Seneca's first movements may have served a further purpose. They may provide the Stoic reply to Aristotle's theory of tragic catharsis discussed in Chapter 1 above. Most interpretations agree that catharsis lightens emotions in the audience by first arousing them whether catharsis is the analogue of purgation by laxative and emetic or of religious purification. If the Stoics think they have a reply this would also explain as regards the later Stoics what has seemed very puzzling why despite their discussion of the theatre they make virtually no explicit reference to Aristotle's brilliant theory. There may possibly be a reference in Diogenes of Babylon although this is a very conjectural suggestion.1 How widely the theory is mentioned outside the Stoic school depends on what we count as catharsis. If any relief of emotion gained by arousing it is to count then there will be plenty of references. But if we are to stick to something reasonably close to Aristotle's idea then I shall argue in Chapter 19 that we cannot find much more than a brief mention in an Epicurean papyrus fragment possibly by Philodemus another in the Neo-Pythagorean Tablet of Cebes and a number of references in the Middle Platonist Plutarch until we come to the extensive discussions in the Neoplatonists.
PART I: Emotions as Judgements versus Irrational Forces
5: The Arts First Movements and Controversies on Drama and Music Aristotle Philodemus and the Stoics
Seneca on Drama and Catharsis
Given the Stoics’ interest in emotions in the theatre we might have expected such a striking theory to be discussed by them again and again. But if catharsis involves the arousal of audience emotion we can find the answer in On Anger 2. 2. 3–6. Here Seneca tells us that the arts do not arouse real emotions in the audience but only first movements. That would be fatal for Aristotle's theory of catharsis if the emotions undergo catharsis through first being aroused. We then see why the later Stoics did not think it necessary to consider Aristotle's catharsis theory any further.
The case against genuine emotion
What is the case against emotion in theatre audiences being genuine? When you see Medea on the stage you do not believe a wrong has really been done; you know it is only an appearance. And you do not believe it would be appropriate to react by clambering on the stage to assault Jason or rescue Medea's children.2 So by Chrysippus’ two criteria your response is not fear or anger or so it may be said.
The Stoics might make this view of audience reaction plausible by pointing to the difference in feel between the case of seeing the enacted threat and seeing that there is a real fire ablaze in the theatre. Not only do we then judge that there is danger at hand and that it would be right to react but correspondingly it feels different. To make the comparison with our attitude to Medea's children fair we should imagine that the fire in the theatre is threatening someone else and that we can do nothing to help.
The case against tragedy provoking real emotion might be supported still further by considering Augustine's discussion of how different pity is in the theatre from pity in real life. We positively welcome there being a situation for us to pity. If we enjoy being aroused in the theatre how can it be real pity we are experiencing let alone real grief or fear? Pleasure from the representation of horrible things is noticed also by Plato and Aristotle.3
How too it has been asked can the response aroused be one of the distressing emotions if it is therapeutic?4 And how can we be feeling real suspense if we feel it each time we hear the same story even though we already know the outcome?5
In the Greek theatre there will have been plenty of occasion for ‘first movements’ if a recent theory about ancient Greek masks is correct. It has been suggested that in early Greek theatre masks acted as resonators for augmenting the many cries that punctuate Greek tragedy: to io aiaiai papapa.6
The case for genuine emotion
There is however a case for the response being genuine emotion at least in many instances. And the Stoics can afford to accommodate both types of case just so long as they can identify the relevant judgements as present when there are genuine emotions. This is not difficult when the emotion aroused concerns the real world rather than the story.
Many of us will have found ourselves at some time in floods of tears at the theatre. But this is sometimes only because we are thinking ‘This could happen (or has happened) to me.’ That we judge would be really bad and reaction is appropriate. The Stoics would not be embarrassed by such examples of emotion since the expected judgements are present although they are about real life not about the enacted situation.
But what about emotions concerning the characters in the plot? The treatment of Medea above has not yet brought out the full situation. In order to do so we must distinguish the different judgements that Chrysippus recognizes in emotion. In distress and pleasure the relevant judgements are that inner contractions or expansions are appropriate and this we can judge even in relation to a fictional episode.
