Christof Rapp
“Sing, oh goddess, the anger of Achilles, son of Peleus …” With these words Homer opens his epic, the Iliad, thus indicating that the tragic entanglements that will unfold in the following more than 15,600 verses originate from a particular emotion, namely mēnis – anger, wrath, rage, or the like. In the Greek text, mēnis (strictly speaking the accusative mēnin) is in fact the first word of the entire epos. Achilles’ anger gives rise to a sequence of reactions that leads not only to the death of many of his fellow Greek heroes and his intimate friend Patroclus, but also, and ultimately, to his own untimely death. Taking up the legacy of early epic, the great Greek tragedies of the fifth century BC represented various types of emotions as motives for the conduct of their tragic characters. However, the heroes of epos and tragedy are not the only ones stirred by emotions: the audience, too, when listening to a rhapsodic performance of an epos or to actors and a chorus in the theater, is co-affected by the passions of the protagonists, by their misfortunes and their fates. While the spectators of tragedy experience painful feelings like pity and fear, the spectators of comedy are prone to laughter and amusement. Quite early, writers started to reflect on the emotive effect of stage play and musical performances. They also discovered that although the spectator of a tragedy is seized by painful emotions, he takes pleasure in exposing himself to the theatrical embodiment of painful misfortunes that he would never wish to experience in real life. Although melodic modes and rhythms have the power to excite our minds, they can also provide a pleasant feeling of relief. Accounting for the various emotional effects of music and theater was therefore seen as a theoretical challenge, one that encouraged philosophers to think more about the nature and power of emotions.
The following survey will give a brief indication of how tragedies are meant to arouse the emotions of an audience and of the types of emotions that tragic heroes experience. Early philosophical reflection on aesthetic emotions occurs in a milieu that is in general interested in the nature of emotions and in the role that emotions play in our ethical life. It is Plato and Aristotle who contribute immensely to philosophical accounts not only of the tragic emotions but of all kind of emotions and mental states involving both pleasure and pain. Plato, however, is highly critical of the emotional impact of poetry, and in his Republic he even suggests banning poetry from the ideal city-state because of its tendency to distort the natural order of our souls. For Aristotle, it is an essential feature of tragedy that it arouses the emotions of pity and fear in the audience; he does not think that emotions as such are something bad or distorting. Aristotle famously summarizes the effect of tragedy by saying that it brings about a katharsis – purgation, purification, cleansing, relief, and so forth. This cryptic remark has been the subject of longstanding debates: Is it the spectator that is cleansed of or released from his tragic emotions? Or is it the tragic emotions themselves that are in a way, cleansed, purified, enhanced? The survey will conclude by briefly looking at an example of how these debates that originate in Greek philosophical aesthetics are taken up in the Roman imperial period, when the Stoa became the prevalent philosophical school and the Stoic ideal of apatheia – the absence of emotional upheaval – started to influence the discussion of aesthetic passions.
The place where the aesthetic use of emotions culminates is clearly the theater. In particular, the Classical format of Greek tragedy, as shaped by the great tragedians Aeschylus, Sophocles, and Euripides, was famous for its overwhelming emotional effect on the audience. Ancient sources quite frequently refer to the emotional reaction of different audiences, to particular actors’ capacities to make the spectators cry or to the emotional effect of dramas. For example, it is reported by Herodotus (6, 21) that one particular play, Phrynichus’ The Fall of Miletus, provoked such excessive lamentation in the audience that the magistrate felt obliged to intervene and ban the play from the stage forever.
In tragedy there are, to begin with,1 a number of aural elements that elicit emotive effects. Greek tragedy included parts that were spoken by the principal actors and parts that were sung by the actors or the chorus; both parts were metrically structured in different poetic meters, and the choral recitations were set to a melody. The aspect of melody is, according to Aristotle (Poetics 6, 1450b15–16), the most pleasurable accessory of tragedy. The ability of melodies or songs to excite emotion was almost proverbial: the mythological sirens, for example, were capable of seducing hardened and widely traveled heroes just by their tunes. We know from Plato and Aristotle that the emotive power of music was classified in accordance with the musical key, the rhythm, and the accompanying instruments; of all the musical instruments that were in use at this time the aulos, an oboe-like instrument, was often held to be the most exciting and exhilarating instrument (Aristotle, Politics VII.7, 1342b3) and hence was banned from Plato’s ideal city (Republic III, 399d3–6). Furthermore, tragedians came to see that the manner of recitation and delivery strongly influenced how the drama affected the audience. Whereas in the early days the tragedians would act themselves in their plays, trained elocutionists and actors proved able to impress the audience by the modulation of their voices:
Delivery [hupokrisis] is a matter of how the voice should be used in expressing each emotion, sometimes loud and sometimes soft and [sometimes] intermediate, and how the pitch accents should be intoned, whether as acute, grave, or circumflex, and what rhythms should be expressed in each case.
(Aristotle, Rhetoric III.1, 1403b26–30; translation based on G. Kennedy)
Aristotle reports that actors who were able to use their voice in the way described were the ones who usually won the poetic contests – apparently due to the emotional impact of this aural device. He also mentions an actor by the name of Theodorus who was able to disguise his voice so effectively that the audience believed it was hearing the voice of the actual character (Rhet. III.2, 1404b21–4); he also attests that the voice and, in general, all aspects of delivery make a major difference in how effective one is at arousing pity in the audience (Rhet. II.8, 1386a33).
Another way of modulating the audience’s emotion that could also, for lack of a better word, be termed “aural” relies on the diction, the selection of words (lexis) in accordance with their rhythm, timbre, and onomatopoetic potential on the one hand and their meaning, imagery, or metaphorical impact on the other. Provided that the audience is always in a similar emotional state to that of the orator or character (Aristotle, Rhet. III.7, 1408a23–4), the latter can use the wording of someone who is actually outraged or indignant in order to provoke the same sort of emotions in the former (ibid. 1408a16–9). Sometimes, it is the mere tone of a phonetic sequence that adds emotive value to the spoken word. The ancient Greek language is particularly creative and flexible when it comes to expressive or plaintive noises and screams. Aeschylus, for example, repeatedly has his characters and chorus scream, “Otototoi” (Agamemnon v. 1072–3) or even “Ototototoi” (Persians v. 1043). The finale of the latter play actually culminates in a sequence of screams like “oioi, oioi,” “io io,” iē iē.” In a memorable scene from Sophocles’ Philoctetes the principal character shouts out “papai, apappapai, papappapappapappapai” (v. 745–6).
