In this article, I’m going to provide a careful analysis of the evidence pertaining to the ancient Stoic taxonomy of anger, with particular attention to the terminology used in Latin and Greek, and how their concepts might be interpreted from a modern perspective. This will be quite an in-depth discussion, employing some technical concepts from ancient philosophy and modern psychology. My hypothesis is that Chrysippus, the third head of the Stoic school, and its greatest intellectual, developed a very early, prototypical form of cognitive-behavioural psychotherapy, which was outlined in a lost work called The Therapeutics (To Therapeutikon), part of a four-volume work called On the Passions (Peri Pathon). Fortunately, modern scholars, most notably Prof. Teun Tieleman, have been able to reconstruct a significant portion of that book, in recent decades, based on surviving testimonia and fragments found in the writings of later authors, such as Cicero and Galen.
The difficulty of translating ancient Greek and Latin into modern English is compounded because the words being translated here are used in a technical sense by the ancient authors, because they’re making fine distinctions between difficult concepts such as types of emotion, and also because the available evidence is quite limited, and even conflicting in some regards. Nevertheless, the Stoics arguably had a more systematic and comprehensive approach to the cause and cure of emotional disturbance than any previous school of philosophy and we can say quite a lot about their way of conceptualizing anger and classifying its different forms.
I’ll begin by giving a brief account of the general Stoic model of “passion” (pathos), a word denoting both desires and emotions. Our main sources will be the doxographers Diogenes Laertius, who provides a famous summary of early Greek Stoic doctrines; and Joannes Stobaeus, who is believed to derive his account from the Stoic teacher Arius Didymus. We will also find useful information provided by Cicero, who first translated early Greek Stoicism into Latin; and Seneca, especially his book On Anger.
The Stoic Dual-Phase Model of Anger
Some Stoics made a fundamental distinction between two distinct chronological phases of the passions. Before anger, or any other passion, arises, there are certain initial involuntary reactions. Chrysippus reputedly spoke of little “bites”, meaning sharp contractions in the chest or stomach, which precede the “inflammation” of the psyche with full-blown passions such as anger. Later Stoics called this the “proto-passion” (propatheia in Greek), while Seneca describes it as the “first movement” (primus motus in Latin). Scholars disagree over the extent to which the roots of this underlying distinction can be traced back to the founders of Stoicism. However, by the Imperial era, it appears to play an important role in the writings of Stoic authors such as Seneca and Epictetus.
First Phase. The involuntary psychological and physiological reactions triggered by an initial impression, such as trembling when you are unexpectedly confronted with a dangerous situation or tears welling up when you first hear of a loved one’s death.
Second Phase. When we give “assent” (synkatathesis) to the initial impression, the proto-passion is transformed into a true passion. Assent therefore marks the transition from proto-passion to passion, and is also a necessary feature of all passions, including anger.
Although synkatathesis is conventionally translated as “assent”, I believe that English word is misleadingly weak. The Greek can refer to casting an official vote, and in Stoic psychology assent to the impression that an action is appropriate forms the corresponding impulse (hormē) to act. It therefore involves a decisive commitment to the impression, not merely entertaining or mildly agreeing with it. The actual use of the word by Stoic philosophers implies, I believe, something more akin to the identification of the initial impression with reality. It’s helpful to think of this not as mild assent but rather as a state of mind in which the impression now shapes one’s experience of reality. For instance, synkatathesis is the difference between lightly holding the thought “I should retaliate”, without acting on it, versus accepting it so decisively that you now feel bound to retaliate and attempt to do so.
Those in the grip of genuine anger do not experience themselves merely as “assenting to the impression” that someone deserves to be punished, which would imply a level of detachment and self-awareness more characteristic of calm reflection than passionate anger. Rather they simply experience someone before them as being, in reality, an enemy who deserves to be punished. I believe that synkatathesis is best understood not as “assent” but as a concept somewhat closer to “cognitive fusion” in Acceptance and Commitment Therapy (ACT), or to Beck’s notion of thoughts that are experienced with poor “cognitive distance” or separation from real events.
