Aspasius is an Aristotelian who criticizes the Stoicizing definition of emotion given by an earlier Aristotelian Andronicus. Aristotle for his own purposes in Rhetoric book 2 had been indifferent whether he defined emotions in terms of beliefs appearances or thoughts. But the Stoics had insisted that beliefs (or judgements) were required and that mere appearances were not enough. Andronicus may have been making a concession to them when he defined emotions in terms of supposition (hupolēpsis) as an irrational movement of the soul through the supposition of bad or good.1 For Aristotle had defined supposition as including three species scientific understanding practical understanding and belief (epistēmē phronēsis doxa).2 It is belief which is relevant here.
PART I: Emotions as Judgements versus Irrational Forces
9: Aspasius and Other Objections to Chrysippus
So far the main objections to Chrysippus’ judgemental analysis of emotion have been those of the Stoic Posidonius reinforced at certain points by the Middle Platonist Didaskalikos. But Posidonius’ are not the only objections to Chrysippus’ judgemental analysis of emotion. Some further possible objections have been canvassed in the examination of Posidonius and others putatively forestalled in the exposition of what Chrysippus meant. But there are other objections again and I shall start with those of Aspasius.
Aspasius vs. Andronicus on appearance vs. belief
Aspasius however contests this. The mere appearance that there is something good can arouse emotion without any assent to turn it into a belief or supposition. This is especially clear he says in the case of appetites for pleasure for example in lust or when we are amused by a witty speech which can amuse us even though we do not suppose there is anything good at hand from the speech. Some people we may add can even laugh at their own reverses although that is not something that fits well with the tradition to be discussed in Chapter 19 of connecting laughter with a sense of superiority.
Aspasius goes further. The appearance that something is pleasant may be more relevant in lust and laughter than the appearance that something is good. And he uses this to support his own definition of emotion as the irrational part of the soul being moved by the pleasant or the distressing and his idea that pleasure and distress are the two most generic emotions. Here is the passage:
When Andronicus says that emotion (pathos) is due to a supposition (hupolepsis) of good or bad first perhaps he did not know that certain emotions are generated simply by appearances (phantasia) without any assent (sankatathesis) or supposition. When something appears pleasant or painful simply through sense perception an emotion sometimes results in the soul. So emotions occur not only with supposition but also in advance of supposition. The appetites (epithumiai) reveal this especially. For one often gets an appetite for something lovely just because it is lovely on the mere sight of it without supposition ever having come first. Again people often get an appetite without any supposition occurring that some good is present as when the irrational part of the soul is stirred by a witty speech. For we do not then suppose that some good is present for us but are just moved by pleasure. Sometimes pleasure does follow on an actual supposition of good and distress on a supposition of bad evidently because the soul is moved on the basis of the good being pleasant and the bad distressing. Perhaps then emotion is the irrational part of the soul's being moved by the pleasant or distressing. For whether emotion follows after appearance or after supposition in any case it supervenes on the pleasant and distressing. This indicates that pleasure and distress are the most generic emotions.3
How many generic emotions?
Aspasius raises one more major question and several minor queries. The major question is whether the most generic emotions under which the others can be classified are six in number or as the Stoics said four or as he himself preferred just two pleasure and distress an answer he finds in Plato.4 This answer is also ascribed to Plato in the Middle Platonist Didaskalikos5 and elsewhere to ‘the ancients’.6 But the choice of two seems strange for an Aristotelian. For in book 2 of the Rhetoric where Aristotle devotes ten chapters7 to defining (depending on how one counts) about ten emotions he recognizes three possible genera: distress pleasure (although in the end no examples are given of this serving as a genus) and something Aspasius neglects: desire. Though an Aristotelian Aspasius seems unaware of this discussion.8
When Aristotle talks of desire here in classifying anger (orgē) as desire accompanied by distress he uses his most general word for desire orexis9 and in classifying love and hate he appeals to wishing (boulesthai).10 Aspasius does not notice that he has reversed Aristotle's definition of anger as desire accompanied by distress when he defines it as a kind of distress accompanied by desire.11 And he violates Aristotle's explicit warning12 that hate does not imply distress.
Aspasius does make some concessions to the role of desire not only that anger is at least accompanied by desire but that gratitude (kharis) is desiderative (orektikē) and that pleasure merely has to occur together with (hama) the desire. Nevertheless he concludes that gratitude is to be classified not under desire but under pleasure.13
Still more embarrassing is Aspasius’ recognition that some instances of appetite and anger involve pleasure some distress and some both. So one cannot classify anger or appetite as a whole as falling under one of his two genera either under pleasure or under distress.14 He sees it does not help that some instances fall under the one or the other. In a classification of kinds the entire species has to fall under the genus. This problem suggests the need for a third genus under which to classify anger and appetite most obviously Aristotle's genus of desire. But Aspasius never tackles the problem.