There are also visual effects that contribute to the emotional effect of tragedy. In his Poetics Aristotle claims that opsis – the visual aspect that is visible to the audience, namely the stage setting, enactment, or spectacle – is the most persuasive, most agitating aspect of tragedy, though it is the one least based on art (Poetics 6, 1450b16–8). Hence, the visual aspect of staging can, in principal, evoke the tragic emotions of pity and fear, along with other emotions that are, according to Aristotle, not really essential to the proper effect of tragedy (Poet. 14, 1453b1–11). The scenery, costumes, and theatrical masks may contribute the visual dimension of tragedy’s effect on the audience. In general, showing a victim’s clothes or other belongings might be enough to arouse the audience’s pity (Aristotle, Rhet. II.8, 1386b1–4). Certain gestures were seen as expressing the characters’ state of mind – and the repertoire of emotional and emotive gestures exhibited by Greek epic and tragedy is most impressive. In the Iliad, for example, Priam and Hecuba beg their son Hector not to counter Achilles’ fatal attack in front of the city walls: Hector refuses to avoid the enemy, and his father starts to pluck and tear the hairs from his head, while his mother bares her breast (Iliad XII, 77–80) in order to arouse her son’s pity. Similar dramatic gestures are common to many tragedies of the Classical period: women exposing their breasts, other characters tearing or cutting their hair, beating their heads, slashing their faces, ripping their clothes, and so forth.2
Another important visual aspect of tragedy is dancing. The chorus of a tragedy would dance, and this dance could express various emotions at the same time – those of the protagonists or those of a witness – and arouse similar emotions and a similar level of excitement in the audience. Aristotle actually says that dance in the theater is a means of imitating types of characters, emotions, and actions (Poet. 1, 1447a27–8). In certain contexts, dance can also express orgiastic frenzy, as it was most probably meant to do in Euripides’ Bacchae, where the god Dionysus exploits the ecstatic state of the women of Thebes to take revenge on King Pentheus. Indeed, the general setting of the Bacchae and the ecstatic dancing of the women of Thebes in this remarkable drama should remind us of the context in which Greek tragedy was originally performed, namely the Dionysian cult and the annual Dionysian festival. Festival-goers were acquainted with the orgiastic nature of the Dionysian cult and they came to the theater not with the detached attitude of the average modern theater-goer, but in anticipation of an exciting spectacle that was likely to get its audience emotionally involved.
Last but not least, the subject matter and the plot of a tragedy can also provoke certain passions in the audience. If, for example, the subject matter was taken from historical events in which some of the ancestors of a public audience were involved, the mere mention of such events would clearly provoke a certain emotional effect. In Aristotle’s attempt to give a general account of tragedy, the plot is said to be central to the arousal of the tragic emotions (see below). What he is thinking of, however, is not just the selection of a theme or plot, but rather the artful organization of material by the tragedian in such a way that the protagonist, without having done any condemnable harm or crime, is completely ruined at a tragic turn in his life. If such a plot is properly organized it will, as a rule, bring about the tragic emotions of pity and fear even in someone who only reads the play.
Apart from the emotions such as pity, fear and the like, by which the emotional impact of tragedy was usually characterized, several other types of emotions play an important role in tragedy, not because they are meant to characterize the audience’s reaction to the tragic happenings, but because they are among the principal motives of the tragic agents.3 For example, the heroes of tragedy and epic often act out of motives like anger, rage, fury, indignation, emulation, love, hate, longing, shame, and so on.
As mentioned at the beginning of the chapter, Achilles’ rage (mēnis) was taken by Homer to be the origin of the chain of events that ultimately led to many tragic ends, including Achilles’ own death. Anger or rage (other Greek words for this emotion are orgē and cholos) thus acquired a reputation for being an exceptionally dangerous and pernicious emotion, because it pushes people to perform actions that they are bound to regret. Anger arises whenever we think we have been insulted, slighted, despised, offended, and “disrespected.” It is a painful experience, as it is meant to violate our feeling that we deserve to be treated more respectfully. It goes against our wish to be seen and to see ourselves as respectable, valuable people. The violation of this wish is painful, and the resulting pain awakens the desire for revenge, because revenge can cause a certain amount of pain in the insulter, thus compensating us for what we have suffered. Anticipating the act and the effect of revenge is pleasant, which prompts Aristotle to characterize anger as the longing for revenge caused by an insult, a longing that is partly painful and partly enjoyable (Rhet. II.2, 1378a30–b4). Achilles, to refer again to the example of the Iliad, was hurt by the insulting behavior of Agamemnon, the Great King of the Greeks (who, in order to demonstrate his omnipotence, took away the maid that Achilles had received as prize), and took revenge, by refusing to support the Greeks in battle anymore. Euripides’ Medea is at first led by a feeling of uncompromising love for Jason that later turns to implacable rage after Jason betrays her. He has her say, “I know that what I am going to do is bad, but anger overpowers my plans – anger which is cause of greatest troubles for mortals” (Medea 1078–1080). In a similar way, other tragic heroes are driven by other kinds of passions: Sophocles’ Ajax is even driven to suicide by his feeling of shame. Seneca’s Oedipus is moved by fear and his Phaedra by love, and so forth.
Early reflections on the aesthetic role of emotions went hand in hand with the beginning of philosophical theorizing about the nature of emotions. For philosophers of the late fifth and the early fourth century BC, the phenomenon of emotion became an increasing challenge in several respects and contexts. One such context was the ongoing dispute with the so-called sophists, who propagated a rhetoric that was crucially based on the excitement of emotions in the audience. For philosophers like Plato and the early Aristotle this was a problem, because they wanted a certain degree of rationality to prevail in public and judicial debates, and the increasing prevalence of the kind of emotional rhetoric favored by the sophists rendered this difficult, if not impossible.
Among the first reflections that have come down to us on the power of tragedy are those of the sophist Gorgias, for whom poetry is nothing but metrically structured speech, who praises in his Encomium of Helen the power of speech (logos) and describes the emotional impact of poetry thus: “There comes upon its hearers fearful shuddering [phrikē] and tearful pity [eleos] and grievous longing [pothos] and at the good fortunes and evil actions of others’ affairs and bodies through the agency of words the soul experiences suffering of its own” (Encomium of Helen 9, tr. J. Dillon/T. Gergel). Gorgias mentions here one emotion involving fear and another entailing a sort of compassionate grief or pity. This pair of emotions was subsequently used to account for the emotional effect of tragedy in general; Aristotle famously includes the pair of pity and fear in his definition of tragedy (see below).
