Jump to ContentJump to Main Navigation
Emotion and Peace of MindFrom Stoic Agitation to Christian Temptation$

Richard Sorabji

Print publication date: 2002

Print ISBN-13: 9780199256600

Published to Oxford Scholarship Online: May 2007

DOI: 10.1093/acprof:oso/9780199256600.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 17 September 2019

Seneca's Defence

Seneca's Defence

Third Movements as Harmonizing Chrysippus and Zeno

(p.55) 3 Seneca's Defence
Emotion and Peace of Mind

Richard Sorabji (Contributor Webpage)

Oxford University Press

Abstract and Keywords

Zeno of Citium, the Stoic founder, had tried out other definitions of emotion. One, defended by Chrysippus, was that emotion involves oscillating, like Medea, between accepting the right value judgement and disobeying it. But disobedience to reason is not the same as mistake. How can it be, and is it ever, combined with mistake? The Stoic Seneca (1st century CE) allows this by distinguishing three movements in anger. The first movement is the appearance that revenge is appropriate and the resulting shock to soul or body. The second is the mistaken assent to the appearance that revenge is appropriate. The third movement — the full emotion — moves from mistake to disobedience with the judgement that revenge is to be pursued, appropriate or not.

Keywords:   Zeno, third movement, mistake vs. disobedience to reason, anger, Seneca, appropriate, oscillating, shock

There is a problem. So far I have presented Chrysippus as identifying emotions with (mistaken) judgements of reason. But his predecessor, Zeno of Citium, takes a view which seems scarcely compatible, and according to Galen, Chrysippus endorses this view.

Zeno's Incompatible View

Zeno says that emotion involves not a mistake of reason, but actual disobedience (apeithēs, Latin non obtemperans) to one's own reason.1 Ancient and modern commentators have sought to play this down, either by glossing disobedience to reason as if it meant no more than acting against right reason, i.e. against a standard which one may not know,2 or, with Pohlenz, as meaning that emotions are reason‐less impulses.3 But disobedience implies going against one's own better judgement and Chrysippus himself insists on this. In his own words, Chrysippus distinguishes Zeno's disobedience to reason and his related perversion or turning away from reason (diastrophē, Latin aversa)4 from merely being misled by error (diēmartēmenōs pheresthai), from overlooking something while in accordance with reason (paridōn ti kata ton logon), from making a slip (esphaltai), and (p.56) from reasoning badly and implausibly (kakōs en tōi dialogizesthai, enantiōs pros to eulogōs).5

Galen supplies examples of such a mere mistake, probably taken from Chrysippus. One is thinking pleasure good,6 another is scorning your children's life to save your country.7 The last reference is to Agamemnon sacrificing his daughter so that the Greek fleet could sail to Troy. The examples are of erroneous value judgements, and it is precisely erroneous value judgements that, in Chrysippus' standard account, constitute emotion. But on Zeno's account more is needed for emotion—an actual disobedience to reason. Galen gives an excellent illustration of disobedience, again probably drawing on Chrysippus, namely the story of Medea, who slew her children to take vengeance on her husband Jason. Unlike Agamemnon, she says she understands how bad the result of her deed will be.8 So her judgements are partly true, which increases still further the contrast with Chrysippus' analysis of emotion in terms of mistaken judgement. What Medea lacks is not true judgement, but the ability to abide by it in the face of other judgements which the Stoics would consider false. Zeno seems further to diverge from Chrysippus' standard account in that he will not be able to class as emotion Agamemnon's desire to avenge the Greeks and his fear that they would not be able to sail, unless Agamemnon is rebelling against his own better judgement.

The very different account of emotion in Zeno, as disobedience rather than mistake, explains another set of ideas which come into the Stoic discussion. Chrysippus describes emotion as involving akrasia, which Seneca renders in Latin by impotens.9 The literal meaning of akrasia is ‘being out of control’, but Aristotle had made it the standard term for going against one's better judgement.10 It is Zeno's rather unexpected idea that all emotion does involve such disobedience to one's own reason, and so Chrysippus is merely following him in making all emotion involve akrasia.

