III.1 Introduction

§1. A number of similarities and differences between Attic drama and other parts of our corpus has naturally given rise to research questions and approaches that both converge and diverge. Tragedy and comedy share with Homeric epic and Pindaric song their poetic form and an original performative and ritual context. Dating from the fifth century BCE, these texts are close in time to the work of Pindar as well as Herodotus and Thucydides. With Thucydides the four playwrights share their Athenian background and thus features of culture and dialect.
§2. The corpus for this volume’s research consists of a selection of three plays per author: Aeschylus’ Agamemnon, Libation Bearers, and Persians; Sophocles’ Ajax, Antigone, and Oedipus King; Euripides’ Bacchae, Hippolytus, and Medea; and Aristophanes’ Birds, Frogs, and Lysistrata. [1] These twelve plays are widely viewed as canonical: they are frequently read, studied, translated, and performed around the world. However, the upcoming chapters will occasionally adduce examples and parallels from other plays by the same authors as well. Since tragedy takes up the lion’s share of my corpus, it will receive more attention than comedy in my discussion, but most observations based on tragic material have also been checked in the comic material. Every chapter includes examples from each of the four playwrights.

1.1 The performative context

§3. The dramas in the corpus were written for embodied performances, consisting of several actors and chorus members who moved, gestured, danced, spoke, and/or sang. Most important for my current examination, in their capacity as characters they interacted, verbally as well as nonverbally. [2] And of course, an audience was present at the performances to hear and see all of this.
§4. The performative context has many implications, three of which are especially relevant to the research in this volume. First, the language used in the plays can be assumed to be closer to real spoken dialogue than texts that were written to be either read or performed entirely by a single performer or single group of performers. At the least, the process of turn-taking (several speakers taking turns-at-talk after one another) found in normal spoken conversation is still present in dramatic dialogue, even if it is obviously stylized. [3] Second, there are particularly salient and crucial differences among the parts of the plays. Dialogues, monologues, and lyric songs—the three main components of the plays—do not only differ in their metrical structures and linguistic features, but they were performed in fundamentally divergent ways, because of variation in the number of the speakers (one or more) and their type (character or chorus). This is considerably different for non-dramatic texts: although these may include passages of direct speech or variations in meter, they were continuously performed, or read, by one person or one group. This observation is related to a third point: while polyphony and multiple stances are regularly signaled implicitly in non-dramatic texts, in drama they are often explicitly present. That is to say, the utterances attributable to different (fictional) sources, i.e. the characters, were actually spoken by different actors embodying these characters.
§5. These preliminary observations make it clear that the tragedies and comedies of this volume’s corpus are approached as discourse [4] meant for performance, rather than as (later) texts that we happen to have available to read. [5] Even though the texts are all that remains from the performances of the fifth century BCE, this written language represents part of an originally multimodal event. [6]
§6. Recent literary and historical research has thrown light on various performative aspects of Classical drama. For example, costumes, masks, props, and gestures played a crucial role in the plot and meaning of tragedies. [7] Numerous studies, in addition, focus on the fifth-century audience, in particular their theatrical competence, [8] and their response to the performances. [9] Tragedy and comedy also carry political significance in their original context. [10] Analyses of the plays’ later reception often focus on performance as well. [11] The work presented in the current volume is complementary to the research on all of these topics, in that the focus here is on the linguistic part of the dramatic performance.

1.2 Themes and findings

§7. The specific contexts of the plays make certain pragmatic phenomena particularly relevant for linguistic analysis. The following chapters deal with four general phenomena that are related to analyze particle use. Let me illustrate these here with an example for each.
§8. First, the considerable situational differences among the various parts of the plays are connected to recurrent patterns of language use. III.2 accordingly discusses several linguistic tendencies in dialogues, monologues, and choral songs. Based on insights from Construction Grammar (see especially Östman 2005), the term “discourse patterns” is applied to describe and analyze these tendencies.