Another context that called for a more refined account of emotions was ethics and theory of action. If one assumes, as Plato and Aristotle did, that the good life and the right conduct is a matter of doing whatever we do for the sake of the good, then the non-rational impulses, including appetites and emotions, which are not by their nature directed toward the good, might be seen as distorting forces in our lives that prevent us from doing what we should do and really want to do. And once non-rational impulses are thought to have such a destructive potential, we must take an interest in learning how to either control or domesticate these non-rational desires; otherwise we would find ourselves being acted upon by these impulses rather than acting ourselves. For philosophers who analyze the question of the good life along such lines it is crucial to work out the intuition that we are driven by both rational and non-rational impulses. In Plato’s philosophy the corresponding project leads to the famous model of a tripartite human soul (see below); in Aristotle’s ethics we find the congenial idea that the human soul is partly rational and partly non-rational, and that the non-rational part of the soul is in its best condition (what Aristotle calls the arēte, i.e. virtue, of character) when it is in accordance with what the rational part of the soul prescribes. The philosophical subdiscipline that explores the soul’s various parts or faculties, the discipline we have come to call “moral psychology,” is also concerned with localizing and outlining the emotions along with other psychological entities such as opinions, desires, appetites, feelings of pleasure and pain, and so on. For reasons like these, the prevailing ancient schools of ethics were directly linked to moral-psychological questions, which, in turn, included as a necessary part the inquiry into the nature of emotions.
Before taking a closer look at these accounts of the nature and the psychological neighborhood of emotions, there is another context for the early philosophical exploration of emotions that deserves to be mentioned: The Greeks would often refer to the emotions as “pathē of the soul,” that is, what the soul experiences or suffers. This was meant in contrast to what the body suffers, that is, sicknesses, injuries, aging, loss of hair, bodily pains and bodily pleasures, etc. However, this parlance of “passions or affections of the soul” by no means excludes emotions from at the same time being alterations of the body. On the contrary, several philosophical schools even used emotions as paradigms for what the soul experiences together with the body. Aristotle, for example, argued that shame is an emotion rather than a disposition, because blushing is a clearly bodily symptom of shame (Aristotle, Nicomachean Ethics IV.9, 1128b13–5). Similarly, the Stoics and the Epicureans referred to the bodily symptoms of emotions – for example, that people alter their facial expression, that they turn red or pale respectively – as evidence that there must be an interaction between body and soul (and that, hence, the soul must be corporeal to facilitate this interaction).4 For reasons like these, emotions also came to play a role in discussions about the relation of body and soul. In addition, the somatic-physiological side of emotions was recognized as an object of study for the field of medicine, and thus the bodily alterations connected with the several types of emotions were described in detail.
For those first attempting to localize the emotions among other psychological phenomena it seemed striking that emotions are typically connected with the feelings of pleasure and pain: some emotions are obviously painful, while others are pleasant. Whether and how pleasures are an ingredient of the good and happy life has been a philosophical controversy since the time of Socrates. In his later works Plato sought to respond to this controversy by differentiating various types of pleasure, since he became aware that not all types of pleasure derive from the satisfaction of bodily needs. As opposed to these bodily pleasures there are, as Plato recognized, certain pleasures that are not directly related to the presence of such a somatic need. In the Philebus, for example, he describes how certain pleasures can derive just from the expectation of a pleasant event:
Socrates: But now accept also the anticipation by the soul itself of these two kinds of experiences (the experience of pleasure or pain, sc.); the hope before the actual pleasure will be pleasant and comforting, while the expectation of pain will be frightening and painful. – Protarchus: This turns out then to be a different kind of pleasure and pain, namely the expectation that the soul experiences by itself, without the body. Socrates: Your assumption is correct.
(Philebus 32b–c; tr. D. Frede)
Pleasures deriving from mere expectation belong to the soul, not to the body. Corresponding to the pleasures mentioned above that come about through expectation, there are also pleasures that derive from mere memory (Philebus 33c; see also Aristotle, Rhet. I.11, 1370a27 et seq.). The acknowledgment of the existence of such psychological pleasures and pains was an important step toward developing an account of emotions. The logical next step was to distinguish types of psychological pleasure and pain on the basis of what we expect or remember. A simple version of this step can also be found in Plato:
Athenian: In addition to these two (pleasure and pain, sc.), he has opinions [doxai] about the future, whose general name is ‘expectations’. Specifically, the anticipation of pain is called ‘fear’, and the anticipation of the opposite is called ‘confidence’.
(Laws 644c–d; tr. T.J. Saunders)
In these lines we can find the idea of differentiating pleasant and painful psychological states on the basis of the prospective (or retrospective, as one may infer from Philebus 33c) opinions or views that bring about these feelings; it turns out that these different states correspond to different emotions like fear on the one hand, and confidence on the other. Plato himself was confident that emotions like “wrath, fear, longing, grief, love, jealousy, malice and other things like that” (Philebus 47e) can be explained as precisely these kinds of “pain within the soul itself” or as mixtures of psychological pleasure and pain.
Aristotle was not so explicit in arguing that emotions are themselves kinds or mixtures of pleasure and pain, but he insisted that emotions are at least firmly combined with feelings of pleasure and pain.5 It seems clear from his definitions of particular types of emotions that he followed Plato in differentiating painful and pleasant states in terms of the expectations or opinions with which they are associated. For example, fear, shame, and indignation are all painful feelings, but while the first comes about through the expectation that a painful or destructive evil is likely to happen in the near future, the second is combined with the anticipation of the possible loss of honor and good reputation, and the third depends on the judgment that the fortune of another person is undeserved. In a similar manner, and equipped with the idea that different objects of our pleasant and painful attitudes individuate different emotions, Aristotle undertakes to define something like 15 different types of emotions in the second book of his Rhetoric.
In sum, Aristotle takes emotions to be intimately connected with states of pleasure and pain. They vary in accordance with our opinions or judgments and, correspondingly, with the objects at which they are directed. Thus, although emotions themselves do not belong to the rational part of the soul, they are nevertheless dependent on certain cognitive efforts, such as having beliefs and making judgments. Emotions are also connected with our desires, as we generally strive to avoid painful states and retain pleasant ones. Some emotions are essentially connected with a specific desire. Anger, for example, always involves a desire to take revenge. Our emotional dispositions are trained and habituated in the course of our education, but while we are praised or blamed for our emotional dispositions (hexeis), we are in general not praised or blamed for experiencing occurrences of emotions, but for experiencing them in a suitable or unsuitable way respectively. For Aristotle, having the virtues of character implies that one always has the right emotions when one should, toward whom one should, and so on. Having the right desires and emotions is what Aristotle takes to be the good state of the non-rational part of the soul. Emotions are crucial, then, for defining virtue, which again is one of the key notions of Aristotle’s ethics.
