse returns with a magnificent stallion; people congratulate him; who knows whether this is not a misfortune? His son, who loves to ride, falls and breaks his hip; people commiserate; who knows whether this is not good fortune? A year later, the nomads invade the country, the able-bodied men go off to fight, and nine in ten of them die. The son, crippled, survives¹²⁰.

The exercise that follows from the tale is one of the most practicable in this whole account. It does not ask one to become convinced that an evil is a good: that is difficult and often false. It asks only that one admit ignorance of long-term consequences.

Where form and content coincide

Something more is at stake here.

Chapter 11 established that any argument can be opposed by an argument of equal force, and that reason alone does not settle moral disagreements. This was presented as a problem: an open door to nihilism. Sextus maintains that the door opens onto something else. If arguments balance, one can stop deciding, and to stop deciding brings peace.

The stance this part adopts, setting out without crowning, is thus one of the ways it presents. The form of the account and its content coincide. The coincidence is convenient, and it is not certain that rigor alone led to it.

The limits

One cannot suspend all judgment and go on acting. One has to choose a dish, a profession, a course of conduct. Sextus knew this. He replied that one follows appearances and customs without adhering to them dogmatically. The reply is elegant, but it is not certain that it describes a possible life. Hume, two millennia later, raised the same objection: nature does not let the skeptic doubt for long, and brings him back to his beliefs as soon as he leaves his study.

The epochē can serve as an alibi. “One cannot know” is the most comfortable refuge there is, and it excuses nearly any inaction. There are situations in which one knows very well, and in which doubt is cowardice.

And it is incompatible with chapters 16 and 17. Committing to an activity pursued for its own sake, or becoming attached to someone, presupposes precisely that one has decided it was worth it. The consistent skeptic can neither love nor devote himself, since both require a judgment of value sustained over time. This is the starkest contradiction in this whole part, and it has no solution: one can suspend judgment, or commit, but not both about the same object.

25. The radical pole: renunciation

There remains the most extreme way, the one that does not move the target but abolishes it. It is presented here as a pole, in its own strength.

The argument

It follows rigorously from everything before it, and that is what makes it serious.

If desire is the source of suffering, if its satisfaction yields only respite, if hope is only fear turned inside out, then the reforms proposed in the preceding chapters are half-measures. They reduce desire, direct it, sort it, or shift it toward less disappointing objects. None attacks its root. The complete solution would be to stop wanting.

Buddhism gives this way its oldest and most measured form. Its diagnosis has three steps. Existence is marked by unsatisfactoriness; this unsatisfactoriness arises from thirst, taṇhā; the extinction of this thirst is possible, and that is what the word nirvāṇa designates (it literally means extinction). Renunciation here is not a mortification but a clarification: understanding that nothing lasts, and therefore that nothing can satisfy for long.

Schopenhauer gives the Western version, which is darker. He presents it not as an exercise but as an event. In one who pierces the illusion that separates individuals, all suffering becomes his own, and this knowledge ends by consuming the will, “so that not the least willing, not the least thirst for individual existence, remains in him.”¹²¹

He adds that this passage has a price. “Almost always, therefore, great suffering must first have broken the will before the denial of willing can come about.”¹²² One does not arrive there by decision: one is led there.

This way carries the diagnosis of Parts II and III to its end, and its coherence with that diagnosis is real.

Renouncing hope

The most unsettling point is that giving up hope is presented here not as a misfortune but as a relief. The thought is old and widespread. Seneca takes it from the Stoic Hecato: “Cease to hope, and you will cease to fear,” because hope and fear walk chained together, like the prisoner and the soldier who guards him¹²³. India made it a famous saying, placed in the mouth of Piṅgalā, a courtesan who waited in vain for her visitors: hope is the greatest suffering, and giving up hope the greatest happiness¹²⁴.

The reasoning joins chapter 15: the suffering of hope comes from uncertainty. One who no longer expects anything can no longer be disappointed.

Spinoza gives this idea its least somber form: the mind has more power over the affects the more it understands things as necessary¹²⁵. This is no longer despair, but the calm of one who stops demanding that the world be otherwise.

Kierkegaard takes the opposite course. He makes despair the “sickness unto death,” the self’s diseased relation to itself, and proposes a way out through a leap beyond reason: faith. The framework adopted here cannot follow this answer. Yet it occupies the same place: where argument stops, something else is asked for.