There is more. Zeno said that emotion is a fluttering (ptoia) of the soul.11 Indeed, Chrysippus and other Stoics consider all emotion to (p.57) be a fluttering.12 This is not merely a verbal borrowing from Plato, who used the term.13 Emotion is said to involve for the Stoics an oscillation of mind, too rapid to notice.14 There is an obvious motive for the oscillation view. Emotion for Zeno and Chrysippus gives one opposed thoughts. But Plato had argued that opposed thoughts cannot be entertained in a unitary soul at the same time.15 For Chrysippus the seat of emotions is unitary, but the idea of an oscillation too rapid to notice avoids the thoughts being entertained at the same time. It fits the case of Medea, who speaks now of the supposed good (revenge), now of the bad, in what she is going to do. The problem is that Agamemnon is not presented as having opposed thoughts, so that the oscillation requirement seems strangely irrelevant to him.

I should acknowledge an excellent interpretation which could solve this first difficulty, if it were an isolated one. It has been suggested that Chrysippus calls emotion disobedient to reason only in this sense, that when right reason is offered to people in an emotional state, they refuse to follow it.16 This is indeed the point that is made by Chrysippus in several passages.17 When we look at some of Chrysippus' further statements, however, I doubt if we can take it that this is all he means.

How did the apparent conflict of theories in Chrysippus come about? We are told that many definitions of the emotions were pronounced orally by Zeno (eirēmenoi) and then written out (gegrammenoi) by Chrysippus.18 We are further told that Chrysippus wrote about (graphein huper) certain definitions of emotion,19 which we know to be Zeno's, and that he was an exegete (exēgeisthai) of these definitions.20 At one point Chrysippus is quoted as referring to the irrational movement which ‘he’ describes, where he is presumably Zeno. Perhaps Chrysippus originally set out merely to expound Zeno. But unfortunately he gives at least the impression of endorsing Zeno's definitions, as when he says that to call emotion contrary to nature is not inappropriate, because all such movements (the reference (p.58) of ‘such’ is unclear) are disobedient (apeithēs) to reason and turn away from it (apostrephesthai), rather than merely reasoning badly and implausibly.21 Again, he says that it has been appropriate (oikeiōs) to apply the term ‘fluttering’(ptoia) to the whole class of emotions,22 adding that this is in respect of the agitated and random movements. Certainly, Galen is sure that Chrysippus was endorsing these definitions of emotion. It is no good, I shall argue in Chapter 6, pleading that they are compatible with Chrysippus' own account of emotion as mistaken judgement. The best defence of Chrysippus would be to argue that he was only expounding, not endorsing, Zeno's definitions. But Posidonius, we shall see, does not think so.

There are still further clashes between Chrysippus' conception of emotion as mistaken judgement and Zeno's definitions as expounded by Chrysippus. For Chrysippus says that Zeno defines emotion as irrational in the sense of being without reason and judgement (khōris logou te kai kriseōs).23 This is not compatible with Chrysippus' idea that emotions are (mistaken) judgements of reason, and Galen, taking him to be endorsing both ideas, accuses him of self‐contradiction.

Galen was writing four hundred years later than Chrysippus, but he was drawing on earlier sources. He may well have used handbooks of Stoic self‐contradictions,24 and he certainly drew on the Stoic Posidonius. Posidonius, a mere hundred years after Chrysippus, makes two complaints against him on this issue. First, he takes Chrysippus to hold that people are always carried away by emotion into turning away (apostrephesthai) from reason, and objects that this fails to capture all the cases. It is important that not only Galen, but also Posidonius, made this objection. Chrysippus, the complaint is, fails to explain variations in emotional reaction. Why do some people experience emotion, think it appropriate to react and reject reason, (p.59) when others do not?25 Someone might say, on Chrysippus' behalf, that it depends on an estimate of size—how big the good or evil appears—or on weakness of soul. But the expected correlations, he complains, are simply unavailable. The central example is that when Agamemnon was afraid the Greeks would be defeated by the Trojans, he suffered from weakness of reason and thought himself confronted with a great evil. But none the less he did not reject reason, but actually sought the rational advice of Nestor. Or if it be urged that he was then no longer afraid, it can still be asked why some people do, and some do not, reject reason, though they have the same judgements and the same weakness of soul.26