So much for Aspasius’ divergence from Aristotle. What now of his disagreements with the Stoics? They treat all emotions as desires using for desire the broad term hormē (impulse). But for one of their four main kinds of emotion they use the term ‘appetite’ (epithumia) which had been understood by Plato and Aristotle to mean a desire for pleasure. Aspasius rebukes them for using the term on the grounds that anger on their own account is not a desire for pleasure and so not (by Platonic or Aristotelian criteria) an appetite but a desire for revenge.15 One Stoic reply might be that they use ‘appetite’ in a wider sense to cover all irrational desire (alogos orexis) for future things seen as good.16 But Chrysippus does offer one definition of appetite in terms of pleasure17 and that might force him to fall back on the idea of revenge being seen as sweet.
The Stoics differ not only from Aspasius but also from Aristotle in their recognition of fear as a fourth genus. What justifies their choice? In recognizing four genera (pleasure and distress appetite and fear) they are presumably influenced partly by Plato who for his own purposes picks out this list of four emotions no less than six times.18 But they also have a rationale of their own: pleasure and appetite are concerned with good distress and fear with bad; pleasure and distress with the present situation appetite and fear with the future.19 Moreover they offer some seventy-odd definitions of other emotions to show how they can be classified under the generic four.20
Why not generic emotions relating to the past?
The rationale in terms of present and future omits the past and so immediately suggests Aspasius’ remaining question: should there not be six genera? This is not how Aspasius himself arrives at the question. Rather he notices a passage in which Plato lists six emotions by adding anger and confidence two items which he accuses the Stoics of failing to accommodate. But Aspasius does not think Plato listed these as generic emotions merely as particularly familiar ones.21 There is a much more obvious way of pressing on the Stoics a case for six generic emotions. They recognize two emotions directed towards the future two towards the present but none towards the past. Indeed it was precisely by including emotions directed towards the past that a modern author built up a larger list of basic emotions classified by reference to good and bad for self and others in the past present and future.22 The Stoics by contrast treat not only anger as a forward-looking appetite but also metameleia (remorse repentance change of heart) as a form of distress directed to the present.23 Why is this?
It might be replied that the Stoics have good reason to omit emotions directed to the past because it is only if you feel yourself still to be in a good or bad situation because of past happenings that you will have any emotion about those happenings.24
This defence of the omission is not right as it stands. In the case already discussed of a fire breaking out I may feel no emotion during the emergency itself. It is only when I am out of the situation that the full horror of it dawns on me. This is a well-known phenomenon. It means that I have the emotion only when I know myself no longer to be in the bad situation. Admittedly emotion here may be based on consideration of what my present state or for that matter my future state might have been. It would be rarer for it to be based on consideration solely of what my past state might have been as when I reflect that if I had had the illness I feared I would have had to go through a horrible operation in order to regain my present level of health. To that extent the past in isolation may typically have less hold on our emotions.
I shall return in Chapter 16 to the lesser hold of the past. For now I shall only point out that concern about bad past events depends on whether we see them as making the present or future bad rather than on whether they are having an effect on the present or future independent of our perception.
This comparative dissociation from the past is sufficient so it was argued in Chapter 7 to protect Chrysippus at least partly from Posidonius’ objection about emotions fading faster than judgements. In many cases of fading emotion (though not I pointed out in all) Chrysippus could have replied the relevant judgements fade too as his analysis requires along with the emotions. For the judgement that lingers is only the irrelevant judgement that bad was done to you in the past not the relevant judgement that you are still in a bad situation.
I also argued in Chapter 5 that dissociation from the past explained Seneca's view that historical narratives do not arouse real emotion. I dissented only because the qualification ‘once upon a time’ may cease to hold our attention in a good piece of story-telling.
Is action always judged appropriate in appetite and fear?