One of the most amazing accounts of the emotional impact of the arts is to be found in Plato’s Republic. He is impressed by tragedy’s power to corrupt even decent people (Plato, Republic X, 605c–d). At least in the context of his Republic, he seems to regard these emotional effects as thoroughly unwelcome. The interlocutors of this dialogue undertake to sketch the essentials of a just city. They quickly come to agree that any city that is not content with the satisfaction of basic needs, but aspires to more sophisticated comforts, requires people who act as what they call “guardians” who are meant to function as an army against outer enemies and as police in inner affairs. Due to their double function the guardians should be fierce and harsh to enemies on the one side and gentle toward their own people and those they know on the other (Republic II 374e–376c). This ideal character profile of the guardians occasions a long excursion concerning the education of the future guardians: Education broadly conceived involves gymnastics for the body and instruction in the fine arts for the soul. Concerning the latter, the dialogue offers a detailed discussion of the pedagogical impact of poetry and music. Young people are not supposed to learn from poets that death is something to be feared or lamented, nor that gods commit disgraceful acts (for, if they hear and absorb false opinions from myths and tales, it will be hard to erase these opinions in the course of their adult lives: 378e). The negative impact on the souls of the young is said to be increased by imitation (mimēsis), that is, when actors speak as though they were someone else. The idea is that if someone keeps imitating certain character traits, those same traits will enter into one’s own habits and nature (Rep. III 395d). Hence, young people should not be allowed to imitate people that are suffering misfortune and are possessed by sorrows and lamentations (395d–e) – as happens in the theater. The consequence the interlocutors of the Republic draw is an uncompromising one: If imitation of behavior of this sort is supposed to foster adults with a timid and cowardly character, it would be better for the city to erase all passages from Homer and the other poets that depict indecent behavior of gods and heroes.
The philosophical background of Plato’s position is provided by his theory of the tripartite soul: Since it is possible for us to suffer and to desire opposed things at the same time, while one single subject cannot have two opposite inclinations in relation to the same thing at the same time, Plato argues (Rep. IV 436a–441c), our soul must be subdivided in accordance with different types of inclinations. First, it seems obvious that the rational part of our soul (logistikon), by which we deliberate and calculate, must be distinguished from the appetitive part (epithumētikon) by which we long for bodily pleasures and for the satisfaction of our needs. The appetitive part, however, can be combated by another non-rational part of the soul, the spirited one (thumoeidēs): this happens, for example, if we get angry about appetites we happen to have. Equipped with this tripartite account of the soul together with the maxim that it is the rational part of the soul that should reign over the non-rational parts, Plato is in a position to explain what is wrong about poetic imitation in general. First of all, the work of imitation is always different from the original it imitates; it is not the thing itself, not the true thing, as it were, and, hence, it does not convey truth. Given that the rational part of the soul is always out for truth and the work of imitation has an inevitable distance to the real and true thing, imitation always consorts with a part of us that is far from reason (Rep. X 603a–b). It can be concluded, hence, that poetic imitation primarily addresses the non-rational, appetitive part of the soul that “leads us to dwell on our misfortunes and to lamentation, and that can never get enough of these things” (604d). Furthermore, the imitative poet arouses, nourishes, and strengthens the non-rational part of the soul and so tends to destroy the rational part of it. A hypertrophic non-rational part of the soul, again, interferes with what we would need most under difficult circumstances, that is, the properly functioning deliberation within the rational part of soul (604c).
In the same pedagogical context, Plato discusses the use of music, of musical modes, rhythms, and instruments. The right sort of rhythm and harmony, says Plato, permeate the inner part of the soul more than anything else; it affects it most strongly and can – in the best case – bring about the well-orderedness of the soul (Rep. III 401d), namely the state in which the irrational impulses do not interfere with the rational command and the soul is, accordingly, said to be virtuous. This is the effect of suitable rhythms and modes, while unsuitable music can have contrary and, indeed, disastrous effects on the soul. For the education of the young guardians in the just city of the Republic there is, in particular, no use for modes of lamentation or soft and relaxed ones (as performed at drinking parties); these criteria exclude certain musical modes, while certain other modes turn out to be useful, namely the ones that either imitate the courageous warrior who is active in battles or imitate a person who acts in peaceful voluntary dealings in a moderate and self-controlled way (399b–c). By the same criteria all poly-harmonic and multi-stringed instruments are excluded and only the lyre and the cithara are left for use in an urban environment, while certain flutes are left for the shepherds in the country. In a similar vein, appropriate rhythms are selected in accordance with which rhythms correspond to someone who lives an ordered and courageous life and to the meter of his words.
Plato’s views about the distorting effects of poetry and music have often been met with criticism; his idea to ban the poets from the just city has even been interpreted as a totalitarian trait in his thinking. However, it must be kept in mind that all these regulations are part of a thought experiment that tries to correlate an ideally ordered city with an analogously ordered state of the individual soul. The order of the soul, in turn, implies that the non-rational part does not rebel against the rational one. This, again, leads to the consideration of factors that can distort the order of the soul. And within this very peculiar context, common cultural institutions of Plato’s time, such as the performance of tragic poetry, come under scrutiny. That the resulting suggestion to censor and to ban major parts of Greek literature is not without irony sometimes shines through, for example when the interlocutors of the Republic “ask Homer and the other poets not to get angry” when they start to excise all the supposedly problematic passages from their poems (Rep. III 387a–b). Finally, all the provisions regarding poetry are made for a pedagogical purpose and with a view to emotionally immature recipients, while the more mature interlocutors themselves are happy to admit that the poems under scrutiny can be entertaining and pleasing.
The most influential treatment of the emotions in ancient aesthetics is that in Aristotle’s definition of tragedy. As already outlined above, tragedies were generally famous for their potential to arouse the audience’s emotions. Aristotle clarified this emotional effect and made it a part of the essential nature of tragedy. Far from criticizing it the way Plato did, Aristotle adds that the arousal of such emotions by a tragedy leads to a sort of purgation, purification, cleansing, or relief (katharsis). The key idea is expressed in the following very compressed formula: “Tragedy, then, is an imitation of an action which is serious and complete […] in the mode of dramatic enactment and not narrative, effecting [perainousa] through (the arousal of) pity [eleos] and fear [phobos] the katharsis of suchlike emotions” (Poetics 6, 1449b24–8).
All the other elements of this definition had been introduced in the preceding passages of the Poetics: that the tragedy is a sort of mimēsis (i.e. imitation or representation), that it imitates actions, that these actions have to be unified or complete in a certain way, that the tragic representation of this action takes place by enactment and not just by report or narration, etc. It is only the final formula about pity, fear, and katharsis that has not been prepared or explicated in the immediate context. The term katharsis (in this peculiar meaning) is not taken up again in the remainder of the text. The emotions of pity and fear, by contrast, are mentioned and commented on in different parts of later chapters; they play a particularly important role when Aristotle discusses the plot or the story of the tragedy, which, as he never tires of repeating (Poet. 11, 1452a38–b1; 13, 1452b33 et al.), must be lamentable and frightening in a quite precise way.
If we focus for a moment on the final formula that introduces the emotional effect of tragedy – the formula that tragic imitation “effect(s) [perainousa] through (the arousal of) pity [eleos] and fear [phobos] the katharsis of suchlike emotions” – it is striking, in the first place, that even this brief formula seems to introduce two logically distinct processes: the arousal of pity and fear, and the katharsis of suchlike emotions. Strictly speaking, the formula does not explicitly refer to the arousal of pity and fear – it only says that it effects katharsis from such emotions through pity and fear. However, in the rest of the Poetics we learn little or nothing about the purgative or relieving effect of tragedy, but quite a lot about how it arouses certain emotions in the first place. Also, the second process, the one of purifying or expurgating, presupposes that there are certain episodes of emotions; this is clear from the phrase “of suchlike emotions.” Even if “suchlike” is broad enough to cover more emotions than those explicitly mentioned, it is inconceivable that the formula “of suchlike emotions” could be understood as not including the aforementioned emotions of pity and fear. And if the process of katharsis actually presupposes that the audience is currently moved by episodes of pity and fear, it is most likely that these are just the same episodes of pity and fear that were brought about by the tragic plot – otherwise it would be strange to assume that the spectators just happen to be in such an emotional state of mind.
We can conclude then that Aristotle’s account of the emotional impact of tragedy comprises two parts or processes, the arousal of pity and fear, and the katharsis of “suchlike” emotions (though it is also common to refer to this whole account as the “cathartic effect of tragedy according to Aristotle”). The first part will be considered in the remainder of this section, the second part, which includes katharsis in its narrower sense, will be discussed below. As for the arousal of pity and fear, some remarks about the nature of these two emotions might be in order before the more technical question of how tragedy manages to steer these emotions in the audience is considered.
As already pointed out, it had been common since the time of the sophist Gorgias to characterize the emotional effect of tragedy by emotions that correspond to what we would approximately translate as pity and fear. Aristotle seems to have assumed that emotions can be individuated by the characteristic objects toward which they are directed and by the opinions or judgments that accompany the pleasant or painful feeling inherent in each episode of an emotion. The emotion, usually translated as “pity,” “compassionate grief,” and the like, namely eleos, was thought to be essentially connected with the undeserved misfortune of other people. Whenever one observes another’s unmerited suffering, one comes to experience a certain share of this suffering. This latter phenomenon, that we ourselves feel bad about what other people undeservedly experience, might be due to an anticipation of harmful things that could happen to us. And this is again why the Greek eleos is often said to be self-referential as opposed to its more altruistic Christian counterparts, such as misericordia. In Aristotle’s definition of eleos this moment is captured by the claim that this emotion is aroused by evils “that one expects himself or one of the people close to him to suffer” (Rhet. II.8, 1385b14–5). And since a mechanism like this seems a likely candidate for building an emotional tie between the spectator and the tragic hero – the spectator’s grief about the fate of the tragic character is inextricably connected with the vulnerability of his own existence – eleos or pity became the embodiment of what people feel when watching a tragedy.
Now, pity is only one partner in a pair of emotions that characterize the tragic experience, the other being phobos or fear. Fear, again, can be straightforwardly characterized by the objects at which this emotion is directed and by the thoughts we usually have about these objects. In general, fear is an unpleasant worry about imminent future harms or evils. In order to actually feel fear we must be of the opinion that we are likely to suffer these evils in the not too distant future. As part of the pair of tragic emotions, pity and fear, phobos is the state of being frightened or scared by what happens to people (here: the tragic characters) whom we take to be similar to us, while pity is the feeling directed at undeserved suffering (Aristotle, Poet. 12, 1453a4–5). And one of the reasons why we think of ourselves as similar to the tragic heroes is that we are similarly vulnerable to the sort of misfortune they suffer and that such misfortune would strike us quite as undeservedly as it strikes the tragic heroes.
How is it that tragedy manages to arouse these two emotions? As pointed out above, tragedy has various emotive means at its disposal: aural aspects, such as the voices of the actors, the chants sung by the chorus, visual aspects, and so on and so forth. For Aristotle, however, these are just accessories. He stresses again and again that the most important aspect of a tragedy is its plot (Poet. 6, 1450a15); the plot (muthos) is also said to be the principle (ibid. 1450a38) or the soul (ibid.) of each tragedy. In a good tragedy it is this plot alone that brings about the peculiar emotional impact of tragedies. This can be proved by the test that a good tragic plot will bring about the tragic emotions even without enactment or scenery, if one only reads it to oneself (Poet. 6, 1450b18–9; 14, 1453b1–6). How is such an effect possible?
The mechanisms underlying the tragic arousal of pity and fear are, according to Aristotle, similar to those at work in a public speech, when an artful orator manages to manipulate the audience’s emotions. The orator who uses Aristotle’s art of rhetoric is equipped with a preliminary definition of each and every type of emotion. From this definition he can infer which objects and opinions are connected with each type of emotion. Making an audience feel a particular emotion is, then, essentially a matter of making them believe that a certain object is given and of making them judge or think about this object in a certain way. In order to arouse, for example, anxiety in an audience, the artful orator will make these people think that a harmful or destructive evil is about to happen in the not too distant future and that they have no means to avert this danger. Insofar as the rhetorical mediation of these emotions applies the audience’s opinions, thoughts, or judgments and insofar as these can be regarded as belonging to our cognitive apparatus, one might say that this is a cognitive or cognitively mediated process (saying it is cognitive, however, does not imply that it is rational: for example, the judgment that some harmful evil is about to happen might be ill founded).
Now, this may hold of oratory, but oratory is quite different from tragedy. However, in this particular question – the question of how the tragic emotions come about – Aristotle explicitly says that to make the actions lamentable and frightening the tragedian has to use the same procedures or propositions (ideai)6 familiar to us from the Rhetoric (Poet. 19, 1456b2–5). In fact he points out that regarding the arousal of emotions the only difference between oratory and tragedy is that the tragedian cannot use explicit statements or hints (didaskalia), unlike the rhetorician (who can argue that someone is, for example, deserving of pity), but has to incorporate these characteristics (that someone is lamentable or something is frightening) into the actions themselves (ibid. 1456b5–7). Again, how is this feasible? How can emotional characteristics be incorporated into the actions themselves? Indeed, Aristotle assumes that it is possible to construe the plot as a mimēsis of pitiable or lamentable and frightening actions (Poet. 9, 1452a1–3); and it is one of the core issues of his Poetics to lay bare the conditions under which a plot can have such an effect. As a minimal requirement the particular actions that make up a plot must be connected in a plausible, likely, or even necessary way (7,1452a12–3). Next, the plot will not be pitiable if it presents a wicked protagonist whose fate turns from prosperity to affliction or vice versa; nor is it pitiable if an entirely impeccable character passes from good fortune to bad. The pitiable sequence of action is rather one that shows a good person passing from good fortune to bad – not because of her wickedness, but because of an isolated mistake (13. 1452b33–1453a10). Aristotle distinguishes between simple and complex plots and points out why it is the latter one that is more apt to bring about pity and fear than the former; a plot can be made more complex and can have a greater emotional impact if it includes dramaturgical effects such as a sudden reversal or the protagonists’ recognition of facts or persons that had been previously unknown (Poet. 11). In addition, the tragedian has to depict his protagonists as being of a certain character; their character becomes obvious from what they do or what they say (Poetics 15). Since the spectators will only feel pity for the protagonists if they regard them to be similar, it is important that they come across as having a good character and decent convictions. Bearing such instructions in mind, the tragedian will be able to compose a plot that is the mimēsis of lamentable and frightening human beings and actions and will in this way bring about the emotional effect that defines tragedy.
When Aristotle introduces the notion of katharsis in his famous formula that through pity and fear tragedy effects the katharsis of such emotions, he seems to take for granted that his readers are familiar with this concept. Aristotle himself nowhere explains this term and the explanation that his Politics announces will be given in the Poetics (Politics VIII.7, 1341b38–40) has not come down to us. The problem is that the notion of katharsis is connected with quite heterogeneous associations. For example, it is used by medical writers to describe the impact of laxatives and emetics as well as various other kinds of excretion of bodily fluids. In a quite different context it is used for religious purgation rituals. One part of the notorious controversies about the meaning of katharsis in Aristotle’s Poetics originates from the question whether one of these contexts – the medical or the ritual one – is relevant for the understanding of katharsis in the Poetics or whether it is rather Aristotle’s ethics that provides the proper background for the interpretation of poetic katharsis. Another part of the scholarly disputes about the right interpretation of this notion derives from a linguistic ambiguity: when Aristotle says that through pity and fear tragedy brings about the katharsis of suchlike emotions, the genitive can either be meant as an objective genitive or as a genitive of separation. In the former case it means that the emotions are the object of an act of purifying, cleansing, or refining (the emotions will persist after the katharsis has taken place – though in a purified or somehow “enhanced” way), while in the latter case the genitive indicates that someone is cleansed or released from her emotions (the particular emotions someone is released from will not persist after the katharsis; it is the nature of katharsis to terminate a particular experience of emotions). The difference between these two readings is immense, but a simple and definitive decision between the two ways to construe the genitive seems to be difficult.
What the difference between the objective and the separative genitive amounts to can be fleshed out by the consideration of two competing and by now almost classic interpretations of the Aristotelian katharsis. The first of these interpretations assumes that the emotions of pity and fear must be the objects of the katharsis and that the enhancement that these emotions undergo is a moral one. This would associate the peculiar effect of the tragedy with ethics and moral psychology rather than with medicine or religious rituals. Such a view can be justified, for example, by the observation that Aristotle’s Poetics establishes many links between tragedy and ethics: tragedy is said to be the mimēsis of human actions, and human action is the subject of ethical decisions. Similarly, the Poetics emphasizes the role of the character of its protagonists and the development of a good and virtuous character seems to be the ultimate end of Aristotle’s ethics. Last but not least, the defining effect of tragedy is an emotional one, and in Aristotle’s ethics it is a person’s emotions that indicate whether she has a good and virtuous or a corrupt character. Looking from such an ethical perspective, the question would be why and how a person would benefit from having her emotions cleansed or enhanced, since for Aristotle the having of emotions as such is not blameworthy. Since Aristotle’s ethics is very explicit that the virtuous person has the right amount of emotions and that it is only the excessiveness of emotions that is blameworthy, it seems to follow easily that, if there is a respect in which the theater-goers’ emotions could need a sort of cleansing, it is the excessive amount of his or her emotions. The idea would be, then, that the poetic katharsis cleans or purifies the spectators’ emotions in that it abolishes their excessive amount and leaves behind the moderate and virtuous amount of emotions. The Enlightenment philosopher Gotthold Ephraim Lessing (1729–1781), one of the main supporters of this kind of reading, actually says that in the course of katharsis the passions are turned into virtuous attainments.
The competing classic reading of the Aristotelian katharsis goes with the alternative construction of the genitive: it assumes that in the course of the katharsis the spectator of the tragedy experiences something like a discharge of his emotional state and will, in the end, be released from emotional excitement – just as a patient can be released from toxic or inflammatory liquids by venesection or similar treatments. The originator of this reading, Jacob Bernays (1824–1881), an uncle by marriage of Sigmund Freud, draws heavily on the medical use of katharsis and argues that poetical katharsis is itself like a medical treatment of a pathological state. Bernays looks for support in this reading to Aristotle’s Politics, where katharsis is mentioned in the context of pathological states and explicitly compared to healing. Also, Bernays and other supporters of the medical interpretation pointed out that, for Aristotle, emotions are bodily alterations and must hence be treated like any other abnormal bodily phenomenon.