Posidonius' second complaint is that Chrysippus had supposed appetite can (he does not say ‘must’) involve holding the conflicting beliefs that something is not advantageous or beneficial (mē sumpheron, asumpheron, ophelos), and yet that one should take it (hekteon, axion antekhesthai, dein lambanein). One should take it, he adds, as a great advantage, benefit and good. This second complaint is quoted by Galen as follows:

Next, Posidonius adds this. ‘But the idea of a person's not merely turning away from reason in cases of appetite, as he [Chrysippus] says, but supposing in addition that, even if something is not advantageous (mē sumpheron), it is still to be taken (hekteon), involves a conflict: moving as if to a great advantage and because of its greatness, even if it is disadvantageous (asumphoron), and thinking it worthy of its greatness to cling to it in this way, even if it brings no benefit (ophelos), but the reverse. Let us grant that those who say it is not advantageous are turning away from reason, and that those who announce they will prove it is disadvantageous are thought to be silly, and that what is pursued is so pursued because of its great benefit. Still, it is unconvincing that, because of supposing it a great good, one thinks it necessary to take it, even if it is the greatest evil, declaiming, “Let me perish; this is now to my advantage.’ ”27

Galen later quotes a passage of Chrysippus which both confirms part of Posidonius' report and will prove very relevant to Seneca's revision of Chrysippus. Chrysippus says:

And that is why we can hear utterances like this too in the case of lovers and people with other strong appetites and of angry people. They want their spirit (thumos) gratified, and they want to be left alone, whether it is better or not (eite ameinon eite mē), and they want us to say nothing to them. And (p.60) this has to be done (poiēteon) come what may (ek pantos ge tropou), even if they are making a mistake and it is disadvantageous (asumphoron) to them.28

A very similar view is ascribed to the Stoics generally, rather than to Chrysippus in particular, in a passage of Stobaeus already mentioned. Emotion is described as all involving disobedience (apeithēs) to reason, and it is added that often (pollakis), though not always, the disobedience is conscious. For people in an emotional state often see (horān) that it is appropriate not to do what they are doing. The same description is given of these people as Chrysippus had used, that they are carried away (ekpheromenoi). It is added that when they learn or are taught to realize (mathein, metadidaskesthai) that it is not appropriate to feel grief or fear, they still do not give up.29

Chrysippus' discussion can be further illustrated from the following quotations in Galen.

Indeed, he [Chrysippus] clearly disagrees with himself not only in these respects, but also when he writes about the definitions of emotion, calls it an irrational movement of the soul contrary to nature and an excessive impulse, and then, in explication of ‘irrational’, says that what is meant is, ‘without reason and judgement’(khōris logou te kai kriseōs). He also adduces as a model for excessive impulse people who are running hard. Both of these conflict with emotions being judgements. We shall recognize this more clearly if we quote his very words. One passage goes like this: ‘First one must keep in mind that a rational animal is by nature such as to follow reason and to go by reason as a guide in action. But often it moves towards or away from things in a different direction, in disobedience (apeithōs) to reason, when pushed too much. Both definitions apply to this movement: movement contrary to nature happens with this kind of irrationality, and so does excess in impulses. For ‘irrational’ here must be taken as disobedient (apestrammenon) reason. It is in application to this movement that in ordinary usage, too, we say that some people are pushed and move irrationally without reason and judgement (aneu logou [kai]kriseōs). We are [not] referring to the case of a person who is misled by error (diēmartēmenōs pheretai) and overlooks something in accordance with reason (paridōn ti kata ton logon), but are speaking above all in application to the movement he [Zeno?] (p.61) describes, when it is not in the nature of a rational animal to move in its soul in this way, but in accordance with reason’.30

Indeed, he [Chrysippus] says in addition, ‘Such states are the sort that are out of control (akrateis), as the people are not in control of themselves (ou kratountōn heautōn), but are carried away (ekpheromenōn), as those who run at full stretch are carried away and are not in control of such a movement.’31

Seneca's Third Movements As a Harmonization

What I now want to suggest is that Seneca's On Anger, written a century after Posidonius, represents the earliest extant attempt to reconcile the two conceptions of emotion. The reconciliation takes the form of treating the disobedience to reason as a chronologically later stage in emotion which follows the mistaken judgements. Moreover, because it would not be plausible that everyone in a state of anger finishes up by disobeying reason in Medea's way, Seneca also has to give disobedience to reason a rather different form. There is no trace of these moves in the surviving fragments of Chrysippus himself. They look like a later device by Seneca or his source.