What this way costs

It has real strength: it is the only way that treats the problem at the root, and it is consistent with everything that precedes it. To reject it as morbid would be a refusal, not an argument.

Three observations can nonetheless be made.

The finding does not entail the conclusion. That desire is a structure of lack is a description. That the solution is to renounce it entirely is an existential option, not a deduction. The same finding licenses commitment, attachment, Epicurean ascesis, Stoic reframing, epochē, or Nietzschean affirmation. Schopenhauer chooses, and it is hard not to see in his choice a temperament raised into a metaphysics.

The price is everything that makes a life inhabited. Part I established it: without lack, nothing is set in motion. A complete extinction of willing frees one not only from suffering but also from attachment, effort, love and creation. One cannot obtain peace by this means and keep the rest. It is an exchange, and it must be named as one.

And the argument proves too much. Pushed to its end, it recommends renouncing not only vain desires but every object whatever, including the two that chapters 16 and 17 showed to escape the mechanism in question. Yet the argument rested on that mechanism: habituation, escalation, disappointment. If some things are not subject to it, the general conclusion no longer follows. Renunciation remains a coherent answer to the problem of acquisition; it has not been shown to be the answer to the problem of life.

The Buddhist tradition, moreover, resists the nihilist reading better than its Western transposition does. The extinction of thirst there is not the extinction of all activity. It comes with an insistence on compassion, and its ideal is not withdrawal but unencumbered action. The difference between wanting nothing and wanting without clinging is slight to state and considerable to live. It is probably there that the way of renunciation becomes practicable.

---

Where philosophy stops

One last section, on what nothing in the preceding chapters can do.

The limit of sensation

Chapter 1 distinguished two sources of the affects: sensation and judgment. All the techniques set out here work on the second. None acts on the first.

This limit is wider than it looks. It does not concern only acute physical pain. It concerns every state whose cause is somatic or biochemical rather than interpretive. In these cases, to be clear, the failure of the exercises is not a failure of application: the tool does not fit the task.

Four situations

Major depression is not an intense sadness that reframing could turn around. One of its features is precisely the inability to feel pleasure, which defeats in advance any method that relies on turning attention toward what remains. Worse, the structure of the exercises turns against the sufferer. To ask someone to change their judgment, when judgment is the very organ affected, adds failure to suffering. It also supplies one more confirmation of what the illness already keeps telling them.

Recent bereavement is not a problem to be solved. The consolations of chapters 18 to 22 (life’s tribute, grateful remembrance, impermanence) are true, and they are useless in the first weeks. They presuppose a distance that grief does not allow. Offered too soon, they work as a request to stop suffering, which is the last thing anyone needs.

Trauma responds poorly to the will. A startle reaction, a flashback, hypervigilance do not stem from a judgment one could suspend. They are responses installed at a level that attention does not reach directly. Chapter 18 noted that distancing fails under acute stress. It fails a fortiori where stress has left a lasting trace.

Chronic pain and serious illness belong to yet another regime. How they are lived can be changed considerably; this is even one of the few fields in which the Stoic lineage has produced measurable results. But they cannot be abolished, and a doctrine that promised to abolish them would be lying.

What follows from this limit

Three consequences follow, and the last matters most.

First, the lineage these exercises claim with the cognitive therapies, which earns them part of their credit, is precisely what forces the distinction. Those therapies are conducted by trained practitioners, over a defined period, with a prior diagnosis and follow-up. An exercise is not a treatment, and the analogy that links the two is also what forbids confusing them.

Second, these practices were developed for ordinary life: annoyance, loss, ambition, the fear of death, boredom. That domain is vast, and it is theirs. Taking them out of it adds nothing to their authority and takes some away.

Finally, recommending these exercises is not a neutral act. A doctrine that holds that disturbance comes from judgment carries, by construction, the suggestion that whoever is not well is judging badly. It is therefore obliged to say where this explanation ceases to hold. Otherwise it produces exactly what it claims to cure: an additional suffering, that of not managing.

This is the only prescription this account allows itself: knowing when these pages are not the right answer, and where the right one lies, is part of what they teach.

---

Where we stand

Nine ways, and one pole that negates them all. They do not agree.