Both readings have problems of their own. The ethical reading seems to presuppose a vulgar reading of Aristotle’s doctrine that virtue is a mean and requires a middle amount of emotions. For Aristotle this means no more than that the virtuous person has emotions in an appropriate way, that is, has them when she should, for whom she should, and for the reasons why she should. The appropriateness depends on characteristics of the agent and of the particular situation; there is, however, no absolute amount of emotions which defines the right mean for all possible agents. The ethical reading, as sketched by Lessing, has to assume such an absolute amount of excessive emotions (in all or the majority of spectators) which tragedy reduces to an average or medium level of emotionality. This latter idea can be found in certain versions of Aristotelianism (it is called metriopatheia – middle emotions as opposed to the Stoic ideal of apatheia: see below), but not in Aristotle’s ethics. And more generally, the ethical reading just presupposes that tragedy is meant to have a moral-pedagogical effect. Why is this so? The fact that Aristotle is the author of famous books on ethics by no means implies that he wishes to assign a moral purpose to each and every cultural practice. Bernays’ pathological reading has also its problems. It draws heavily on a passage in Aristotle’s Politics where katharsis is compared with healing (Pol. VIII.7, 1342a8–11); however, this is just a comparison and the passage is not even about tragedy. Also, why should we think that Aristotle regards emotions like pathological states or like something by which one gets polluted or intoxicated? Bernays and his followers keep emphasizing that Aristotle describes emotions as bodily alterations. This is true, but it does not imply that katharsis is primarily seen as a physiological and not as psychological effect. After all, the arousal of the emotions of pity and fear did not make use of the somatic aspects of these emotions; why, then, should we expect the solution and the calming down of the same emotions primarily to utilize the somatic features of an emotion?
Scholars who try to work out alternatives to the ethical and to the pathological reading usually refer to the passage in the Politics that was the crown witness for the pathological interpretation by Bernays. This passage in book VIII starts off with the question of whether and how melodies and rhythms can be used for pedagogical purposes, and notes – very much like Plato in the Republic – that melodies and rhythms can imitate emotions and types of character and can thus also affect the emotional state of the listener. However, he slightly broadens the scope of philosophical consideration of music in that he takes into account that music can not only be useful for pedagogical purposes, but also for recreation, amusement, relaxation, and therapy. In the passage mentioned above he does mention the quasi-pathological case of people who fall into a state of mystic or religious frenzy and can be relieved by certain melodies; however, he does not generalize this particular case, but rather uses it to illustrate the more widespread non-pathological case:
Those who are influenced by pity or fear, and every emotional nature, must have a like experience, and others in so far as each is susceptible to such emotions, and all are in a manner relieved [katharsis] and their souls eased [kouphizesthai] together with a feeling of pleasure. In a similar manner the melodies that refer to actions bring about an innocuous pleasure to mankind.
(1342a11–6; tr. B. Jowett, slightly altered)
People who undergo episodes of pity and fear and other emotions can be eased and relieved by certain melodies. This seems to be the sort of impact of passionate melodies that Aristotle wishes to highlight and which he describes as katharsis or ease. It comes together with a feeling of pleasure; and just like the pleasure that is effected by a certain type of melody, namelu the melodies that are referred to actions, this kind of pleasure is an innocuous one. All this is very much compressed, but Aristotle seems to think that such a pleasant relief from certain strained passionate states is one of the possible benefits that we gain from the use of music. The use of katharsis in this passage seems to refer to the quality of experience we have when we are relieved from such states, and does not imply, as both the ethical and the pathological reading do, that anything has been enhanced or improved (as, for example, the quality of emotions, the character or the physical state of the spectator). Still, the entire passage is about music and not about tragedy, so that it is uncertain to what extent this use can be applied to the katharsis of the drama; but at any rate it offers an example of a use of katharsis that is closely related to the pleasant experience we undergo when states of pity and fear are solved or calmed down – and this is comparable to the setting that we can infer from the katharsis formula in the Poetics. In contrast to Plato’s stance in the Republic to the effects of music and poetry, it is also significant to highlight that there are pleasures – aesthetic ones – that are at least “innocent.” And once one undertakes to explore the possibility that the Aristotelian katharsis of tragedy could, in principle, refer to a pleasant relief from the painful emotions of pity and fear, it is worthwhile to compare the passage from the Politics with Aristotle’s description of the tragic pleasure that he gives in chapter 14 of the Poetics: “for it is not every pleasure, but the appropriate one, which should be sought from tragedy. And since the poet ought to provide the pleasure which derives from pity and fear by means of mimesis, it is evident that this ought to be embodied in the events of the plot” (Poet. 14, 1453b10–4; tr. S. Halliwell).
The kind of pleasure that is considered appropriate for tragedy is the one that derives from pity and fear by means of mimēsis. Unfortunately, this formula is ambiguous again. It could mean that we feel the pleasure that is produced by mimēsis – even while we are feeling pity or fear. However, pity and fear are supposed to be painful experiences – does Aristotle try to say that in the peculiar case of aesthetic experience people take pleasure in what would be painful in real life? This is one way of reading this passage. An alternative reading would try to align this formula with what we know from Politics VIII, that is, that the relief from pity and fear is a pleasant experience. Accordingly, one could argue that the pleasure mentioned here in the Poetics is the pleasure that follows from the termination of pity and fear; if something like this is meant here, the passage would by and large describe the same process that we know from the katharsis formula: that tragedy first arouses the emotions of pity and fear and finally relieves us of them, which is – as we can add now – an easing and, hence, pleasant experience.
In the Roman world Horace’s Ars Poetica holds a similar place as Aristotle’s Poetics in Greek culture. In the famous slogan that the poet either wishes to profit or to delight (aut prodesse … aut delectare), or to deliver at once both the pleasures and the necessaries of life (v. 333–4), he clearly endorses a pleasant and emotional effect of poetry. It is not sufficient for the work of poetry, he says, to be beautiful; it should also be capable of moving the hearers and of bringing their souls to wherever the poet wants – to laughter, crying, mourning, joking, anger, and so forth (v. 98–107). He also addresses the question of how to depict a hero, such as Achilles, as irascible and inexorable (v. 120–2). It clearly follows, then, that the topic of emotions and of the emotional impact of poetry is present and prominent in Horace; even so, emotions are not as essential as in Aristotle’s katharsis formula. More precise references to the Aristotelian emotions of pity and fear can also be found in the texts of the Roman rhetoricians (e.g. Cicero, De Oratore II. 189 and III. 215; Quintilian, Institutio oratoria VI. 2, passim).
When it comes to philosophical reflections about the arts and their impact on the human soul, the Greek legacy manifested itself through the continuing influence of the great philosophical schools, the Academy, Aristotle and the Peripatetics, Epicurus and Epicureans, and, above all, the Stoa. And though many of the great Roman philosophers and intellectuals see themselves as eclectic, selecting and adapting ingredients of various philosophical schools, the influence of variations of Stoic philosophy is often prevalent – especially with regard to the theory of emotions. The Stoic attitude to emotions is different from Aristotle’s, most notably in that they do not distinguish between appropriate and inappropriate, moderately or excessively felt emotions, but see all emotions, all pathē, as similarly bad and contrary to the highest good. For the early Stoa, an emotion is an excessive impulse which is contrary to the dictates of reason (Long/Sedley 65 A). It is generated when we give our approval to an impulse for ends that are not strictly speaking good or virtuous. For example, when we approve of a desire for something that seems to be pleasant in the future, the resulting emotion will become excessive and will disrupt our rationality, since it cannot be stopped or controlled anymore – just as someone who starts running, but then will not be able to stop at will. This is why emotions are seen as a disturbance within our soul; they are opposed to the ideal of an emotionally undisturbed life: apatheia – the absence of any emotion. Also, whereas Plato and Aristotle tend to deal with phenomena of irrationality by appealing to several parts of the soul, the early Stoa tries to model the emotions as occurring by way of conflict within the same monolithic soul that is also responsible for making rational judgments; even the emotions themselves are called “fresh beliefs” or judgments (Long/Sedley 65 C) – since, after all, they depend on our explicit approval. Emotions are irrational, then, not because they originate in a non-rational part of the soul, but, first, in that they rest on a false judgment (false, because they approve the impulse and its end as if they were good, though they are definitely not), and, second, in that they have the tendency to get out of control. It is for this latter reason that in the context of Stoic philosophy episodes of emotions are also compared to or even equated with temporary states of madness.7
In the light of this orthodox account of Stoic emotions, with its ideal of a life that is free of disturbances of the soul, one should expect that Stoic philosophers would never endorse a practice like tragedy which is famous for its power to arouse strong emotions in a great number of people. This, at least, is the theory; the historical reality, however, seems to be much more complex. Chrysippus, a main representative of early Stoic philosophy, is said to have been an admirer of Euripides’ tragedies; according to Diogenes Laërtius (VII.180), Chrysippus quoted almost the complete text of Euripides’ Medea in one of his works. A possible reason for this interest might be that he found in the character of Medea (who is aware that she is going to do something wrong, but nevertheless stays to be overpowered by her vengefulness) the kind of inner conflict that, according to his own teaching, leads to the generation of emotions.
Probably the most interesting figure for the question of how Stoic philosophy might be reconciled with an interest in tragedy is Seneca, who was both a Stoic philosopher and a great dramatist. As a philosopher he dedicated an entire treatise to the emotion of anger (De ira), in which he vividly describes the disastrous, sometimes even war-inducing effects of this particular emotion. As a dramatist he seems to follow the example of Euripides who had a reputation for depicting his heroes as being caught up in a whirlwind of passion. Because of the apparent conflict between the goals of a Stoic philosopher (as he presents himself most impressively in De ira) and the emotional effect of tragedies, it has even been doubted that Seneca the philosopher and Seneca the tragedian was actually one and the same person. On closer examination, however, it seems that there are several connections between Seneca’s philosophy and his dramas – though the precise nature of this relation is still a matter of dispute.8 To begin with, it might be rewarding for a Stoic philosopher to use the tragic storyline as a kind of cautionary tale by depicting the tragic heroes as acting in a state of uncontrolled emotions. This might be called the “apotropaic” function of tragedy: to warn the audience against the disastrous consequences of surrendering oneself to irrational passions. At least, some of Seneca’s dramas, such as Medea, Phaedra, Agamemnon, and Thyestes, might be meant to have such an effect. This deterrent function, however, would offer only a relatively general connection with the goals of Stoic philosophy. Some interpreters therefore tried to identify a more substantial echo of the Stoic philosophy in some of the comments and warnings expressed either by the chorus or by certain characters of a drama. For example, in the Medea there is a nurse who describes the visible symptoms of Medea’s emotional state in a very vivid way that comes close to a philosophical characterization (v. 382–390).
Still, one could object that the emotions that are enacted on stage will probably have a similar emotional impact on the spectators and that, from a Stoic point of view, such an excitement of the soul must be an unwelcome consequence. Concerning the liberal arts in general, Seneca complains that none of them liberates us from our fear, relieves us from our desires, or delimits our libido (Epistulae 88, 3). How can he then approve of a practice that is likely to increase our emotional excitement? This is the place where one of Seneca’s most significant contributions to the Stoic account of emotions comes in. Seneca (probably influenced by the unorthodox views of the Stoic philosopher Posidonius, who tried to combine the Stoic theory of emotions with Plato’s model of the tripartite soul) suggests that in the generation of a particular episode of emotion we have to distinguish three successive stages, the first of which consists in an involuntary impulse (ictus); it is only after this first impulse, at the second stage, that we are able to (voluntarily) assent or dissent to this spontaneous movement; by the third stage, the emotions have already got out of our control (De ira II.4). What happens to the spectator of a stage play, Seneca explains (De ira II.2), is like the first stage of an emotion, which is not yet an emotion itself, but the beginning of or the prelude to a full-fledged emotion. What we experience in theater is only this first impulse – like the shiver when being splatted with cold water, like a sudden vertigo when facing a profound abyss, like spontaneous blushing or sexual erection – and this impulse, he says, is no more anger than it is sorrow (tristitia, corresponding to pity) or fear (timor). It is only when we succumb to these impulses (and assent to them) that a full-fledged emotion is generated. Seneca’s three-stage model of emotions thus opens the possibility that we are affected in the theater in some particular way and that we deal with this initial affection in one way or the other without having the kind of emotion that the Stoics see as irrational. The gap thus provided between the initial affection that the theater-goer inevitably experiences and the full-fledged harmful stage of an emotion leaves room for ideas about aesthetic pleasures and pains and the virtually beneficial impact of tragedy.
Stanford (1983) provides a very readable survey on the role of emotions in Greek tragedy. For discussions of particular types of emotions, see Konstan (2006). A stimulating treatment of some aspects discussed in connection with Plato’s Republic can be found in Moss (2007). For the study of Aristotle’s Poetics, Halliwell (1986) and Halliwell (2002) are indispensable reads; Halliwell (2003) offers a very condensed summary of problems concerning the interpretation of Aristotelian katharsis. Historical testimonies of the notorious debates about Aristotelian katharsis are collected in Luserke (1991). The specific pleasures that are involved in tragedy are treated in Heath (2003). For the philosophical impact of Seneca’s tragedies one should consult Chaumartin (2014), Fischer (2014), Gill (1997), Schiesaro (1997) and Staley (2010).