What Seneca does is to distinguish at least three stages or ‘movements’ in anger, of which the second and third are the relevant ones here.32 In the second stage the mind assents to the appearance of injustice and there is a will to the effect that it is appropriate (oportet) for me to be avenged, or for him to be punished. The third stage is introduced as one in which emotions are carried away (efferantur). At the second stage there was a moral mistake of reason, but in this third stage one wills (vult) in disobedience even to that erroneous application of reason. For one goes on to decide, ‘I must be avenged, come what may (utique)’, and omits the ‘if it is appropriate’(si oportet) which one's reason had a moment earlier recognized.

Seneca's description of this third stage as overcoming reason(evicit rationem) corresponds to Chrysippus' talk of disobeying reason and turning away from it. His ‘uncontrolled’ (impotens) corresponds to Chrysippus'akratēs. His ‘come what may’, omitting ‘if it is appropriate’, corresponds to the description given by Chrysippus of lovers and angry people who want to be left alone ‘whether it is better or not’, and who say that they do what they are doing ‘come what may (ek pantos ge tropou), even if they are wrong and it is disadvantageous (p.62) to them’.33 Finally, his ‘carried away’(efferantur) represents Chrysippus'ekpheromenoi.34

By splitting the process up into a second and third stage, Seneca shows how we can accommodate both of the two conceptions of emotion without contradiction. The mistaken beliefs that there is good or bad at hand and that it is appropriate to react come at the second stage. The disobedience to your own reason follows, but it does not take the form it has in Medea of going against a clear recognition of the case against revenge. Rather it takes the form of rejecting that residual appeal to appropriateness (oportet) which was preserved even by an erroneous application of reason at the second stage. Seneca enables us to answer Galen's question how emotion can be ‘without reason and judgement’, if it consists of judgements of reason. The answer which emerges from Seneca is that in the third stage anger lacks the judgement which reason had earlier supplied with its reference to appropriateness. There is, however, still a judgement operating at the third stage, because willing (vult) is conceived, as we saw in Chapter 2, as a kind of judgement. Only the judgement is now something like ‘I must be avenged, come what may.’35 There is a small oddity because, strictly speaking, judgements which constitute willing ought themselves to be judgements of appropriateness: it is appropriate to be avenged ‘come what may’. If Seneca had noticed this, it would simply have enhanced his account of disobedience to reason, for such a judgement would be even more obviously at odds with the earlier judgement ‘It is appropriate to be avenged, because I have been injured.’ Seneca's account even leaves room for the possibility of an oscillation or fluttering. But it will be an oscillation not like Medea's between a correct and an incorrect judgement, but between an incorrect judgement and an even more incorrect judgement.

Seneca's postulation of a third movement is very true to life. An observer can even time the shift from ‘revenge is appropriate’ to ‘I must be revenged’. Modern psychologists have observed the same phenomenon,36 and it bears some relation to what they call ‘flooding’.

(p.63) On one interpretation, Seneca makes a concession at On Anger 2. 3. 4, that disobedience to reason does not quite always follow the initial mistaken judgements (putavit, voluit), provided those judgements are retracted at once. But (on this interpretation) the account of emotion remains unaffected because the mistaken judgements will not then be called ‘emotion’, so that for emotion disobedience to reason becomes a conceptually necessary condition. It is more likely, however, that putavit and voluit are used loosely for mere appearance, so that the point is only the familiar one that appearance is not to be called emotion. For this is the point to which Seneca immediately returns in 2. 3. 5, with no hint that he has just introduced a fresh innovation, and opinio was used with similar looseness in 2. 2. 2.

Seneca's distinction of second and third movements runs as follows:

In order that you may know how emotions (adfectus) (1) begin, or (2) grow, or (3) are carried away (efferri), (1) the first movement is involuntary (non voluntarius) like a preparation for emotion and a kind of threat. (2) The second movement is accompanied by will (voluntas), not an obstinate one, to the effect that it is appropriate (oporteat) for me to be avenged since I am injured, or it is appropriate for him to be punished since he has committed a crime. (3) The third movement is by now uncontrolled (impotens), and wills (vult) to be avenged, not if it is appropriate (si oportet) but come what may (utique), and it has overthrown (evicit) reason.37

I have been arguing concerning Seneca's third movements that Seneca is defending Chrysippus. Before I come to his first movements, I should finish explaining how I see Zeno's relation to Chrysippus.

Chrysippus' Departures From Zeno

I have claimed there is a conflict between Chrysippus' main account of emotion and Zeno's, that is, between mistaken judgement and disobedience to recognized truth. In a way, I think Pohlenz was right when he said there was a big difference between Zeno and Chrysippus, but his claim got discredited because he selected the wrong difference. He thought that Zeno distinguished an irrational part of the soul, as Posidonius claims, like that in Plato's Republic, and assigned emotional impulse to that.38 But it is not clear Posidonius (p.64) is right, since, as others have pointed out, even Galen acknowledges that Zeno's meaning is uncertain, and he inclines to think that Zeno sided neither with Plato's division of the soul in the Republic into rational and irrational parts, nor with Chrysippus' rival insistence on the unity of the seat of emotions.39 Zeno may rather have followed Plato's earlier view in the Phaedo, in which a unitary soul is seen as opposing the body, rather than manifesting internal divisions.40 This still differs from Chrysippus, who gives no such role to the body.

There probably is, then, a divergence between Zeno and Chrysippus, even if not the one Pohlenz suggested. And there are still further differences too. For one thing, the ideal of freedom from emotion (apatheia) will mean different things, if ‘emotion’ means different things. Again, Chrysippus was pioneering when he made emotions identical with judgements. For this there had been almost no parallel, except for Plato's suggestion recorded in Chapter 1 above, and confined to the case of fear—that fear is expectation of evil.41 Zeno does give a role to judgements in emotion, but the role is not one of identification. We saw in Chapter 2 that for Zeno distress and pleasure occur on the occasion of (epi) judgements.42 Such a role for judgements was already commonplace before Zeno, as seen in Chapter 1. Zeno does not even regard erroneous judgements as a sufficient cause of emotion. Not only have we seen this in the case of Agamemnon sacrificing his daughter, but Zeno spotted the point, which I shall address in Chapter 7, that the emotion of distress may fade over time, even though the judgement persists that you have been harmed.43

Zeno and Chrysippus differ in how they relate the various ingredients in emotion: impulse (hormē), judgement and, thirdly, contraction and expansion. They seem to agree that emotion is an impulse. For Chrysippus, appetite and fear at least are impulses to avoid or reach for something. Zeno characterizes all emotion as excessive impulse (pleonazousa hormē),44 and Chrysippus explains the excessiveness in terms of disobedience, turning away from reason, (p.65) akrasia, and being carried away.45 But Chrysippus intellectualizes his account by identifying impulse with judgement—there is no sign of this in Zeno—and by making contractions and expansions a mere concomitant. Zeno, in contrast, is said, in passages already referred to, sometimes to define pleasure and distress as being expansions and contractions on the occasion (epi) of judgements.46 Emotions are for Zeno also excessive impulses, but how they can be impulses as well as expansions or contractions is never clarified, since later Stoics at least did not consider expansions and contractions to be impulses. Zeno's impulses are likely to occur, like the emotions with which they are identified, on the occasion of judgements, rather than in Chrysippus' manner actually being judgements.

Chrysippus also intellectualizes, I shall argue in Chapter 7, when he tackles Zeno's problem about distress fading while judgements persist. I believe he argues that one at least of the judgements will have faded too. To argue this, he has to distinguish more than one judgement as being involved in an emotion, and we saw in Chapter 2 that that is indeed another of Chrysippus' innovations. Chrysippus also appears to have innovated in distinguishing eupatheiai, the sage's good states of feeling, from pathē, emotions, in spite of their having the same judgemental structure, and in introducing the idea of wanting with reservation.

Chrysippus being such a radical innovator should not blind us to the fact that Zeno was in his time an innovator too. But that is something I am describing in another publication.47


(1) Stobaeus 2. 88. 8 Wachsmuth; Cicero Off. 1. 136(SVF 1. 205).