Engagement assumes that one can act; the dichotomy of control applies when one cannot. Attachment to other people increases exposure; the loosening of the ego and renunciation reduce it. Work on judgment assumes a subject acting on itself; the doctrine of non-self holds that there is no subject. Suspending judgment brings peace by ceasing to decide; commitment and love require having decided. The economy of desire reforms wanting; renunciation abolishes it. The Stoic consents to an order; the Buddhist undoes the idea that there is one.

These disagreements are real, and they bear on what matters. No synthesis will be drawn from them.

Only one thing is common to all of them, and it comes straight from chapter 15: none proposes to aim at happiness.

That is the answer the paradox imposed. Since the target recedes when aimed at, every tradition moves it. Some move it outward, by engaging attention in an activity or in other people. Others move it inward, by reforming the way what happens is received. Contemplation suspends aiming for a moment; renunciation goes further and abolishes it.

What this inquiry permits one to affirm therefore comes down to this, which is little and is not nothing: happiness is never an object, always an effect. It arises on the way to something else, and the only practical question is what that something is.

---

---

Conclusion

What holds

On human nature. Three levels of explanation fit together without difficulty: learning by consequences, the selection that fixed its architecture, and the error signal that implements it. In human beings, the symbol extends them as far as the capacity to die for an idea. As description, this framework is solid; as proof, it is weaker than it looks. The thesis that every action aims at an affective state cannot be refuted by any observation, and that should raise suspicion. It is a fruitful grammar, not a law. Butler’s objection still holds: the pleasure taken in a thing presupposes a desire for that thing, and so cannot explain it.

On freedom. What is called deciding is the point at which a configuration of affects, information and circumstances produces its effect. The central argument holds: a decision is either caused or it is not, and neither branch yields a sovereign agent. The regress has no end: no one is self-made. But the compatibilist objection was not overturned, only sidestepped. And the analysis of the reactive attitudes shows that blame and gratitude probably never depended on what people believed about causation.

On morality. Once transcendent foundations are set aside, the only candidate for the title of moral fact is sensibility. It is a fact that already comes charged with a valence, because a felt state is constitutively good or bad for the one who undergoes it. This is the most fragile point in the whole structure. The passage from fact to value is not demonstrated; it is assumed. Kant’s objection (pure reason can ground a moral law without the help of any affect) has not been refuted. It has been brought down to the rank of its rivals, which all posit an axiom. It has also left behind a demand for impartiality that naturalism does not produce on its own. Two difficulties remain open. The natural roots of moral feeling are partial, not universal, so universality is a rational construction grafted onto an affect that did not grow toward it. And nothing licenses the step from “sensibility is the only moral fact” to “only my sensibility counts.”

On happiness. We methodically pursue a state that the pursuit pushes away. Habituation extinguishes what is obtained, escalation renews what is demanded, and hope contains fear. This paradox is not an objection from outside: the system produces it itself. Pascal, Butler, Sidgwick and Mill had formulated it before anyone measured it. Mill’s conclusion remains the most useful: happiness is found along the way, by those whose minds are fixed on something else.

Chapter 14 isolated a further distinction, and it governs the way out. The word covers at least three things: well-being as experienced moment by moment, the satisfaction felt when evaluating one’s life, and meaning, which is measured not in moments but over a trajectory. These three measures regularly diverge. A life can be pleasant and empty, or painful and full.

On the ways. Nine ways respond to the paradox without agreeing, and one pole denies them all. The ways from without (activity pursued for its own sake, attachment to other people) substitute one object for another. They escape the treadmill because an activity or a relationship is not acquired; it is exercised. The ways from within (judgment, the dichotomy of control, the economy of desire, the loosening of the ego, the relation to time and death, discipline, the suspension of judgment) work on how what happens is received. Nearly all of them proceed by subtraction. The former assume that one can act; the latter hold when one cannot.

Between the reform of desire and its abolition, an intermediate position was noted: contemplation, which suspends willing intermittently without destroying it. And a single feature is common to all the ways: none of them aims at happiness. The paradox left no other way out.

An answer that refuses the others

One answer is missing. It deserves to be named last, because it refuses everything that precedes it.

Camus starts from the same diagnosis. The world offers no given meaning; we demand one; from this confrontation arises what he calls the absurd. He identifies two escapes from it and rejects both. The first is suicide, which removes the human term from the equation. The second, which he calls philosophical suicide, consists in granting oneself a meaning by a leap out of reason, as Kierkegaard does with faith. In both cases the tension is dissolved instead of held.

His answer is revolt: holding lucidity and refusal together without reconciling them. Not consenting. Not hoping either, since hope would be another form of flight. Sisyphus watches his rock roll back down. He knows his effort is futile, and he goes back down to fetch it. Camus holds that one must imagine him happy.

Of all the ways surveyed here, it is the only one that refuses agreement with reality. The Stoics consent, the Taoists attune, the Buddhists dissolve, Schopenhauer renounces, the skeptics suspend. Revolt stands up, and does so without hope, which is its hard point. It expects no victory. It is a way of standing, not a strategy.

The objection is serious, and it targets this whole account as much as Camus. Revolt presupposes someone to revolt, and an act that is truly that person’s own. Part II dismantled almost exactly that. A refusal is as caused as a consent; there is no more freedom in the clenched fist than in the open hand. Nietzsche escaped the problem by holding that emancipation consists in becoming the author of the meaning of one’s necessity. But the interpretation is caused too, and the difficulty remains.

This way still deserves to be kept, for a reason that is not logical. It names something the others leave aside: the possibility of holding a position without having been soothed into it. All the others promise calm. This one does not. For anyone who judges that calm sometimes costs too much, it is the only one left.

What remains open

It is better to list the tensions than to dissolve them. Dissolving them would be exactly the move this inquiry has spent four parts unlearning.

Desire: engine or servitude? This is the central tension, and it is unresolved. Desire produces suffering; it also produces effort, creation and attachment. Every technique that reduces it brings calm, and can also extinguish. The distinction between desires for acquisition, which habituate, and desires for activity, which do not, shows where to prune. It does not say how far. And the line probably does not fall in the same place for everyone.

Engagement or detachment. This is the sharpest tension of Part IV, and it is structural. To attach oneself to a work or to someone is to increase one’s exposure deliberately. Detachment reduces that exposure, and in the same gesture cuts off what gave things their value. Stoicism tries to hold both (loving without demanding, acting with reservation). It is not certain that this position can be lived rather than merely stated. Skepticism cannot even attempt it: suspending judgment and devoting oneself to something exclude each other.

Fact and value. Grounding ethics in sensibility does not bridge the logical gap between what is and what ought to be. No other ethics bridges it either. That is not a consolation; it is an observation about the state of the question.

Acceptance and revolt. One can consent to one’s lot or rise against it. Both can be defended; neither can be demonstrated. It is a choice of how to exist.

Merit and protection. Removing fault also removes achievement. And a purely preventive justice, stripped of the idea of proportion, no longer has any internal brake. No satisfactory answer has been found.

The self that works on itself. Almost all the practical ways presuppose an agent capable of reforming itself. The thesis of the self without a center has just dissolved that agent. The difficulty was flagged but not addressed.

The three measures of well-being. Experienced pleasure, retrospective satisfaction and meaning regularly diverge. Several of the ways presented clearly favor the last two at the expense of the first. The traditions that propose them never make this choice explicit, and no argument establishes which of the three should prevail.

Where philosophy stops. Depression, recent bereavement, trauma and chronic pain lie beyond these exercises. They sometimes resist them all the more the harder one insists. The limit has been named, but it has not been mapped. There is no clear criterion for knowing in advance on which side a given suffering falls.

The scope. This account treats the manner of living as an individual matter. It sets aside the collective conditions of that life: institutions, law, the organization of work. That is not neutrality. A book that speaks only of what each person can make of what happens to them implies that this is where what matters most is decided. That is true only within the limits that chapters 16 and 19 marked out. The conditions of absorption are unequally distributed, and the dichotomy of control becomes an anesthetic when applied to what could be changed. This boundary was chosen to keep the book focused. It does not say that the rest matters less.

And a general suspicion. At least three times, a framework proved able to absorb any observation: the thesis of finality in chapter 2, internal evaluation in chapter 7, the affective loop in chapter 5. A system that explains everything is a system that risks nothing. This ease belongs on the debit side rather than the credit side.