The judgement that pursuit or avoidance is appropriate is more problematic in relation to fiction. But with ghost stories we have a method of avoidance which we may judge appropriate viz. closing the book. Of course we may also transfer our judgement of appropriate behaviour from the fictional to the real world. Thus the ghost story may lead someone to approve avoiding dark corners the erotic story to approve selecting them.
There is another possibility that has been brought out by Paul Woodruff.7 We may well judge that it would be appropriate for Medea's children to be rescued—but by someone in the play. I would add that this is not very different from our situation in real life when we are caught as pitying or fearful but helpless spectators of the peril of others although we still judge action would have been appropriate or would be for other people.
The remaining type of judgement is the judgement of good and bad. The problem with the Medea was that the audience knows nothing bad has really happened. Of course audiences judge that something bad has happened in the play to the fictional character Medea so there is a judgement made. I shall come back to the effect of that qualification ‘only in the play’.
There is also a special way in which our judgements of good or bad may relate to the real world because of Aristotle's point that poetry is more universal than history.8 It may lead us to think ‘How true!’ because it shows the sort of thing that would happen rather than the accidents of what actually happened.
I have so far been talking of fiction but I do not think the situation is very different when we see hear or read historical stories. It is not a difference that inner expansion or contraction can be judged appropriate. Nor is there a present opportunity of intervening. So pursuit and avoidance are unlikely to be judged any longer appropriate although they might have been at the time. The main difference is that historical events can be judged really good or bad. But this difference should not matter to the Stoics for they hold that emotion involves the judgement that things are now or will be bad or good. It is not enough to judge that they were bad once unless the situation is still bad. The qualification ‘once upon a time’ is therefore as important as the qualification ‘only in the play’ and I shall come back to its effect too.
In looking for relevant judgements I have not so far relied on the case in which people start believing the misfortune in the story is real and perhaps even their own. This would give rise to the relevant judgements. But it might have been less common given the distancing effect of ancient Greek theatre production.
So far I have argued that although Seneca denies that response to fictional drama or historical narrative is genuine emotion the Stoics could in many cases find relevant judgements at work and so could have allowed that often there was genuine emotion. What was harder to find was judgements of ‘good’ or ‘bad’ in relation to the plot unqualified by a ‘once upon a time’ or ‘only in the play’. This difficulty is matched by some of the objections already raised to the idea that our reaction to the plot and its characters can be genuine emotion. Let us reconsider those objections.
One objection I spoke about is the difference in feel from real-life emotions even when the real-life fear is for someone else and you know that intervention is impossible. But I am not sure that there is always a difference of feel. When there is not the audience's feeling may become unbearable and then they can reduce it by reminding themselves that this is only a play or only something that happened once upon a time. This suggests that something else besides judgement is important: attention. Whether our response to the dramatic predicament is genuine emotion may depend on how far we are attending to the qualification that the bad event has happened only in the play or only once upon a time.
There remain the objections about reactions to the plot being enjoyable or therapeutic when the real emotions would merely be distressing and the argument from the possibility of suspense despite the story being familiar. There have been attempts both modern9 and ancient to answer the first two difficulties. To take the ancient answers one is Aristotle's theory of catharsis which says that tragedy lightens and pleases us by catharsis of such emotions as pity and fear. Aristotle also mentions many other pleasures involved in art and drama10 as well as this specific tragic pleasure.11
There is a further point. Even in real life the ancients repeatedly noticed people find pleasure in grief in jealousy in anger12 and in such disgusting things as corpses13 although Augustine dismisses the last as satisfying curiosity rather than giving pleasure.14 It was equally noticed as I shall bring out in Chapter 15 that in real life concern with the lot of others can be therapeutic through putting your own situation into perspective.
As regards suspense I believe its character is changed once you are familiar with the plot according to whether you know the outcome will be happy or tragic and for whom. This either makes available a pleasant relief or creates a need to cultivate resignation. In so far as the character of the suspense is not changed by familiarity this may once again be because you are not attending to the known outcome or reminding yourself of it.