§9. For example, the particle τε turns out to be much more frequent in choral songs than in other parts of the plays. The following lyric excerpt from Euripides (see III.2 §41, (t8)) illustrates this phenomenon: the song contains more τε instances than is usual in tragedy.
       (Χο.)  †οὐδὲ γὰρ ἐν† πόλεσι δίπτυχοι τυραννίδες
                μιᾶς ἀμείνονες φέρειν,
475           ἄχθος τ’ ἐπ’ ἄχθει καὶ στάσιν πολίταις·
                τεκόντοιν θ’ ὕμνον ἐργάταιν δυοῖν
                ἔριν Μοῦσαι φιλοῦσι κραίνειν.
Euripides’ Andromache 471-477
(Ch.) For cities, likewise, double kingship is worse than single to endure, grief piled on grief for the citizens and the cause of faction. When two poets produce a hymn, the Muses are wont to work strife between them.
On the one hand, τε does have one clear syntactic function which operates in virtually all cases in the drama corpus, and that is the connective one. On the other, this single, widespread function does not adequately explain why τε’s distribution over different kinds of discourse in the plays is variable. That is to say, the variation in the distribution is significant enough to require an explanation. Apart from signaling a connection, τε must provide another pragmatic contribution to the discourse. III.2 discusses this contribution, which is connected to shared encyclopedic and cultural knowledge. The chapter presents analyses of eleven particles with this approach, using distribution as input for interpretation.
§10. The prominence of dialogue in the plays provides an excellent opportunity to consider how speakers build on the words of previous speakers. The ways in which speakers explicitly and implicitly refer to other speakers’ utterances fall within the domain of dialogic syntax, a recently developed linguistic framework (see Du Bois 2014). If speakers deliberately stress certain similarities between their utterance and previous ones, we can speak of “resonance”—the pragmatic phenomenon that III.3 delves into. The highlighted similarities may be lexical, semantic, syntactic, morphological, phonological, metrical, and/or pragmatic. Playwrights may use such resonance for goals that lie beyond the specific interaction of the characters, such as stressing a main theme of a play.
§11. Particles may indicate how a speaker uses such resonance. Consider the following Sophoclean passage (see III.3 §81, (t16)):
        Με.     ἤδη ποτ’ εἶδον ἄνδρ’ ἐγὼ γλώσσῃ θρασὺν
                   ναύτας ἐφορμήσαντα χειμῶνος τὸ πλεῖν,
                   ᾧ φθέγμ’ ἂν οὐκ ἐνηῦρες, ἡνίκ’ ἐν κακῷ
1145            χειμῶνος εἴχετ’, (...)
1150 Τευ.    ἐγὼ δέ γ’ ἄνδρ’ ὄπωπα μωρίας πλέων,
                   ὃς ἐν κακοῖς ὕβριζε τοῖσι τῶν πέλας.
Part of Sophocles’ Ajax 1142-1151
Me. In the past I have seen a man of reckless speech urging sailors to sail during a storm. But one heard no word from him when he was in the grip of the storm's attack; (…)
Te. And I have seen a man full of stupidity, who harried others in their time of troubles.
In his reply to Menelaus, Teucer echoes the other man’s linguistic construction (“I have seen a man…”). [12] The particle combination δέ γε clarifies how the speaker intends this echo to be understood: as a hostile new step in the discourse, in reaction to the previous insult. The two particles occur together several times in pragmatically similar contexts.
§12. III.4 applies the framework of Conversation Analysis. Recent scholarship has identified many regularities in modern talk-in-interaction. A close look at the stylized dialogues in tragedy and comedy reveals that such general patterns play a role there as well. The use of certain particles is clearly related to turn-taking strategies, conversational structure, and the communicative actions performed by turns of speaking.