On the manner

An account that ends without a verdict may disappoint. The reasons for ending this way should therefore be given.

The traditions it drew on each recommended it in their own way. The skeptics made uncertainty a path rather than a failure. The Stoics located freedom in the interval between impression and assent, that is, in not concluding too quickly. Epictetus advised examining what leads up to an undertaking and what follows from it before beginning. Marcus Aurelius wrote for himself, correcting himself as he went, without ever claiming to have finished.

There is also a simpler reason. A closed doctrine of how to live presupposes as settled what is not: whether desire should be reduced or cultivated, whether necessity should be accepted or refused, whether the self should be loosened or affirmed. These questions are not decided by argument. They are decided by a life, and not in the same way for everyone.

The problem, posed from beginning to end, remains whole: how to live, once it is granted that what is pursued recedes as one advances?

The traditions have offered nine answers, and one pole that denies them all. They do not agree. That is exactly where the question stands, and it is already better than pretending it is settled.

---

A note on sources

The quotations come from works in the public domain. The ancient Greek and Roman sources are Epictetus’s Enchiridion and Discourses as reported by Arrian; Marcus Aurelius’s Meditations; Seneca’s Moral Letters to Lucilius and his treatises; Epicurus’s Letter to Menoeceus, Principal Doctrines and Vatican Sayings; Lucretius’s On the Nature of Things; Cicero’s Tusculan Disputations and On Ends; Seneca’s Natural Questions; the fragments of Hierocles preserved by Stobaeus; the fragments of Democritus and Heraclitus; Plutarch’s On Tranquility of Mind; Diogenes Laertius’s Lives of the Eminent Philosophers; Sextus Empiricus’s Outlines of Pyrrhonism; Aristotle’s Nicomachean Ethics; Plato’s Gorgias and Symposium; and the Magna Moralia attributed to Aristotle. The Chinese and Indian sources are the Tao Te Ching and the Huainanzi, the Dhammapada, the Saṃyutta Nikāya, the Majjhima Nikāya, the Milindapañha and the Bhāgavata Purāṇa. The modern sources are Montaigne’s Essays; Pascal’s Pensées; Spinoza’s Ethics and correspondence; Butler’s Sermons; Hume’s A Treatise of Human Nature; Adam Smith’s The Theory of Moral Sentiments and The Wealth of Nations; Kant’s Groundwork of the Metaphysics of Morals, Critique of Practical Reason and Critique of Judgment; Bentham’s An Introduction to the Principles of Morals and Legislation; Mill’s Autobiography and Utilitarianism; Sidgwick’s The Methods of Ethics; Schopenhauer’s The World as Will and Representation; and Nietzsche’s Twilight of the Idols.

Where the precise reference of a passage (letter, book or paragraph number) could not be established, the note says so rather than claim an accuracy that has not been verified.

Contributions from outside the traditions are introduced as such in the text. They include work on the reward prediction error signal and on the dissociation between wanting and liking; variable-ratio reinforcement schedules; hedonic adaptation; the divergence between an experience and its memory; the dissociation between pleasure and meaning; split-brain patients; Libet’s experiment; the intuitionist model of moral judgment; the lineage linking Stoicism to cognitive therapies, and the evaluation of the protocols derived from them; the notion of optimal experience and its distribution between work and leisure; and longitudinal surveys on the determinants of well-being. These findings shed light on mechanisms. They validate no philosophical thesis, and the introduction explained why this reservation is not a mere formality.

---

Footnotes

1. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Letter to Menoeceus. ↩

2. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Reward prediction error model: Schultz, Dayan and Montague (1997). ↩

3. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/The wanting / liking dissociation: work by Kent Berridge and Terry Robinson on incentive salience. ↩

4. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Brickman and Campbell (1971), the origin of the notion of the hedonic treadmill. ↩

5. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/The “peak-end” rule and the divergence between the experiencing self and the remembering self: work by Daniel Kahneman and colleagues. ↩

6. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/David Hume, A Treatise of Human Nature, Book II, Part III, Section 3. ↩

7. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Baruch Spinoza, Ethics, III, scholium to proposition 2. ↩

8. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Baruch Spinoza, Letter 58, to Schuller. ↩

9. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Benjamin Libet (1983), on the readiness potential that precedes conscious intention. The interpretation is contested. ↩

10. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Work by Roger Sperry and Michael Gazzaniga on split-brain patients, and on the “interpreter” of the left hemisphere. ↩

11. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Daniel Wegner, experimental work on manufacturing the sense of agency. ↩

12. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Plato, Gorgias, 509e. ↩

13. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Friedrich Nietzsche, Twilight of the Idols, “The Four Great Errors.” ↩

14. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/The formulation known as the “basic argument,” due to Galen Strawson. ↩

15. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Baruch Spinoza, Ethics, V, scholium to proposition 4. ↩

16. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Sextus Empiricus, Outlines of Pyrrhonism, I, 12. ↩

17. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Social intuitionist model of moral judgment: Jonathan Haidt’s work on “moral dumbfounding.” ↩

18. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Arthur Schopenhauer, The World as Will and Representation, Supplements, ch. XIX. ↩

19. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Immanuel Kant, Groundwork of the Metaphysics of Morals, second section. ↩

20. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Ibid.* “Respect” as a feeling produced by reason, and the “fact of reason,” are set out in the Critique of Practical Reason, Part I, Book I, chs. I and III. ↩

21. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Jeremy Bentham, An Introduction to the Principles of Morals and Legislation, ch. 1. ↩

22. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Ibid.*, ch. XVII, note. ↩

23. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Mih%C3%A1ly Csíkszentmihályi’s work on optimal experience. ↩

24. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Studies comparing the correlates of hedonic well-being with those of the sense that life has meaning: notably Baumeister and colleagues. ↩

25. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/John Stuart Mill, Autobiography, ch. V. ↩

26. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, On the Shortness of Life, XVII. ↩

27. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Arthur Schopenhauer, The World as Will and Representation, IV, § 57. ↩

28. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Ibid.*, IV, § 56. ↩

29. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Blaise Pascal, Pensées, bundle “Diversion” (Lafuma 136; Brunschvicg 139). ↩

30. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Ibid.* ↩

31. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Blaise Pascal, Pensées (Lafuma 47; Brunschvicg 172). ↩

32. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Baruch Spinoza, Ethics, III, Definitions of the Affects, explanation of definition 13. ↩

33. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, IV, 3. ↩

34. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Dhammapada*, I, 1. ↩

35. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Sa%E1%B9%83yutta Nikāya*, XXXVI, 6 (Sallatha Sutta, “The Arrow”). ↩

36. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Principal Doctrines, IV. ↩

37. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Pierre Hadot, work on ancient philosophy as a way of life and on “spiritual exercises.” ↩

38. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/John Stuart Mill, Autobiography, ch. V. ↩

39. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Aristotle, Nicomachean Ethics, X, 4: pleasure is added to activity “as the bloom of youth to the young.” ↩

40. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Work by Mihály Csíkszentmihályi on optimal experience and its conditions. ↩

41. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Adam Smith, An Inquiry into the Nature and Causes of the Wealth of Nations, Book V, ch. 1. ↩

42. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Mih%C3%A1ly Csíkszentmihályi and Judith LeFevre, on what they called the “paradox of work.” ↩

43. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Principal Doctrines, XXVII. ↩

44. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Vatican Sayings, 34. ↩

45. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Vatican Sayings, 23. ↩

46. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Aristotle, Nicomachean Ethics, VIII, 1; Books VIII and IX. ↩

47. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Magna Moralia*, II, 15 (attribution to Aristotle disputed). ↩

48. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Hierocles the Stoic: the image of the concentric circles is known through Stobaeus, Anthology, IV, 27, 23. ↩

49. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, II, 1; VIII, 34. ↩

50. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 3; Discourses, III, 24. ↩

51. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 3. ↩

52. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Aristotle, Nicomachean Ethics, VIII, 3. ↩

53. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 5. ↩

54. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 45. ↩

55. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, IV, 7. ↩

56. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 1. ↩

57. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, VI, 13. ↩

58. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 98, 8. ↩

59. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 24 and 91. ↩

60. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Plutarch, On Tranquility of Mind. ↩

61. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Plutarch, On Tranquility of Mind, 8. ↩

62. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 96. ↩

63. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/A lineage claimed by Albert Ellis (rational emotive therapy) and Aaron Beck (cognitive therapy). ↩

64. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/The so-called third-wave approaches: acceptance and commitment therapy, mindfulness-based therapies. ↩

65. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 1. ↩

66. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 1, 5. ↩

67. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Cicero, On Ends, III, 22. ↩

68. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 8. ↩

69. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 107, 9. ↩

70. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, X, 21. ↩

71. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Baruch Spinoza, Ethics, V. ↩

72. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, VII, 26; and II, 1. ↩

73. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Discourses, I, 18. ↩

74. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 43. ↩

75. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Baruch Spinoza, Ethics, IV, proposition 54. ↩

76. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Letter to Menoeceus; see also Principal Doctrines, XXIX. ↩

77. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Vatican Sayings, 33. ↩

78. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Vatican Sayings, 25. ↩

79. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Vatican Sayings, 71. ↩

80. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Discourses, IV, 1, 175. ↩

81. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Vatican Sayings, 68. ↩

82. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, VII, 27. ↩

83. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Democritus, fragment B 191 (Diels-Kranz numbering, which varies across editions); Plutarch, On Tranquility of Mind. ↩

84. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Letter to Idomeneus, reported by Diogenes Laertius, Lives of the Eminent Philosophers, X, 22. ↩

85. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Vatican Sayings, 55. ↩

86. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Sa%E1%B9%83yutta Nikāya*, XXII, 59 (Anattalakkhaṇa Sutta). ↩

87. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Milindapa%C3%B1ha*, II, 1, 1 (dialogue between Nāgasena and King Milinda). ↩

88. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Heraclitus, Fragments, 91. ↩

89. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Work by Roger Sperry and Michael Gazzaniga on split-brain patients; hemispherectomy cases in pediatric epileptology. ↩

90. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Derek Parfit, Reasons and Persons, on personal identity as a relation that holds by degrees. ↩

91. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 13. ↩

92. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 33. ↩

93. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Diogenes Laertius, Lives of the Eminent Philosophers, VII, 101-102. ↩

94. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Democritus, fragment B 244 (Diels-Kranz). ↩

95. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 21. ↩

96. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, VII, 69; IV, 48. ↩

97. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus, Letter to Menoeceus; Principal Doctrines, II. Symmetry argument: Lucretius, On the Nature of Things, III. ↩

98. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Michel de Montaigne, Essays, I, 20, “That to Philosophize Is to Learn to Die.” The formula comes from Cicero, Tusculan Disputations, I, 30, who took it from Plato’s Phaedo. ↩

99. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Michel de Montaigne, Essays, III, 12, “Of Physiognomy.” ↩

100. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Plato, Symposium, 207a-209e. ↩

101. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Aristotle, Nicomachean Ethics, I, 11. ↩

102. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epicurus's will, reported by Diogenes Laertius, Lives of the Eminent Philosophers, X, 16-21. ↩

103. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, V, 23; IV, 36. ↩

104. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, VII, 18. ↩

105. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 1; and On the Shortness of Life. ↩

106. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, VI, 36 (a recurring motif in Books VI to XII). ↩

107. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, IV, 32-33. ↩

108. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Natural Questions, I, preface. ↩

109. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, IV, 3. ↩

110. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Majjhima Nikāya*, 118 (Ānāpānasati Sutta). ↩

111. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Marcus Aurelius, Meditations, III, 2. ↩

112. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Arthur Schopenhauer, The World as Will and Representation, III, § 34 and § 38. ↩

113. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Immanuel Kant, Critique of Judgment, § 2 and § 5. ↩

114. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 80. ↩

115. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 12. ↩

116. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 18. ↩

117. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Epictetus, Enchiridion, 29. ↩

118. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 11 and 25. ↩

119. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Sextus Empiricus, Outlines of Pyrrhonism, I, chs. IV and XII. ↩

120. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Huainanzi*, ch. XVIII, “In the World of Men.” ↩

121. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Arthur Schopenhauer, The World as Will and Representation, IV, § 68. ↩

122. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Ibid.* ↩

123. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Seneca, Moral Letters to Lucilius, 5, 7-8. ↩

124. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/*Bh%C4%81gavata Purāṇa*, XI, 8, 44. ↩

125. https://github.com/SimonKreis/essay-on-human-nature/blob/HEAD/Baruch Spinoza, Ethics, V, proposition 6. ↩