The Stoics defined wisdom as requiring “freedom from hasty assent”. Epictetus, for instance, tells his students that when they experience a troubling impression, such as those provoking anger, they should take a moment to examine it, withhold “assent”, and speak to it saying: “You are just an impression and not entirely what you claim to represent.” That therapeutic strategy implies that passions grip us, and we lose control, when we rush to judgment and hastily equate our initial impressions with reality. Chrysippus, who was a long-distance runner, famously compared being overwhelmed by a passion, such as anger, to someone running so fast he can’t easily stop or avoid obstacles.
If we apply the generic Stoic two-phase model of passion to anger, it looks something like this:
First Phase. Proto-anger, you experience an impression of injury such as “That man has harmed my reputation”, which automatically triggers initial “bites” or unpleasant feelings and urges, comparable to the reactions of animals and small children to perceived threats, which we might compare to the physiological fight-or-flight response or the experience of raw affect.
Second Phase. Anger, which the Stoics compare to the psyche becoming temporarily inflamed, as though with a fever, occurs when an impression of threat or harm, is identified with reality, and the judgment added that it is appropriate or necessary to retaliate. Seneca gives the example: “It is appropriate for me to be avenged because I have been injured” (On Anger, 2.4).
Cicero states that even when we no longer judge the event to be bad, “the grieving distress will be removed, thought bites and certain little contractions of the intellect will remain”, making it clear that he’s describing something akin to the Stoic concept of proto-passions. Nowhere do the Stoics actually use a term like “proto-anger” but doing so is a reasonable way to make “proto-passion”, the more general term used by them, more specific to this emotion.
In On Anger, Seneca actually describes three movements, or phases, in which the full passion may be followed by further escalation and loss of control (On Anger, 2.4).
Third Phase. Seneca says that anger now wants revenge not “if it is appropriate,” but “regardless,” having “overcome reason.”
Scholars dispute the significance of this final phase. Whereas Richard Sorabji interprets it as the escalation of anger, Margaret Graver thinks it refers to deterioration beyond anger into brutal cruelty.
The Stoics were, in any case, aware that emotions can best be viewed as processes, divisible into distinct chronological phases. In addition to a beginning and middle of passions, such as episodes of anger, the Stoics also discussed their end. Chrysippus observed that, for instance, both laughter and grief tend to run their course and naturally come to an end eventually. In other words, under normal circumstances passions are typically self-limiting episodes. The early Stoics, possibly Zeno himself, appear to have referred to passions being caused by “fresh” (prosphatos) impressions, and naturally wearing off over time. There is some scholarly disagreement over the interpretation of this concept - whether “fresh” means recent or occurring. However, I believe that the Stoic claim that passions require “fresh” impressions may resemble the habituation model of anxiety, according to which anxious arousal abates naturally over time as a result of prolonged exposure to the trigger. For instance, when the impression of an insult is fresh in my mind, my anger may be at its peak. However, my anger, at least in some regards, may abate naturally over time as the impression becomes a progressively older and more distant memory.
Passions would, according to this model, continue as long as we appraise them as appropriate or necessary, in given circumstances. One clear application of this in the ancient world has to do with common debates about how long — how many days or months — it is appropriate to grieve over someone who is deceased. Grief may end naturally when the impression of loss is no longer “fresh” in the mind, in other words, unless it is perpetuated by the belief that it is necessary to continue grieving for longer. Anger, likewise, may end naturally when the initial impression of insult or injury is no longer fresh, but it may be perpetuated for even longer if we believe that we should feel angry, behave angrily, or continue to angrily ruminate about the offence.
Stoic Taxonomy of Anger
The terminology used by the Stoics to classify different forms of anger is hierarchical.
Passion
At the highest level, for the Stoics, anger is a species of passion (pathos). The word pathos covers both emotions and desires. It implies passivity, or being affected by something, and also mental disturbance. Pathos, indeed, is the root of our word pathology.
For this reason, when Cicero first attempted to translate the term from Greek into Latin, he wrote:
What the Greeks call pathē: I could have called them “diseases” [morbi], and that would have been a word-for-word translation, but it would not accord with our usage. — Cicero, Tusculan Disputations, 3.7
He later adds:
What the Greeks call pathē, we prefer to call “disturbances” [perturbationes] rather than “diseases” [morbi]. — Cicero, Tusculan Disputations, 4.10
Cicero’s point is that pathos can mean disease, and that fits some of the examples given by the Stoics but seems slightly excessive for others, so he chooses to refer to them as emotional “disturbances” instead. However, the English words “passion” and “emotion” risk losing sight of that connotation completely. In short, the Stoics are not referring to all emotions but specifically to emotional disturbances.
Indeed, the early Greek Stoics specify that the word pathos is used by them to mean unhealthy, irrational, and excessive desires and emotions.
According to Zeno, “passion” is an irrational and contrary-to-nature movement of the soul, or an excessive impulse. — Diogenes Laertius
They say that a “passion” is an excessive impulse, disobedient to the determining reason, or an irrational movement of the soul contrary to nature; and that all passions belong to the ruling [executive] faculty of the soul. — Stobaeus
“Excessive” (pleonazousa) does not always mean that the impulse feels too intense but, more importantly, that it goes beyond what reason determines as appropriate.
Cleanthes famous piece of versified dialogue between Reason and Anger illustrates the excessive demandingness of anger, which threatens to usurp reason.
Reason: “What, Anger, is it that you want? Tell me that.”
Anger: “I, Reason? To do everything I wish.”
Reason: “Why, that is royal; but still, say it once again.”
Anger: “In whatever way I desire, that this will come about.”
Chrysippus famously quoted Medea to illustrate the way in which anger goes beyond the bounds of reason: “I understand what evils I am about to do, but rage [thymos] is stronger than my deliberations.” Medea was so utterly enraged by her husband’s betrayal of her that, contrary to her own better wisdom, she retaliated against him by murdering their children. Epictetus, incidentally, described Medea as “the shipwreck of a noble mind”. There’s an obvious connection here with the Stoic virtue of temperance (sophrosune), through which our impulses, urges, and actions are brought into alignment with reason and wisdom. In other words, the passions, including anger, are intemperate.
Chrysippus was well-known for popularizing the Stoic doctrine that the passions are judgments (krises). I believe this justifies the claim that the Stoics introduced the earliest major form of cognitive psychotherapy. He distinguished between underlying beliefs, of a general nature, which lead to the specific judgments that form desires and emotions. These judgments are usually said to consist in assenting to impressions (phantasia). The concept is not unlike that of schemas leading to specific appraisals and automatic thoughts, in modern cognitive therapy.
The passions are divided into four main genera:
distress (lupē), which can encompass grief, sadness, frustration, and emotional pain in general, and is based upon the judgment that something bad/harmful has occurred.
fear (phobos), which is based on the judgment that something bad/harmful is impending.
craving (epithumia), which is usually translated as “desire” but can also mean appetite or craving and is based on the judgment that something good/beneficial is impending.
pleasure (hēdonē), which refers to the emotion rather than the physical sensation and is based on the judgment that something good/beneficial has occurred.
Desire
Anger is a species of craving (epithumia), which is, in turn, a species of passion. It’s important to note that epithumia is conventionally translated as “desire” but that does not fit the Stoic usage, which defines it as an irrational, excessive and unhealthy form of desire. We therefore see craving defined as follows by the Stoics:
Craving is an irrational striving. — Diogenes Laertius
Craving is a striving which is disobedient to reason; its cause is believing that a good is approaching and that when it is here we shall do well by it… — Stobaeus
In other words, the Stoics define the word epithumia or “craving” as a specific form of emotional disturbance that is more irrational and excessive than healthy desires, which accord with reason and wisdom. Once again, it is tempting to compare this to the concept of “rigid demands” in Albert Ellis’ rational emotive behaviour therapy (REBT).
Anger, therefore, is an irrational striving (orexis), which believes that revenge is potentially within its grasp, and evaluates the punishment of others as something intrinsically good and beneficial, worthy of being desired.