I have made three suggestions. Sometimes Seneca is right: the audience experiences only first movements. Sometimes the audience experiences real emotions about things outside the play but that creates no problem for the Stoics since the expected judgements are there. What I think Seneca disallows and disallows wrongly is that there can be genuine emotions about characters in the historical or fictional plot. The stumbling-block to acknowledging this is that the judgement about the situation being bad is qualified by ‘only in the play’ or ‘once upon a time’. It is only an appearance that the actual situation is now bad or will be. But the answer is that attention is important as well as judgement. In so far as you do not attend to the qualifications your mind will be focused on judgements no different from those of real-life emotion. If the resulting emotions are (shorter-lived this is because the qualifications can quickly return to mind.
Seneca on music
So much for Seneca's view that drama and literature arouse first movements rather than real emotion. But will his view apply as he applies it to music?
Posidonius has his own justification for saying that wordless music does not affect judgements as we shall see. But a different justification that might have more initial appeal would be that wordless music does not have a story line as the play Medea does. Consequently there is no obvious subject-matter for one to assess as good or bad or calling for reaction and so no obvious subject-matter for the relevant judgements.
Just as with drama and literature defenders of Chrysippus might attempt two opposite lines of defence. One would be to protest that judgements are indeed involved when we are moved by wordless music. The other would be that suggested by Seneca: admittedly judgements are not involved but then neither are genuine emotions.
The view that music does not directly produce either emotion or judgement has been maintained not only by Seneca but by his Epicurean predecessor Philodemus15 and in modern times by Eduard Hanslick and Peter Kivy.16 Others have maintained one or other of these denials without maintaining both.17 The debate is very well traced by Martha Nussbaum.18
It certainly does need to be recognized that music often expresses emotion without arousing it. We can hear music as sad without being saddened. We may find ourselves exploring the emotion rather than feeling it.19 But the question is whether emotion is sometimes aroused and if so whether judgements are then aroused as well.
There are two parts to Seneca's view: music produces neither emotion nor judgement. He chooses an example in which emotion assent and judgement may well be out of place. The soldier in peacetime may be stirred by the blast of a trumpet and may even reach for his sword. But if he knows there is no enemy he has no cause to assent to the idea that action is called for. Equally we may well feel that there need be no emotion.
But are not emotions sometimes aroused by wordless music? And if they are can we defend Chrysippus’ identification of emotion with judgement by the alternative strategy of identifying judgements? Geoffrey Madell has recently argued we cannot. Music in his view produces emotion but no judgement.20 His excellent discussion however seems to me to support a more complex conclusion. To develop his example we may experience a longing for the music to reach a resolution and something like joy at its doing so or disappointment at its not doing so.21 These emotions which are directly about the music may cause emotions which are not about the music by reminding us of a real-life longing or by awakening a feeling of disappointment or joy directed beyond the music.
When we long for the music to reach a resolution defenders of Chrysippus can resort to the alternative strategy and say contrary to Madell's conclusion that a judgement is involved. We judge such a resolution to be good. Madell has supplied the materials we need for answering the difficulty that wordless music gives us no subject-matter about which to make a judgement. Similarly with the secondary emotions that are not about the music if we can identify an object for them. We may then make a judgement of goodness or badness about the object of our real-life longing or attach our disappointment or joy to an object which we judge to be good or bad.
The cases in which a judgement is likely to be missing are not these but rather those in which we do not recognize the object of our real-life longing22 or in which we cannot attach our further disappointment or joy to a real-life object. In those cases we may as we say feel as if something was good or bad and that may be enough to set up first movements. But feeling-as-if is not judging. These are the cases which threaten Chrysippus’ identification of emotion with judgement. For although judgement is absent it looks as if emotion is present. We cannot plausibly re-employ the device we used above contrasting our reaction to Medea with our reaction to a fire in the theatre in order to show that our responses to music are not emotions in this kind of case. They feel too like emotions.