(2) Stobaeus 2. 89. 14–16, Wachsmuth, on ‘contrary to nature’; on ‘disobedient to reason’ see Inwood Ethics and Human Action, 156, 158; Annas, Hellenistic Philosophy of Mind, 105. David Sedley has drawn my attention to Plutarch's gloss on ‘irrational’ as ‘contrary to the dictates of reason’(para ton hairounta logon). But this too would imply a wholly unconscious mistake of reason, not disobedience.

(3) Pohlenz Die Stoa, vol. 1, pp. 141–53, with nn. in vol. 2; id., ‘Zenon und Chrysipp’, Nachrichten der Gesellschaft der Wissenschaften zu Göttingen, phil.‐hist. Kl., Fachgruppe 1. 2 (1938), 173–210, repr. in his Kleine Schriften, vol. 1 (2 vols., Hildesheim, 1965), 187–99.

(4) Themistius In DA 107. 17–18(SVF 1. 205); Cicero Tusc. 4.11; 4.47 (SVF 1. 205).

(5) Galen PHP 4.2.12 and 24; 4.4.21 and 23; 5.4.14, pp. 240–2, 254–6, 316 de Lacy.

(6) Ibid. 4. 4. 29, p. 256 de Lacy.

(7) Ibid. 4. 2. 26–7, pp. 242–4 de Lacy.

(8) Ibid. 4. 2. 27, p. 244 de Lacy.

(9) Ibid. 4. 4. 24, p. 256 de Lacy; Seneca On Anger 2. 4. 1.

(10) Aristotle, esp. Nicomachean Ethics 7. 1–10. It merely reflects a shift of context that the Stoics are writing about emotion, Aristotle about action, being against one's better judgement.

(11) Stobaeus 2. 39. 8 Wachsmuth (SVF 1. 206).

(12) Ibid. 2. 88. 11–12 Wachsmuth (SVF 1. 206); Galen PHP 4. 5. 6, p. 260 de Lacy.

(13) Plato Republic 439 D.

(14) Plutarch On Moral Virtue 446 F–447 A (SVF 3. 459), translated in ch. 18 below.

(15) Plato Republic 436 B–439 E, partly translated in ch. 20.

(16) John Cooper, ‘Posidonius on Emotions’, in Juha Sihvola and Troels Engberg‐Pedersen (eds.), The Emotions in Hellenistic Philosophy (Dordrecht, 1998), 71–111. An earlier version was summarized by Annas, Hellenistic Philosophy of Mind, 118–20.

(17) Galen PHP 4. 6. 27, p. 274 de Lacy; Stobaeus 2. 88. 8–2. 90. 6 Wachsmuth.

(18) Galen PHP 4. 7. 2, p. 280 de Lacy.

(19) Ibid.4.2.8, p.240 de Lacy.

(20) Ibid. 4. 2. 13 and 19, pp. 240–2 de Lacy.

(21) Galen PHP 4. 4. 17, p. 254 de Lacy.

(22) Ibid.4.5.6, p.260 de Lacy.

(23) Ibid. 4.2.8; 4.2.12; 4.2.24; 4.4.5; 4.5.4, pp. 240–2, 250, 260 de Lacy.

(24) At PHP 4. 4. 1, p. 250 de Lacy, Galen says he may put Chrysippus' self‐contradictions into a single treatise, if he gets the time. Von Arnim, SVF vol. 1, preface, pp. xiii–xiv, suggests that Galen was thinking of drawing on a handbook. Teun Tieleman (Galen and Chrysippus: Argument and Refutation in the De Placitis 2–3, 140–1) confirms this conjecture by pointing out that the contradiction introduced at 4. 4. 1 is derived from several treatises by Chrysippus not otherwise mentioned by Galen. Since the contradiction is also not in Plutarch's treatise on Stoic self‐contradictions this supports von Arnim's view that Galen was not drawing on Plutarch. Von Arnim thought that both Plutarch and Galen were drawing on an earlier treatise. On the other hand, Boys‐Stones has offered evidence that Plutarch's is the earliest treatise on Stoic self‐contradictions, followed by that ascribed to Taurus at Aulus Gellius 12. 5. 5. See ‘Plutarch on koinos logos: Towards an Architecture of De Stoicorum repugnantiis’, Oxford Studies in Ancient Philosophy, 16 (1998), 299–329.

(25) Galen PHP 4. 5. 26–44, pp. 264–70 de Lacy.

(26) Ibid. 4. 5. 40–1, p. 268 de Lacy.

(27) Ibid. 4. 5. 42–3, pp. 268–70 de Lacy.

(28) Ibid. 4. 6. 27, p. 274 de Lacy.

(29) Stobaeus 2.88.8–2.90.6 Wachsmuth, translated in LS 65A. There is reference also to fluttering (ptoia), to rejecting reason (apostrephesthai), and to the state not being like deception (exēpatēmenoi), although the example given of deception (believing that the atoms are first principles) is not a moral one. The further suggestion of the passage, that the people described are carried away ‘as if by a disobedient horse’, may represent Posidonius' addition. But Christopher Gill has suggested to me that it may be already anticipated by Chrysippus at PHP 4. 2. 27, p. 244 de Lacy.

(30) Galen PHP 4. 2. 8–12, p. 240 de Lacy.

(31) Ibid.4.4.24, p.256 de Lacy.

(32) Seneca On Anger 2. 4. 1.

(33) Galen PHP 4. 6. 27. I am grateful to Miriam Griffin for this last confirming comparison. Cf. Epictetus Diss. 3.24.23; 4.1.67, ex hapantos.

(34) Galen PHP 4. 4. 24, p. 256 de Lacy; Stobaeus 2. 89. 17 Wachsmuth.

(35) My interpretation is different from that of P. L. Donini, ‘Pathos nello stoicismo romano’, Elenchos (1995), 195–216.

(36) See Daniel Goleman, Emotional Intelligence (London, 1996), esp. 60–7 (first published 1995).

(37) Seneca On Anger 2. 4. 1.

(38) Pohlenz, Die Stoa, 1. 141–53, with nn. in vol. 2; id., ‘Zenon und Chrysipp’, 187–99. Pohlenz relies on Posidonius as recorded by Galen PHP 5. 6. 33, p. 332 de Lacy.

(39) Galen PHP 5. 6. 42, p. 334 de Lacy; see Inwood Ethics and Human Action, 292 n. 14; Ian Kidd, ‘Zeno's Oral Teaching’, in Dory Scaltsas (ed.), Zeno of Citium (forthcoming).

(40) See Teun Tieleman, ‘Zeno of Citium and Psychological Monism: The Evidence Reconsidered’, in Scaltsas (ed.), Zeno of Citium (forthcoming).

(41) Plato Protagoras 358 D; Laches 198 B; Laws 644 C–D.

(42) Galen PHP 4.3.1–2; 5.1.4; cf. 4.2.5–6, pp. 246–8, 292, 240 de Lacy.

(43) Cicero Tusc. 3. 75; Galen PHP 4. 7. 3–5, p. 280 de Lacy (SVF 1. 212).

(44) Diogenes Laertius Lives 7. 110; Stobaeus 2.39.5; 2.44.5; 2.88.8 Wachsmuth (SVF 1. 205–6).

(45) Galen PHP 4.2.8; 4.2.14–18; 4.4.24; 4.5.10–15; 4.6.35, pp. 240–2, 256, 260–2, 276 de Lacy.

(46) Ibid. 4.3.1–2; 5.1.4; 4.2.5–6, pp. 246–8, 292, 240 de Lacy.

(47) Richard Sorabji, ‘Zeno and Chrysippus on Emotion’, in Scaltsas (ed.), Zeno of Citium (forthcoming). Zeno prefers the four emotions often picked out by Plato as the basic ones, and gives a unitary account of them by making them all impulse, and by making their cause to be judgement, not mere appearance. If we can trust Cicero, Ac. Post. 1. 38 (SVF 1. 207), he concluded from the role of judgement that emotion was voluntary. His spotting of the problem of emotion fading faster than judgements is noted above. Zeno is also credited with introducing the idea that everything is indifferent except character and rationality: Stobaeus 2. 57. 18 Wachsmuth (SVF 1. 190). Some things are naturally preferred or dispreferred and are to be ‘selected’, though not ‘chosen’: Stobaeus 2. 84. 21 Wachsmuth; Cicero Ac. Post. 1. 37(SVF 1. 192–3).