§13. An example is the use of δέ—a particle that marks discourse boundaries of various kinds—at the beginning of a turn. When δέ is in turn-initial position, the directly preceding change of speaker already establishes that a new unit has just started, namely a new turn of speaking. The particle therefore marks a specific discourse boundary in this initial position. The following Aeschylean dialogue (see III.4 §§34-36, (t8)) contains a case in point:
715 Δα.      τίνι τρόπωι; λοιμοῦ τις ἦλθε σκηπτὸς ἢ στάσις πόλει;
                  Βα.οὐδαμῶς, ἀλλ’ ἀμφ’ Ἀθήνας πᾶς κατέφθαρται στρατός.
       Δα.      τίς δ’ ἐμῶν ἐκεῖσε παίδων ἐστρατηλάτει, φράσον.
       Βα.      θούριος Ξέρξης, κενώσας πᾶσαν ἠπείρου πλάκα.
Aeschylus’ Persians 715-718
Da. How has it happened? Has our state been stricken by a virulent plague, or by civil strife?
Qu. Not at all; what has happened is that our entire army has been destroyed in the region of Athens.
Da. And tell me, which of my sons led the army there?
Qu. The bold Xerxes; he emptied the whole expanse of the continent.
δέ in this environment marks its host turn as the first part of a new question-answer pair within a series of such pairs. That is, the particle combines with the signal conveyed by the conversational structure, in order to indicate how the turn is intended to fit into its surrounding co-text. III.4 analyzes this and similar relations between particle use and conversational phenomena.
§14. Another phenomenon especially relevant in multi-party discourse is the linguistic reflection of emotional states of mind. This process usually involves several verbal and nonverbal signals at the same time. Clues about the play’s plot and semantic or syntactic patterns are used in III.5 to detect how particle use relates to calmness, agitation, and the specific agitated state of anger.
§15. In the dialogue from Aristophanes cited in (t4) (see III.5 §§57-58, (t12)), for example, several cues make it clear that the two speakers are angry:
556bis     Πα.    οὐ μὲν οὖν με προσεδόκας,
557                   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.
Aristophanes’ Frogs 556bis-560
Innkeeper. Hah! You didn’t think I’d recognize you again with those buskins on. Well? I haven’t even mentioned all that fish yet.
Plathane. Right, dearie, or the fresh cheese that he ate up, baskets and all.
γε is particularly frequent in contexts like this one. III.5 explains the connection of this particle to certain affective environments. In general, the chapter illuminates how particles can contribute to expressions of calmness or agitation without ever being the only signs of a particular feeling.
§16. The analysis of how particle use relates to several pragmatic phenomena in dramatic texts provides new tools for understanding particles. [13] The approach taken here makes it possible to explain not only the pragmatic functions of certain particles, but also why they occur in the particular co-texts and contexts they do.
§17. A better understanding of particles, in turn, leads to a stronger basis for interpreting dramatic texts. Taking into account the distribution of particles across different kinds of discourse enables a reader to notice when a certain particle is in fact peculiar to its context, a realization that affects interpretation. For example, the presence of τε in stichomythia, where it is atypical, may carry a sense of solemnity (see III.2). Similarly, sensitivity to a speaker or writer’s uses of resonance and the role of particles within this strategy illuminates content. By considering the presence of certain particles or particle combinations together with resonating elements in an utterance, we may better appreciate how, for example, an utterance works as an insult or sarcastic joke (see III.3). Further, examining particle use in relation to conversational features helps explain the communicative actions of characters. Particles may for instance signal that a speaker wants to keep the floor. Even the fictional language of Greek drama shows that dialogues are constructed jointly by several speakers (see III.4). Finally, characters’ emotional states of mind are connected to the ways in which they structure their discourse. This includes patterns of particle use, even though particles do not directly indicate specific emotions (see III.5).
§18. In sum, particles were crucial to the communicative strategies that the fictional characters represented on stage used to interact with each other, as well as to the strategies used by the authors themselves to communicate with their original fifth-century audiences. If viewed through the analytic tools offered here, particles can once again be of aid to the modern readers of these fascinating literary works.


[ back ] 1. Unless otherwise noted, the editions from the Thesaurus Linguae Graecae Online are used: Page 1972 for Aeschylus, Lloyd-Jones and Wilson 1990 for Sophocles, Diggle 1984 and 1994 for Euripides, Wilson 2007 for Aristophanes. The translations from the most recent Loeb editions are cited, unless indicated otherwise: for Aeschylus Sommerstein 2008a, 2008b; for Sophocles Lloyd-Jones 1997, 1998; for Euripides Kovacs 1998, 2001, 2002, 2005; for Aristophanes Henderson 1998, 2000, 2002.
[ back ] 2. See I.1 §§12-17 on the connectedness of verbal and nonverbal communication.
[ back ] 3. See e.g. Schuren 2015 for a recent study on Euripides that takes turn-taking into account.
[ back ] 4. See I.1 §§12-21 for an introduction to the concept of discourse, and to our use of it in this monograph.
[ back ] 5. This general view on Greek drama has been common at least since Taplin’s influential publications on tragedy (especially 1978 and 1989). See also e.g. Bain 1977 and Revermann 2006 (on drama generally), Foley 2003 (on tragedy), Budelmann 2000 (on Sophocles), Hesk 2007 (on Aristophanes). Meineck 2011 (on tragedy) combines this performative view with insights from neuroscience.
[ back ] 6. See also part II on particle use in Homer and Pindar, where the originally multimodal performance event of these poems is likewise emphasized. This basic approach is outlined in II.1.
[ back ] 7. On costumes in Euripides, see e.g. Worman 1999. On tragic masks, see e.g. Wiles 2007 and Meineck 2011. On props in tragedy, see e.g. Sommerstein 2010 (on Aeschylus) and Mueller 2015. On gestures in tragedy, see e.g. Csapo 2002, and Mueller 2011. On gestures in tragedy and comedy, see e.g. Boegehold 1999. On various methodologies for analyzing all these performative aspects in tragedy, see e.g. Powers 2014. On comic masks, see e.g. Wiles 2008. On several performative aspects in Aristophanes, see e.g. Slater 2002 and Revermann 2006a.
[ back ] 8. See Revermann 2006b.
[ back ] 9. See e.g. Gruber 1986 on the audience response to Aristophanic and other comedies; Ruffell 2008 on audience response to Greek drama in general.
[ back ] 10. See e.g. Goldhill 2000 on tragedy and comedy; Slater 2002 on Aristophanes; Carter 2007 on tragedy; Sommerstein 2010 on Aeschylus.
[ back ] 11. On modern performances of tragedy, see e.g. McDonald 1992; Foley 1999; Meineck 2013 (specifically on the chorus). On modern performances of Aristophanic comedy, see e.g. Van Steen 2000, and papers in Hall and Wrigley 2007 (eds.).
[ back ] 12. Throughout the chapters, formulations such as “Teucer says to Menelaus” are used as convenient shorthands for those such as “Sophocles writes as Teucer’s utterance” or “the audience will have heard the actor performing Teucer say to the actor performing Menelaus.”
[ back ] 13. Several of these phenomena are also linguistically analyzed, albeit usually without an explicitly pragmatic framework, by e.g. Hancock 1917 on stichomythia in tragedy and other genres; Werres 1936 on swearing expressions in Aristophanes; Earp 1944 on the style of Sophocles, and 1948 on the style of Aeschylus; Ireland 1974 on stichomythia in Aeschylus; Stevens 1976 on colloquial expressions in Euripides; Mastronarde 1979 on contact among tragic characters; Stanford 1983 on emotions in tragedy; Pfeiffer-Petersen 1996 on repetition in Sophocles; Kloss 2001 on pragmatic humor in the language of Aristophanes; J. Barrett 2002 on tragic messenger speeches; Van Wolferen 2003 on Euripidean prologues; R.J. Allan 2009 on narrative modes in Euripides; Van Emde Boas 2010 on conversational phenomena in Euripides; Willi 2010 on register variation in drama and other genres; Goldhill 2012 on the language of Sophocles; Rutherford 2012 on tragic language; and Schuren 2015 on Euripidean stichomythia. These works contain remarks on particle use, but this is not their main focus.