We shall find cases of every type cases where emotion and judgement are both present both absent or—and this is the case which hurts Chrysippan Stoicism—one present one absent. Posidonius is looking for an example of the last when he says that wordless music can change emotion without changing judgements. When his example is repeated by Augustine and others we find that the emotion first aroused and subsequently calmed in Posidonius’ story was lust. The youths in the story were trying to break down the door of a chaste woman.23 But this emotion Chrysippus would happily argue involves judgements of sex as a supposed good to reach for. Indeed on Stoic principles the action as well as the emotion requires a judgement of appropriateness.24 So although I believe Posidonius is right to cite wordless music because it does provide examples of emotion aroused without judgement his own example is not the right one to show this.
On the other hand if emotion is aroused as we shall see had been suggested by Damon Plato and the Stoic Diogenes of Babylon merely by some kinship between the music and an emotion it is less likely that an object for judgement and assent will be identified.25 Diogenes does not further analyse the kinship. But examples could be music imitating a sad voice or the gay beat of a dancing rhythm.26 In these cases we may merely feel as if there was something good or bad and as if inner reactions would be appropriate. Certainly we have no good reason to assent that this is really so.
I conclude that Chrysippus can sometimes be defended by Seneca's strategy of showing emotion and judgement both absent sometimes by the alternative strategy of showing them both present. But there is a residue of cases which support Posidonius in which wordless music arouses emotion without judgement. In Chapter 10 I shall discuss the possible neurophysiological basis of this phenomenon. But I should say straight away that we might reasonably expect a neurophysiological basis because of what others have referred to: the universal cooing exchanges with infants long before they have words and the infant's ability to learn intonation also before words.27
Theories designed to make cause and effect alike
The problem of how music can cause emotion has been made very much harder in both ancient and modern times by the erroneous assumption that cause and effect must be like each other.28 This assumption has created a difficulty for those who think about wordless music which involves no judgements or appearances when they confront the evidence that emotion does involve judgements or appearances. How of two things so unlike can one be the cause of the other? The problem I believe is a false one because the causal assumption is false. None the less it has provoked much the same permutations in the ancient and in the modern solutions.
The erroneous causal assumption is shared both by the Stoic Posidonius and by his near contemporary the Epicurean Philodemus. I shall postpone until Chapter 8 justifying my interpretation of Posidonius. and for now I simply assume it. I shall assume inter alia that the argument is Posidonius’ own even though it is voiced by Galen. I believe Posidonius supposes that cause must be like effect. Both in his argument about music and in the immediately following argument about imagination29 he takes it that the irrational affects the irrational and the rational the rational. He is quite deliberately talking about wordless music: rhythms and scales. This does not incorporate rational judgement. Consequently in his view the emotions it arouses (epegeirein) or calms cannot themselves consist of rational judgements. The same view as Nussbaum shows has been taken in the modern literature.30 The passage runs as follows:
Why was it for heaven's sake—I shall put this question too to the followers of Chrysippus—that when Damon the musician came up to a woman playing the Greek oboe (aulein aulētris) in the Phrygian mode to some young men who were drunk and doing frantic things and ordered her to play in the Dorian mode they immediately stopped their frenzied antics? For surely they are not taught to revise the beliefs of the rational element by the oboe music. Rather they are aroused (epegeiresthai) or calmed in respect of the emotional element of the soul which is irrational through irrational movements. For help or harm comes to the irrational through irrational things and to the rational through knowledge or ignorance.31
Posidonius believes that his own account can satisfy the requirement that cause be like effect. According to Posidonius the emotional factor in the soul (to pathētikon) which corresponds to Plato's two irrational parts engages in the emotional movements (pathētikai kinēseis) which I shall discuss in Chapter 7. Posidonius’ emotional movements I shall argue there are spatial movements of the physical soul just as we have seen in Chapter 2 the expansions and contractions of the physical soul postulated by Chrysippus are spatial movements. That is why Posidonius can argue that the emotional movements of the soul follow (hepesthai) the condition of the body:32 both are spatial entities. It is now easy to see how musical rhythms which involve spatial movements of the air could produce emotional movements in the soul via the body. In fact Plato who is Posidonius’ model had already discussed certain analogous effects on the soul. Sensations (aisthēseis) are themselves movements which pass through the body to the soul and when they are very violent as they are in children or in our first ancestors when newly created then they can even disturb the circular movements of the rational part of the soul and stop people thinking.33 No wonder sounds can set up movements of soul. Plato defines sound in the same work as the shock (plēgē) inflicted by the air on the brain and blood through the ears and passed on as far as the soul. Hearing he adds is the resulting movement starting from the head and ending around the seat of the liver.34 Melody and rhythmical rocking put babies to sleep because the movement from outside masters the inner movement that was causing fear and Bacchic dancing to the oboe cures patients by the same means.35
Posidonius is thinking in this same tradition and supposing that the spatial movements involved in sound set up spatial movements in the emotional part of the soul. Such causation respects the principle that effect is like cause.