Anger
It’s a striking fact that most ancient Greek and Roman philosophers, across different schools, agreed that anger is best defined, broadly speaking, as the desire for revenge, retaliation or punishment. Seneca, of course, defines anger as the “desire for revenge” (cupiditas ulciscendi) in On Anger (1.2.3). The Greek Stoics likewise defined anger specifically as the desire for timōria, which is usually translated as “revenge” but has a slightly broader meaning and can also be rendered as “retaliation”.
Anger is a desire for retaliation against someone who is thought to have wronged one inappropriately. — Diogenes Laertius, 7.113–114
The Greek can also mean that the target of one’s retaliation is thought to have acted unjustly. In modern psychology, we may also speak of the urge to aggress, or to behave aggressively, of which the desire for retaliation may be considered one specific form.
The definition of anger as a “desire for revenge” implies that it is a response to a preceding sense of perceived injury and negative affective state — revenge is stereotypically a form of retaliation for a perceived injury, and associated feelings of pain. Arguably, this makes anger resemble a coping response. The Stoic definition can be interpreted as making anger a response to the evaluative belief that one has been harmed, and also to the proto-anger, which is itself a primitive aversive experience — one we could perhaps describe as raw affect.
Cicero describes the Stoic theory as holding that anger is a “desire to punish”, which is somewhat broader in scope than “desire for revenge”. There may be some Stoic theory missing here, which would distinguish rational from irrational punishment. We find a clear statement of a distinction that may be relevant here in Plato’s Protagoras. Punishment, according to this view, is irrational, like the behaviour of a “wild beast”, when it focuses on the past, and seeks to do harm for its own sake, out of retaliation. Rational punishment (kolasis) looks to the future, and aims to correct behaviour, and improve the person being punished.
It’s worth noting that the Stoic model may imply that some form of irrational pleasure is experienced, or at least hoped for, when revenge achieves what it desires. As we’ve seen, the passion of irrational, excessive, and unhealthy “pleasure” is experienced when something good or beneficial is judged to be present. Both Stobaeus and Diogenes Laertius say the Stoics describe an irrational form of pleasure called “joy at another’s misfortune” or “malicious pleasure” (epichairekakia), which could potentially correspond with the desire for revenge. This Greek word literally means “joy over something bad [happening to another].” The Stoics themselves define it as “pleasure at another’s misfortunes” — it means virtually the same thing as the familiar German word Schadenfreude.
Cicero translates it into Latin as malevolentia, meaning “ill-will” or “malicious pleasure”, which he defines as follows:
Malevolentia is pleasure arising from another person’s misfortune, without any advantage to oneself. — Cicero, Tusculan Disputations, 4.20
This can be compared with a passage in Seneca’s On Anger, which states that a rational person may sometimes punish others but “not because he takes pleasure in anyone’s punishment — for such inhuman savagery is far removed from the Sage.” He then adds:
If a good man does not rejoice in punishment, neither will he rejoice in that passion for which punishment is a pleasure; therefore anger is not natural. — Seneca, On Anger, 1.6.4-5
The wise man he says will punish wrongdoing out of duty, and not out of anger (On Anger, 1.12).
In other words, the Stoics perhaps meant that when someone has irrational anger and an unhealthy desire for revenge, they frequently look forward to experiencing a malicious pleasure in the punishment and suffering of their victim. The wise do not rejoice in the suffering of others, and therefore they do not judge anger, or the desire for revenge, to be good.
The Subtypes of Anger
The Stoics further divide the main categories of desire and emotion into subtypes. Diogenes Laertius gives only two subtypes of anger:
Thymos is “anger beginning” or incipient anger, the initial flash of emotion, which the Greeks typically associate with agitation and a “hot” reaction to some provocation.
Mēnis is “a kind of anger that has grown old, nurses a grudge, and is watchful for its opportunity”, it is long-standing and spiteful anger, on the lookout for opportunities to take revenge.
This distinction may be intended to distinguish between subtypes of anger based upon episode length. Thymos may perhaps refer to a burst of anger, which flares up temporarily, whereas mēnis is more settled and prolonged anger, presumably often accompanied by rumination, which matures into resentment and plots revenge.
It’s tempting to equate thymos with the proto-passion of anger, and it’s possible that is how the word was used by some Stoics. However, the textual evidence leans somewhat in favour of concluding instead that it refers to the moment of initial assent and the earliest phase of the passion of anger proper.
Diogenes Laertius says that the concept of mēnis is clearly illustrated by the following quote from Homer, which describes why it is considered dangerous to provoke the anger of powerful rulers such as Agamemnon:
For even if he should digest his rage [cholos] that very day, nevertheless afterwards he nurses his grudge [kotos], until he brings it to fulfilment. — Iliad, 1.81-82
As we’ll see, the Stoics elsewhere utilize both these Greek words for anger, although they are not part of the basic taxonomy given above. Here, I take Diogenes Laertius to be implying that cholos is comparable to thymos and kotos to mēnis. In other words, the passage illustrates how some people may “swallow” or suppress their initial surge of anger arousal, or conceal it from others, only to ruminate about it, nursing lasting resentment, and awaiting an opportunity for revenge.
This contrast seems to distinguish acute and chronic episodes of anger, the implication being that acute anger can be very intense, accompanied by high physiological arousal, but that in chronic anger, which waits for an opportunity to retaliate, the urge to express itself must be restrained, and the initial “heat” reduces to a “simmer” over time. A very similar contrast appears in a fragment preserved under the name of Arrian and traditionally associated with the teachings of Epictetus. Although the fragment does not itself use the word mēnis, its description seems remarkably close to the Stoic definition of mēnis as anger that has “grown old”. It also explicitly contrasts a conspicuous form of thymos with a quieter form of hostility in which the desire for retaliation persists much longer:
Some men of great thymos [megalothymoi] behave gently, quietly doing, as though without anger [aorgētōs], the very same things as those who are violently carried away by thymos. We must therefore be on our guard against the error of such men too, since it is far worse than that of someone who is intensely angry [orgizesthai]. For the latter quickly get their fill of revenge [timōria], whereas the former prolong it for a long time, like those suffering from a low-grade fever.
Recall the language of psychological “bites” and “inflammation” attributed to Chrysippus. It is tempting to speculate that these belong to a broader cluster of medical metaphors for the passions, and especially for anger, that recurs throughout Stoic literature. On this reading, an insult or provocation is first experienced as a kind of psychological “bite”, rather like an insect bite. This is a tiny injury in itself, but one that automatically produces the initial affective reaction, or inflammation of the wound, that we might call proto-anger. If we assent to the impression that we have been injured, though, and no remedy is applied, this initial reaction can become progressively more “inflamed”, swelling into the fever of full-blown anger. The passage attributed to Arrian seems to apply the same imagery into the chronic phase. Acute anger may burn itself out relatively quickly, whereas a more persistent state of resentment can linger like a “low-grade fever”. The initial swelling and heat have subsided; the person may even behave calmly and “as though without anger”, yet the underlying passion has not disappeared. Acute physiological arousal has naturally diminished, but the desire for retaliation persists and may flare up again when circumstances provide an opportunity.
A more detailed account of the Stoic taxonomy of anger is given by Stobaeus who expands upon their taxonomy of anger as follows:
thymos is “anger just beginning”
pikria is “anger which breaks out immediately”
cholos is “swollen anger”
mēnis is “anger stored up for a long time”
kotos is “anger which watches for an opportunity for vengeance”
It’s more challenging to interpret the finer distinctions being made here. It’s possible, for instance, that the word cholos is meant, in contrast with the other terms, to denote a form of anger that neither breaks out immediately nor lingers on for a long time, like a low-grade fever, but rather remains acutely inflamed for some time. The limited information available makes that interpretation speculative, though.
Cicero appears to translate a version of this Greek Stoic taxonomy of anger into Latin.
ira is the “desire to punish someone who seems to have injured one unjustly”, clearly translates orgē, the general Stoic term for anger in Greek.
excandescentia, which means “flaring up”, is defined as “anger just arising and newly existing”, Cicero actually added “which in Greek is called thymōsis”, a term cognate with thymos.
odium is usually translated as “hatred” or “animosity” and defined by Cicero as “long-established anger”, which may plausibly be intended to translate the Greek mēnis.
inimicitia means “enmity” or “hostility” and is defined here as “anger watching for the time to take revenge”, which may plausibly be intended to translate the Greek kotos.
discordia, meaning “discord” or “conflict”, is here “a more bitter anger conceived deep in the mind and heart”, which seems unlikely to be intended as a translation of pikria as it is not described as a brief outburst, but sounds more like an enduring state of anger.
These terms seem to neatly confirm that most of the Greek taxonomy provided by Stobaeus was known to the Romans to be Stoic doctrine. The only really problematic part of Cicero’s account is his description of discordia, which doesn’t quite match up with the Greek pikria. Curiously, though, Cicero describes discordia as an acerbior or “more bitter” type of anger and pikria literally means “bitter” in Greek. It may therefore be more accurate to think of Cicero as providing a parallel, albeit very similar, taxonomy to that of Stobaeus, rather than a direct translation.
Seneca actually admits that whereas the Greek Stoics distinguished between many different species of anger, Latin lacks perfectly corresponding terms. He does give some Latin terms for anger but of more interest is his breakdown of various ways of expressing anger through our words and actions (On Anger, 1.4.2–3).
Frequent and persistent anger. Some forms of anger are “no less persistent than frequent”, which could plausibly mean that they involve repeated outbursts and outward expression, taking the various forms below.
Anger discharged in shouting. Some “subside within shouting”, and are exhausted through noisy verbal expression alone.
Verbally abusive anger. Some pour themselves out into the bitterness of words and insults, so not necessarily raising the voice but saying hurtful things.
Physically violent but verbally restrained anger. Some are savage “with the hand” but sparing of words.
Complaining and rejecting/withdrawing anger. Some go no further than “complaints and withdrawal”, which could refer to sulking, avoidance, and indirect verbal aggression or perhaps even passive-aggressive behaviour.
Deep, inwardly directed anger. Some are “deep and weighty and turned inward”, so they are barely expressed at all but rather stored up inside.
The last example is particularly notable because whereas ancient philosophy generally tends to focus on direct outward expression of anger (“anger-out”), Seneca here seems to recognize the internalization of anger (“anger-in”) as a mode of expression.
The Dual-Appraisal Model of Anger
Chrysippus reputedly introduced a model of the passions according to which they are caused by two distinct judgments. These judgments take more general beliefs and apply them to specific situations. Today we might say that more general cognitive schemas are activated by specific triggers, which lead to situation-specific appraisals. According to Richard Lazarus’ highly-influential “transactional” model of stress and emotion, these are cognitive appraisals of the problem we face and our ability to respond to it, which I will call “primary appraisal of threat” and “secondary appraisal of coping”. Chrysippus’ model is quite similar to Lazarus’, as Prof. Margaret Graver has noted in Stoicism and Emotion.
The First Judgment, assents to the impression that something good/bad is either impending or present. Because it relates to how good or bad, helpful or harmful, something is appraised as being, I would call this the evaluative judgment, and compare it to Lazarus’ primary appraisal of threat. We’re told Chrysippus also refers to the passions as judging misfortune to be “intolerable” (anupomonētos) or “unbearable” (akarterētos) in some cases.
The Second Judgment, then assents to an impression regarding how it is appropriate, or necessary, to respond to what has happened, or is about to happen.
There are some problems of translation here but the word the Stoics sometimes use, kathekon, conventionally means “fitting”, “appropriate” or “a duty” and as they also say the judgment that something is appropriate is both “irrational” and “excessive”, it is tempting to compare this type of belief to Albert Ellis’ concept of irrational beliefs about rigid demands, such as “should” and “must”, etc. In other words, I believe the Stoics may be interpreted saying that unhealthy passions are always accompanied by beliefs about how we “must” respond, although they phrase it as excessive and irrational belief in some course of action, such as retaliation, being normative or right. Certainly, the Stoics believe that when the second judgment is identified with reality, it compels us to behave emotionally, i.e., to perpetuate, act out, and express the emotion.
If we apply this to anger, it would look like this:
First Appraisal. The Stoics typically refer to anger as a response to a perceived unjust injury, inflicted by another person, which is evaluated as intrinsically bad. We’re told Their theory may be flexible enough to allow them to acknowledge a range of perceived threats as potential triggers for anger, though.
Second Appraisal. In the case of anger, the Stoics describe this as the belief that “revenge” (or retaliation) is appropriate or necessary.
Interestingly, Chrysippus reputedly held that it is difficult to reason with someone and change their mind about their first judgment while their passions are still inflamed. He said it was more effective at that time to challenge their second judgment, i.e., their opinion about what is most appropriate or necessary to do. The Stoics might say it’s easier to dispute the normative judgment that “It is appropriate for me to seek revenge” than the evaluative judgment that “Your disrespect has harmed me / is extremely bad.” In other words, Chrysippus appears to have distinguished between therapy that works proactively to challenge evaluative beliefs and crisis intervention, during an episode of anger, which works reactively and mainly challenges beliefs about how it is appropriate to behave, or emotional coping.
Chrysippus on Anger
At this point, it’s worth quoting in full a remarkable passage about anger believed to actually be from the Therapeutikon of Chrysippus.
For in this way we take leave of ourselves and become outside ourselves, and are completely blinded when we are frustrated; so that sometimes, if we have a sponge or a piece of wool in our hands, we pick it up and throw it, as though we were going to accomplish something by means of it. And if we happened to have a knife or some other object, we would use that in much the same way.
Chrysippus continues:
Often, in this sort of blindness, we bite the keys and beat the doors when they do not open quickly; and if we stumble against stones, we behave punishingly [timōrētikōs] toward them, breaking them and throwing them away somewhere, while saying the most outrageous [or absurd] things on each such occasion.
And concludes:
From such cases one might understand the irrationality present in the passions, and how at such times we are blinded, as though we had become different people from those who had previously reasoned.
The word timōrētikōs is particularly striking here because it comes from the same family as timōria, retaliation or punishment. This is the very desire that the Stoics have used to define anger. Chrysippus sounds like he could be describing a person who has displaced their anger onto an inanimate object as though it were an enemy deserving of severe punishment.
It seems likely that the examples given above by Chrysippus relate to what people judge appropriate or necessary behaviour. Although he concludes by making a general point about the passions, he employs the imagery of anger, such as a person so enraged and frustrated that he even “punishes” or “retaliates” against a stone. Cicero explains that when people are distressed by grief their judgment about what it is appropriate to do leads them to wail and tear their hair, etc. (Tusculan Disputations, 3.61). This is important because it shows that the Stoic “second judgment” was not merely the belief that one ought somehow to feel distressed: it could govern characteristic forms of emotional expression.
What struck me immediately when I first read the passages from Chrysippus’ Therapeutikon above was that most of his examples of irrational behavior in anger appear to be forms of emotional venting. People often consider venting a natural way of expressing their anger. It has also been proposed as a deliberate therapeutic procedure, called “catharsis”, although studies have generally failed to support its efficacy for treating anger. To be clear, Chrysippus does not think this type of behaviour is therapeutic. On the contrary, he views “biting keys”, saying outrageous things, beating at doors, throwing stones, etc., as an irrational and unhealthy way to respond when enraged. These examples suggest that beliefs concerning the appropriateness or necessity of aggressive emotional expression may could been natural targets for Chrysippean therapy. The original Therapeutikon may even have recommended the disputation of beliefs in the necessity or usefulness of irrational behaviours we would now describe as “venting”, although the surviving fragment does not state this explicitly.
State versus Trait Anger
Modern psychologists also make a distinction between emotional states, which are temporary and relatively short-lived, and traits, which are longer-lasting dispositions to experience states. The same distinction was made by the Stoics with regard to the passions in general. They use the Greek word euemptōsia to refer to the “proclivity” or “susceptibility” for certain passions, including anger.
The proclivity toward anger is called orgilotēs, which literally just means “anger-proneness”, but is sometimes translated as “irascibility” — it is the trait that predisposes someone to experience states of anger. Stobaeus also mentions an anger-related proclivity called akrocholia perhaps meaning “quick tempered”. It’s difficult to tease apart the terminological distinction the Stoics wanted to make between being orgilotēs and akrocholia. At face value, the word orgilotēs simply describes someone as being prone to anger in general. The word akrocholia was used by Aristotle to denote someone who is quick to anger in general, but whose anger also fades quickly (Nicomachean Ethics, 4.5). However, we can’t be certain if the Stoics intended to use the word in the same way.
Cicero expresses essentially the same distinction in Latin and may, perhaps, even be translating orgilotēs and orgē, as follows:
In some it is called “irascibility” [iracundia], which differs from anger [ira]: it is one thing to be irascible, another to be angry.” — Cicero, Tusculan Disputations, 4.27
Seneca uses the same Latin terminology and explains that someone may be anger-prone or “irascible” (trait) without currently being angry (state), and vice versa (On Anger, 1.4.1). He compares this to saying that someone may either be drunk without being a drunkard, or be a drunkard without being drunk. Of course, they could also be both drunk and a drunkard, just as someone may be both irascible and angry at the same time.
The Stoics also distinguished these emotional proclivities from a deeper and more serious pathological condition called a nosēma, literally a “disease” of the psyche. Cicero explains that if the initial “heat and agitation of the mind” becomes established through time, it can settle into the “veins and marrow” and develop into what the Stoics called a nosēma (Tusculan Disputations, 4.23–24). He elsewhere defines hatred (odium) as ira inveterata, “inveterate” or long-established anger (Tusculan Disputations, 4.21), and Seneca likewise says that a hostile passion can become “incorrigible when it has hardened from anger into hatred” (On Anger, 3.41.3). The Greek Stoic taxonomy likewise distinguishes hatred (misos) from anger, defining it as a desire that another person should fare badly characterized by “development and prolongation”. Anger, then, can apparently cease to be merely a desire for retaliation and, if a grudge is nursed and allowed to harden, it may develop into a persistent hostile attitude, or hatred of another person.
Conclusion
It’s worth mentioning in conclusion that Stoic psychotherapy (therapeia) aimed to a large extent at apatheia, or freedom from irrational, excessive, and unhealthy passions. However, the Stoics also have a taxonomy of healthy emotions (eupatheiai). The healthy alternative to “craving” is conventionally, if rather awkwardly, translated as “wishing” (boulēsis), meaning rational and appropriate desire — it’s often best just to translate this term as “rational desire”.
The concept of rational desire also has its subtypes and within them we find the terms “goodwill” (eunoia) and “kindness” (eumeneia), which seems to refer to moments of good intent and their more lasting and settled form respectively. These may describe the healthy desires and emotions that most directly oppose anger. Indeed, Marcus Aurelius, treats eumenēs, a cognate term for eumeneia, as the antidote to anger.
Consider that a kind disposition [eumenēs] is unconquerable, provided it is genuine and not an affected smile or hypocritical. — Meditations, 11.18
Marcus doesn’t mean that kindness always conquers others, but that nobody can prevent our sincere attempts to act with kindness and benevolence. The wise desire to help others, insofar as nothing prevents them, by correcting or even punishing wrongdoing, and ideally to turn enemies into friends. Kindness is therefore the healthy emotion or attitude that the Stoics believe should replace anger and the desire for revenge.




This is a fascinating deep dive. The first phase is so important. I think we can develop the ability to stay in that first phase and then choose to avoid the second, rather then let the cascade continue. No matter how "just" we may believe our anger, it has no value and can only harm ourselves and others. Most of the time, perhaps all of the time, the source of our anger can be found within, if we just stop and analyze. One of the best maxims is this: "I never give anyone else the power to make me angry." Thanks again for a great article.
Sometimes the most dangerous anger is the one that can't retaliate against its actual cause, and gets discharged downward onto someone weaker. A partner, a child, anyone who can't push back.
What follows almost inevitably is rationalization: the target gets recast as "provoking it" or "needing to learn," and displacement gets laundered as discipline.
Then the loop tightens. Next time, the discharge onto the same weaker person comes faster.