The Stoics would object to Aspasius that in confining the generic emotions to two pleasure and distress he has omitted those emotions which involve judgements about the appropriateness of action. But Aspasius raises a counter-objection about the relation of anger to thoughts of retaliation. He objects that fathers can be angry with their sons without thinking that revenge would be appropriate at all.25 The same sort of point was to be made later by the Latin-writing Church Father Lactantius about people who are angry with their wife children or pupils.26 And earlier than either thinker the Epicurean Philodemus had written a treatise advocating a kind of anger that seeks correction indeed but not revenge.27 I believe these objections about anger with no desire for action are correct and there are other cases too. In appetite or fear action may not be judged appropriate because it seems hopeless or alternatively immoral. This objection leads on to a more general one and at this stage I shall pass beyond Aspasius.
Will emotion depart when reaction is judged inappropriate?
The more general objection is this. If emotion involves judging that reaction is appropriate would it not become too easy to rid oneself of emotion? For one would have only to drop the judgement that reaction is appropriate in order to be rid of the emotion and is not this implausible? Indeed no one need ever seek help with getting rid of an emotion for in wanting to be rid of it they would already have judged their reactions inappropriate and so should already be rid of it. Wanting riddance is riddance—or so it might seem.
A good answer28 to this version of the problem is that in wanting riddance one may merely judge it inappropriate that one judges reaction appropriate which is not yet to be rid of the latter judgement. But the question still remains: if one does get rid of the latter judgement that reaction is appropriate will one necessarily be rid of the emotion? I think the answer comes out differently for different emotions.
In the case of anger it makes a difference whether angry reaction is rejected as undignified or as counter-productive. The Stoic advice to see how ugly anger makes you trades on indignity and I am not sure how often this works. But if one judges angry reaction counterproductive then one may be rid of two constituent judgements at once not merely of one. Imagine the context of deliberation for example on a committee in which one hears of things one judges very bad. None the less if one is above all committed to reaching a sensible decision and seeking revenge would get in the way of that one may reject seeking revenge as counter-productive. And if seeking revenge is on that account rejected as inappropriate a second judgement will also be rejected for securing revenge will not be seen as a good. This rejection of two judgements increases the chance that one will not get angry at hearing of the bad things. At least in some cases then the judgement that reaction is appropriate appears necessary for anger.
What about pleasure and distress? Here too it is harder than might be expected to confront Chrysippus with cases of emotion in which the judgement of appropriateness is missing. For his two judgements are again linked although in a different way. This time the ‘inappropriate to react judgement’ tends to depend on the ‘not bad’ or ‘not good’ judgement as a prerequisite. This is because the relevant reactions in these cases are involuntary and hidden expansions and contractions. It is not easy to convince oneself that this inner involuntary reaction is inappropriate at least so long as one does not see it harming other people. For overcoming distress therefore it may often be necessary for the other judgement to change the judgement about how bad one's situation is. That may be the easiest way to make the contraction or expansion seem inappropriate.
Admittedly it is not the only way. When Seneca tells the grieving Marcia that she is neglecting her family29 this might make her contractions seem to her inappropriate. It is intended of course to dispel the grief as well. But if it were to achieve only half its purpose discrediting the contractions without dispelling the grief then the approval of contractions would have been shown unnecessary for the persistence of grief.
If we turn finally to the examples of lust and fear we find a different situation again. It is here much easier than with pleasure and distress to convince oneself that it is inappropriate to react. But this is much less likely than in the case of anger to make the emotion absent itself. Augustine we shall see in Chapter 26 gives a most graphic account of how impervious lust can be to the reflection that reaction is inappropriate. The most that could be urged on Chrysippus’ behalf is that in all cases of appetite or fear action (flight for example) would be judged appropriate in the absence of countervailing reasons. I do not think it plausible that in the frightened person there must persist some idea that avoidance is appropriate. They may only wish that the situation could somehow cease to confront them or not even that—they may recognise it as inevitable.
It may be protested that fear is different from appetite for fear admittedly need not involve a desire for avoidance whereas an appetite is a desire and one who desires to act for example illicitly must think the action appropriate in some way e.g. as exciting however inappropriate he thinks it overall. But I do not think that recognizing the action would be exciting is necessarily to judge it appropriate in any way.
Do these objections mean that Chrysippus ought to retreat to equating emotion with the single judgement that there is good or bad in the offing? No: Chrysippus cannot afford this for he has four reasons for requiring a second judgement that were explained in Chapter 2.
Judgement necessary causally or as constituent of the concept?
An extra weakness has emerged from the discussion of whether judgements of appropriateness are necessary for emotion. Sometimes they are but it remains to be shown whether they are necessary only causally or as Chrysippus supposes necessary as a constituent of the concept.30 For that further supposition he relies on our intuition and on the difficulty of finding alternative constituents.
Desire as an alternative constituent?