In fact the principle that effect must be like cause and be non-rational if the cause is non-rational is erroneous. A hot water-bottle which involves no rational judgements can induce in me the rational judgement that the world is not such a bad place after all.
There is also a further oversight. For if effect really did need to be like cause Chrysippus could plead against Posidonius that the requirement is satisfied. For the musical modes and judgements are alike in Stoic theory in both having a physical basis and the physical bases could interact. The physical basis of one could interfere with the physical basis of the other thereby destroying the current judgement and perhaps leaving intact an earlier one. The like-to-like principle could also work in other ways by reminding someone of earlier judgements made in a similar musical context.
Philodemus and Seneca
Philodemus comes to a different conclusion on the basis of the same erroneous assumption that effect must be like cause. Melody is non-rational (melos alogon) and so cannot change emotions. Much less can it have stopped a civil war as had been claimed. It is rather the words in music that arouse emotion.36 Once again analogous views have been taken in the modern debate.37 Philodemus gives the following account.
For no melody being qua melody irrational rouses the soul from stillness and quiet and leads it to the natural disposition of character. Nor does it calm the soul and settle it in a state of rest from thrusting and rushing to whatever it may be. Nor can it turn the soul from one impulse to another nor lead its existing disposition in the direction of increase or decrease.38
Nor did [Diogenes of Babylon] provide examples of that sort of thing but of thought. He threw in melody only by way of assertion and incidentally. Nor did [Diogenes] show Ibycus and Anacreon and others like them corrupting young men by melodies but by thoughts. For they broke the young men if at all by the words which Persaios39 spoke of whereas a melody which accompanies as a vocal quality cannot be inflammatory [reading purōdēs]. So we shall not agree that lovers are pleased if they want this because of melodies which resemble their state. We shall say it is rather through words and thought and that Aristophanes shows that the ancients use a weakened voice as they called it in antiquity and prostitute themselves with their eyes not with melodies. If he said it was with melodies we shall assign [reading prostaxomen] him tears. For so far as concerns the power of melodies they neither summon people to what he so confidently says they do nor do they summon men and women to private [reading idias] copulations nor men in the bloom of youth to womanizing. For neither has he nor have the comic writers shown any such thing in the case of [reading epi] the works of Agathon and Democritus. They merely claim this. Nor did the actor Nicandros provide evidence [reading tekmērion] of this through his actions though if anything he did go astray. Not only is music indeed unable to assuage misfortune in love since that is the task of speech only but it distracts (makes anepiblētoi) people into switching their attention just like sex and drunkenness.40
For indeed the philosopher ought to have written down how irrational melodies can halt a rational dispute and ought to have persuaded us in that way that the melodies of Thaletas and Terpander halted the disputes of the Laconians.41
On Philodemus’ view music effects the therapy recommended by Epicurus as opposed to the Stoics. Instead of changing people's judgements on rational questions wordless music merely distracts people and makes them inattentive (anepiblētoi) to their emotional concerns.
Already we have two permutations. There is Posidonius’ view that emotion does not require judgement and Philodemus’ that melody per se does not cause emotion. Seneca would side with Philodemus. In his view music produces first movements instead of emotion. But Seneca's own argument for this would be stymied by the assumption that effect must be like cause. For Seneca acknowledges that music produces the appearance that there is good or bad. Indeed this appearance is needed in order to produce the first movements. And the appearance is just as propositional as a judgement and so just as unlike the wordless music that causes it. Turning the effect into an appearance with first movements does nothing to show how it can be like its cause.42
Zeno of Citium
The Stoics in the person of their founder Zeno provide material for yet a third permutation. Zeno said that music is after all rational: it has logos.43 This suggests a way of accommodating the requirement that effect should be like cause. As we have seen Zeno believed that emotions are rational though in the sense of being caused by rather than identical with judgements. They can in their turn be caused by music he could have argued because music is also rational.
As it stands this argument is no good. For music has logos not in the relevant sense of involving judgements of reason but in the different sense of having a structure intelligible to reason. However in her forthcoming Gifford lectures Martha Nussbaum has argued that music is a form of symbolic representation. She sees this as showing how music can be the cause of the judgements which are involved Bin emotion. I must leave it to her to expound the argument which is both subtle and fascinating.44 But what I want to say is that it is in my view unnecessary. For if effect need not be like cause the music which causes judgements need not itself contain anything like judgements.
Two reactions to Diogenes of Babylon
Philodemus is certainly and Posidonius is probably reacting to the earlier Stoic Diogenes of Babylon. Whether or not either Philodemus or Posidonius knew of the other's work each position is a reversed image of the other. According to Posidonius non-rational music can change the emotions and he concludes that emotions at any rate in these cases are non-rational. Philodemus starts from the opposite assumption that emotions are rational and concludes that since melody is non-rational (melos alogon) it cannot change the emotions.
Philodemus though an Epicurean is in one way closer to Chrysippus’ intellectualism than is his Stoic opponent Diogenes. Not only does he take emotions to be rational but he insists that disagreement over harmony is due to differing beliefs (doxai).45 He accuses Diogenes in the same breath of taking a very non-intellectualist view and saying that such disagreement is due to irrational perception. In fact this misrepresents Diogenes46 who following Speusippus ap. Sextum M 7. 145–6 had taken an intermediate view. It is only high and low pitch that are perceived by irrational perception. Harmony is perceived by a perception that is scientific (epistēmonikē).47
Whether or not Posidonius knew the work of Philodemus he surely knew of Diogenes’ views.48 Diogenes was to some extent the heir of Damon and Plato. Ten fragments of Damon are collected in the standard Diels-Kranz Fragments of the Presocratic Philosophers and Damon is quoted by Plato for his views.49 It looks as if Damon had not only insisted before Plato that music affects character but had also supplied Plato with his view that music achieves this by its imitation (mimēsis) of or likeness (homoiotēs) to character.50 Diogenes twice quotes Plato's Laws51 and reports Plato52 and Damon53 as saying that music can be used to inculcate justice. Diogenes argues that music can affect emotions.54 Inter alia it can affect erotic emotions and lead to erotic virtue.55
As to how music affects emotion Diogenes’ discussion could have helped to reveal to Posidonius the difficulty of defending Chrysippus’ judgemental view. Diogenes holds that our dispositions can be calmed or intensified by becoming like or akin to (homoion oikeion) the music56 although he appears to have conceded presumably as a correction of Plato and Aristotle that the likeness was not an imitative one (ou mimētikas).57 The reason as was argued above why such cases create a difficulty for the judgemental view is that when emotion is aroused merely through the kinship of the music to an emotion we are not likely to be able to make a judgement. At most we may feel as if there was something good or bad at hand.
Later history of the argument from music
Elements of Posidonius’ argument on music also had a later history. According to Plutarch Pythagoras’ use of music to comfort the soul shows that he was aware that the soul has a non-rational ingredient.58 The story of the Greek oboist calming the youths recurs in different versions in Ammonius and Elias in Greek and Cicero Quintilian Augustine and Boethius in Latin.59 Augustine uses it as we shall see in Chapter 26 to make a quite different claim about lust: if music can calm lust lust cannot be under the command of the will.
An eighteenth-century statement by Gibbon observes the same empirical facts although it does not draw the same philosophical conclusions:
Experience has proved that the mechanical operation of sounds by quickening the circulation of the blood and spirits will act on the human machine more forcibly than the eloquence of reason and honour.60
From the book: