III.5 Reflecting emotional states of mind: Calmness versus agitation

5.1 Introduction

§1. Drama texts incorporate a multitude of characters’ voices, embodied by different actors. [1] These multivocal performative aspects furnish ideal opportunities to explore various pragmatic phenomena in the plays’ language, three of which are discussed in III.2, III.3, and III.4. The characters’ verbal expressions of their emotional states of mind are an essential component of dramatic interactions, and thus deserve pragmatic analysis as well. Particles are among the linguistic features that may reflect emotional states of mind; it is the goal of the present chapter to illuminate how they do so. [2]
§2. The concept of emotions is notoriously difficult to define and describe consistently. [3] As Sanford and Emmott 2012:191-192 outline, theories of emotion tend to either “emphasize the role of an experiencer’s judgment (appraisal),” or “highlight the role of an experiencer’s body (…) in producing an emotion.” The authors note that there are many variants of both theories. [4] Often the cognitive and somatic sides are combined: a person’s appraisal of an event influences her bodily reaction, that is, her tendency to be drawn toward or away from an object. [5] On top of the definition problem, the term “emotion” and its conceptualization are highly culture-dependent, as many scholars point out. [6] The ancient Athenians will therefore have conceptualized emotions differently from modern English-speaking scholars. However, the terms “emotion” and “emotional state of mind” will be used here nevertheless, as this chapter does not aim to investigate historical or philosophical elements that inform the cultural side of emotions.
§3. A caveat specific to the study of emotionality in drama is that the speakers are fictional characters. The emotions themselves and the ways of revealing them may very well be based on idealizations and stereotypes, rather than on actual human experience. However, any interaction of the characters on stage has to be at least recognizable and comprehensible to the audience—an assumption that underlies also III.2, III.3, and III.4. [7] Even if these interactions are stylized, they are grounded in reality. The actors’ masks expressed only part of the characters’ emotions; any nuance and alteration had to come from language, prosody, and gesture.
§4. This chapter complements previous research on emotions in Greek literature by analyzing linguistic signs of calmness and agitation, in particular by means of particle use. The way in which characters (in fact the playwrights) organize the flow of discourse is explored, in particular the use of particles, under these states of mind. In addition, examples of particle use are discussed that are influenced by the more global temperament of certain tragic characters (see §§78-87 below).
§5. The chapter mainly discusses broad states of mind, rather than specific emotions such as sadness or fear. The reason is that calmness and agitation [8] are often clearly indicated by the context, and reflected in discourse organization and particle use, whereas specific emotions tend to require specific clues, such as the explicit mentioning of the feeling. The focus here is not on such semantic expressions, but on the pragmatic and performative side of communication. However, anger is discussed separately: the broad co-text and context usually indicate the presence of anger better than other specific emotions, if no explicit label is given.
§6. The expression of calmness and agitation is closely linked to the interaction on stage; while some emotions are more interactive than others (see §53 below on anger), it is always crucial to take into account the interactional context for the interpretation of linguistic patterns. A speaker’s specific communicative goal in a certain situation may require the display of a higher or lower degree of arousal (see §25 below).
§7. Another important general point to keep in mind is that linguistic features associated with a certain emotional state, whether a particle, a syntactic construction, or a strategy like repetition, do not encode this emotion all by themselves, but only in combination with certain co-texts and contexts. That is, the same linguistic feature may carry a different function in another co-text and context. The chapter illustrates this multifunctionality for several linguistic features: see §28 on the order of subordinate and main clauses, §67 on appositions, §77 on repetition, §§34-42 on μέν, §§51-63 on γε, and §§85-86 on δέ. In other words, some functions unrelated to emotionality will also be discussed, in order to clarify the role of the different co-texts and contexts.
§8. The chapter starts with an overview of the scholarship, both on emotions in ancient Greek texts (§§9-21) and on calmness versus agitation in general (§§22-25), and a discussion of my use of its insights and methods. 5.2 analyzes linguistic reflections of a calm state of mind (§§27-43), and of an agitated state of mind (§§44-50). 5.3 devotes attention to γε in different comic contexts (§§51-63). This is followed by two tragic case studies that contrast calm discourse to agitated discourse (§§64-87), and by the conclusions (§§88-95).

5.2 Approaches to emotions

5.2.1 Emotions in ancient Greek texts

§9. The elaborate scholarship on emotions in ancient Greek literature in general, and in Classical drama in particular, can be divided into two groups. Most studies adopt a historical, literary, or philosophical approach, and focus on the macro-levels of cultural differences between Greek and English emotional terms, or the literary meaning of emotions. A smaller number of works delves into the micro-level of the linguistic expressions of emotions. This section gives examples of both kind of research.
§10. Recent publications on the ancient Greeks’ conceptualization of emotions include Konstan 2006, Cairns 2008, Fulkerson 2013, and Sanders 2013 and 2014. [9] Konstan explores the differences between those conceptualizations and our modern ones. [10] Emotion terms such as ὀργή (~ “anger”), φιλία (~ “friendship”), and φθόνος (~ “envy”) can in fact not be translated directly into English emotional labels. Greek terms may be more widely applicable, and specific English terms may be absent from the Greek vocabulary. However, Cairns argues that in the cases of pity, jealousy, and pride, Konstan sees larger differences between the Greek and the English conceptualizations than are in fact justified, because Konstan focuses too much on English labels for specific emotions. For example, the fact that classical Greek has no term for “pride” does not mean that the phenomenon was absent from classical Greek society or literature, in Cairns’ view. Similarly, Sanders 2013 argues that sexual jealousy did exist in classical Greece, in contrast to Konstan’s claim.
§11. Fulkerson 2013 is a monograph on regret in antiquity, describing, like Konstan’s work, differences between the ancient and modern conceptions of this emotion. The author analyzes sources from Greek epic, tragedy, historiography, New Comedy, and several Latin genres. She argues, for example, that status was more important in the ancient conceptualization of remorse than in the modern one.
§12. Sanders 2014 argues for complementing the lexical approach to ancient emotions with a socio-psychological approach. This author offers an analysis of envy and jealousy in Classical Athens, describing the Athenians’ experience, the expression, and the literary representation of these emotions. Tragedy and comedy are among the sources in which he analyzes φθόνος. Sanders outlines “scripts,” that is, stylized cognitive scenarios, which correspond to certain emotion terms. [11] He concludes that φθόνος for example corresponds to twelve different scenarios, which are similar to those covered by English envy, jealousy, and rivalry.
§13. Visvardi 2015 adopts a literary approach to ancient Greek emotions, with a specific focus on the tragic chorus. The author analyzes choruses that enact fear and pity, and compares them to the depiction of emotions in Thucydides’ Histories. Both the tragedians and Thucydides, she argues, display sensitivity to the motivational power of collective emotion in the Athenian institutions.
§14. Numerous scholars focus on the literary function of specific emotions in tragedy, analyzing individual plays, sometimes even in relation to individual characters. [12] For example, Thumiger 2013 analyzes the connections between eros and madness in several tragedies. She concludes that erotic emotion never brings comfort but only has negative consequences. This has to do with genre conventions, as well as with the strict individuality connected to eros. In the same volume, Sanders discusses several emotions of the Euripidean Medea, and argues that sexual jealousy is part of her motivation. Other examples are Gerolemou 2011 on female madness in tragedy, and Provenza 2013 on the portrayal and the function of madness in Euripides Heracles. All of these studies offer rich analyses of the specific emotion(s) in the play(s) they discuss; however, they do not examine the syntax, pragmatics, or discourse organization connected to emotional expressions.
§15. As for the linguistic expression of emotion, scholars focus on different possibilities. [13] First, emotions can be expressed with interjections, which is the topic of Nordgren’s work on Greek drama (2012 and 2015). The category called expressive interjections, such as αἰαῖ, οἴμοι, and φεῦ, [14] express surprise, pain and vexation, lamentation, or joy, each of them with several specific nuances. [15] These interjections can be considered an unmediated way of emotion expression.
§16. Second, one can spotlight semantics, that is, when characters explicitly name their feelings. This is, for instance, the main focus of Schnyder 1995 on fear in Aeschylus. The author describes the vocabulary of fear used in several plays, including metaphors, thereby identifying differences in the fear vocabulary between Aeschylus, on the one hand, and epic and lyric, on the other. [16]
§17. More important to the current investigation are the levels of syntactic constructions and of discourse organization. The 2015 article by Luraghi and Sausa focuses on the linguistic constructions associated with different emotion verbs in Homer. The authors argue that verbs denoting anger, hate, and envy, which typically take a nominative-dative construction, “are construed as complex and potentially interactive, with experiencers that have agent properties” (16). The people or other agents that caused these feelings “are conceptualized as likely to react” (16). Verbs of longing, loving, and desiring, in contrast, often take a nominative-genitive construction: the subject is construed as having no control over the event, and no cause is mentioned. This group of emotion verbs, then, “does not imply any interaction.” (18) Regarding other verbs, “the NomDat construction is associated with verbs of social interaction, while the NomGen construction is associated with verbs of hitting, touching, reaching or trying to reach.” (21) Luraghi and Sausa’s results show that Homeric Greek displays syntactic differences between the representation of anger and similar emotions, on the one hand, and less interactive emotions, on the other hand. [17] Their case study thus illustrates that the expression of emotions influences grammatical choices in ancient Greek literature.
§18. The influence of emotions on discourse organization in tragedy is discussed by Mastronarde 1979 and Stanford 1983. Mastronarde’s remarks are part of his argument about contact among tragic characters; for Stanford, conversely, emotions are his main focus. Both authors observe, for example, that antilabe may express agitation. [18] Mastronarde further discusses “suspension of syntax” beyond an utterance in stichomythia, that is, turns of speaking that are syntactically incomplete, and are potentially completed in a later turn; in Euripides this technique may emphasize a character’s strong feelings, as it emphasizes the “self-absorbed continuation of her own thoughts” after an utterance by another speaker (62). In other words, Mastronarde shows that discourse organization, such as the build-up of syntactic structures, may be connected to characters’ emotional states of mind. Stanford identifies several linguistic and stylistic markers of specific emotions, such as hyperbole, used to express anger, or single-word repetition, used to express excitement. [19] This scholar analyzes both the emotions experienced by characters or choruses and those aroused in audiences.
§19. As far as particles and emotion expression in ancient Greek are concerned, little has been written so far. Some remarks can be found in the ancient grammarians’ writings. The treatise Περὶ Ἑρμενείας (On style) [20] states that δή contributes πάθος to a certain passage in Homer. [21] Similarly, Apollonius Dyscolus (second century CE) claims that γε may intensify the emotion expressed by καλῶς γε. [22] In modern times commentators offer notes on particles’ emotional quality. They observe, for example, that καί in questions may indicate the speaker’s surprise, doubt, or indignation when used at the start of questions. [23] However, commentators usually do not clarify which co-textual and contextual features are relevant to their interpretation of a certain particle instance. In fact these features are crucial: particles never express an emotion or emotional state of mind by themselves.
§20. All in all, several approaches are available to emotions and emotional states of mind in ancient Greek. The present chapter fills a gap in previous investigations, in that it analyzes how certain emotional states of mind relate to particle use. That is, such states do not only influence linguistic choices in the direct description of emotions with verbs, discussed by Luraghi and Sausa, but also linguistic choices concerning discourse organization, which can be considered indirect manifestations of emotional states. Since particles only carry out their functions in combination with co- and contextual features, it is important to look at the connection between characters’ emotional states and discourse organization more generally, including the contribution of particles. As it will turn out, stretches of discourse larger than sentences need to be taken into account, as well as the interaction among characters.
§21. In this analysis one should keep in mind that most utterances tend to perform multiple communicative actions simultaneously. The expression of emotions may be an utterance’s main pragmatic goal, or may be a secondary goal accompanying a different main goal. The former is generally the case when a speaker utters no more than an interjection. [24] We encounter the latter situation when, for example, a threat or even the description of an outrageous past event conveys the speaker’s anger. [25] Similarly, a carefully composed argumentative speech may show signs of the speaker’s calmness while her main goal is to try to persuade the addressee. [26] Any utterance may be influenced by the speaker’s emotional mindset to a higher or lower degree; there is no black-or-white distinction between utterances that express emotions and utterances that do not do so. I will therefore examine utterances that carry the reflection of some emotional state of mind while achieving various paragmatic goals.

5.2.2 Calmness versus agitation beyond ancient Greek

§22. Scholars of emotions often use a dimension of arousal for distinguishing among different emotions. [27] This dimension represents a continuum between calmness (or “deactivation”; see Reisenzein 1994) and agitation (or “activation” in Reisenzein’s terms). The other dimension is that of valence (positive-negative). Juslin 2013 uses the following image, based on the work of Russell 1980, to show the possible distribution of emotions in a two-dimensional view. [28] The degree of arousal (calm-agitated) forms the vertical axis, the degree of pleasure or displeasure the horizontal axis: [29]
Figure 1: Two-dimensional emotions model, from Juslin 2013:256
In order to fully distinguish between emotions that occupy a similar location in this two-dimensional space, Reisenzein 1994 proposes to embed this theory within a cognitive theory that takes into account the appraisals themselves. He mentions the examples of disappointment, envy, and shame, which involve roughly the same proportions of displeasure and activation, but are caused by different (interpretations of) situations. [30]
§23. Though it is not my aim to classify the emotions found in tragedy and in comedy, the dimension of arousal and the notion of appraisal causes are relevant to the upcoming analyses. The valence dimension (see §22) will not concern us here, because positive emotions are less frequently expressed in the dramatic corpus, and are usually less clearly indicated on the linguistic level than negative emotions are. [31] The presence of positive emotionality is therefore harder to detect for us as readers. In contrast, the degree of calmness or of agitation brought about by a negative emotion tends to be identifiable, even if it is less clear which particular emotion is expressed that corresponds to a lower or higher degree of arousal. A linguistic expression might not make it clear exactly, for instance, whether the speaker is desperate, frustrated, or annoyed, but it usually communicates whether or not she is agitated. The contexts also tend to make clear the appraisals on which emotional states of mind are based. These appraisals, such as the interpretation of a past action as an insult causing anger, are among the nonverbal indications of a certain state of mind.
§24. Whether or not a speaker seems calm, in other words, can be inferred from the broad co-text and context. Relatively long speaking turns give more indications of a speaker’s calmness or agitation than short utterances. [32] Nevertheless, also in short turns contextual cues such as an utterance’s main goal or a speaker’s social status provide hints about calmness. For example, a high-status speaker who gives an order or piece of advice to an inferior usually does not show agitation, because she has no reason to do so. Another context that is connected to calmness is an official speech by a leader figure (see e.g. (t2) below). Different contexts, then, can provide cues about a character’s degree of arousal.
§25. Note that the literary functions of calmness in tragedy and comedy are different. Tragic calmness may betray characters’ ignorance about advancing disaster (such as Oedipus in (t15) below), [33] demonstrate the secure power of gods, [34] or be needed for narrating a story, such as a messenger speech. [35] Comedy, conversely, does not contain “ironic” calmness before misfortune; the high social status that some comic characters adopt is not as absolute as that of tragic gods; and long narratives are rare. In both genres, nevertheless, calmness is associated with relatively long speeches that have a persuasive goal. [36] However, whereas in tragedy it is often a matter of life or death if someone is persuaded, [37] in comedy such speeches may concern absurd or mocking topics, such as comparing the pólis to a woollen fleece, [38] or presenting birds as gods. [39] Despite these differences in underlying reasons, in my tragic and comic corpus the speakers of such long speeches must generally suppress their excitement, if present, in order to be persuasive.

5.3 Reflections of calmness and agitation

§26. Calm discourse has different linguistics characteristics from agitated discourse, and particle use is among the evidence for this difference. The reflections of calmness and agitation can be perceived on several linguistic levels.

5.3.1 Calmness

§27. One of the linguistic features that may be connected to calmness is the syntactic build-up in which a subordinate clause, or more, precedes its main clause. This type of composite sentence, in other words, begins with a clause that cannot syntactically stand on its own, but creates the expectation that something will follow. A few examples are found in Socrates’ calm utterances in the following passage. He explains the clouds’ appearance to Strepsiades, after the latter has expressed his surprise about that:
345      Σω. ἀπόκριναί νυν ἅττ’ ἂν ἔρωμαι.
345bis Στ. λέγε νυν ταχέως ὅτι βούλει.
            Σω. ἤδη ποτ’ ἀναβλέψας εἶδες νεφέλην κενταύρῳ ὁμοίαν
                  ἢ παρδάλει ἢ λύκῳ ἢ ταύρῳ;
347bis Στ. νὴ Δί’ ἔγωγ’. εἶτα τί τοῦτο;
            Σω. γίγνονται πάνθ’ ὅτι βούλονται· κᾆτ’ ἢν μὲν ἴδωσι κομήτην
                 ἄγριόν τινα τῶν λασίων τούτων, οἷόνπερ τὸν Ξενοφάντου,
350           σκώπτουσαι τὴν μανίαν αὐτοῦ κενταύροις ᾔκασαν αὑτάς.
           Στ. τί γὰρ ἢν ἅρπαγα τῶν δημοσίων κατίδωσι Σίμωνα, τί δρῶσιν;
           Σω. ἀποφαίνουσαι τὴν φύσιν αὐτοῦ λύκοι ἐξαίφνης ἐγένοντο.
Aristophanes Clouds 345-352
So. Now answer some questions for me.
St. Ask away, whatever you like.
So. Have you ever looked up and seen a cloud resembling a centaur, or a leopard, or a wolf, or a bull?
St. Certainly I have. So what?
So. Clouds turn into anything they want. Thus, if they see a savage with long hair, one of these furry types, like the son of Xenophantus, they mock his obsession by making themselves look like centaurs.
St. And what if they look down and see a predator of public funds like Simon, what do they do?
So. To expose his nature they immediately turn into wolves.
The exchange of questions and answers is a parody of Socratic style, a way of speaking that generally does not require agitation. Especially Socrates is calm, since he possesses more knowledge than Strepsiades. [40] On multiple occasions in this scene Socrates utters subordinate clauses before their main clauses: an order that projects more to come through the syntactic incompleteness of the subordinate clauses. [41] In Socrates’ utterance in 348-350 two such clauses (ἢν μὲν ἴδωσι κομήτην /ἄγριόν τινα τῶν λασίων τούτων and οἷόνπερ τὸν Ξενοφάντου) and a participial clause (σκώπτουσαι τὴν μανίαν αὐτοῦ) interrupt their main clause (κᾆτ’… κενταύροις ᾔκασαν αὑτάς). In 352 the participial clause ἀποφαίνουσαι τὴν φύσιν αὐτοῦ comes earlier than its main clause λύκοι ἐξαίφνης ἐγένοντο. In addition, the imperative ἀπόκριναι (“answer”) in 345 semantically projects the utterance of an object such as “my question.”
§28. Parallels from Aristophanes as well as tragedy suggest that this order of clauses is often found in calm situations. [42] The phenomenon may relate to a calm state of mind, because the speaker lingers on one hypotactic construction for a relatively long time. [43] This is not to say that it is the function of this clause order to “express” calmness; rather, it may be one of the effects of calmness that a speaker opts for this order in certain cases. [44] Moreover, to be more precise, subordinate clauses preceding their main clauses are not confined to calm contexts; but when this order occurs in agitated contexts, other features tend to be present to show the different state of mind (see e.g. Aristophanes Frogs 561, cited in (t12) below). As with the analysis of particles or other linguistic features, there is no one-to-one mapping of a certain emotional mindset and the language uttered; it is necessary to take into account the co-text and context beyond one sentence (see §7 above).
§29. Note that the mentioning of the projection that this clause order produces, or, conversely, of incremental style (see 5.3.2 below on agitation), is meant to refer to hearers’ online reception of utterances. This kind of reception is different from the map view of readers, who easily connect a main clause to a subordinate clause that occurred several lines earlier, for example. For hearers the meaning of an utterance is incrementally updated, with every discourse act adding new information. It is for hearers, of course, that the ancient playwrights composed their poetry.
§30. Another feature connected to calmness is pragmatic projection, mainly through vocatives and priming acts. [45] Vocatives near the beginning of a speaking turn are pragmatically not complete on their own, but produce the expectation that several more acts, addressed to the particular addressee, will follow; therefore they have the potential to work as a floor-holding device, in order for the speaker to utter an elaborate turn. Vocatives are often combined with a calm attitude. [46] Priming acts are short discourse acts that start a multi-act move; “priming” refers to the cognitive priming of the concept or referent that the act mentions. [47]
§31. The following excerpt from Assemblywomen contains an example of a priming act. Here Praxagora begins a lengthy monologue after sending away an incompetent speaker from the assembly platform: [48]
170    (Πρ.) αὐτὴ γὰρ | [49] ὑμῶν γ’ ἕνεκά μοι λέξειν δοκῶ |
                   τονδὶ λαβοῦσα. | τοῖς θεοῖς μὲν εὔχομαι |
                   τυχεῖν κατορθώσασα τὰ βεβουλευμένα. |
                    ἐμοὶ δ’ | ἴσον μὲν τῆσδε τῆς χώρας μέτα |
                   ὅσονπερ ὑμῖν· (…)
Aristophanes Assemblywomen 170-174
(Pr.) To judge from what I’ve seen of your abilities it seems best that I put on this garland and make a speech myself. I beseech the gods to grant success to today’s deliberations. My own stake in this country is equal to your own, (…)
Praxagora’s calm state of mind can be inferred from the official, serious content of this utterance and the parts of her speech that follow. [50] In 173 she uses a priming act ἐμοὶ δ’, in order to project a series of acts related to “me.” The position of μέν tells us, retrospectively, that ἐμοὶ δ’ forms a separate discourse act. Note that at the same time this act involves a syntactic projection beyond an act: the dative in the act ἐμοὶ δ’ is syntactically incomplete on its own. Additionally, even though γε may also occur in a later than act-peninitial position, it is probable that αὐτὴ γάρ in 170 is a priming act as well, which emphasizes Lysistrata’s decision to take the floor. [51] Priming acts in drama are strikingly frequent in calm contexts, such as official speeches in Aristophanes, and certain monologues in tragedy. [52] These acts can be a sign of calmness in that the speaker promises, so to say, to stick with a certain concept or referent for at least one more discourse act. [53]
§32. As my discussion of (t2) shows, particles in and after priming acts are relevant to our interpretation in two ways. The priming act itself, on the one hand, tends to contain a particle, in this case δέ, that carries out its normal function by signaling how the act relates to the preceding discourse. The examples collected suggest that δέ is indeed the most frequent particle in priming acts. The next act, on the other hand, often includes a particle, in this case μέν, that retrospectively enables us as readers to see the boundary of the priming act—which the original audience was probably able to perceive via an intonational break. [54]
§33. The following passage from Aristophanes Birds illustrates the connection between priming acts and calmness in a rather different context: not an official speech, but a 2-line utterance. Three gods (Poseidon, Heracles, and the so-called Triballian) have come to Peisetaerus in order to discuss a settlement between him and the gods; Peisetaerus demands Zeus’ girl Princess for himself in this negotiation. The three representative gods now discuss this proposal:
1679bis Ηρ. παραδοῦναι λέγει.
1680      Πο. μὰ τὸν Δί’ οὐχ οὗτός γε παραδοῦναι λέγει,
                    εἰ μὴ βαβάζει γ’ ὥσπερ αἱ χελιδόνες.
              Πε. οὐκοῦν παραδοῦναι ταῖς χελιδόσιν λέγει.
              Πο. σφώ νυν διαλλάττεσθε καὶ ξυμβαίνετε· |
                    ἐγὼ δ’, | ἐπειδὴ σφῷν δοκεῖ, | σιγήσομαι.
Aristophanes Birds 1679bis-1684
He. He [i.e. the Triballian] says, hand her over.
Po. No, by Zeus, he’s not saying hand her over; he’s just twittering like the swallows.
He. [55] All right, he’s saying hand her over to the swallows.
Po. Very well, you two negotiate the terms of a settlement; if that’s your decision, I’ll keep quiet.
Poseidon’s turn in 1680-1681 contains the swearing expression μὰ τὸν Δί’ as well as two instances of γε, revealing a certain agitation (see §§45-47 below). However, after the squabbles, in 1683-1684 he gives in and leaves the final negotiation to the other two gods, declaring that he will “keep quiet” (σιγήσομαι). That is, he indirectly states that he will refrain from further agitation: he has decided to be calm from now on. [56] In 1683 he uses καί to combine διαλλάττεσθε “be reconciled” and ξυμβαίνετε “come to an agreement,” two semantically similar words, the second of which may be considered a specification of the first. This use of καί is in fact most frequent in long monologues, where the speaker feels at leisure to formulate a concept in several slightly distinct ways; [57] the construction may therefore reflect calmness. Moreover, Poseidon’s utterance features a subordinate clause (ἐπειδὴ σφῷν δοκεῖ) before its main verb σιγήσομαι in 1684. It also contains the priming act ἐγὼ δ’. [58] I interpret these features as discourse-organizational effects of Poseidon’s calmness.
§34. Let us now consider more specific manifestations of calmness than subordinate clauses preceding their main clauses and forms of pragmatic projection. The case study of Aeschylean Agamemnon’s particular use of μέν shows the influence of specific co-textual and contextual elements on how a particle may reflect calmness. [59] This example is followed by a discussion of several uses of μέν that do not reflect calmness, in order to clarify how co-text and context determine the emotional quality of particles.
§35. Agamemnon in his eponymous play utters μέν 7 times, or with a frequency of 1.5% (out of a total of 473 words), higher than any of the other 20 main characters in the 9 tragedies of my corpus. [60] This statistic alone cannot account for specific pragmatic goals; after all, particles are multifunctional. What is remarkable about Agamemnon’s use of μέν is that he, despite his fondness for the particle, only employs certain uses of it. [61]
§36. Let us consider the following instance from the king’s answer to Clytaemnestra’s long welcome speech:
          Αγ. Λήδας γένεθλον, δωμάτων ἐμῶν φύλαξ,
915           ἀπουσίαι μὲν εἶπας εἰκότως ἐμῆι·
                 μακρὰν γὰρ ἐξέτεινας. ἀλλ’ ἐναισίμως
                 αἰνεῖν, παρ’ ἄλλων χρὴ τόδ’ ἔρχεσθαι γέρας.
                 καὶ τἄλλα μὴ γυναικὸς ἐν τρόποις ἐμὲ
                 ἅβρυνε, (…)
Aeschylus Agamemnon 914-919
Ag. Daughter of Leda, guardian of my house, you have made a speech that was like my absence—you stretched it out to a great length; but to be fittingly praised is an honour that ought to come to me from others. For the rest, do not pamper me as if I were a woman; (…)
With μέν in 915, Agamemnon projects the continuation of his discourse. [62] He warns Clytaemnestra immediately at the start of his speech that he will not only assess the length of her preceding monologue, but also, as it turns out, the content: her praise and her suggestion to walk on the purple fabric were in his view excessive. There is no δέ that answers this μέν in the discourse acts that follow it (916-919), which suggests that μέν’s projecting signal here works more globally than to mark semantic or syntactic juxtaposition. In particular, the signal here pertains to the level of acts rather than content. As a signal of discourse organization, this μέν implies that Agamemnon feels he can go on speaking for some time without any problems. Thus the use of this particle displays calmness, and perhaps also an authoritative tone. Along these lines E. Fraenkel 1950 remarks: “[t]he king is, at least up till now, completely composed, he speaks with the gracious dignity of a great gentleman.” [63] With this calmness Aeschylus in turn invites the audience to infer that Agamemnon does not suspect his upcoming murder. [64]
§37. Later in the play the king uses μέν together with a δέ in the next act, again to structure the presentation of his discourse, without indicating semantic contrast. Clytaemnestra has now persuaded her husband to walk on the fabric:
          (Αγ.) καὶ τοῖσδέ μ’ ἐμβαίνονθ’ ἁλουργέσιν θεῶν
                   μή τις πρόσωθεν ὄμματος βάλοι φθόνος.
                   πολλὴ γὰρ αἰδὼς δωματοφθορεῖν ποσὶν
                   φθείροντα πλοῦτον ἀργυρωνήτους θ’ ὑφάς.
950             τούτων μὲν οὕτω, τὴν ξένην δὲ πρευμενῶς
                   τήνδ’ ἐσκόμιζε· (...)
Aeschylus Agamemnon 946-951
(Ag.) and as I walk on these purple-dyed , may no jealous eye strike me from afar! For I feel a great sense of impropriety about despoiling this house under my feet, ruining its wealth and the woven work bought with its silver. Well, so much for that. This foreign woman—please welcome her kindly.
In this case μέν occurs in the last act of a move in which Agamemnon comments on the act of treading on the robes; the first act of the next move, about Cassandra, contains δέ. This construction, with μέν and δέ belonging to different moves, differs from the construction in which the two particles together imply semantic contrast. [65] Fraenkel 1950 ad 950 considers τούτων μὲν οὕτω “a dry, businesslike formula of transition.” [66] Again, the careful articulation of discourse organization befits a calm state of mind. For the audience, well aware of what is about to happen to the king, the irony of his calmness may have heightened the tension.
§38. One μέν construction that Agamemnon does not use is that involving a strong semantic contrast between a μέν act and an immediately following δέ act. Sophoclean Antigone favors this particle construction, however. [67] Even though this use of μέν… δέ does not signal calmness or agitation, it will be discussed here in order to clarify the role of co-textual and contextual features in our interpretation of particle constructions. In Antigone’s utterances, the μέν… δέ construction conveys the speaker’s stance, more specifically her disalignment with her addressee Ismene, and even hostility towards her. [68] That is, this use is more connected to the emotional dimension of pleasure-displeasure, by conveying a negative stance, than to the dimension of arousal.
          Ισ. οἴμοι τάλαινα, κἀμπλάκω τοῦ σοῦ μόρου;
555    Αν. σὺ μὲν γὰρ εἵλου ζῆν, ἐγὼ δὲ κατθανεῖν.
          Ισ. ἀλλ’ οὐκ ἐπ’ ἀρρήτοις γε τοῖς ἐμοῖς λόγοις.
          Αν. καλῶς σὺ μὲν τοῖς, τοῖς δ’ ἐγὼ ’δόκουν φρονεῖν.
          Ισ. καὶ μὴν ἴση νῷν ἐστιν ἡ ’ξαμαρτία.
          Αν. θάρσει. σὺ μὲν ζῇς, ἡ δ’ ἐμὴ ψυχὴ πάλαι
560          τέθνηκεν, ὥστε τοῖς θανοῦσιν ὠφελεῖν.
Sophocles Antigone 554-560
Is. Ah me, am I to miss sharing in your death?
An. Yes, you chose life, and I chose death!
Is. But I did not fail to speak out!
An. Some thought you were right, and some thought I was.
Is. Why, our offence is equal!
An. Be comforted! You are alive, but my life has long been dead, so as to help the dead.
Three times in this passage (555, 557, and 559) Antigone directly follows an act containing σὺ μέν with an act containing δέ and a first-person pronoun. [69] In this way she highlights her disalignment from her sister, that is, the contrast between their respective views—a rhetorical strategy that conveys emotional distance and hostility. [70]
§39. Ismene uses μέν… δέ constructions within one line as well, but never to stress an opposition to her sister. In 99 she even uses it to emphasize that Antigone is dear to her: ἄνους μὲν ἔρχῃ, τοῖς φίλοις δ’ ὀρθῶς φίλη “in your going you are foolish, but truly dear to those who are your own.” [71] The contrast expressed by the “local” μέν… δέ construction, then, may also be used to express emotional nearness.
§40. When μέν… δέ constructions contrast actions or situations, instead of decisions or opinions, they do not imply hostility, even if they do contrast the second and first person. For example, the angry Philoctetes in Sophocles Philoctetes 1021 and 1025-1026 uses the μέν… δέ construction twice in order to contrast his addressee (Odysseus) with himself, within a long monologue. However, he contrasts their respective situations, rather than their decisions; in this way he conveys more bitterness than hostility with these parts of the discourse. The differing nature of the co-texts, then, leads to a interpretation that is pragmatically different from the particles in Antigone’s hostile utterances.
§41. Another example that clarifies the difference between μέν… δέ used to convey disalignment and μέν… δέ expressing another kind of contrast is the following passage from Aristophanes Women at the Thesmophoria. [72] The character Euripides has just managed to distract the archer who was guarding Euripides’ kinsman, and so has found an opportunity to rescue him.
          Ευ. Ἑρμῆ δόλιε, ταυτὶ μὲν ἔτι καλῶς ποιεῖς.
                 σὺ μὲν οὖν ἀπότρεχε, παιδάριον, ταυτὶ λαβών·
                 ἐγὼ δὲ λύσω τόνδε. σὺ δ’ ὅπως ἀνδρικῶς,
1205         ὅταν λυθῇς τάχιστα, φεύξει καὶ τενεῖς
                 ὡς τὴν γυναῖκα καὶ τὰ παιδί’ οἴκαδε.
Aristophanes Women at the Thesmophoria 1202-1206
Eu. Trickster Hermes, just keep on giving me this good luck! You can run along now, kid; and take this stuff with you. And I’ll release this one. As soon as you get loose you’d better run like a man away from here and head back home to your wife and kids.
After acknowledging that the god Hermes provided good luck (1202), Euripides first orders one of his slaves to leave, marking this order with μέν (1203) as only part of what he wants to say. [73] He then describes his own intended action in a δέ act (ἐγὼ δὲ λύσω τόνδε, 1204): he will free his kinsman. Another δέ act follows (σὺ δ’ ὅπως ἀνδρικῶς, 1204), addressed to the kinsman, in which Euripides tells him, too, what he should do. [74] The acts thus juxtapose actions by different people; they do not refer to different decisions or opinions, as Antigone’s hostile μέν and δέ acts did.
§42. The following passage from Frogs contains a further exploitation of μέν and δέ in successive acts, in this case with a metapoetic goal. No particular emotional state of mind is detectable from the context; again, the specific co-textual and contextual features determine the interpretation of the particle construction. The juxtaposition marked by μέν… δέ here combines with a figure of speech for parodic effect, as well as with a priming act. The god Dionysus has just told his half-brother Heracles that he wants to bring back Euripides from Hades.
          Ηρ. εἶτ’ οὐ Σοφοκλέα πρότερον ὄντ’ Εὐριπίδου
                 μέλλεις ἀναγαγεῖν, εἴπερ ἐκεῖθεν δεῖ σ’ ἄγειν;
           Δι. οὐ πρίν γ’ ἂν Ἰοφῶντ’, ἀπολαβὼν αὐτὸν μόνον,
                 ἄνευ Σοφοκλέους ὅ τι ποιεῖ κωδωνίσω.
80             κἄλλως ὁ μέν γ’ Εὐριπίδης πανοῦργος ὢν
                 κἂν ξυναποδρᾶναι δεῦρ’ ἐπιχειρήσειέ μοι· |
                  ὁ δ’ | εὔκολος μὲν ἐνθάδ’, | εὔκολος δ’ ἐκεῖ. |
Aristophanes Frogs 76-82
He. If you must resurrect someone, then why not Sophocles, who’s better than Euripides?
Di. No, first I want to get Iophon alone by himself and evaluate what he produces without Sophocles. Besides, Euripides is a slippery character and would probably even help me pull off an escape, whereas Sophocles was peaceable here and will be peaceable there.
Slings 2002:101 argues that the repetition of εὔκολος in 82 make this verse “not only a tribute to Sophocles’ character, but also to his style,” because Sophocles regularly makes use of anaphora, that is, repetition of words at the beginnings of successive clauses. [75] Perhaps, I would add, Aristophanes also means the juxtaposition of a μέν act and a δέ act to allude to antithetic Sophoclean style. [76] The conspicuous act boundary after ὁ δ’, possibly accompanied by a prosodic break, further helps the line stand out. Together with the anaphora this priming act—typical of calm, carefully structured discourse—makes the sentence appear like an official, important message. This form may have contributed to turning Frogs 82 into “probably the most famous anaphora from Aristophanes” (Slings 2002:101).
§43. To sum up my observations in this section: a speaker’s calmness tends to be linguistically reflected in subordinate clauses preceding their main clauses, pragmatic projection through vocatives and priming acts, and certain uses of μέν. The particle δέ often figures in priming acts. However, μέν and δέ also occur in constructions that are unrelated to a certain emotional state of mind. The difference depends on several co-textual and contextual elements.

5.3.2 Agitation

§44. Let us now have a look at the way tragic and comic speakers linguistically express agitation (combined with negative feelings), the general state of mind at the other end of the arousal dimension. Commentators describe the following passage from Aeschylus Seven against Thebes, for example, as conveying “alarm” (Tucker 1908 ad loc.) or “intense emotion” (Hutchinson 1985 ad loc.). [77] Here a messenger informs the chorus (Theban women) of Eteocles and Polyneices’ death:
805    Αγγ. ἅνδρες τεθνᾶσιν ἐκ χερῶν αὐτοκτόνων.
          Χο. τίνες; | τί δ’ εἶπας; | παραφρονῶ φόβωι λόγου.
Aeschylus Seven against Thebes 805-806
Me. The men have died at each other’s hands.
Ch. Who? What are you saying? Your words are frightening me out of my mind.
In his discussion of this example Stanford 1983 adds precision to the commentators’ descriptions: he remarks that “jerky syntax often indicates emotion” (99). [78] Differently formulated, the three discourse acts in 806 are not syntactically integrated, but form syntactic units on their own. This incremental style is reminiscent of the so-called “adding style” that is often mentioned in connection to Homeric syntax. As Bakker 1997c:147 discusses, this style is defined by “the absence of syntactic anticipation.” [79] It is my impression that several successive acts without syntactic integration are typical of agitated tragic utterances. [80]
§45. As for particles, γε and δῆτα occur more frequently in agitated than in calm contexts. In general, these particles typically appear in dialogues with short turns, that is, in communicative situations in which speakers come to the foreground in their identity as communicators. [81] In such situations there is a clear focus on who is saying something, apart from the attention to what is said. The expression of agitated emotion, by its nature, tends to highlight the presence of the speaker, and therefore forms a suitable contexts for the two particles, just as dialogues in general do.
§46. Examples of γε in an agitated context are found at the end of a monologue by the dying Heracles in the Women of Trachis. [82] The hero in his rhesis speaks to the gods as well as to the people gathered at his deathbed: his son Hyllus, the chorus of Trachinian women, and an old man serving as a doctor.
          (Hρ.) ἀλλ’ εὖ γέ τοι τόδ’ ἴστε, κἂν τὸ μηδὲν ὦ,
                   κἂν μηδὲν ἕρπω, τήν γε δράσασαν τάδε
                   χειρώσομαι κἀκ τῶνδε. προσμόλοι μόνον,
1110            ἵν’ ἐκδιδαχθῇ πᾶσιν ἀγγέλλειν ὅτι
                   καὶ ζῶν κακούς γε καὶ θανὼν ἐτεισάμην.
Sophocles Women of Trachis 1107-1111
(He.) But know this for certain, even if I amount to nothing and I cannot move, I shall chastise her who has done this, even in this condition! Let her only come near, so that she may be taught to proclaim to all that both in life and in death I have punished evildoers!
Heracles on his deathbed thinks (wrongly) that his wife Deianeira had intended to kill him, and he desperately wants to punish her for that. Here, γε works to highlight a specific part of the utterance: “but know this for certain” (ἀλλ’ εὖ τοι τόδ’ ἴστε) or only εὖ in 1107, “her who has done this” (τὴν δράσασαν) in 1108, and “evildoers” (κακούς) in 1111. As Jebb 2004 [1892] ad 1111 remarks, “[t]he γε is very expressive”; he uses italics to render this in his paraphrase. The particle, in other words, works similarly to prosodic emphasis. [83] Heracles indicates that he considers his angry message and his wife’s (presumed) responsibility highly important. Through this highlighting the particle indirectly betrays Heracles’ emotional agitation. More specifically, these γε instances mark agitation in form of anger (see §§53-58 below). After 61 lines of mainly expressing pain, sadness, and desperation, but not anger (1046-1106)—without any γε—Heracles now turns his attention to the person he considers responsible for his situation, even though she is absent from the scene. That is, he turns to his feelings of anger.
§47. Recall from other chapters that γε’s pragmatic function is to highlight a specific part of the utterance as highly relevant, according to the speaker, and as implicitly contrasted to something else; indeed the highlighted part often relates to the speaker’s stance. [84] In this highlighting function, γε is comparable to the paralinguistic signs of prosodic emphasis or an exclamation mark. [85] That is to say, γε in itself does not express any emotion per se, but its use is very suitable for agitated contexts, just as prosodic emphasis and exclamation marks are.
§48. As a display of Heracles’ agitation other than γε, the passage also contains a double repetition of καί, in both 1107-1109 (three instances) and 1111 (two instances), which emphasize the speaker’s statements (“even if… even if… even” and “both in life and in death”). [86] Moreover, in the online reception of these lines by hearers, it is not immediately clear whether the κἄν clauses syntactically belong to τόδ’ ἴστε, and therefore follow their main clause, or to χειρώσομαι (1109), and therefore precede it. That is, even if they are retrospectively constructed with the latter, in their moment of utterance this structure is not yet apparent. The appearance of γε in both main clauses that frame the two κἄν clauses suggests that the prosodic emphasis is anyway on the pieces of information in the two main clauses. Regardless of the hypotactic structure, as received while hearing the lines or in retrospect, the first act κἂν τὸ μηδὲν ὦ (1107) does not project a second one of similar structure. This repetition therefore creates the impression that these remarks have great relevance to Heracles—which fits the emotionally agitated context.
§49. δῆτα, which in monologues is even more rare than γε, usually occurs in contexts of agitation. Consider the following speech by Euripidean Heracles, after he has unwillingly killed his own wife and children. He has elaborately described the miseries throughout his life, and now cries out:
(Ηρ.) τί δῆτά με ζῆν δεῖ; τί κέρδος ἕξομεν
          βίον γ’ ἀχρεῖον ἀνόσιον κεκτημένοι;
Euripides Heracles 1301-1302
(He.) Why then should I live? What advantage shall I have if I possess an accursed and useless life?
Barlow 1996 ad loc. calls this rhesis (1255-1310) “a speech of despair”; she considers 1301-1302 its “emotional climax.” [87] Indeed one can hardly imagine a more desperate question than “why then should I live?” δῆτα indicates a logical connection to the preceding discourse; at the same time, since it normally occurs in dialogues, where questions are immediately answered, δῆτα provides a tone of urgency, and draws attention to the speaker. [88] It thereby reflects and emphasizes the speaker’s agitation and desperation. [89] Menge 1999 [1914]:246 implies this emotional implication of δῆτα questions in his general translation of τί δῆτα into German: “was denn nur?” [90]
§50. Agitated speakers, then, frequently employ an “incremental” style, where subsequent acts are not projected beforehand. Additionally, the particles γε and δῆτα are connected to agitation. γε highlights a specific part of an utterance and thereby emphasizes what the speaker is agitated about. δῆτα questions in monologues convey a sense of urgency, and appeal to hearers, since they are expected to be immediately answered. The particle γε may also reflect agitation in a more specific way, which is what we will turn to next.

5.4 The different emotional and interactional associations of γε in Aristophanes

§51. The connection of γε to agitated contexts merits closer attention, because more specific relations between the particle and certain emotional and interactional contexts can be identified. In fact the particle does not just fit agitation in general, but, as already suggested concerning (t10), at least in drama it is associated with angry contexts. In addition, γε tends to occur in contexts of stancetaking, which cross-cut those of anger: stancetaking contexts may or may not involve anger or agitation in general. This section analyzes examples from both of these—sometimes overlapping—contexts; together these two uses account for most of the γε instances of in tragedy and comedy. [91] Here the focus is on Aristophanes, since γε is much more frequent there than in tragedy; the functions mentioned here do however also occur in tragedy. [92]
§52. I connect the particle’s unparalleled high frequency in Aristophanes to the playwright’s tendency to let speakers refer explicitly to their own subjective opinions, judgments, attitudes, and feelings—often in potential contrast to those of others—rather than to speaker-external topics, such as those in arguments, narratives, or gnomic expressions. Although both tragedy and comedy contain all these communicative actions, comedy tends to draw more attention to the speaker of a message, but tragedy more to the message itself. [93] This generalization may be connected to Taplin’s claim that comedy tends to pay more attention to particulars, and tragedy more to general aspects: comedy, as he puts it, “cannot universalise for long without falling over a heap of dung” (1986:173). [94] The comic emphasis on particular things fits well with a linguistic emphasis on individual speakers. In addition to this, another reason for γε’s much higher frequency in comedy seems to be the fact that comic language requires a higher degree of inferential activity from the audience than tragic language, because humor tends to be inferentially complex. γε par excellence invites inferences about elements that are implied, but not explicitly spelled out.

5.4.1 γε in angry contexts

§53. Anger is a specific kind of agitated emotion. [95] It is also an interactional emotion: it involves not only someone experiencing the emotion, but also an external agent to whom the angry person attributes responsibility for causing the anger. As Konstan 2006:45 puts it, anger “involves a judgment of intentions. That is why we do not normally get angry at stones: they can hurt us, but cannot insult us (…). Nor can we take revenge on them.” Anger is inherently a reaction to a (real or supposed) action by someone else, unlike other feelings such as joy, happiness, sadness, grief, despair, or even fear, which can all be felt without the influence of other people. [96]
§54. Konstan warns that the ancient Greek conception of anger differs from the modern English one. That is, Aristotle describes one of the ancient Greek terms for anger, ὀργή, as “a realizable desire for revenge” (2006:64) in reaction to a slight that involves contempt. Aristotle sharply distinguishes between anger on the one hand—a personal, temporary reaction to an intentional insult, with the possibility of revenge—and hatred or enmity on the other hand—a lasting, general attitude towards someone. [97] In English, by contrast, these two concepts overlap to a great extent, according to Konstan. Here we focus only on linguistic reflections of the personal, temporary feeling of anger, but the particular causes of this feeling are not analyzed; thus the appraisal causes may include not only contemptuous insults, but also general intentional harm.
§55. In tragedy and comedy anger plays a crucial role: it is often a driving force in the unfolding plot. [98] As Allen 2003 points out, anger was a central concept in fifth-century Athens, especially “in the ethical discourses that produced Athenian definitions of the good citizen, justice, and just behavior.” (78) The author detects a positive view on male anger in Aristophanic comedies: there anger is treated as the source of Athens’ independence, greatness, and egalitarianism (84). Tragedy, on the other hand, often revolves around “the angry woman” according to Allen (84), which is connected to the Athenian fear that anger would enter into the household, where it would be destructive. [99]
§56. The expression of anger in tragedy and comedy, then, tends to have different functions in terms of plot and character, just like the expression of calmness (see §25 above); nevertheless, in both genres the use of γε can be connected to this emotion. In fact γε appears in contexts of strong emotion since Homer; [100] for Aristophanes this usage has also been observed. [101] γε’s function naturally fits angry contexts, for what is at stake when someone is angry is that she has a different opinion from someone else, at the very least. [102] Indeed, in both tragedy and comedy γε is particularly frequent in angry or otherwise agitated contexts. [103] In general, as discussed in III.2 §§58-63, γε is significantly more frequent in dialogues than elsewhere. [104] This distribution makes it likely that γε’s functions are connected to the speaker’s personal involvement.
§57. In this scene from Aristophanes Frogs, two innkeepers (called “innkeeper” and Plathane in Henderson’s translation) [105] in the Underworld are furious with Heracles for eating an enormous amount of food without paying. They obviously do not realize that the person looking like Heracles standing in front of them is actually Dionysus in a costume.
556bis Πα. οὐ μὲν οὖν με προσεδόκας,
                  ὁτιὴ κοθόρνους εἶχες, ἀναγνῶναί σ’ ἔτι;
                  τί δαί; τὸ πολὺ τάριχος οὐκ εἴρηκά πω.
           Πλ. μὰ Δί’ οὐδὲ τὸν τυρόν γε τὸν χλωρόν, τάλαν,
560             ὃν οὗτος αὐτοῖς τοῖς ταλάροις κατήσθιεν.
           Πα. κἄπειτ’ ἐπειδὴ τἀργύριον ἐπραττόμην,
                   ἔβλεψεν εἴς με δριμὺ κἀμυκᾶτό γε
           Ξα. τούτου πάνυ τοὔργον· οὗτος ὁ τρόπος πανταχοῦ.
           Πα. καὶ τὸ ξίφος γ’ ἐσπᾶτο, μαίνεσθαι δοκῶν.
Aristophanes Frogs 556bis-564
In. Hah! You didn’t think I’d recognize you again with those buskins on. Well? I haven’t even mentioned all that fish yet.
Pl. Right, dearie, or the fresh cheese that he ate up, baskets and all.
In. And when I presented the bill, he gave me a nasty look and started bellowing.
Xa. That’s his style exactly; he acts that way everywhere.
In. And he drew his sword like a lunatic.
In this case the speakers’ angry mood is not demonstrated in an incremental syntactic style (in 557 and 561 a subordinate clause intervenes within the main clause) but the dialogue does contain a striking number of γε instances. The speaker’s anger can be inferred from the content of the dialogue; linguistically it is here reflected especially in their particle use.
§58. Tsakmakis 2010 interprets the use of γε in this scene as a marker of coherence. In this way, γε “highlights the common ground of the communication” (351). The use of γε in 562, for example, is in his view “a rhetorical strategy intended to make the new information appear consistent with existing contextual information. Consequently, the new information will appear less unbelievable” (351). However, in this angry context the innkeeper and Plathane are probably not primarily interested in emphasizing that the narrated events (eating cheese in 559, bellowing in 562, and drawing a sword in 564) form a coherent story. [106] After all, both interlocutors have witnessed the events. They rather want to stress the outrageous nature of Heracles’ behavior. γε’s local function is to single out τὸν τυρόν (559), κἀμυκᾶτο (562), and τὸ ξίφος (564)—implying, for example, that these events were extremely unexpected or outrageous. The women also emphasize stealing cheese, bellowing, and drawing one’s sword as successive stages of escalation. [107] In 564 it is rather καί, in my view, that marks the link with the preceding utterance (here the speaker’s own previous turn in 561-562), as well as a climax in the upcoming utterance. [108] More globally γε attests to the speakers’ agitated state of mind, in this case anger. Moreover, the particle’s distribution across tragedy and comedy makes Tsakmakis’ interpretation improbable (see §56 above). If its function would be the marking of coherence, we would expect to find γε equally often in calm as in agitated contexts. [109]

5.4.2 γε in stancetaking contexts, with or without agitation

§59. Communicative situations involving anger, then, form highly suitable contexts for the pragmatic functions of γε. However, the particle also appears in contexts without anger or another kind of agitation. In these situations γε’s function simply is to imply a contrast with others’ views. [110] The unifying factor in these cases is an explicit expression of stance. [111] Angry utterances generally tend to involve some stancetaking by the speaker as well, but this does not have to be made explicit.
§60. In the following passage from Birds, Tereus the Hoopoe has asked Peisetaerus and Euelpides, two Athenian visitors, what kind of city they are looking for. Peisetaerus has just answered that he would love to live in a city where his friends would insist on inviting him to parties. Tereus reacts:
135     Επ. νὴ Δία ταλαιπώρων γε πραγμάτων ἐρᾷς.
                  τί δαὶ σύ;
136bis Ευ. τοιούτων ἐρῶ κἀγώ.
136ter Επ. τίνων;
Aristophanes Birds 135-136
Te. My word, it’s miserable troubles you [i.e. Peisetaerus] long for! And what about you [i.e. Euelpides]?
Eu. I long for much the same.
Te. Namely?
Tereus reacts to Peisetaerus’ tastes by taking a clear stance: he finds them “miserable.” This qualification is meant ironically, as the “troubles” that Peisetaerus longs for are in fact pleasant. That is, the utterance implies a contrast between its literal meaning and the conveyed ironic meaning. With γε Tereus highlights this implicit contrast, thereby emphasizing the irony. [112]
§61. Commentators observe emotions other than anger in this passage, which do not necessarily imply any agitation. Van Leeuwen 1902 ad loc. interprets Tereus as laughing (ridens) while saying this; Dunbar 1995 ad loc. considers Tereus’ tone here “surprised.” These feelings may involve a high degree of arousal, i.e. agitation, with a positive attitude, rather than a negative one as in the case of anger. However, the context here does not give clear indications about Tereus’ level of agitation. In any case γε in Aristophanes does not in itself express anger; it can be employed in more friendly and calm contexts, in order to emphasize part of a stancetaking expression or answer. Probably these utterances do not have to involve some form of agitation.
§62. γε also occurs in contexts that clearly relate to both explicit stancetaking and anger. An example is found in the utterance directly following (t13): Euelpides’ answer to Tereus’ question.
           Επ. τί δαὶ σύ;
136bis Ευ. τοιούτων ἐρῶ κἀγώ.
136ter Επ. τίνων;
           Ευ. ὅπου ξυναντῶν μοι ταδί τις μέμψεται
                 ὥσπερ ἀδικηθεὶς παιδὸς ὡραίου πατήρ·
                 “καλῶς γέ μου τὸν υἱόν, ὦ Στιλβωνίδη,
140            εὑρὼν ἀπιόντ’ ἀπὸ γυμνασίου λελουμένον
                  οὐκ ἔκυσας, οὐ προσεῖπας, οὐ προσηγάγου,
                  οὐκ ὠρχιπέδισας, ὢν ἐμοὶ πατρικὸς φίλος.”
Aristophanes Birds 136-142
(Te.) And what about you [i.e. Euelpides]?
Eu. I long for much the same.
Te. Namely?
Eu. A city where a blooming boy’s father would bump into me and complain in this fashion, as if wronged: “A fine way you treat my son, Mr. Smoothy! You met him leaving the gymnasium after his bath, and you didn’t kiss him, didn’t chat him up, didn’t hug him, didn’t fondle his balls—and you are my old friend!”
Euelpides explains what would present an “ideal problem” for him: namely, the possibility that the father of an attractive boy might complain if Euelpides did not kiss his son. The quoted man feels wronged and insulted by Euelpides’ hypothetical behavior: one can infer that the imagined father is angry. [113] The quotation also involves irony or sarcasm, since the cited speaker could never mean καλῶς—an explicit expression of stance—in a serious way (cf. 137 μέμψεται, “complain,” “blame”). [114] γε after καλῶς in 139 highlights the contrast between the literal meaning of this evaluative adverb, and the implied negative meaning. The particle also helps to signal that the announced quotation is starting: in Aristophanes the particle frequently occurs in turn-initial position in utterances expressing stance. [115]
§63. γε’s functions are associated, then, with several emotional and interactional contexts. Highlighting the speaker’s own view, often in implicit contrast to others, is particularly suitable for utterances that express anger. The connection to the speaker’s opinion also fits contexts of explicit stancetaking that are not accompanied by anger; in these cases other high-arousal emotions, perhaps on the pleasure side of the emotional field (see Figure 1 in §22 above) may be expressed. The implicit contrast that γε hints at can be part of an ironic expression, where the literal meaning of a word contrasts with its conveyed meaning. Irony always involves stancetaking; it may or may not be combined with anger.

5.5 Two tragic case studies of calm versus agitated discourse

§64. This section spotlights calm and agitated discourse that simultaneously illustrate several of the findings discussed so far. The first case study compares utterances by the same character in different states of mind, the second involves the emotional inclinations of a play’s two main characters.

5.5.1 Sophocles’ calm versus agitated Oedipus

§65. Two speeches by Oedipus in Oedipus King well exemplify the two ends of the arousal spectrum. At the play’s beginning, when Oedipus is unaware of his troubles, he appears calm, but at the end he becomes extremely desperate, that is, agitated in a certain (negative) way. The difference in these emotional states is reflected in several linguistic differences.
§66. At the very start of the play, Oedipus utters the following lines.
          Οι. Ὦ τέκνα, Κάδμου τοῦ πάλαι νέα τροφή,
                 τίνας ποθ’ ἕδρας τάσδε μοι θοάζετε
                 ἱκτηρίοις κλάδοισιν ἐξεστεμμένοι;
                 πόλις δ’ ὁμοῦ μὲν θυμιαμάτων γέμει,
5               ὁμοῦ δὲ παιάνων τε καὶ στεναγμάτων·
                 ἁγὼ δικαιῶν μὴ παρ’ ἀγγέλων, τέκνα,
                 ἄλλων ἀκούειν αὐτὸς ὧδ’ ἐλήλυθα,
                 ὁ πᾶσι κλεινὸς Οἰδίπους καλούμενος.
                 ἀλλ’, ὦ γεραιέ, φράζ’, ἐπεὶ πρέπων ἔφυς
10             πρὸ τῶνδε φωνεῖν, τίνι τρόπῳ καθέστατε,
                 δείσαντες ἢ στέρξαντες; ὡς θέλοντος ἂν
                 ἐμοῦ προσαρκεῖν πᾶν· δυσάλγητος γὰρ ἂν
                 εἴην τοιάνδε μὴ οὐ κατοικτίρων ἕδραν.
Sophocles Oedipus King 1-13
Oe. Children, latest to be reared from the stock of Cadmus, why do you sit like this before me, with boughs of supplication wreathed with chaplets? and why is the city filled at the same time with incense, and with the sound of paeans and lamentations? Thinking it wrong to hear this from the report of others, my children, I have come myself, I who am called Oedipus, renowned to all. Come, aged man, tell me, since it is fitting you should speak for these, what is your state, one of fear or one of longing? Know that I am willing to render every kind of aid; I would be hard of heart if I felt no pity at such a supplication.
These words situate the audience in the play’s opening state of affairs and enables them to infer Oedipus’ specific state of mind at that point. He is here presented as a thoughtful and compassionate king, ready to help his people in times of need. He wants to inform himself well before taking action. As is usual for speakers at the beginning of tragedies, [116] the king appears calm, despite references to his own sense of worry and pity (κατοικτίρων, 13). [117]
§67. The speech contains an elaborate participial clause preceding its main clause, as well as examples of pragmatic projection through vocatives, ἀλλά, and a priming act. [118] In 6-7 a long participial clause, with the intervening vocative τέκνα in 6, precedes the finite verb ἐλήλυθα. In 1 the vocative ὦ τέκνα produces pragmatic projection: it is not complete on its own, but pragmatically requires the attachment of an utterance addressed to this addressee. [119] ὦ τέκνα projects more than one act: τίνας ποθ’ ἕδρας τάσδε μοι θοάζετε, and ἱκτηρίοις κλάδοισιν ἐξεστεμμένοι. The extension of the already incomplete vocative with the apposition Κάδμου τοῦ πάλαι νέα τροφή in line 1 delays the fulfillment of the projection in this case. That is, even though appositions may be part of a purely incremental style, here the construction contributes to prolonging the vocative’s pragmatic projection.
§68. Another vocative, ὦ γεραιέ “old man,” occurs in 9, again triggering the expectation of subsequent acts—now addressed to the old priest rather than the group of young suppliants. In this case the main clause φράζ’ “tell me” follows immediately afterwards. This verb in itself semantically projects a complement clause clarifying what the addressee should tell, even though the utterance would have been syntactically complete if it had ended here. [120] Moreover, the particle ἀλλά before the vocative enhances the pragmatic projection: whether we read it as an act on its own or as one act together with ὦ γεραιέ, ἀλλά always marks some shift in the discourse. [121] The shift to a different addressee in this case creates the pragmatic expectation that several upcoming discourse acts will be addressed to this person, not just one imperative. [122]
§69. Oedipus’ speech also contains a priming act: πόλις δ’ | ὁμοῦ μὲν θυμιαμάτων γέμει (4). As in the Aristophanic example of a priming act discussed in §31 above, the particle μέν retrospectively demonstrates the discourse-act boundary directly after δέ. The ὁμοῦ μέν and ὁμοῦ δέ acts following it both pertain semantically to “the city”: together they form a multi-act move. [123] With δέ Oedipus presents πόλις δ’ as a new step in the discourse, without making a more specific connection explicit. [124]
§70. Now let us turn to Oedipus’ later agitation. The following passage is part of the speech (1369-1415, 46 lines in total, 295 words) that he utters shortly after he has learned of his troubles and blinded himself.
          (Οι.) ἐγὼ γὰρ οὐκ οἶδ’ ὄμμασιν ποίοις βλέπων
                   πατέρα ποτ’ ἂν προσεῖδον εἰς Ἅιδου μολών,
                   οὐδ’ αὖ τάλαιναν μητέρ’, οἷν ἐμοὶ δυοῖν
                   ἔργ’ ἐστὶ κρείσσον’ ἀγχόνης εἰργασμένα.
1375           ἀλλ’ ἡ τέκνων δῆτ’ ὄψις ἦν ἐφίμερος,
                   βλαστοῦσ’ ὅπως ἔβλαστε, προσλεύσσειν ἐμοί;
                   οὐ δῆτα τοῖς γ’ ἐμοῖσιν ὀφθαλμοῖς ποτε·
                   οὐδ’ ἄστυ γ’, οὐδὲ πύργος, οὐδὲ δαιμόνων
                   ἀγάλμαθ’ ἱερά, τῶν ὁ παντλήμων ἐγὼ
1380           κάλλιστ’ ἀνὴρ εἷς ἔν γε ταῖς Θήβαις τραφεὶς
                   ἀπεστέρησ’ ἐμαυτόν, αὐτὸς ἐννέπων
                   ὠθεῖν ἅπαντας τὸν ἀσεβῆ, τὸν ἐκ θεῶν
                   φανέντ’ ἄναγνον καὶ γένους τοῦ Λαΐου.
Sophocles Oedipus King 1371-1383
(Oe.) For I do not know with what eyes I could have looked upon my father when I went to Hades, or upon my unhappy mother, since upon them both I have done deeds that hanging could not atone for. Then, could I desire to look upon my children, since their origins were what they were? Never could these eyes have harboured such desire! Nor to look upon the city, or the wall, or the statues of the gods or the temples, from which I, who had enjoyed the greatest luxury in Thebes, had in misery cut myself off, commanding with my own lips that all should drive from their houses the impious one, the one whom the gods had shown to be impure and of the race of Laius.
This rhesis is highly emotional. Commentators speak of Oedipus’ dread in remembering his incest, [125] his “incommunicable anguish,” [126] and his “desperate state of mind.” [127] Kamerbeek 1967 ad 1398-99 describes the whole speech as “Oedipus’ most pathetic rhesis.” Emotions such as desperation and anguish can be said to belong to the agitated pole of the arousal dimension.
§71. The use of particles is connected to Oedipus’ emotionality. The entire speech contains, notably, four γε (1377, 1378, 1380, 1386) and two δῆτα (1375, 1377). The γε instances locally highlight specific parts of the discourse that are particularly connected to Oedipus’ curse (the city of Thebes) or his current misfortune (his eyes), and globally reveal his state of mind. δῆτα in 1375 makes its host question resemble a dialogic turn of speaking, thus engaging potential hearers. Oedipus goes on to answer his rhetorical question himself with οὐ δῆτα in 1377: here the particle provides a strong emphasis on the negation.
§72. Numerous other linguistic features likewise relate to the agitation. Besides the semantic markers παντλήμων (1379), τοὐμὸν ἄθλιον δέμας (1388), and ἀνδρὸς ἀθλίου (1413), the abundance of first-person references (29 in total, 10% of all words) and negations (19 in total, 6% of all words) is remarkable. [128] The two devices work together to push the speaker into the foreground, the first-person references by directly pointing to the speaker, negations by displaying his subjective influence on his way of expression. [129] Oedipus now knows that all eyes are on him, because he himself has been the center of the story which he had been unraveling.
§73. Regarding syntax and discourse structure, this passage does not contain moves with a priming act followed by several acts fulfilling its projection, as the calm speech at the play’s beginning did. Rather, the discourse structure mainly consists of a “spontaneous” adding of acts onto each other, as in ἐγὼ γὰρ οὐκ οἶδ’ | ὄμμασιν ποίοις βλέπων | / πατέρα ποτ’ ἂν προσεῖδον | εἰς Ἅιδου μολών, | /οὐδ’ αὖ τάλαιναν μητέρ’ (“For I do not know | [looking] with what eyes | I could have looked upon my father | when I went to Hades, | or upon my unhappy mother,” 1372-1373). Here the syntax is complete after προσεῖδον, and does not project another act. Oedipus could have formulated the remarks about Hades and his mother in a separate construction; instead he adds them as increments to the already syntactically complete remark about his father. [130]
§74. Other examples of such syntactic increments in (t17), which do not project the structure beforehand, include οὐδ’ ἄστυ γ’, | οὐδὲ πύργος, | οὐδὲ δαιμόνων / ἀγάλμαθ’ ἱερά (“nor to look upon the city, or the wall, or the statues of the gods or the temples,” 1378-1379) and αὐτὸς ἐννέπων / ὠθεῖν ἅπαντας τὸν ἀσεβῆ, | τὸν ἐκ θεῶν /φανέντ’ ἄναγνον | καὶ γένους τοῦ Λαΐου (“commanding with my own lips that all should drive from their houses the impious one, the one whom the gods had shown to be impure and of the race of Laius,” 1381-1383). The syntax would not have required these acts to be placed where they are. Therefore they give the impression that Oedipus only thinks of them at the very moment of utterance.
§75. The multi-act structure | τῶν ὁ παντλήμων ἐγὼ | / κάλλιστ’ ἀνὴρ εἷς ἔν γε ταῖς Θήβαις τραφεὶς | / ἀπεστέρησ’ ἐμαυτόν | (1379-1381) seems an exception to the incremental style in this speech, as the second act intervenes between the subject ἐγὼ in the first act and its syntactically projected verb ἀπεστέρησα in the third act. However, the intervening act is not a subordinate clause preceding and therefore projecting an entire main clause, but an apposition, not syntactically projected or required, within a subordinate clause. The multi-act structure is incremental in a different way: each of these acts adds information about Oedipus himself in a piecemeal fashion, which enhances the dramatic focus on his fate. Additionally, παντλήμων “utterly miserable” (1379) semantically does not project a word like κάλλιστ’ “in a most beautiful way” (1380) so closely afterwards.
§76. One may object, justifiably, that an accumulation of increments may still form an elaborate structure of acts that syntactically, semantically, and/or pragmatically belong together. However, my point is that the earlier parts of such moves do not project or require the addition of the later parts. That is, we may identify multi-act moves such as ἀλλ’ ἡ τέκνων δῆτ’ ὄψις ἦν ἐφίμερος, | / βλαστοῦσ’ ὅπως ἔβλαστε, | προσλεύσσειν ἐμοί; | (1375-1376) as coherent units. But in the moment of their utterance their combined structure is in fact incremental, literally: “but was my children’s sight desirable, then? | [the sight that had] originated in such way as it had, | [was it desirable] for me to look at?” Line 1375 could have stood on its own; the two acts in 1376 only later turn out to belong, syntactically, semantically, and pragmatically, to the preceding line as well. The entire move appears more fragmented than a move in which earlier parts project the later ones.
§77. Note that in both Oedipus’ calm and his agitated speeches he uses repetition in successive acts—ὁμοῦ-ὁμοῦ in the calm speech (lines 4-5), τόν-τόν in the agitated one (line 1382). These instances, however, serve different purposes. The ὁμοῦ acts are part of a carefully composed move, and pragmatically projected beforehand by a priming act, whereas the unannounced repetition of the definite article τόν contributes to an incremental style, and thus to the image of an agitated speaker. [131] Lexical repetition by itself, then, is not a sign of either calmness or agitation, but can be employed for different pragmatic goals. It is multifunctional and dependent on its specific context, just as other linguistic features, including particles, are (see §7 above).

5.5.2 Euripides’ agitated Pentheus versus calm Dionysus

§78. So far I have discussed how a certain emotional state of mind at particular moments affects linguistic output. The second tragic comparison of calmness and agitation concerns how a character’s speech style reflects a more permanent emotional state, that is, when emotionality is a feature of someone’s temperament. As Revelle and Scherer put it, “personality is to emotion what climate is to weather” (2009:304). Because of the relation of γε to anger the particle also reflects, in specific cases, the more global feature of a tragic character’s irascibility. The case of Pentheus in Euripides Bacchae illustrates γε’s connection to anger as well as to temperament. His opponent, the god Dionysus, generally stays calm: this is reflected in his particular uses of δέ, among other features.
§79. Pentheus is presented as particularly short-tempered: Dodds 1960 [1944] for example mentions that “Pentheus is flurried, irascible, full of an unhealthy excitement” (xliv). [132] The character is fond of γε: he utters 11 instances, which means 1.0% of his 1119 words in the whole play. Most other tragic characters use the particle less often: for example, Clytaemnestra in Aeschylus Agamemnon 0.3% (6 instances in 1927 words in total); Antigone in the eponymous play by Sophocles 0.7% (8 instances in 1220 words); Hippolytus in his eponymous play by Euripides 0.8% (13 instances in 1644 words); Medea in hers 0.4% (14 instances in 3447 words). [133] These frequencies are of course only part of the information at our disposal, and do not tell us how a character uses a certain particle. Differences in frequencies across characters can, however, form a starting point for an analysis.
§80. In Pentheus’ case, the high frequency of γε can be connected to his fiery temper. However, being angry by itself does not always entail that a character expresses this feeling in words, which is the aspect relevant to particle use. That is, if a character expresses her anger mainly through nonverbal means, then this feeling may not have clear reflections in her language use. It may seem surprising, for instance, that Euripidean Medea does not utter many instances of γε, but in fact she does not very often express her anger verbally. She spends most of her words for other communicative actions, such as lamenting her fate (e.g. lines 111-114, 144-147, 160-167), arguing her case (214-266), or explaining her plans (364-409, 764-810). Even when speaking directly to Jason she hides her anger in one of her speeches (869-905). In short, Medea expresses her anger especially in her nonverbal actions, but uses language to do other things, for which γε is less fitting. Pentheus, in contrast, destined as he is to lose the battle with Dionysus, can only rage with words.
§81. The other main character of Bacchae, Dionysus, tends to remain calm. Dodds 1960 [1944]:xliv notes this general calmness, and writes ad 621-622: “amid the physical turmoil of the earthquake and the moral turmoil of the baffled Pentheus, the Stranger’s calm marks him as something supernatural; it is like the sinister calm at the heart of a typhoon.” Dionysus utters γε with a relatively low frequency of 0.6% (9 times in 1549 words). It is δέ that he uses more often than any other tragic character: its frequency in Dionysus’ utterances is 4.1% (63 instances in 1549 words). [134]
§82. The following dialogue shows how emotional differences manifest in divergent particle use. Over the course of the exchange, Pentheus becomes increasingly agitated, while the Stranger (in fact Dionysus) remains calm throughout.
          Πε. πρῶτον μὲν ἁβρὸν βόστρυχον τεμῶ σέθεν.
          Δι. ἱερὸς ὁ πλόκαμος· τῶι θεῶι δ’ αὐτὸν τρέφω.
495    Πε. ἔπειτα θύρσον τόνδε παράδος ἐκ χεροῖν.
          Δι. αὐτός μ’ ἀφαιροῦ· τόνδε Διονύσωι φορῶ.
          Πε. εἱρκταῖσί τ’ [135] ἔνδον σῶμα σὸν φυλάξομεν.
          Δι. λύσει μ’ ὁ δαίμων αὐτός, ὅταν ἐγὼ θέλω.
          Πε. ὅταν γε καλέσηις αὐτὸν ἐν βάκχαις σταθείς.
500    Δι. καὶ νῦν ἃ πάσχω πλησίον παρὼν ὁρᾶι.
          Πε. καὶ ποῦ ’στιν; οὐ γὰρ φανερὸς ὄμμασίν γ’ ἐμοῖς.
          Δι. παρ’ ἐμοί· σὺ δ’ ἀσεβὴς αὐτὸς ὢν οὐκ εἰσορᾶις.
          Πε. λάζυσθε· καταφρονεῖ με καὶ Θήβας ὅδε.
          Δι. αὐδῶ με μὴ δεῖν, σωφρονῶν οὐ σώφροσιν.
505    Πε. ἐγὼ δὲ δεῖν γε, κυριώτερος σέθεν.
Euripides Bacchae 498-505
Pe. First I shall cut off your delicate locks.
Di. My locks are sacred: I grow them long in the god’s honor.
Pe. Next, hand over that wand.
Di. Take it from me yourself: I carry it, but it belongs to Dionysus.
Pe. We will keep you penned up inside and under guard.
Di. Dionysus himself will free me when I so desire.
Pe. Sure, when you stand surrounded by bacchants and call on him.
Di. Yes, even now he is near and sees what I am undergoing.
Pe. Where is he? To my eyes he is not in evidence.
Di. He’s with me: since you are a godless man you do not see him.
Pe. Seize him! He’s treating me and Thebes with contempt!
Di. And I forbid it: I am sane and you are not.
Pe. I say bind him, and I have more authority than you.
Pentheus’ use of γε keeps pace with his mounting frustration. In 499, γε marks a hostile use of resonance: Pentheus echoes his interlocutor’s ὅταν, thereby extending its syntactic dependence on the main clause “the god himself will free me” (λύσει μ’ ὁ δαίμων αὐτός, 498); he also picks up the reference to his interlocutor from the previous utterance. The addition of γε in resonating utterances emphasizes the speaker’s hostile goal in echoing his opponent’s words, and thereby implies anger or hate. [136] Line 505 contains another γε, in this case preceded by turn-initial δέ. As discussed in III.3, in contexts of resonance the combination of these two particles mark the echo from the preceding utterance as a hostile one. [137] In other words, Pentheus here employs γε as well as δέ… γε in such a way as to convey anger and hostility.
§83. Pentheus’ utterance in 501 is a question starting with a καί act, which may imply indignation. [138] This implication fits the king’s growing anger. His use of γε to highlight ὄμμασιν ἐμοῖς, “my eyes”—implying a contrast to what his addressee is claiming—accordingly receives a hostile function: the utterance implies that Pentheus’ addressee is lying about the god’s presence.
§84. In contrast to Pentheus, Dionysus stays calm in this scene. He adopts a solemn speaking style, which is reflected in his frequent use of αὐτός when it is semantically redundant (494, 496, 498, 502), [139] and in starting several turns without turn-initial contextualization cues, even though these are not answers to questions (see 494, 496, 498). [140] In 500 the subordinate clause at the start (ἃ πάσχω) and the intervening participial phrase (πλησίον παρών) project that their main clause (ὁρᾶι) will follow.
§85. Both Pentheus and Dionysus use δέ, but in notably different constructions. These uses illustrate that in combination with several co-textual and contextual features, δέ may also contribute to the linguistic reflection of calmness or agitation. Dionysus twice uses δέ not in the first act of his turn, but later; this is in fact his usual habit. [141] When he utters δέ in turns longer than one line, which he often does, the particle also rarely appears in turn-initial position. [142] Pentheus, in contrast, usually uses δέ in turn-initial position. [143]
§86. I associate these different δέ constructions with different pragmatic goals. On the one hand, δέ marking an act boundary later in an utterance simply signals a new step in the discourse, without explicitly relating the host act to the preceding one; this use is for example frequent in long narratives. [144] On the other hand, δέ in the first discourse act of an utterance marks the start not just of a new act, but of an entire new adjacency pair, usually a new question-answer pair. [145] This turn-initial construction is therefore especially frequent in a list of questions, such as Pentheus’ interrogation of the Stranger at Bacchae 460-486, where he utters 7 instances of turn-initial δέ. The one-line questions in that interrogation scene usually give little information about the speaker’s emotional state; nevertheless, in the case of Pentheus’ list of questions the god, who possesses the desired information, is calmer than the king, who tries to get it out. [146] Dionysus’ preference for non-turn-initial δέ, then, implies that he tends to mark discourse connections of various kinds with the neutral signal of δέ, without making particular relations explicit; this reflects less communicative pressure, and therefore calmness. Pentheus usually employs the particle in its turn-initial construction, which is connected to interrogating, in his case in an agitated way. That is, δέ in itself indicates neither calmness nor agitation, but specific constructions of the particle in combination with other features do reflect these states of mind.
§87. The tragedians, then, represent some characters as more prone to certain emotions than other characters, and this leads to differences in particle use. In the case of Bacchae, Pentheus’ frequent use of the particle γε and δέ in its turn-initial construction reflects his irascible personality. Dionysus’ less frequent use of γε and his inclination for δέ in later than turn-initial position relates to his general calmness.

5.6 Conclusions

§88. In this chapter we have seen that particles do not directly express emotional states by themselves, but often play an important role in facilitating our interpretation of an utterance’s emotional qualities. Calmness and agitation are two opposing states of mind that reveal themselves in divergent patterns of discourse organization and particle use. These linguistic tendencies are found in both tragedy and comedy, although the literary functions of the two emotional states tend to differ in the different genres.
§89. Calmness may be reflected in subordinate clauses that precede and thereby project their main clauses, as well as in pragmatic projection, especially through vocatives and priming acts. Several particles may play a role in priming acts, with δέ being the most frequent. Certain constructions with μέν tend to be found in calm contexts: they demonstrate that speakers who trust that they can keep the floor are at leisure to pay ample attention to structuring their discourse. In other words, I connect a discourse organization that involves several kinds of “stretching”—a strategy that all of these features instantiate—to calmness. Certain uses of δέ also relate to calmness, by indicating emotionally neutral discourse boundaries within longer turns of speaking.
§90. In agitated contexts speakers tend to utter discourse acts that appear to be spontaneously added to one another. The particles γε and δῆτα especially fit such contexts as well. γε highlights a part of an utterance that expresses the speaker’s subjective views, attitude, or feelings, potentially in contrast to those of other people. That is, the particle fits utterances that express the speaker’s stance. Moreover, the contribution of γε is comparable to that of the emphatic prosody represented by an exclamation mark in English—an emphasis especially appropriate in contexts of high emotional arousal. δῆτα, which appears less frequently than γε, generally works to signal that a question arises from the preceding discourse; additionally it appears to directly reflect a speaker’s sense of disquiet. The emotional prosody it conveys would be akin to a loud and desperate pronunciation of the entire utterance.
§91. γε can be more specifically connected to the agitated state of anger. The particle’s pragmatic function, to highlight one element and stress the speaker’s own views in contrast to others’, makes it particularly useful for angry contexts. This interpretation again invites an analogy with the exclamation mark: like γε, this paralinguistic device does not signal anger by itself, but is expected to occur more frequently in angry contexts.
§92. As discussed in IV.4, taking stance usually involves some degree of emotional involvement. At the same time, expressing an emotion tends to involve an evaluation, positioning, or alignment by the speaker. [147] That is, even though expressing anger and expressing stance do not entail each other, they are communicative actions that can be combined. Both are more frequent in comedy, with its greater emphasis on particular speakers, than in tragedy, which focuses more on communicated content (see §52 above); the frequency of these actions in Aristophanes partly explains why this playwright uses γε so often.
§93. The results of this chapter show that Greek drama does not only provide information about the cultural and literary significance of emotional states of mind, which previous investigations have focused on. The texts are also an important source for the manner in which emotionality influences language use. By focusing on this influence, with special attention to calmness and agitation, the current analyses complement the approaches to ancient emotions that have been prevalent in recent research.
§94. All of these observations mean, once more, that it is necessary to look beyond the clause or the sentence, in order to fully understand the use of particles, and their pragmatic contributions to the discourse. [148] Particles do not just modify the single discourse act in which they occur; they also, at the same time, reflect the degree of emotional arousal of a passage, or aspects of a character’s general emotional tendencies. In general, this chapter shows that the organization of discourse in drama (such as syntactic structure, pragmatic projection, and the marking of transitions) is connected to characters’ calmness or agitation. That is to say, emotionality and discourse organization in drama are not only compatible, but also to some extent interdependent. More specifically, emotional arousal reveals itself through alterations in linguistic patterns, such as patterns of particle use.
§95. Since a speaker’s emotional state of mind is relevant to this kind of linguistic choices, as this chapter has demonstrated, the study of discourse organization should not ignore the interactional level of discourse, including characters’ attitudes towards their utterances and towards their addressees. Calmness and agitation are not an optional addition to the “dry” meaning of dramatic utterances; they are an inherent part of the embodied performance. [149] Both of these levels of emotional arousal have their own functions in the interaction among the characters, as well as in the literary communication between playwright and audience. The direction of investigation started in the current chapter therefore promises to illuminate the ancient dramas in many more ways.


[ back ] 1. Voice in other genres is discussed in II.3 §28 (on the voice of a character in Homer) and especially IV.4 §§15-44 (on direct and indirect voices of characters in Herodotus and Thucydides).
[ back ] 2. It is not my goal here to examine how the audience’s emotions are aroused by a performance. Though these emotions are part of the communicative process of drama, they are less directly relevant to particle use. For emotional responses of the ancient audience, see e.g. Budelmann and Easterling 2010 (on tragedy); Grethlein 2010:88 (on Aeschylus Persians); Munteanu 2011 (on comedy) and 2012 (on tragedy). At Oxford University, the current project “Adults at play(s)” (started in 2014) investigates the psychology of dramatic audiences, using both Greek and Shakespearean tragedy as research corpus. For modern readers’ emotional responses to narrative, and their mental representations of characters’ emotions, see Sanford and Emmott 2012:191-232.
[ back ] 3. In general, emotions are seen as “relatively brief and intense reactions” to changes in a person’s environment (Altenmüller, Schmidt, and Zimmermann 2013:344). The transitory nature of emotions is also mentioned in the definitions by e.g. Caffi and Janney 1994:327; Mortillaro, Mehu, and Scherer 2013:4; Owren, Philipp, Vanman, Trivedi, Schulman, and Bachorowski 2013:175; Schwarz-Friesel 2013:70.
[ back ] 4. On (problems in) defining and classifying emotions, see also e.g. Reisenzein 1994; Wierzbicka 1999; Altenmüller, Schmidt, and Zimmermann 2013; Juslin 2013; Schwarz-Friesel 2013:43-87.
[ back ] 5. See Sanford and Emmott 2012:192-193, with reference to Arnold 1961:177.
[ back ] 6. See e.g. Stanford 1983:21-46; Wierzbicka 1999:esp. 3-4; Konstan 2006:passim; Cairns 2008; Theodoropoulou 2012:433; Schwarz-Friesel 2013:59; Sanders 2014:passim. Along the same lines, Sidnell and Enfield 2012:321 argue more generally: “the language you speak makes a difference in the social actions you can perform.”
[ back ] 7. See e.g. III.4 §§3-4 for discussion of the unrealistic nature of tragic and comic dialogues and why we can nevertheless use Conversation Analysis to analyze them. More generally, Hogan has convincingly shown in numerous publications (e.g. 2010; 2011) that literature is a crucial source for learning about emotions and how they are expressed. He illustrates (2010:188-194) the usefulness of literature for emotion research by analyzing the levels of (1) literature’s existence, (2) universal genres such as romantic and heroic narratives, and (3) individual literary works.
[ back ] 8. I use “agitation” as a shorthand for “a relatively high degree of arousal”; see §§22-23 below. That is, “agitation” does not refer to one particular emotion, but is used as the contrary of the equally broad term “calmness”; whether the agitation is linked to a positive or a negative emotion needs to be understood or inferred from the context. In practice, I will be concerned almost entirely with negative emotions; see §22 below.
[ back ] 9. See §2 above. Stanford 1983 (see §18 below) also notes the difficulty in studying ancient Greek emotions, because terms and concepts cannot be translated one-to-one from the Greek language and culture to ours. See especially pp. 23-27 for problems surrounding “pity.” However, it is not Stanford’s main goal to examine the differences in emotion conceptualizations between us and the ancient Greeks; he focuses instead on the expressions and literary functions of emotions in tragedy.
[ back ] 10. See also Konstan 2001 on the differences between the modern concept of pity and those of the Greeks and Romans. On the emotions of the Romans, see also Kaster 2005.
[ back ] 11. See II.4 §46 on the notion of scripts, with references.
[ back ] 12. On the literary functions of emotions in Roman tragedy, see e.g. the book-length study of Seneca’s revenge tragedies by Winter 2014. Instead of a literary perspective, Munteanu 2012 adopts a philosophical and historical one on pity and fear in several Greek tragedies. That is, her analyses aim to illuminate the ethical and social implications of these emotions. The papers in Chaniotis 2012 (ed.) and Chaniotis and Ducrey 2013 (eds.) similarly analyze emotions in various ancient sources as a historical phenomenon.
[ back ] 13. On the importance of analyzing language when analyzing emotion in general, see e.g. Argaman 2010; Schwarz-Friesel 2013. On different linguistic and paralinguistic cues working together to convey emotions in any language, see e.g. Bazzanella 2004:62-63; Caffi and Janney 1994:348; Selting 1994; Van Lancker Sidtis 2008; Foolen 2012. A study of imitative constructions (onomatopoeic interjections) and their use to express emotions in Finnish is Jääskeläinen 2013. See also Theodoropoulou 2012 on the various linguistic means to express emotions, especially metaphors, and how these are employed in ancient texts.
[ back ] 14. On “expressive interjections” as a subcategory of Greek interjections, and on their function of expressing the speaker’s mental state, see Nordgren 2015:17-19. On the connection between interjections and emotions in a modern language, see e.g. Schwarz-Friesel 2013:154-162 on German.
[ back ] 15. See Nordgren 2015:93-164.
[ back ] 16. Stanford 1983 on tragedy also pays attention to the semantic level of emotion expression.
[ back ] 17. See Foolen 2012:353 for reference to cross-linguistic research on the use of prepositions in relation to emotion words; the different prepositions attest to different construals of the emotions, just as the different syntactic constructions in Homeric Greek do.
[ back ] 18. See Mastronarde 1979:59 and Stanford 1983:99.
[ back ] 19. On hyperbole especially expressing anger in tragedy, see Stanford 1983:101; on single-word repetition indicating emotional agitation (especially connected to grief), see pp. 93-95.
[ back ] 20. As discussed in I.2 §58, this work has been attributed to Demetrius of Phaleron, fourth to third century BCE, but was possibly written later.
[ back ] 21. See I.2 §58 and De Jonge 2015:994 on this observation.
[ back ] 22. See I.2 §75. Apollonius does not clarify which emotion is expressed; he simply speaks of ἔκπληξις (“consternation,” “amazement,” “agitation”).
[ back ] 23. See §83 with note 138 below. See also Hancock 1917:29 on this use of καί. This use is discussed in III.3 §45, with note 33; III.4 §36 with notes 72 and 73.
[ back ] 24. Examples are Aeschylus Persians 1043; Sophocles Ajax 891; Euripides Hippolytus 310bis; Aristophanes Frogs 653. On interjections in Greek drama, see §15 above.
[ back ] 25. As for example in (t12) below from Aristophanes Frogs.
[ back ] 26. For examples of long speeches intended to persuade that reflect the speaker’s calmness, see note 52 below.
[ back ] 27. Beside the dimensional approach to emotion classification, the other main approach is the categorical approach, in which one has to decide the number and naming of emotional categories—a debated issue. See e.g. Altenmüller, Schmidt, and Zimmermann 2013:341-343; Schwarz-Friesel 2013:62-69 on both the categorial and the dimensional approaches.
[ back ] 28. For an overview of several two-dimensional emotion models, see Barrett and Russell 2009.
[ back ] 29. Sometimes a third dimension is added, such as tension, intensity, or control. See e.g. Caffi and Janney 1994:338 on a three-dimensional model of “affective experience”; see also Schwarz-Friesel 2013:69 on the three “parameters” to describe the “wesentlichen Eigenschaften von Emotionen.”
[ back ] 30. See also e.g. Sanford and Emmott 2012:191-193 on the relevance of appraisal in theories of emotion.
[ back ] 31. An exception is Aeschylus Libation Bearers 235-245, where Electra realizes that Orestes has returned alive and is standing in front of her. Blass 1906 ad loc. speaks of “leidenschaftliche Freude” (“passionate joy”), Garvie 1986 ad loc. describes Electra as “beside herself with joy.” An Aristophanic example is the joyful greeting in Lysistrata 78-81.
[ back ] 32. This is why most parallel examples given in the footnotes are from relatively long turns of speaking, both calm and agitated ones. This is not meant to imply any influence of the degree of the speaker’s agitation on her utterance length. Short and long utterances can both express calmness as well as agitation. In the case of long utterances, there simply tend to be more linguistic indications available from which to infer the speaker’s state of mind.
[ back ] 33. Examples of tragic characters, who are unaware of upcoming doom and speak in a calm way, are Agamemnon in Aeschylus Agamemnon 810-854; the Pythia in Eumenides 1-33; Oedipus in Sophocles Oedipus King 1-13, cited in (t15); Andromache in Euripides Andromache 183-231 and 269-273.
[ back ] 34. Examples of gods speaking calmly in tragedy are Athena in Sophocles Ajax 1-13; Dionysus in Euripides Bacchae (throughout: see §§78-84 below), Aphrodite in Hippolytus 1-57.
[ back ] 35. Tragic messenger speeches reflecting calmness include Sophocles Antigone 1192-1243; Euripides Medea 1136-1230.
[ back ] 36. Apart from Aristophanes Assemblywomen 171-240 by Praxagora (see (t2) and notes 48 and 50 below), speeches meant to persuade that reflect calmness include Sophocles Antigone 683-723 by Haemon; Oedipus at Colonus 1181-1203 by Antigone; Philoctetes 1314-1347 by Neoptolemus; Euripides Alcestis 280-325 by Alcestis; Bacchae 266-327 by Teiresias; Hecuba 299-331 by Odysseus; Hippolytus 433-481 by the nurse, 983-1035 by Hippolytus; Medea 522-575 by Jason; Aristophanes Acharnians 496-556 by Dicaeopolis, Wasps 548-559 by Lovecleon.
[ back ] 37. For example, both Phaedra’s nurse in Euripides Hippolytus 433-481 and Tecmessa in Sophocles Ajax 485-524 want to convince their masters not to commit suicide. Haemon in Sophocles Antigone 683-723 tries to persuade his father Creon not to kill Antigone. Andromache in Andromache 183-231 and Hippolytus in Euripides Hippolytus 983-1035 need to save their own life by arguing their cases. Hecuba in Euripides Hecuba 803-805 does not want to save anyone but tries to get help in taking violent revenge on her enemy Polymestor. See also McDonald 2007:475: “In Greek tragedy, many speeches justify murder.”
[ back ] 38. Lysistrata 574bis-586 in Lysistrata’s speech.
[ back ] 39. Birds 481-538 in Peisetaerus’ speech, with some interruptions.
[ back ] 40. Starkie 1911 and Dover 1968 perhaps imply this in their comments ad 345, by noting Socrates’ pedagogic style.
[ back ] 41. See IV.3 §24 on the difference between “ascending” (subordinate clause first) and “descending” periods (main clause first) in historiography.
[ back ] 42. Examples of elaborate subordinate clauses preceding their main clauses, or intervening within it, and certain similar structures from calm tragic monologues (see §24 with note 32 above) include Aeschylus Agamemnon 841-842, 846-847, 848-850, and 854 by Agamemnon; Eumenides 9-11 by the Pythia (two participial phrases preceding the main clause); Sophocles Antigone 170-174, 178-181, 182-183, 198-206, and 209-210 by Creon, 701-702 by Haemon (a genitive participial phrase preceding the comparative on which it depends); Philoctetes 70-71 by Odysseus; Euripides Andromache 29-30 by Andromache, 209-210 by Andromache; Bacchae 13-20 by Dionysus (a highly elaborate participial clause preceding its main clause); 288-290 by Teiresias (two subordinate clauses preceding their main clause); Euripides Children of Heracles 158-160 by the herald; Hecuba 802-805 by Hecuba; Heracles 1326-1328 by Theseus; Hippolytus 3-6, 21, 24-28 (an elaborate participial phrase preceding the main clause), and 34-40 by Aphrodite, 451-456 by the nurse; Medea 526-528 by Jason. [ back ] Sophocles Women of Trachis 1114-1115 by Hyllus also contains a subordinate clause preceding its main clause that may reflect calmness; it is however not part of a long speech, but of a 6-line utterance. [ back ] Other Aristophanic examples in calm contexts include Acharnians 520-522, 526 (a participial clause intervening between subject and finite verb), 541-543; Assemblywomen 517-518 (a second subordinate clause intervening within a first), 518-519 (several adverbial phrases preceding their main verb); Birds 1001-1004 (two participial clauses preceding their main clause, in this case even interrupted by a different speaker), 1007-1009 (a participial phrase preceding the subject and main verb), 1355-1357, 1360-1361, 1368-1369; Clouds 404-405; Frogs 31; Wasps 552-553 (an object preceding its finite verb and subject).
[ back ] 43. Perhaps the association that I propose between subordinate clauses preceding their main clauses and calmness also has to do with speech planning. On speech planning in general, see e.g. Ferreira and Swets 2002; Konopka and Brown-Schmidt 2014. Ferreira and Swets 2002 point out, on the basis of psycholinguistic experiments, that it is partly under the speaker’s control whether to plan an entire utterance in advance, or to speak incrementally. These authors write that “the extent to which planning occurs (…) depends on the intentions that motivate the speech” (80). Similarly, Konopka and Brown-Schmidt 2014 suggest (16) that complex messages probably require a more holistic planning before speaking than simple messages. The research on message planning so far, these authors note, “suggests considerable flexibility in the process of message planning and considerable sensitivity and perspective-taking on the part of the speaker when designing messages in different conditions and for different listeners” (17). It may be inferred, I suggest, that calmness and agitation may also play a role in the extent to which speakers engage in speech planning, and this may have an effect on their use of syntactic projection produced by the order of clauses.
[ back ] 44. A function of uttering a subordinate clause before its main clause may be to hold the floor; see III.4 §30 for discussion and an example. Compare also Auer 2000 on certain pre-posed subordinate clauses in spoken German.
[ back ] 45. On the term and concept of pragmatic projection, see Auer 2002, and II.2 §§50-57.
[ back ] 46. Examples of vocatives in calm contexts projecting more than one act are Aeschylus Agamemnon 914 by Agamemnon; Sophocles Ajax 1 by Athena; Antigone 162 by Creon; Electra 1-2 by the old servant; Oedipus King 1 by Oedipus (see (t15) with discussion in §§67-68 below); Oedipus at Colonus 1 by Oedipus, 14 by Antigone; Philoctetes 3-4 by Odysseus; Women of Trachis 49 by the nurse; Euripides Alcestis 1 by Apollo; Medea 49 by the tutor, 869 by Medea, 908 by Jason; Aristophanes Assemblywomen 834 by the herald. [ back ] Cataphoric demonstratives announce their referent, and thereby contribute to the creation of elaborate coherent units as well: see e.g. Aristophanes Clouds 429, which contains both a turn-initial vocative (ὦ δέσποιναι) and a cataphoric demonstrative (τουτί). This utterance implies a certain degree of calmness because of its content’s importance to the speaker. Shortly before, in 412, the elaborate turn-initial vocative uttered by the chorus of clouds also projects more discourse acts; the speakers are here arguably calm because of their exaggeratedly divine status and appearance. See III.4 §30 for more examples of syntactic and pragmatic projection, including a cataphoric demonstrative. On pragmatic projection realized by vocatives and other means in tragedy, see §§67-68 below.
[ back ] 47. On the use and functions of priming acts in Homer, see De Kreij 2016, and II.2 §§63-79 on Homer and Pindar. On priming acts in Herodotus and Thucydides, see IV.2 §§38-41 and IV.3 §§107-116.
[ back ] 48. Praxagora’s monologue in 171-188 is 20 lines long; after interruptions it is followed by another 11.5 lines (192bis-203), then 8.5 more lines (204bis-212), and 97 more lines (214-240). All these parts together contain 415 words.
[ back ] 49. A vertical bar indicates a (relevant) discourse-act boundary. See II.2 §26.
[ back ] 50. On 171-172 Murphy 1938:87 writes, in connection to the rest of the speech, that Praxagora’s “solemn prayer to the gods to prosper her plans indicates to her audience the gravity of the situation and the importance of her subject.” See pp. 109-110 for Murphy’s analysis of the argumentative build-up of 171-240.
[ back ] 51. On priming acts with γάρ starting a move in Thucydides and Herodotus, see IV.3 §108.
[ back ] 52. Other priming acts in calm, official speeches in Aristophanes include Acharnians 509 (ἐγὼ δὲ | μισῶ μὲν…), 513 (ἀτάρ, | φίλοι γὰρ…); Assemblywomen 84 (ἡκκλησία δ’, | εἰς ἣν…); Wasps 678 (σοὶ δ’ | ὧν ἄρχεις…). Priming acts in other calm contexts, namely orders or advice by high-status Aristophanic characters, include Assemblywomen 509 (καὶ μέντοι | σὺ μὲν…); Birds 837 (ἄγε νυν | σὺ μὲν…), 1363 (σὺ γὰρ | /τὸν μὲν…); Frogs 31 (σὺ δ’ οὖν | ἐπειδὴ…). See note 55 below for priming acts in short utterances in Aristophanes. [ back ] Examples of priming acts in calm tragic monologues (at least 25 lines; see III.2 §17) include Aeschylus Agamemnon 854 (νίκη δ’, | ἐπείπερ…); Libation Bearers 279 (νῦν οὖν | σὺ μὲν…); Seven against Thebes 24 (νῦν δ’ | ὡς); Sophocles Ajax 487 (ἐγὼ δ’ | ἐλευθέρου μὲν…); Antigone 722 (εἰ δ’ οὖν, | φιλεῖ γὰρ…), 1226 (ὁ δ’ | ὡς ὁρᾷ σφε; NB this is in a messenger speech; the form and function of the priming act resemble those in Homer; see II.2 §§64-71; and see III.2 §28 on messenger speeches’ similarity to epic); Electra 577 (εἰ δ’ οὖν, | ἐρῶ γὰρ…), 951 (ἐγὼ δ’ | ἕως μὲν…); Oedipus at Colonus 377 (ὁ δ’, | ὡς); Oedipus King 222 (νῦν δ’, | ὕστερος γὰρ…), 258 (νῦν δ’ | ἐπεὶ…); Philoctetes 1343 (ταῦτ’ οὖν | ἐπεὶ…); Euripides Alcestis 313 (σὺ δ’, | ὦ τέκνον μοι), 323 and 325 (καὶ σοὶ μέν, | πόσι,/ γυναῖκ’ ἀρίστην ἔστι κομπάσαι λαβεῖν,/ ὑμῖν δέ, | παῖδες…); Andromache 6 (νῦν δ’, | εἴ τις…), 209 (σὺ δ’ | ἤν τι…); Bacchae 268 (σὺ δ’ | εὔτροχον μὲν…), 274 (δύο γάρ, | ὦ νεανία), 1323 (νῦν δ’ | ἄθλιος μέν…; this speaker, Cadmus, is in grief, but not agitated); Children of Heracles 23 (οἱ δ’ | ἀσθενῆ μὲν…), 819 (μάντεις δ’, | ἐπειδὴ…; NB this is in a messenger speech); Hecuba 51 (τοὐμὸν μὲν οὖν | ὅσονπερ…), 326 (ἡμεῖς δ’, | εἰ…), 546 (ἡ δ’, | ὡς…); Heracles 1331 (θανόντα δ’, | εὖτ’ ἂν…); Hippolytus 47 (ἡ δ’ | εὐκλεὴς μὲν), 1025 (νῦν δ’ | ὅρκιόν σοι…); Medea 244 (ἀνὴρ δ’, | ὅταν…), 526 (ἐγὼ δ᾽, | ἐπειδὴ…), 529 (σοὶ δ’ | ἔστι μὲν…), 1141 (κυνεῖ δ’ | ὁ μέν…; this is in a messenger speech), 1156 (ἡ δ’, | ὡς ἐσεῖδε…; this is in a messenger speech; see the remark above on Sophocles Antigone 1226), 1177 (εὐθὺς δ’ | ἡ μὲν…; this is in a messenger speech). [ back ] Other tragic examples of priming acts, in calm contexts but outside of long monologues, include Sophocles Ajax 1 (ἀεὶ μέν, | ὦ παῖ…) in a 13-line utterance by Athena; Electra 15 (νῦν οὖν, | Ὀρέστα…) in a 22-line utterance by the old slave; Philoctetes 1140 (ἀνδρός τοι | τὸ μέν…) in a 6-line utterance by the chorus; Women of Trachis 52 (νῦν δ’, | εἰ…) in a 12-line utterance by the nurse; Euripides Andromache 269 (δεινὸν δ’ | ἑρπετῶν μὲν) in a 5-line utterance by Andromache (see §25 with note 31 above on tragic characters’ ignorance of upcoming disaster); Hecuba 900 (νῦν δ’, | οὐ γὰρ…) in a 7-line utterance and 1243 (ἐμοὶ δ’, | ἵν’…) in a 12-line utterance, both by Agamemnon. [ back ] See IV.3 §§107-116 for the observation that priming acts are frequent in Herodotus’ and Thucydides’ narrator text. This may be analogous, as the historians are usually not noticeably agitated about writing their chronicles.
[ back ] 53. Concerning structurally similar constructions in other languages, see e.g. Ochs Keenan and Schieffelin 1976; Salmon 2010. Admittedly, Ochs Keenan and Schieffelin consider the “Referent + Proposition construction” in spoken American English to be “a form of “unplanned” speech” (1976:248), which seems not to fit calm, well-thought-through utterances; however, the construction they discuss involves material that is not syntactically integrated in what follows. In my corpus of Greek drama priming acts do usually form a syntactic whole with subsequent acts. Salmon analyzes double-subject sentences in spoken Brazilian Portuguese, of the form “This president, taxes are getting higher.” This construction does not involve syntactic integration either. Interestingly, Salmon considers this construction to be “a tool of style, of rhetoric” (2010:3441) and a reflection of the speaker’s “attention to the informational needs of the audience” (3437).
[ back ] 54. This boundary may however also be indicated by another signal, such as a subordinating conjunction.
[ back ] 55. It is unclear whether it is Peisetaerus or Heracles speaking this line; see Dunbar 1995 ad loc. and the apparatus of Wilson 2007. Both options would fit the context. Wilson opts for Peisetaerus; Henderson 2000 in his translation for Heracles.
[ back ] 56. See Schröder 1927 ad loc., who interprets Poseidon’s attitude as feeling too noble to deal further with that terribly rhetorical speaker Peisetaerus; the god gives in and keeps quiet (“Pos. hat kaum hingehört: zu vornehm, mit dem entsetzlich redegewandten weiter sich einzulassen gibt er nach und schweigt.”).
[ back ] 57. See III.2 §36.
[ back ] 58. Other examples of priming acts in calm short utterances in Aristophanes are Assemblywomen 57 (official question of 3 lines; κάθησθε τοίνυν, | ὡς…), 601 (interested question of 1.5 lines; πῶς οὖν | ὅστις…), 610 (rhetorical argument of 2 lines; νῦν δ’, | ἔσται γὰρ…), 728 (thoughtful decision of 2 lines; ἐγὼ δ’ ἵν’…); Lysistrata 111 (official question of 2 lines, also containing syntactic projection across line-end; ἐθέλοιτ’ ἂν οὖν, | εἰ…), 120 (official proclamation of 3.5 lines; ἡμῖν γάρ, | ὦ γυναῖκες, εἴπερ…); Wasps 764 (compromise of 3 lines; σὺ δ’ οὖν, | ἐπειδὴ…). Calm short utterances can also form part of longer speeches: see note 52 above for priming acts in such contexts.
[ back ] 59. Tragic characters are more suitable for such a spotlight than comic ones, because tragic characters generally have a more consistent personality, with more complex life histories, which may lead to specific communicative goals. References to Agamemnon in this section only involve the character in Aeschylus Agamemnon, not in other plays.
[ back ] 60. For the comparisons of particle use across characters, the utterances of the following 21 tragic characters have been taken into account (in parentheses the total numbers of words spoken per character): Aeschylus: Agamemon (473) and Clytaemnestra (1927) in Agamemon, Electra (929) and Orestes (1864) in Libation Bearers, Xerxes (266) and the queen (1097) in Persians; Sophocles: Ajax (1629) and Tecmessa (1269) in Ajax, Antigone (1220), Creon (2117), and Teiresias (447) in Antigone, Creon (882), Oedipus (4258), and Teiresias (518) in Oedipus King; Euripides: Dionysus (1549), Pentheus (1119), and Teiresias (615) in Bacchae, Hippolytus (1644) and Phaedra (1088) in Hippolytus, and Medea (3447) and Jason (1225) in Medea. For characters that have relatively few lines to speak, their particle frequencies are influenced more by the specific scenes they appear in than by their personality or life-history features. This is especially striking in the case of Teiresias in Oedipus King: his utterances have a much higher frequency of γάρ (2.9%, 15 instances) and γε (1.2%, 6 instances) than usual. Emotionality, however, does play a role in these high frequencies: Teiresias in this play is relatively often engaged in angry stichomythia, where γάρ and γε are particularly at home (see III.3 §§76-79 on γε and §§95-98 on γάρ). For other analyses of particle distributions in tragedy and comedy, concerning different communicative situations rather than different characters, see III.2. See III.2 §69 for the average frequencies of μέν in different parts of tragedies and comedies, regardless of the specific speakers.
[ back ] 61. Apart from the uses mentioned here, he utters μέν in 924 to imply that others may think differently (ἐμοὶ μέν, with a potential counterpart left implicit; see II.2 §60 on this use); and in 932 in a 1-line utterance to acknowledge that the conversation goes on, projecting further utterances within the dialogue rather than further acts within the utterance (see III.4 §33 on this use). Both of these uses differ from the ones discussed here, as they do not concern the discourse organization within one turn of speaking.
[ back ] 62. On the general function of μέν in tragedy and comedy, see III.2 §§69-72; on μέν as a marker of projection, see II.2 §§46-62; on μέν in Herodotus and Thucydides, see IV.3 §§125-146. A similar μέν instance by Agamemnon, also at the start of a speech, is found in 810.
[ back ] 63. Groeneboom 1966 [1944] ad loc. speaks of a cold stateliness.
[ back ] 64. See §25 with note 33 above on demonstrating ignorance about one’s upcoming doom as one of the literary functions of ostensible calmness in tragedy.
[ back ] 65. On μέν at the end of moves, often followed by δέ at the start of the next move, see II.2 §§49-56 on Homer; IV.3 §§65-66 and IV.5 §32 and §§35-36 on historiography. Agamemnon utters similar μέν acts in 829 and 846, also here followed by δέ acts.
[ back ] 66. See IV.3 §§125-129 and IV.5 §32 on οὗτος forms + μέν, sometimes followed by δέ acts, in discourse transitions in historiography.
[ back ] 67. References to Antigone in this discussion only involve this character in Sophocles Antigone, not in other plays. Like Aeschylus’ Agamemnon, Antigone is relatively fond of μέν: her utterances have a frequency of 0.9% (11 instances in 1220 words). Only those of Agamemnon (see §35 above), the queen in Aeschylus Persians (1.2%), and Jason in Euripides Medea (1.0%) have a higher frequency of μέν; the other 17 main characters in the 9 tragedies (see note 60 above) utter the particle less often.
[ back ] 68. See IV.4 on stance in general, with §§61-63 on (dis)alignment and its linguistic expression.
[ back ] 69. Griffith 1999 ad loc. notes that 555-560 refer back to the sisters’ earlier argument at 71-81. Also in this earlier dialogue Antigone explicitly contrasts her own decisions to those of her sister, using μέν and δέ to signal the oppositions.
[ back ] 70. The use of antithesis and chiasmus (regardless of which particles are used) may also betray the influence of rhetoric on Sophocles’ style. See Slings 1997b on such figures of speech (which he argues are not always literary language) in Sophocles and other authors. See e.g. McDonald 2007 and Pelling 2005 on the relation between rhetoric and tragedy in general; McDonald points out that rhetoric is more present in Sophocles than in Aeschylus, and even more in Euripides.
[ back ] 71. See Griffith 2001:127-129 on differences between Antigone’s and Ismene’s speech styles. See also Finley 1939, who argues that Sophocles Antigone is “in style the most antithetical (...) of all extant Greek tragedies” (58).
[ back ] 72. Apart from the one in (t8), Aristophanic examples of such μέν-δέ constructions constrasting actions of first and second person without hostility include Assemblywomen 351-352, 509-510; Frogs 495-497; Peace 1122.
[ back ] 73. μέν and οὖν do not form a cluster in this case, but each carry out their own separate function. οὖν here marks a conclusion from the preceding discourse or situation: now that the archer is gone, the next planned actions can go on. See III.2 §§82-83 on μὲν οὖν as a cluster or as a combination of separate functions in drama; see IV.3 §§144-146 on μὲν οὖν in Thucydides.
[ back ] 74. Another example of σὺ μἐν… ἐγὼ δέ in Aristophanes, contrasting actions rather than decisions, is found in Assemblywomen 509-510.
[ back ] 75. See Slings 1997b:176-192 on anaphora in Sophocles Electra and several other authors.
[ back ] 76. On antithesis in tragedy, see e.g. Navarre 1900:106; Finley 1939. See also note 70 above.
[ back ] 77. Even though “intense emotion” does not necessarily refer to an intense emotion with a high degree of arousal (see §22 with Figure 1 above), in this case it is likely that this is what Hutchinson means.
[ back ] 78. A similar point is made by Mastronarde 1979:62 about suspension of syntax beyond a turn of speaking in Euripides; see §18 above. Also Schuren 2015:37 on Euripidean stichomythia notes that a “rapid succession of short questions [i.e. within one line] creates the impression of urgency, curiosity, and shock.” This scholar provides several parallels of three or four “sentences in a single stichomythic line” in Euripides (2015:36n151).
[ back ] 79. On the connection that has been made between “adding style” and parataxis, see IV.3 §82.
[ back ] 80. Hutchinson 1985 ad loc. refers to two lines that are very similar in their syntactically incremental structure: Aeschylus Agamemnon 268, and Sophocles Philoctetes 1231. As in (t9), in both these cases the speakers themselves refer to their emotional state of mind as a clarification for why they are asking for more information. A similar utterance in Aristophanes is found at Lysistrata 830bis with three discourse acts in less than one line: τί δ’ ἐστίν; | εἰπέ μοι, | τίς ἡ βοή; (“But what is it? Tell me, what is the shouting?”). The speaker’s apparent agitation, as expressed by the structure of her utterance, fits the parody of a (tragic) war situation, in which one suddenly sees an enemy approaching. [ back ] Examples of syntactic increments in agitated tragic monologues include the structures in Aeschylus Agamemnon 1266-1268 by Cassandra; Sophocles Ajax 1003-1007 by Teucer; Philoctetes 932-933 and 949-951 by Philoctetes; Women of Trachis 1086-1090 by Heracles; Euripides Alcestis 689-691 by Pheres, 781 by Heracles; Andromache 388-392 by Andromache; Medea 1327-1332 by Jason. A similar example from Aristophanes is Birds 1199-1201 (a 3-line utterance).
[ back ] 81. See III.2 §58.
[ back ] 82. Other examples of γε in agitated tragic monologues include Aeschylus Agamemnon 1267 (a generally accepted conjecture) and 1279 by Cassandra; Euripides Andromache 385 and 408 by Andromache.
[ back ] 83. Note also the two exclamation marks in Lloyd-Jones’ translation. On the comparison of γε to exclamation marks, see §47 and §§90-91 below; III.4 §64; IV.4 §44. Kamerbeek 1959 ad loc. merely calls γε in 1107 and 1111 “emphatic,” and the instance in 1108 “between emphatic and limitative.” Davies 1991 ad 1107 refers to E. Fraenkel 1977:37, who cites several semantically similar loci, including Sophocles Antigone 1064, which includes γε; however, neither commentator pays attention to the particle.
[ back ] 84. For discussions of γε in drama, see III.2 §§58-63 (distribution and general function), III.3 §§76-79 (in contexts of resonance), and III.4 §§62-64 (in answers and stancetaking utterances). Its uses in other genres is related but slightly different: see II.5 §§27-50 on γε in Homer, and IV.4 §§40-44 and IV.5 §20 and §34 on γε in Thucydides and §62,§§67-68, and §73 on γε in Herodotus. On γε in contexts of stancetaking, regardless of the speaker’s emotional state of mind, see also §§59-63 below.
[ back ] 85. On the possibly emotional meaning of exclamation marks, it is interesting to note that Argaman 2010:92-94 includes “outstanding graphical means” such as certain forms of punctuation in her list of potential markers of emotional intensity in written Hebrew. Indeed, Argaman finds in her experiment that these features are more frequent in subjects’ written expressions of more intense happiness than in those of less intense happiness. She does not, however, discuss the use, relevance, or statistical significance of these graphical means.
[ back ] 86. See also III.2 §37 on this highlighting effect of repeated καί in certain contexts.
[ back ] 87. Also Bond 1981 ad 1255-1310 observes the strong emotionality of the speech, especially after line 1279. On Sophocles, Goldhill 2012:43 similarly suggests that δῆτα, especially when repeated, may indicate “emotional expressivity.”
[ back ] 88. See III.3 §86 on δῆτα in questions marking a logical connection; see III.3 §88 on δῆτα questions in monologues creating the impression of a quasi-dialogue.
[ back ] 89. Other examples of δῆτα questions in emotional tragic monologues than in Oedipus’ speech include Aeschylus Agamemnon 1264 and 1286 by Cassandra; Sophocles Ajax 518 by Tecmessa; Philoctetes 1348, 1352, and 1367 by Philoctetes; Euripides Andromache 404 by Andromache. The instance in Euripides Children of Heracles 162 occurs in a rational, non-emotional speech, but at this point of his speech the speaker, the herald, imagines hypothetical arguments against his own opinion (see W. Allan 2001 and Wilkins 1993 ad loc.); obviously he wants to present those counter-arguments as less rational. [ back ] Other examples of δῆτα questions in strongly emotional contexts (outside of long monologues) include Aeschylus Prometheus Bound 747 by Io (a 5-line utterance); Sophocles Antigone 230 by the guard (a 14-line utterance), 449 by Creon (a 1-line utterance; καὶ δῆτα is called “indignant” by Griffith 1999 ad loc.); Euripides Electra 967 by Orestes (a 1-line utterance); Hippolytus 806 (a 5-line utterance); Aristophanes Frogs 1399 by Euripides (a 1-line utterance).
[ back ] 90. Wiesner 1999:356 also suggests “denn (nur)” as a German translation for δῆτα in questions. A Google search of “was denn nur?” (23 December, 2014) shows entries with several question marks and/or exclamation marks, as well as contributions provided with explicit descriptions of the author’s desperate feeling (e.g. “[ich] werd langsam wahnsinn[i]g” at www.urbia.de/archiv/forum/th-3517711/aaah-werd-langsam-wahnsinng-was-denn-nur-los.html). An expression such as “what the hell?” seems to work as an English paraphrase of “was denn nur?”
[ back ] 91. Two other uses of γε in drama, that can also overlap with contexts of anger and/or of stancetaking, are the one in resonance contexts, discussed in III.3 §§76-79, and the one at the beginning of answers, discussed in III.4 §62.
[ back ] 92. The frequency of γε in Aristophanes, more than 1.07% of all words, is higher than in any other author; see I.5 and III.2 note 212 for discussion and clarification. See III.2 Table 11 for more accurate frequencies in drama, although concerning only parts of the plays.
[ back ] 93. As one illustration, consider the example of εὖ γε (or εὖ γ’), an expression with little semantic content but a strong connection to the speaker’s personal view: it occurs once in Aeschylus, 5 times in Sophocles, 6 times in Euripides, and 17 times in Aristophanes. Stevens 1976:8 considers the expression “clearly colloquial.” The subjectivity of speakers can be connected to their voice: see IV.4 §§40-44 on γε in connection to that in historiography.
[ back ] 94. Taplin discusses several differences between tragedy and Old Comedy, mainly concerning the relation of the play’s worlds to the audience’s world, and the related use of theatrical self-reference. He concludes that the two genres “are in essence fundamentally different” (1986:173).
[ back ] 95. See e.g. Kuppens 2009:32 (on anger in modern humans in general) and Figure 1 above for anger as an emotion with a relatively high degree of arousal.
[ back ] 96. On anger in general, see e.g. Kuppens 2009. On taking into account the appraisal causes of emotions in distinguishing between them (see also §22 above on such classification), see e.g. Reisenzein 1994:537. In this vein, anger typically presupposes that an action of some external agent has taken place in order to trigger the emotion. Kuppens 2009 notes that the agent who is blamed for a “goal-incongruent” situation does not necessarily have to be external, but typically is so. See Konstan 2006:38-40 on emphasis placed by Aristotle on emotions in social interactions, such as anger, rather than emotions arising without others’ intentions, such as sadness. See also §17 above on the study by Luraghi and Sausa 2015 on emotion verbs in Homer: they find that verbs involving active reactions to other agents are construed with different grammatical constructions than verbs that only involve the experiencing subject. On fear in general, and the various stimuli that may cause it, see Öhman 2009.
[ back ] 97. See Aristotle’s definition of ὀργή at Rhetoric 2.1.1378a, cited by Allen 2003:79 and Konstan 2006:41. See Konstan 2006:46-76 for discussion of the differences, according to Aristotle. On another Greek term for anger, that is, μῆνις, which plays a central role in the Iliad and Odyssey, see e.g. Clay 1983:esp. 65-68 (specifically on the Odyssey); Frisk 1946; Muellner 1996 (specifically on the Iliad and Hesiod’s Theogony); Watkins 1977. On μῆνις as well as other Greek terms for anger, such as χόλος, see Cairns 2003; Considine 1966; Irmscher 1950. Specifically on κότος and χόλος in Homer, see Walsh 2005.
[ back ] 98. This is the case also for the Iliad (see e.g. Irmscher 1950; Muellner 1996; Walsh 2005) and, as Clay 1983 argues, the Odyssey.
[ back ] 99. See also Konstan’s discussion (2006:57-64) of the anger of Medea and Hecuba in Euripides’ plays, and Gerolemou 2011 on “mad women” in tragedy.
[ back ] 100. See e.g. Monro 1882a:258 on γε in Homer: the particle “sometimes emphasises a word as a strong or appropriate one, or as chosen under the influence of a feeling (anger, contempt, etc.).”
[ back ] 101. On γε in Aristophanes connected to emotionality in general, see Neil 1901:188: “After the first word in a sentence, γε emphasizes the word and gives an emotional or ‘pathetic’ colour to the whole phrase.”
[ back ] 102. On the pragmatic function of γε in drama, see §§46-47 above, with references to other chapters.
[ back ] 103. Examples of γε in angry contexts other than the ones discussed in this section include Aeschylus Libation Bearers 190 (though γε here, printed by Page 1972 and Sommerstein 2008a, is Porson’s conjecture of δέ, and not accepted by Blass 1906, Garvie 1986, Groeneboom 1949, and Murray 1955 [1937]); Sophocles Antigone 70, 538, 739, 745, 747, 762; Electra 298, 341, 518, 520, 536; Oedipus King 361, 363, 365, 369, 372 (δέ… γε; see III.3 §§80-83), 376, 383, 393; Oedipus at Colonus 1352, 1354; Women of Trachis 1107, 1106, 1111 (on these three cases see (t7) above with discussion), 1127; Euripides Hippolytus 1080; Medea 495, 514, 608; Aristophanes Birds 892, 894, 1208, 1210, 1216, 1220, 1575; Frogs 845; Lysistrata 529 (twice), 530; Wasps 416, 422 (see ὀργῆς in 424), 486.
[ back ] 104. See III.2 §58 on the distribution of γε over the different parts of tragedy and comedy.
[ back ] 105. Tucker 1906 argues that there is only one innkeeper with her maid in this scene, and attributes some of the utterances that others give to the second innkeeper to Dionysus instead.
[ back ] 106. Tsakmakis notes (2010:347n11) that Bakker 1988:97-98 “rightly observes that γε is always related to a fact,” which may lead one to think that truth is important in γε-utterances. However, Bakker’s description exclusively concerns γε in Homer, whereas Tsakmakis discusses its use in Aristophanes; this does not need to be exactly the same as in Homer. See II.5 §§27-50 for the construction ὅ γε in Homer. Moreover, even when there is emphasis on facts, emotional or otherwise subjective implications can be attached to that emphasis. Note that in Aristophanes, γε is often found in stancetaking (see III.4 §63), which cannot be called pure “facts.”
[ back ] 107. The mentioning of these events are preceded by that of eating bread (551), meat (553, with γε), garlic (555), fish (558), and followed by a description of the women’s fear (565bis, with γε), and of Heracles’ departure with their mattresses (567, with γε).
[ back ] 108. See IV.2 §§106-107 and §§122-132 on καί marking a narrative peak or a climax in Herodotus and Thucydides.
[ back ] 109. Tsakmakis notes (2010:345n2) that his discussion of γε is influenced by Kroon’s 2009 description of the Latin particle quidem, which she argues to be a signal (both backward- and forward-looking) of conceptual unity across several discourse acts. This description seems to work well for quidem, which according to Kroon (155) is relatively rare in dialogic contexts; this does not mean, however, that it would translate well to Greek γε, which favors dialogues and agitated contexts. Tsakmakis does note that “[t]he communicative situation is extremely important” (352n20), which includes the “degree of involvement,” but in my view this observation does not influence his analysis enough.
[ back ] 110. On γε implying contrast in Homer, see II.5 §§27-50; in historiography, see IV.4 §§40-44 and IV.5 §20, §§67-68, and §73.
[ back ] 111. On stancetaking, see especially IV.4. On linguistic features of utterances that express stancetaking in drama, see III.4 §§63-67.
[ back ] 112. Note the exclamation mark in Henderson’s translation. Dunbar 1995 ad loc. paraphrases γε’s contribution with italics in her paraphrase of the utterance.
[ back ] 113. Dunbar 1995 ad loc. explicitly describes this hypothetical father as angry. In line with the emotional tone, Henderson uses an exclamation mark in his translation of the γε utterance, as well as Dunbar 1995 and Van Leeuwen 1902 ad loc. in their paraphrases. Hartung 1832:372 also cites this instance, using boldface to convey the highlighting function of γε.
[ back ] 114. On the irony or sarcasm in this utterance, see Bothe 1829, Dunbar 1995, Kock 1864, and Van Leeuwen 1902 ad loc.
[ back ] 115. See III.4 §63. In Birds 1327, Dunbar 1995 ad loc. and Neil 1901:190 consider the presence of γε an argument for reading a change in speaker. See also Neil 1901:190 on the general association of γε with speaker changes in Aristophanes. [ back ] A very similar expression of stance with γε in a hypothetical quotation is found in 1442 of the same comedy, where all commentators refer back to 139. The expression in 1442 is not ironic.
[ back ] 116. Other calm speakers are found at the beginning of Aeschylus Eumenides (Pythia), Seven against Thebes (Eteocles); Sophocles Ajax (Athena), Electra (old slave), Philoctetes (Odysseus), Women of Trachis (Deineira; she describes her fear, but in a calm way); Euripides Andromache (Andromache; she speaks of her misery, but without agitation), Bacchae (Dionysus), Children of Heracles (Iolaus), Hecuba (Polydorus), Hippolytus (Aphrodite).
[ back ] 117. Bollack 1990, Dawe 2006 [1982], Van Herwerden 1866, and Jebb 2004 [1893] do not remark on the speaker’s emotions in this passage. Ritter 1870 speaks of Oedipus’ “Wahre Liebe und innige Theilnahme für die Bittenden” (“true love and heartfelt sympathy for the suppliants”; ad 6-8), and Kamerbeek 1967 mentions his “readiness to be helpful” (ad 6-7), remarks that can perhaps be connected to a relatively calm state of mind.
[ back ] 118. Although the speech is only 75 words long and it is therefore hard to say anything about frequencies of linguistic items, perhaps another reflection of calmness is that the frequencies of first-person references (5 instances, that is 6.7%) and of negations (3 instances, that is 4.0%) are closer to the average frequencies in Sophoclean monologues of 25 lines or longer (6.5 and 4.0%, respectively) than to those in Sophoclean dialogues (7.3 and 4.9%, respectively). See §72 and note 126 below for the strikingly high frequencies of these items in Oedipus’ agitated speech.
[ back ] 119. See §30 above. In Pindar vocatives serve to redirect the audience’s attention (note its co-occurrence with ἀλλά here), and as such typically start new moves (see II.3 §70).
[ back ] 120. In this case the pragmatic projection is less strong than in the case of the vocative, because Oedipus has already asked the questions to which he would like his addressee to respond (lines 2-5).
[ back ] 121. On the functions and uses of ἀλλά in drama in general, see III.2 §§64-68.
[ back ] 122. On ἀλλά similarly introducing a multi-act move in Herodotus, see IV.5 §63.
[ back ] 123. See II.2 §71 for a similar example from Homer, where likewise a μέν act and a δέ act follow a priming act. As discussed there, Denniston 1950 [1934]:371 notes that such μέν and δέ acts often form one bigger unit together, but not that the preceding act may project this entire structure. See IV.2 §108 for καί as a priming act followed by a μέν act, a construction that is frequent in Herodotus and especially Thucydides.

In this case, the projection is simultaneously syntactic as well, because πόλις is the subject of the following finite verb γέμει, but since no full syntactic clause intervenes between them, the syntactic projection is less striking. That is, syntactically the structure πόλις δ’ ὁμοῦ μὲν θυμιαμάτων γέμει is simply one independent clause distributed over two discourse acts, rather than preceded or interrupted by another, dependent clause, as in the stronger cases of syntactic projection.
[ back ] 124. On the possible exploitation of the “neutral” connection signaled by δέ in tragedy, see III.2 §§24-25.
[ back ] 125. Van Herwerden 1866 ad 1376.
[ back ] 126. Jebb 2004 [1893] ad 1415.
[ back ] 127. Kamerbeek 1967 ad 1389-1390.
[ back ] 128. The average frequency of first-person references in Sophoclean monologues is 6.5%; of negations 4.0%. See III.2 Tables 13 and 17. See note 118 above.
[ back ] 129. On the relation of negations to the speaker’s explicit presence, see III.2 §66.
[ back ] 130. Only the first act of this structure does project, semantically, an object for οἶδ’ “I [don’t] know,” which is fulfilled in the second and third acts (“[looking] with what eyes | I could have looked upon my father”).
[ back ] 131. Other examples of lexical repetition reflecting emotional distress are the several word doublings spoken by Philoctetes in Sophocles Philoctetes 1169-1217, such as πάλιν πάλιν in line 1169 and φονᾷ φονᾷ in line 1209. In this lyric dialogue between Philoctetes and the chorus, some lines are distributed among several speakers (antilabe), which Stanford 1983:99 identifies as another sign of “emotional excitement.”
[ back ] 132. For example Pentheus’ speech at 215-262 reflects anger (even though no γε is present here); Dodds 1960 [1944], Oranje 1984, and Seaford 1996 ad loc. remark on this anger, and all call the passage a “tirade” (so does Mastronarde 1979:23 on syntactic reflections of contact in tragedy).
[ back ] 133. See note 60 above for the 21 tragic characters taken into account in my comparisons of particle frequencies. Oedipus in Sophocles Oedipus King is the only other character with a higher γε frequency than Euripides’ Pentheus: 1.1% (48 instances in 4258 words). Oedipus speaks much of his text within angry dialogues.
[ back ] 134. Of the 21 tragic characters analyzed (see note 60 above), the next highest δέ frequencies are those of Clytaemnestra in Aeschylus Agamemnon, the queen in Aeschylus Persians, and Teiresias in Euripides Bacchae (all three 3.7%); those of Electra in Aeschylus Libation Bearers (3.6%); and those of Hippolytus in his eponymous play (3.4%). Although the frequencies by themselves do not yet tell us which communicative strategies these characters favor, since they may prefer different uses of the same particle, the quantitative comparison does make it clear that Dionysus utters δέ strikingly often. The next step is then to analyze in which way exactly he tends to use the particle; see the discussion of (t17) below for examples of such analyses.
[ back ] 135. On turn-initial τε (a rare position for this particle), see III.4 §27, with note 50.
[ back ] 136. On this function of γε in contexts of resonance in tragedy and comedy, see III.3 §§76-79. On this instance of γε, Oranje 1984 notes that Pentheus speaks “mockingly” (60); he calls γε “emphatic” (61n154). Rijksbaron 1991 ad loc. cites Oranje and reads γε as implying assent through emphasis, while limiting that assent. Seaford 1996 ad loc. interprets the utterance as sarcastic (Pentheus “means, sarcastically, that the invocation will occur in prison”). On another note, Elmsley 1821 considers γε an argument against taking this utterance as a question (see III.4 §64 on the rarity of γε at the start of questions).
[ back ] 137. See III.3 §§80-83, including discussion of this particular example. Again, Oranje 1984:62 notes the anger conveyed by Pentheus’ words. Similarly, Wecklein 1903 ad loc. and Oranje 1984:77 note the sarcasm in Pentheus’ utterance in 796, which contains another γε. On another particle implying emotion, see III.2 §87 on ἦ in tragedy.
[ back ] 138. On this instance of καί marking a contemptuous or indignant question, see Dodds 1944 ad loc. and Oranje 1984:61n155. On this use of turn-initial καί in drama in general, see note 23 above.
[ back ] 139. In 498 αὐτός also hints at his double identity, since he is in fact “himself.” αὐτός typically refers to gods or heroes. On the pragmatics of αὐτός, particularly in Homer, see Bonifazi 2012:137-183.
[ back ] 140. On turn-initial contextualization cues in tragedy and comedy, see III.4 §12.
[ back ] 141. The other one-line utterances by Dionysus with δέ in non-turn-initial position are Bacchae 464, 474, 484, 647, 833, 841, and 1345.
[ back ] 142. Dionysus utters δέ in non-turn-initial position in turns longer than one line in Bacchae 4, 6, 10, 11, 13, 23, 28, 33, 37, 48, 50, 62, 461, 617, 618, 621, 622, 624, 626, 627, 630, 632, 633, 634, 636, 638, 657, 659, 788, 847 (in the manuscripts this instance is turn-initial, but all editors accept Musgrave’s transposition of 848 to before 847; indeed the text makes more sense like this, and as Seaford remarks ad loc., in this way “δέ acquires its proper place”), 850, 853, 854, 859, 861, 917, 924, 944, 947, 948, 960, 965, 966, 975, 976, 1333, 1335, 1336, 1338, and 1341. He utters it in turn-initial position (that is, in the first discourse act of a turn) only four times: in 490, 654, 813, and 815.
[ back ] 143. Pentheus utters turn-initial δέ in Bacchae 465, 467, 469, 471, 473, 481, 485, 505 (see (t17); δέ… γε), 663, 830, 832, and 941.
[ back ] 144. On the “neutral” signal of δέ within turns, see III.2 §§24-25; on its relatively high frequency in messenger speeches, see III.2 §§27-28. A high frequency of δέ may also trigger an association to epic; see III.2 §26 on that. In angry or generally agitated contexts δέ seems to be less frequent. Speakers there tend to connect their discourse acts in a less neutral way, or use an incremental style without marking the start of new acts at all (see §§44 above). For example, in Oedipus’ emotional speech at Sophocles Oedipus King 1369-1415 (on which see §§70-77 above), δέ is even completely absent. Heracles’ emotional speech at Sophocles Women of Trachis 1046-1111 (on which see §46 above) has a low δέ frequency of 1.3% (5 instances in 387 words; the average δέ frequency in Sophoclean monologues is 2.6%; see III.2 Table 2); Hermione’s angry 8-line utterance at Euripides Andromache 261-268 has 1.9% (1 instance in 52 words; the average δέ frequency in Euripidean dialogues [maximum of 4 lines per turn] is 2.4%, in monologues [minimum of 25 lines] 3.7%); Jason’s angry speech at Medea 1323-1350 has 2.3% (4 instances in 171 words; the average δέ frequency in Euripidean monologues is 3.7% see III.2 Table 2).
[ back ] 145. See III.4 §§34-35 on this function of turn-initial δέ, and III.4 §15 for discussion of the concept of adjacency pair.
[ back ] 146. A similar stichomythic interrogation scene is Sophocles Oedipus King 1015-1046, where the questioner Oedipus utters 4 turn-initial δέ in his questions; he is clearly much more emotionally shaken by the answers than the shepherd who calmly gives his information. Compare in Aristophanes the question-answer scene of Socrates and Strepsiades cited in (t1) above. As in the Oedipus scene, here as well the one who possesses more knowledge (Socrates) is the calm, high-status character; however, in this case he is the questioner, as befits the philosopher’s style (see §27 above).
[ back ] 147. See Du Bois 2007 and IV.4 §§46-51 for evaluation, positioning, or alignment as the three components of stancetaking. See §22 above on emotions involving a judgment. See IV.4.4.3 on the possible coexistence of epistemic and emotional stance in historiography.
[ back ] 148. This holds also for the observations about connections between particle use and other general pragmatic phenomena, which can be found in the other research chapters of this monograph.
[ back ] 149. See e.g. Foolen 2012:360-364 on the connection between emotion and embodied cognition in general. An indication for this connection is, for example, the finding that emotion-laden words are processed differently from non-emotional words.