One alternative constituent that has been suggested is desire. Chrysippus agrees that all emotion involves desire (hormē) but this is because the Stoics equate hormē in humans with a judgement of appropriateness as explained in Chapter 2.31 The modern view is that desire is distinct from a judgement of appropriateness. This opens up the possibility of trying to define emotion as a judgement of goodness and badness coupled with a desire to react rather than with a judgement approving reaction. There have been attempts by modern philosophers to define emotions quite generally in terms of desire.32
I have not personally been convinced. Certainly in distress we do not desire a sinking feeling. At most we may wish that things were otherwise. But that is a very qualified sort of desire and need not arise at all if there is no question of their being otherwise. In pleasure we need not necessarily desire the pleasure to be prolonged; we may think its duration just right. We may welcome it while it is there but welcoming is not the same as desiring which implies a lack. There is also the difficulty33 since not all desire is emotional of specifying which desires constitute emotions. So I shall not pursue this option further even though it has provoked much discussion.
Does every decision to pursue or avoid involve emotion?
Chrysippus’ analysis faces another problem: is not his pair of judgements too common? In most cases when we decide to pursue or avoid something we judge something to be good or bad and judge it appropriate to react—not of course with inner expansions and contractions but with pursuit and avoidance. Yet these decisions do not all involve emotion.34 I believe Chrysippus would be prepared to say that they do. According to the theory of indifference which will be more fully explained in Chapter 12 it is very different to judge something good or bad and to judge it as merely preferred or dispreferred. To see it as preferred and appropriate to reach for is merely to ‘select’ it and this attitude as explained in Chapter 2 involves no emotion. But the judgement of good or bad already gives you a certain emotional attachment. Similarly with the judgement that pursuit or avoidance is appropriate you can reduce your liability to emotion or at least to initial shock by adding the qualification discussed in Chapter 2 ‘if God wills’ or ‘if nothing interferes’. So the judgements involved in ordinary decisions to pursue or avoid something may very well constitute emotion. I am not sure that this Chrysippan response should remove our doubts. But at least this much might be conceded to Chrysippus: the evaluative judgements involved in our ordinary deliberative decisions do make us at least liable to emotion in a way that judgements of preferred indifference and desiring with reservation would not.
Good bad and novel?
In seventeenth—and eighteenth-century accounts of the passions there is an interest in the passion of wonder35 omitted from the Stoic lists. It may be for that reason that the definitions of passion speak of the apprehension of something good bad or novel. Unfortunately not all novelty provokes emotion but Chrysippus’ definition as it stands does not cover a sense of wonder and wonder provides a case in which there need not be even an appearance of good or bad.
The Neoplatonist Plotinus would take issue with Chrysippus’ account of emotion on at least two points. First he denies that judgements of good or bad are always sufficient for grief anger or appetite but not because he like Chrysippus desiderates a second judgement that it is appropriate to react. Rather his point is that the reference to judgement does not yet explain the contributions to emotion of body soul and compound since it confines itself to a contribution of the soul.36
A further disagreement concerns not a pathos but a eupatheia although it is something that we would classify as an emotion: the love involved in mystical contemplation:
This contemplation is intellect in love (nous erōn) when it goes out of its mind drunk with nectar. Then it falls in love made simple and transformed into happy feeling (eupatheia) by satiety.37
This love according to Plotinus can involve no judgement because judgement involves a duality which is excluded by mystical union.38 For the Stoics by contrast eupatheia involves judgement just as much as pathos does. In particular this is true of the will involved in eupathic love.
How much of Chrysippus’ analysis survives?
The objections of the last three chapters have been quite enough to show that Chrysippus’ judgemental analysis of emotion will not do as it stands. And yet it still contains a lot of truth. Nearly all emotion involves at least the appearance of good or bad. Typically the appearance is endorsed. Very often there is an additional appearance or endorsement that one of the four reactions picked out by Chrysippus is appropriate. Often this second appearance or endorsement is as necessary to the emotion as the first whether or not the necessity in this second case is merely causal. Moreover these truths are sufficient to ensure the utility of Chrysippus’ account for therapy by showing us what we need to be rid of if we are to free ourselves from unwanted emotion. Only in some cases will Chrysippus’ oversights impair therapy—where his definition does not fit the emotion at all. But for example the question whether the necessity of a given judgement is merely causal will not affect the efficacy of procedures that get rid of that judgement.
None the less there are holes in the judgemental analysis and there will be corresponding failures in therapy. To understand this we need to consider in the next chapter recent neurophysiological research on emotion.
From the book: