The Expression of Approval in Euripides amd Aristophanes

ATHENAEUM Studi di Letteratura e Storia dell’Antichita ` pubblicati sotto gli auspici dell’Universita `di Pavia VOLUME CENTONOVESIMO II ———— 2021 Estratto SANDRA RODRI ´GUEZ PIEDRABUENA The Expression of Approval in Euripides and Aristophanes COMO - NEW PRESS EDIZIONI - 2021 ISSN 0004-6574 ATHENAEUM Studi di Letteratura e Storia dell’Antichita ` DIRETTORI GIANCARLO MAZZOLI (onorario) - DARIO MANTOVANI (responsabile) COMITATO EDITORIALE FEDERICO CONDELLO - STEPHEN HARRISON - FABIO GASTI LUIGI PELLECCHI - CHRISTOPH PIEPER ELISA ROMANO - JOHN SCHEID - FRANCESCA SCHIRONI CHRISTOPHER SMITH - DANIELLE VAN MAL-MAEDER COMITATO REDAZIONALE ALESSIA BONADEO - ALBERTO CANOBBIO - MARCO FRESSURA DONATELLA ZORODDU (coordinatrice) PERIODICITA `SEMESTRALE COMITATO SCIENTIFICO INTERNAZIONALE Michael von Albrecht (Ruprecht- Karls-Universita ¨t Heidelberg) Mireille Armisen-Marchetti (Universite ´de Toulouse II - Jean Jaure `s) Francisco Beltra ´n Lloris (Universidad de Zaragoza) Francis Cairns (Florida State University) Carmen Codon ˜er Merino (Universidad de Salamanca) Michael H. Crawford (University College London) Jean-Michel David (Universite ´Paris I Panthe ´on-Sorbonne) Werner Eck (Universita ¨tKo ¨ln) Michael Erler (Julius-Maximilians- Universita ¨tWu ¨rzburg) Alessandro Garcea (Universite ´Paris IV Sorbonne) Pierre Gros (Universite ´de Provence Aix-Marseille 1 / Acade ´mie des Inscriptions et Belles-Lettres, Paris) Jeffrey Henderson (Boston University) Michel Humbert (Universite ´Paris II Panthe ´on-Assas) Wolfgang Kaiser (Albert-Ludwigs- Universita ¨t Freiburg) Eckard Lefe `vre (Albert-Ludwigs- Universita ¨t Freiburg) Matthew Leigh (St Anne’s College, Oxford) Carlos Le ´vy (Universite ´Paris IV Sorbonne) Clelia Mora (Universita `di Pavia) Jan Opsomer (KU Leuven) Ignacio Rodrı ´guez Alfageme (Universidad Complutense de Madrid) Alan H. Sommerstein (University of Nottingham) Theo van den Hout (University of Chicago) Juan Pablo Vita (ConsejoSuperiordeInvestigaciones Cientı ´ficas, Madrid) Gregor Vogt-Spira (Philipps-Universita ¨t Marburg) Paul Zanker (Ludwig-Maximilians- Universita ¨tMu ¨nchen / SNS Pisa) Bernhard Zimmermann (Albert- Ludwigs-Universita ¨t Freiburg) INDICE DEL VOLUME Articoli A. BELTRAMETTI,Dietro la trama compatta della rivolta ionica. L’io di Erodoto e le memorie corali [Behind the Compact Plot of the Ionian Revolt. The Ego of Herodotus and the Choral Memories] p. 357 S. RODRI ´GUEZ PIEDRABUENA,The Expression of Approval in Euripides and Aristophanes ......... » 379 M.B. CHARLES -M.SINGLETON,An Elephant Bone from Tamuda, Morocco. Contextual Observations and Historical Implications ................................................................................. » 404 S. ACERBO,Leggere e scrivere le tradizioni mitiche. Il lavoro del mitografo nella sezione teogonica della Biblioteca di Apollodoro [Reading and Writing the Mythical Traditions. The Role of the Mythographer in the Theogony of Apollodorus’s Library] .................................................... » 417 M.J. REQUENA -A.SAPERE,Entre la tiranı ´ a y la hegemonı ´ a de masas. Ima ´ genes de la democracia en la Vida de Pericles de Plutarco [Between Tyranny and Hegemony of the Masses. Images of Democracy in Plutarch’s Life of Pericles] ........................................................................ » 438 P. REDONDO REYES,Claudio Ptolomeo y el origen del lenguaje [Claudius Ptolemy and the Origin of Language ].................................................................................................................. » 491 L. PELLECCHI,Retorica di scuola, argomentazione forense e processo nella declamatio minor 336 dello Ps. Quintiliano [School Rhetoric, Legal Argument and Procedure in Ps. Quint. decl. min. 336 ].................................................................................................................. » 513 J. OLLER GUZMA ´N,La «Pompeya del desierto». 200 an ˜ os del redescubrimiento del Mons Smaragdus por Fre ´ de ´ ric Cailliaud [«Pompey of the Desert». 200 Years from the Rediscovery of Mons Smaragdus by Fre ´ de ´ ric Cailliaud ]......................................................................................... » 555 Note e discussioni R. XIAN,A Note on Aratus’ Phaenomena 100-107 ............................................................. » 597 P. PERSANO,Un’urna romana inedita da Villa Miollis [A Roman Unpublished Urn from Villa Miollis ]...................................................................................................................... » 602 P. PAOLUCCI,Estetica dei trucioli [The Beauty of Vergilian Centos]....................................... » 608 J. HØYRUP,Observations and Reflections Concerning the First Volume of a New Edition of the Liber abbaci ............................................................................................................... » 618 M.H. CRAWFORD,Two Copies of the Digestum Nouum .................................................... » 628 Recensioni V. AZOULAY,The Tyrant-Slayers of Ancient Athens. A Tale of Two Statues (F. Muscolino) ..... » 631 N. BARRANDON,Les massacres de la Re ´ publique romaine (R. Scuderi) ................................... » 636 F. BORDONE (ed.): Paolino Nolano, Per la morte di un fanciullo [carm. 31] (S. Audano) ...... » 640 A. BORGNA,Ripensare la storia universale. Giustino e l’Epitome delle Storie Filippiche di Pompeo Trogo (C. Bearzot) ...................................................................................................... » 644 F. BUDELMANN -T.PHILLIPS (eds.), Textual Events. Performance and Lyric in Early Greece (E. Cerbo) ....................................................................................................................... » 648 P. FEDELI -G.ROSATI (a c. di), Ovidio 2017. Prospettive per il terzo millennio (S. Martorana) » 652 V. GIGANTE LANZARA (a c. di): Arato di Soli, Fenomeni (M. Rossetti) ................................ » 655 F. GRELLE -M.SILVESTRINI -G.VOLPE -R.GOFFREDO,La Puglia nel mondo romano. Storia di una periferia. L’avvio dell’organizzazione municipale (G. Mennella) ................................... » 658 H.-CH. GU ¨NTHER (ed.), A Short Companion to Tibullus and the Corpus Tibullianum (E. Merli) » 663 TH. HARRISON -E.IRWIN (ed. by), Interpreting Herodotus (E. Biondi) ................................ » 668 G. HAWES (ed.), Myths on the Map. The Storied Landscapes of Ancient Greece (S. Acerbo) ..... » 671 M. HUMAR,Rhetorik der Verunsicherung. Affekt-Strategien in den platonischen Fru ¨ hdialogen (F. Ferrari) ...................................................................................................................... » 675 S. KACZKO,Archaic and Classical Attic Dedicatory Epigrams. An Epigraphic, Literary, and Linguistic Commentary (L. Floridi) .................................................................................................. p. 679 CH. LILLINGTON-MARTIN -E.TURQUOIS (eds.), Procopius of Caesarea: Literary and Historical Interpretations (M. Cristini) ......................................................................................... » 684 J. MARTI ´NEZ-PINNA,Roma y los latinos. ¿Agresividad o imperialismo? (G. Bandelli) ................ » 688 G.M. MASSELLI,Status causae tra dottrina e prassi scolastica (A. Casamento) ......................... » 693 C. MOATTI, Res publica. Histoire romaine de la chose publique (K.-J. Ho ¨lkeskamp) .............. » 696 P. MORET,Des noms a `la carte. Figures antiques de l’Ibe ´ rie et de la Gaule (S. Panichi) ........... » 699 J.K. NEWMAN,Catullus as Love Poet (A.M. Morelli) .......................................................... » 707 P. ORSI,I taccuini I, riproduzione anastatica e trascrizione dei Taccuini 1-4, a c. di G. LAMAGNA - G. MONTEROSSO (S. De Maria) ........................................................................................ » 712 S. PANICHI,La Cappadocia ellenistica sotto gli Ariaratidi ca. 250-100 a.C. (L. Ballesteros Pastor) » 715 P. POCCETTI (ed.), Oratio obliqua. Strategies of Reported Speech in Ancient Languages (R. Utard) » 719 A. ROLLE,Dall’Oriente a Roma. Cibele, Iside e Serapide nell’opera di Varrone (F. Muscolino) .» 721 Semanas de Estudios Romanos, Vol. XVIII. In memoriam Paul B. Harvey, Jr. (L. Polverini) .... » 726 F. SILVESTRELLI,Le duc de Luynes et la de ´ couverte de la Grande Gre ` ce (F. Muscolino) ........... » 729 E. ZINGG (ed.): Isokrates, Archidamos (C. Bearzot) ............................................................ » 733 Notizie di Pubblicazioni S. DESTEPHEN -B.DUME ´ZIL -H.INGLEBERT (e ´d.), Le Prince chre ´ tien de Constantin aux royaute ´ s barbares (IV e -VIII e sie ` cle) (A. Marcone) ........................................................................ » 735 U. FANTASIA,Ambracia dai Cipselidi ad Augusto. Contributo alla storia della Grecia nord-occiden- tale fino alla prima eta `imperiale (C. Biagetti) ................................................................ » 737 E. FRANCHI -G.PROIETTI (ed.), Conflict in Communities. Forward-looking Memories in Classical Athens (D. Bonanno) .................................................................................................. » 738 W. HU ¨BNER,Athena am Sternhimmel bei Proklos. Astrologie im Dienste neuplatonischer Philosophie (F. Ferrari) .............................................................................................................. » 740 N. KAVVADAS,Jerusalem zwischen Aachen und Bagdad.Zur Existenzkrise des byzantinischen Christentums im Abbasidenreich (L. Troiani) ......................................................................... » 740 F. KNAUSS,Die Kunst der Antike. Meisterwerke der Mu ¨ nchner Antikensammlungen (M. Brunetti) » 742 C. KOEHN,Justinian und die Armee des fru ¨ hen Byzanz (A. Marcone) .................................. » 743 R. LUCA,Labirinti dell’Eros. Da Omero a Platone (F. Cannas) ............................................ » 744 T. PHILLIPS - A. D’ANGOUR (ed. by), Music, Text, and Culture in Ancient Greece (L. Bravi) .» 747 Pubblicazioni ricevute ..................................................................................................... » 749 Elenco dei collaboratori dell’annata 2021 ......................................................................... » 751 Indice generale ............................................................................................................... » 753 Elenco delle pubblicazioni periodiche ricevute in cambio di «Athenaeum» e donate alla Biblioteca di Studi Umanistici dell’Universita `di Pavia ............................................................ » 756 Athenaeum 109/2 (2021) THE EXPRESSION OF APPROVAL IN EURIPIDES AND ARISTOPHANES ABSTRACT. The expression jakx &|e >kena| can be literally translated as «you spoke beautifully», «you spoke well». The departure point of this paper is to examine how this formula works beyond its propositional meaning. The distribution of jakx &|e >kena| and variants is examined in terms of their position on the sequence organisation of turns at talk. Likewise, the formulae are analysed with regard to two key factors in Politeness Theory, namely Power and Distance, i.e. the difference in the status and the degree of familiarity between participants, respectively. The formula jakx &|e >kena| appears mostly in the second part of an adjacency pair or at third position, viz. as a post-expansion to an adjacency pair. In addition, jakx &|e >kena| is mostly uttered by characters with greater power to ones with less power in the interaction. As a result of this research, it is possible to explore the difference between agreement and approval and the formulae associated with these two different speech acts. Since approval is expected to be uttered by characters with greater power in the interaction, the formulae associated with it can work as taxemic markers, that is, markers of non-symmetric interactions between participants of different status. 1. Introduction Jakx &|e >kena| can be literally translated as «you spoke beautifully», «you spoke well». This paper examines whether this expression, together with some variants, has any other function beyond its propositional meaning, that is, whether it has a procedural meaning, and if so, what meaning it has and on what factors it depends. 1.1. Corpus The analysed corpus includes Euripides’ work, as well as Aristophanes’, excluding the fragments and the Rhesus, due to its dubious authenticity (Ritchie 1964). I will focus on the following expressions: jakx &|e >kena|, jakx &|ke *cei|, jakx &|, et #ke *cei| and et #ei #pa|, to which I will refer as ‘jakx &|ke *ceim formulae’ hereinafter, for the sake of simplicity. Texts were drawn from the TLG browsing tool, although I resorted to the main ad-locum commentaries when necessary. There are 59 instances of jakx &|ke *ceim formulae. A further group is included which is labelled as ‘varied formulae’. This group primarily consists of variants of the former expressions, see 1) and 2). Example 1) is closer to the formulae under study than example 2): Athenaeum 109/2 (2021), pp. 379-403 * I am most thankful to the reviewers, to Emilia Ruiz Yamuza, and to Professor Mastronarde for improving my work. Any errors or infelicities remain exclusively my own. This research has been funded by the Spanish Government through the project GRAPAGA (PGC 2018-095147-8-100). 1) roi *s\ et #ke *kejsai jai +sa +sx &md\ at \vx &,ce *qom, soiat &s\ e >rerhai" lmglomet *resai va *qi| You have spoken well, old man, and I am confident that these children will do just as you say: our favour will be remembered. (Eur. Heracl. 333-334, trans. by Allan 2001, p. 83). 2) cemmai &ale +msa *d\ ei #pa| a \kk\ a \lg *vama. The things you have said are noble but impossible. (Eur. Heracl. 464, trans. by Allan 2001, p. 91). I have examined jakx &|ke *ceim formulae in contrast with the analysis of o \qhx &| [e >kena|/ke *cei|] («that’s right,» literally «you spoke rightly»), as well as of la *kirsa («most of all», «certainly»), e >rsai («it shall be so»), and e \paime *x/ai \me *x(«to praise»). In total, 176 examples have been analysed (Table 1). Table 1. Corpus jakx &|e >kena| (13) jakx &|ke *cei| (12) jakx &|(10) et #ke *cei| (18) et #ei #pa| (6) varied formulae (23) o \qhx &|[e >kena|/ke *cei|] (9) e >rsai (23) la *kirsa (21) e \paime *x(14) ai \me *x(27) 59 117 176 1.2. Previous Approaches P.T. Stevens considered these expressions colloquialisms (1937, p. 187; 1976, pp. 54-55). He picked up ideas by J.H. Quincey (1966), who focused on jakx &| e >vei «it holds good» and concluded that it conveys acceptance, and specifically gratitude. J.H. Quincey extended his conclusion to account for the meaning of the expressions containing ke *ceim/ei \pei &m, herein analysed, which are much more frequent. As a result, jakx &|ke *ceim formulae were assumed to convey thanks: «they function in exactly the same way: they express gratitude by conferring praise» (Quincey 1966, p. 142). The expression of thanks can be conveyed by virtue of «reluctance» (Brown- Levinson 1987, pp. 190, 210), «self-abasement» (Brown-Levinson 1987, pp. 67, 247) or «debt». Examples of the latter are found 1but thanking in Ancient Greek is mostly conveyed in laudatory terms, for instance by the verb (e \p)aimx &«to praise», as well as by the expression et #poiei &m+ participle/x <|(Stevens 1937, p. 187). This S. Rodrı ´guez Piedrabuena, The Expression of Approval in Euripides and Aristophanes380 1For expressions including va *qi|, cf. Soph. Phil. 557-558; see also Lloyd 2006, p. 231. could be the reason why the formulae here revisited were traditionally interpreted as expressions of gratitude 2. Another relevant question relates to the expression of what is traditionally understood as «polite refusal». I will refer to the examples of polite refusal as «dispreferred SPPs», which stands for «dispreferred S[econd] P[air] P[art]s» (see Sidnell 2010). Actions as conceived in Conversation Analysis (CA) are organised in minimal units known as «adjacency pairs» (Sidnell 2010, pp. 63-66; Berger 2017, p. 12). Each adjacency pair consists of an FPP (first pair part) or starter and an SPP (second pair part) or completer (ten Have 2007, p. 136), by virtue of what is known as «conditional relevance» (Schegloff 1968, p. 1083). In other words, an FPP determines and constrains the number of possibilities for a corresponding SPP, the absence of which is pragmatically salient. A classic example is that of the ‘questionanswer’ pair. The main features of an adjacency pair are: FPP and SPP are adjacent (although not necessarily contiguous); they are uttered by different speakers; and they have a set order in that FPP constrains the relevance of and the preference for a particular SPP (Schegloff-Sacks 1973, pp. 74, 295-296; Heritage 1984, p. 246). Thus, a particular SPP is not just a matter of relevance but of preference: «the response to the first pair part that promotes the accomplishment of the activity is, typically, ‘preferred’» (Sidnell 2010, p. 77). An example of this is given by the ‘invitation-acceptance’ pair. Scholarshave traditionallystruggled to account separately for the expressions conveying polite refusal and, on the other hand, for those expressing acceptance or gratitude, since it was considered a contradiction that antithetical functions could be conveyed by one and the same expression 3. Thus, even though J.H. Quincey considers «jakx &|e >vei as a refusal formula», including ja *kkirsa, he strives to establish an origin of the formulae for acceptance or gratitude which is different from the origin of the formulae for the expression of polite refusal. He considers that the latter «has as its full form jakx &|e >vei loi, impersonal», whereas the acceptance formula would have «as its full form jakx &|e >vei, personal, in the sense ‘[Your gift, etc.] is good, all right’» and that there is eventually an overlap between the two types, presumably of a different origin.It seems that the data in supportof this theory are insufficient: only one instance Athenaeum 109/2 (2021) 381 2For example, Barrett 1964, p. 294 (ad Eur. Hipp. 715) translates «thank you for your words» (the text has other readings). Likewise, Collard 2005, p. 364, considers these formulae expressions of gratitude only and compares them with lat. bene facis as in Ter. Eun. 186; Ad. 601, 604. With regard to bene facis as an expression of gratitude, see Don. ad Eun. 186; Bagordo 2001, p. 114, who treats it as translation from Gr. jakx &| poiei &mformulae; Unceta Go ´mez 2010, p. 629. 3Amati 1901, p. 142; Stevens 1937, p. 188; Quincey 1966, p. 135; Bond 1981, p. 377; 133 ad Eur. HF 275: «a one-word formula of approval [...]; Wilamowitz is premature in taking it as a polite refusal, like the colloquial e \paimx &: the refusal begins at 277». However, Wilamowitz is right (see Section 5). On the other hand, Quincey 1966, p. 147, interprets the formula as an expression of gratitude. with dat. loi is provided for the case of polite refusal (Quincey 1966, pp. 135, 141- 142). The remaining examples are the same as in the case of acceptance. Furthermore, this assumption could have distorted the identification of examples that express polite refusal, asthe author claims that «onlythe acceptance formula wasconsidered dignified enough to be admitted to tragedy» (Quincey 1942, p. 135, see also 142). The results from the corpus here analysed do not validate this idea: there are 30 examples of dispreferred SPPs, out of the 176, 24 of which belong to Euripides and 6 to Aristophanes. Since roughly twice as many of Euripides’ plays have been preserved, there is no great difference proportionately (see Table 4). 2. Method The above-mentioned expressions have been analysed under the theoretical framework of Politeness Theory as advanced by Brown-Levinson 1987 4.Thistheoretical framework provides general principles which help to perceive the interpersonal or procedural meaning beyond the literal translation or propositional meaning of a given expression (Watts 2003, pp. 72-175). R.J. Watts conceived this opposition as suggested by Halliday 1978: an expression can convey an ideational or interpersonal meaning. The author gives the expression «nice day today» as an example, which can convey a propositional meaning, relating to the weather conditions; or a procedural meaning in the context of an adjacency pair as a greeting formula 5. I have analysed the distribution of jakx &|ke *ceim formulae according to the following parameters: 1) text type (Dialogue, Rhesis, Chorus); 2) gender (Male, Deity, Female); 3) distance (Positive, Negative, Non-applicable 6); 4) power (Positive, Negative, Non-applicable) 7. The P (power) factor corresponds to the vertical axis: it accounts for the difference in status among participants in the interaction (see Kerbrat-Orecchioni S. Rodrı ´guez Piedrabuena, The Expression of Approval in Euripides and Aristophanes382 4I do not disregard the fact that the theory by Brown-Levinson 1987 has been variously contested (Eelen 2001; Watts 2003, pp. 98-99), to the point of rejection. In this respect, Berger 2017, p. 16, states that their theory is «as criticised as influential». Nonetheless, beyond what Eelen 2001 and Watts 2003 try to claim, their models of linguistic politeness are actually compatible and complementary to Brown and Levinson’s, as will be seen later, see Dickey 2016; Leech 2014, p. 43: «these different approaches tend to be presented as in competition with one another, as if in choosing one we have to reject the others [...]. But often the matters that appear to be differences of opinion are no more than differences of emphasis or perspective». 5Watts 2003, p. 173. This approach is analogous to that carried out by Zakowski 2014, pp. 174-175, concerning the expression ei \pe *loi. 6Whenever the degree of familiarity between participants cannot be easily determined. 7Whenever the interaction is symmetric, that is, when there is no difference in the status of the speakers or, if there is any difference, it is not easy to determine. 1992, pp. 71-139). The D (distance) factor identifies with the horizontal axis of the interaction. The horizontal axis is related to the conversational history of the participants. It can include the extent to which participants know each other, their emotional or social bonds (e.g. mother-son), as well as the interactive context (home-of- ficial ceremony) (see Kerbrat-Orecchioni 1992, pp. 39-69; Denizot 2011, p. 126). In a very simplified scheme, there is a gradation which ranges from the formal to the familiar. We can determine whether a positive or negative D factor should be considered by checking the first entrance of a character on stage: we would consider a negative D factor whenever the character already on stage is able to identify the arriving character. Four dyads emerge by combining P and D variables (see Figure 1). Dyad I + Power + Distance Dyad II + Power - Distance Dyad III - Power + Distance Dyad IV - Power +- Distance Figure 1. Fourfold dyadic Power-Distance system (see Brown-Levinson 1987, p. 250) These dyads can account for the reason why, among other aspects, nurses and tutors in tragedy present an interactive behaviour which is different from that of the remaining characters of low status. That is precisely because the D factor is negative (Dyad II), which means that they have a high degree of familiarity with their masters, due to their years in service and experience. The ‘Dyad II’ with a low D minimising the P factor, as described here, is also attested in the Roman comedy in the interactions between (young) masters and slaves 8. The first two dyads represent non-symmetric interpersonal relationships: there is a participant who possesses more power in the interaction. Under this theoretical framework, a taxeme is defined as a marker of non-symmetric interactions between participants of different status. There are fairly manifest verbal taxemes, such as the forms of address (Held 1999, pp. 21-36). Other aspects related to CA can be interpreted as taxemes. For instance, issues concerning turn-allocation: who is to begin and end a conversation, who assigns the next turn, who holds the floor longer, who interrupts, etc. (Denizot 2011, p. 129). It is possible to determine the procedural meaning of jakx &|ke *ceim formulae by analysing their distribution according to this fourfold dyadic system. Athenaeum 109/2 (2021) 383 8See Barrios-Lech 2016, pp. 229-231. Thus, the D factor can alter any initially predicted pattern, as advanced by Venegas Lagu ¨e ´ns 1991, p. 123, in her study on linguistic characterisation in Gaskell: «when can the rules of conversational conduct be broken? When the relation between the speakers is intimate, the possibility of predicting their verbal conduct is low», see Venegas Lagu ¨e ´ns 1991, p. 131. ceding to his biased arguments. The only two examples of o \qhx &|ke *cei| uttered by Socrates are special: one belongs to the minority of third sequence-closing assessments (Pl. R. 543c.4: o \qhx &|, e >ugm, ke *cei|. a \kk\ a >c\, e \peidg +sot &s\ a \peseke *ralem...) while the other is a uniquum, namely an overt dispreffered reaction (Pl. R. 578b. 5: ot \je *s\ o \qhx &|ke *cei|). In this regard, there are only two examples of the unmitigated expression ot \jakx &|ke *cei| in the Platonic corpus. One of them appears in a hypothetical conversation by Socrates (Pl. Ap. 28b.6: i >rx| a /mot #mei >poi si|...) while the other one is uttered by Alcibiades (Pl. Prt. 336b.7), who abruptly intervenes self-selecting turn when Callias is rebuking Socrates. This can be considered a tool for characterisation. 4. Results Regarding the text type, jakx &|e >kena| and variants appear mainly in dialogue 33. When they appear in rheseis, they are frequently placed at the beginning of a replying rhesis. This hints at the potentially interactive functions that these expressions have. The difference between a positive and negative D factor is minor and, consequently, the inferences gained from the D factor are small: characters address jakx &| ke *ceim formulae no matter whether they are close to the addressee or not. Thus, the D factor is positive in 42.37% of the cases 34, that is, characters are not acquaintances; the D factor is negative in 47.46%, that is, characters do know each other 35. Non-applicable cases are 10.17%, in which the degree of familiarity between participants cannot be easily determined 36. Female characters rarely employ jakx &|ke *ceim formulae when addressing male characters: 79.66% of the jakx &|ke *ceim formulae are uttered by male characters or S. Rodrı ´guez Piedrabuena, The Expression of Approval in Euripides and Aristophanes390 33 There are only two exceptions, namely jakx &|e >kena| in Eur. IT 597 (rhesis) and in Eur. Ion 648 (choral distich). 34 a)jakx &|e >kena|: Eur. Alc. 1104; Heracl. 726; IT 597 (Orestes-Iphigenia, NB: the anagnorisis happens later); Hel. 141; Bacch. 953 (Pentheus-Dionysus [as captive]); IA 829; b)jakx &|ke *cei|/ke *cese: Ar. Ach. 765; yNub. 1289; Av. 1614; Lys. 1102; Plut. 481; Plut. 1092; yEur. Hel. 441 (Menelaus - Old Woman [a gatekeeper]); c)jakx &|: Ar. Av. 651; Eccl. 1092 (Old Woman - Young Man); Ran. 888; Eur. Ion 417 (Xuthus- Ion [as servant of the shrine]); d)et #ke *cei|: Eur. Or. 1524; Ar. Ach. 1013; yAr. Eq. 492 (Sausage seller - Servant [v.l. Demosthenes]); Av. 1124; Lys. 1095; Plut. 198-199; Plut. 800; e)et #... ei #pa|: Eur. Bacch. 824 (Pentheus-Dionysus [as captive]). 35 a)jakx &|e >kena|:yEur. Hipp. 715; Hec. 1007; El. 640; IT 766; IT 909; Ion 648; Hel. 158; b) jakx &|ke *cei|/ke *cese: Ar. Lys. 184; Lys. 180; Lys. 1182; Ran. 169; Ran. 643; c)jakx &|: Ar. Nub. 848; Vesp. 785; Ran. 512; Eur. HF 599; Or. 1216; d)et #ke *cei|: Eur. Hipp. 503; Phoen. 985; Or. 783; IA 1445; Ar. Pax 1051; Ran. 1480; e)et #... ei #pa|: Eur. HF 606; IT 118; IT 1325; Hel. 1063; Or. 638. 36 a)jakx &|:yAr. Eq. 23 (Slave B - Slave A, v.l. Nicias-Demosthenes); b)et #ke *cei|:yEur. Or. 111; Or. 173; yAr. Nub. 1092; Pax 934; yEccl. 279. deities 37; only 20.34% are uttered by female characters, and when they are, they are addressed to characters who are also female 38. As for their dyadic distribution, they appear predominantly in the non-sym- metric Dyads I and II, that is, they are uttered by a participant with a positive P factor, and they are addressed to a participant of lower status in the interaction. Thus, characters who utter jakx &|ke *ceim formulae are of a higher status than their addressees in 67.8% of the cases 39; only 32.2% of jakx &|ke *ceim formulae are uttered in contexts in which the interaction is symmetric or not easily determined (22.03%) 40 or in which the speaker has no power in the interaction (10.17%) 41. The latter is exceptional and happens when the D factor is negative, that is, when participants are familiar with each other, since the P factor is neutralised. Thus, the examples found in these cases are formulae uttered by sons to fathers 42, by daughters to mothers 43, or by tutors/nurses to their superiors 44. In these cases, characters do not use the Euripidean variant jakx &|e >kena|, but other less formal variants. Athenaeum 109/2 (2021) 391 37 a)jakx &|e >kena|: Eur. Alc. 1104; Heracl. 726; Hec. 1007 (a); El. 640; IT 597; IT 766; IT 909; Ion 648 (Chorus-Ion/Xuthus); Hel. 158; Bacch. 953; IA 829; b)jakx &|ke *cei|/ke *cese: Ar. Ach. 765; yNub. 1289; Av. 1614; Lys. 1102; Ran. 169; Ran. 643; Plut. 481; Plut. 1092; yEur. Hel. 441; c)jakx &|:yAr. Eq. 23 (Slave B - Slave A, v.l. Nicias-Demosthenes); Nub. 848; Vesp. 785; Av. 651; Ran. 512; Ran. 888; Eur. HF 599; Ion 417; Or. 1216; d)et #ke *cei|: Eur. Phoen. 985; Or. 783; Or. 1524; Ar. Ach. 1013; yEq. 492; yNub. 1092; Pax 934; Pax 1051; Av. 1124; Lys. 1095; Ran. 1480; Ar. Plut. 198-199; e)et #... ei #pa|: Eur. HF 606; IT 118; IT 1325; Hel. 1063; Bacch. 824; Or. 638. 38 yEur. Hipp. 715; Hipp. 503; Hel. 141; yOr. 111; Or. 173; IA 1445; Ar. Lys. 180; Lys. 184; Lys. 1182; yEccl. 279; Eccl. 1092; Plut. 800. The exceptions are Ar. Plut. 800 (Chremylos’ Wife - Plutus) and Ar. Eccl. 1092 (Old Woman - Young Man). However, the remaining exceptions can contribute to the characterisation of exceptional characters and situations: Eur. Hel. 141 (Helen-Teucer); Ar. Lys. 1182 (Lysistrata- Athenians). 39 a)jakx &|e >kena|: Eur. Alc. 1104 (Admetus-Heracles, as host and ne *mo| respectively); Heracl. 726; yHipp. 715; Hec. 1007; El. 640; IT 597; IT 766; IT 909; Hel. 158; Bacch. 953 (Pentheus-Dionysus [as captive]); IA 829; b)jakx &|ke *cei|/ke *cese: Ar. Ach. 765 (Dikaiopolis-Megarian [a beggar]); Av. 1614; Lys. 1102; Lys. 1182; Ran. 169; Ran. 643 (Xanthias [disguised as Heracles] - Aeacus); yEur. Hel. 441; c)jakx &|: Ar. Nub. 848; Vesp. 785; Av. 651; Eccl. 1092; Ran. 512 (Xanthias [disguised as Heracles] - Maid); Eur. HF 599; Eur. Ion 417; Or. 1216; d)et #ke *cei|: Eur. Hipp. 503; Or. 783; Or. 1524; IA 1445; Ar. Ach. 1013; yEq. 492; Av. 1124; Lys. 1095; Plut. 198-199; e)et #... ei #pa|: Eur. IT 118; IT 1325; Hel. 1063; Bacch. 824 (Pentheus-Dio- nysus [as captive]); Or. 638. 40 a)jakx &|ke *cei|/ke *cese:yAr. Nub. 1289; Lys. 184; Lys. 180; Plut. 481; Plut. 1092; b)jakx &|:yAr. Eq. 23; c)et #ke *cei|:yEur. Or. 111; Or. 173; yNub. 1092; Pax 934; Ran. 1480; yEccl. 279; d)et #... ei #pa|: Eur. HF 606. 41 jakx &|e >kena|: Eur. Ion 648; Hel. 141; b)jakx &|: Ar. Ran. 888; c)et #ke *cei|: Eur. Phoen. 985; Ar. Pax 1051; Plut. 800. For Eur. Or. 1522-1527, see this Section below. 42 Eur. Phoen. 985 (Menoeceus-Creon: et #ke *cei|,pa *seq). 43 Eur. IA 1459 (Iphigenia-Clytemnestra: lg +rt *c\" ot \jakx &|ke *cei|). This lies outside the corpus since it is in the negative form. 44 Eur. Ion 1022 (Tutor-Creusa: ot \jet #so *d\ ei #pa|" jai +rt +ca +q sot \lo +mwe *cei|). This lies outside the corpus since it is in the negative form. When we compare these expressions with apparently similar expressions, such as o \qhx &|[e >kena|/ke *cei|] and la *kirsa, we find that in the latter the P factor is not significant. This indirectly implies that the distribution of the P factor is not random in the jakx &|ke *ceim formulae. Thus, a positive P factor is only 46.67% in the case of o \qhx &|[e >kena|/ke *cei|] and la *kirsa 45, whereas 53.34% are instances in which the interactions are symmetric or not easily determined (36.67%) 46, or the characters are of lower status (16.67%) 47. As already pointed out (see Section 1.2), jakx &|ke *ceim formulae were considered to convey thanks, similar to (e \p)aimx &48. However, there are significant differences in their distribution 49. First, e \paimx &and ai \mx &are more frequent in rheseis, especially the latter 50. Second, the speakers are indistinctly male and female. Third, the interactions are not necessarily non-symmetric, that is, the speaker does not have a positive P factor as frequently as in the case of jakx &|ke *ceim formulae. No distribution pattern found in the jakx &|ke *ceim formulae works in (5): 5) ai \mx &Uoi &bom ot \jai \mot &ra pqi *m ot %mev\ ot ’pos\ g \le *kgre paido +|a \podi *dxri *loi. I p r a i s e Phoebus, even though I did not before, since he gave me back the child about whom he never cared. (Eur. Ion 1609-1610, Creusa-Apollo) The speaker is a female character addressing not just a male character but a deity. The text type is rhesis and is not strictly interactive, as Apollo does not interact with Creusa on stage – Apollo is notoriously a no-show at the end of the play, so S. Rodrı ´guez Piedrabuena, The Expression of Approval in Euripides and Aristophanes392 45 a)o \qhx &|(e >kena|/ke *cei|): Eur. Ion 745; Ion 1025; Bacch. 838 (Pentheus-Dionysus [as captive]); Ar. Nub. 679; Ran. 672; Plut. 1033; b)la *kirsa:E.Med. 677; Med. 944; Tr. 63; Hel. 1415; Or. 235; Or. 1108; Ba. 812; Ar. Pax 834. 46 a)o \qhx &|(e >kena|/ke *cei|): Eur. Or. 100; yOr. 110; Ar. Ach. 397; b)la *kirsa:E.Hec. 989; Hec. 1004; Ion 1293; E. IA 363; Ar. Nub. 253; Nub. 672; Ecc. 768; Pl. 826. 47 la *kirsa:E.Heracl. 641; Heracl. 794; IA 835; Ra. 125; Ecc. 1128. 48 I have found the following instances in my sample, in which e \paimx &can be potentially understood as a thanking formula, see Quincey 1966, pp. 133-155: Eur. HF 1235 (Bond 1981, p. 377; Amati 1901, p. 143); Tro. 53, 718 (e \pz *mer\); Or. 1092, 1672; IA 440; Ar. Ran. 508. Likewise, there are the following examples of ai \mx &: Eur. Med. 1157; HF 275; ySupp. 388; Ion 1609-1610; El. 396 (but see Cropp 1988, p. 126); IT 1023, 1486; Hel. 1232; Phoen. 614, 1683; Tro. 890; Bacch. 944; IA 506, 824, 655. The following examples are unclear, since it is possible to understand the verb literally («to praise»): Eur. Hipp. 483; Eur. Or. 786. 49 The following examples are excluded because they are not interactive (either the character addresses himself, the chorus is the speaker but does not interact with the characters or the context is of a general reflection): Ar. Ach. 485; Thesm. 1213; Ran. 596. Similarly, Eur. Heracl. 436; El. 1247; Hipp. 37 are excluded. 50 The verb in its simple form ai \mx &seems to belong to a higher register, since it is not attested in Aristophanes. There are similar instances, such as a \pohmz *rjx/hmz *rjx, as the former is prosaic whereas the latter is poetic, cf. Ku ¨hner-Gerth, II, pp. 352-353, 442-443, x343. it is even difficult to say that Creusa is actually addressing him rather than talking about him. Be that as it may, the verb can be conceptually or literally interpreted as «to praise» or we can interpret it in terms of giving thanks. There are certain examples (e.g. Ar. Lys. 198; Eur. Med. 908) in which the verb features as expressing approval and adopts the function of jakx &|ke *ceim formulae (see below). However, these cases are rare and are found more often than not in indirect speech, that is, not in contexts in which the verb works as performative (prototypically direct speech with a first-person singular form in the present). Note the examples 6) and 7). The verb in 6) is also in the negative, something barely attested in the case of jakx &|ke *ceim formulae: 6) e \peqqo *gram d\ oi =le +mx <| jakx &|ke *coi, oi =d\ ot \je \pz *motm Some shouted in answer that he spoke well, others did not approve. (Eur. Or. 902) 7) o <qy &|, e \paimei &vot #so| See? He also approves 51. (Ar. Av. 1616) The common point is that both jakx &|ke *ceim formulae and the verb (e \p)aimx & can feature as dispreferred SPPs, as discussed in Section 5 below 52. Jakx &|ke *ceim formulae usually appear in conversational contexts 1) and 2), as described above (see Section 3), and convey approval in 94.91% of the cases 53,of which dispreferred SPP (see Section 5) are 16.95%. In contrast, o \qhx &|[e >kena|/ke *- cei|] and la *kirsa do not regularly appear in the conversational contexts 1) and 2) 54. Only 23.33% of the cases could be potentially interpreted as expressing approval and dispreferred SPPs constitute 13.33% of the cases. \Oqhx &|[e >kena|/ke *- cei|] differs from jakx &|ke *ceim formulae and e >rsai (see Table 6) in the sense that it generally keeps its propositional or literal meaning, in line with the cognate o \qhoe *peia («correctness of diction»). Some examples could well mean «you are right» as if acknowledging or conceding, perhaps by virtue of the symmetry of the participants (e.g. Eur. Ion 745; Or. 100; yOr. 110; Bacch. 838). The adverb la *kirsa neither conforms to the patterns established for jakx &| Athenaeum 109/2 (2021) 393 51 Other examples are Eur. Cyc. 431; Hipp. 1319; IT 1073 and Phoen. 481. 52 See Amati 1901, p. 142. The instances of e \paimx &in this sample are: Eur. HF 1235; Ar. Ran. 508... the examples of ai \mx &are: Eur. Alc. 1093; Hipp. 483; HF 275; IT 1023; Or. 786; IA 506. 53 The only exceptions are Eur. Ion 648 (distich); HF 599, and Ar. Eccl. 1092. See nt. 13 above. 54 The only possible examples are: a)o \qhx &|(e >kena|/ke *cei|): Eur. Ion 1025; Bacch. 838; Ar. Ran. 672; Plut. 1033; b)la *kirsa:E.Med. 944; Or. 1108; Ar. Ecc. 1128. ke *ceim formulae nor is it used to convey approval in conversational contexts 1) and 2). It usually appears in adjacency pairs as an affirmative answer to a polar question (a Yes/No question). In these contexts, la *kirsa is employed to convey an emphatic «yes». Accordingly, P.T. Stevens renders la *kirsa/g %jirsa as «certainly,» «of course» (Stevens 1937, p. 187). It can also convey an ironic assertion which, consequently, results in refusal 55. This ironic dimension is absent in the rest of the formulae herein analysed, which could explain the reason why la *kirsa, in turn, is never used in contexts of dispreferred SPPs (see Section 5). The incompatibility of la *kirsa with the expression of polite refusal is understandable when one acknowledges that la *- kirsa could potentially be misinterpreted, given its hint of irony. Conversely, the distribution pattern of e >rsai does conform to that of the jakx &|ke *ceim formulae (see Table 6). Table 6. Distribution of e >rsai in Euripides and Aristophanes Parameters Results Procedurality (interactive contexts 1 and 2) Procedural (86.95%) 56 Non-procedural (13.04%) 57 Power Positive (52.17%) 58 Negative (26.08%) 59 Non-applicable (21.74%) 60 Distance Positive (21.74%) 61 Negative (69.56%) 62 Non-applicable (8.69%) 63 Gender Male/Deity (82.61%) 64 Female (17.39%) 65 S. Rodrı ´guez Piedrabuena, The Expression of Approval in Euripides and Aristophanes394 55 E.g. E. Or. 235; E. IA 835, 363; Ar. Ecc. 768. E. Ion 1293 is not ironic, as Creusa is explicitly opposing Ion in a hostile manner. 56 Eur. Alc. 328; Alc. 1152; Med. 731; Heracl. 567; Hec. 898; Supp. 348; HF 332; Ion 425; Tro. 87; El. 650; IT 1206; Hel. 744; Hel. 1075; Hel. 1262; Hel. 1294; Phoen. 748; Or. 1598; IA 149; Av. 446; Nub. 431. 57 Eur. Supp. 1182; Ion 413; Or. 1041. 58 Eur. Alc. 328; Heracl. 567; Hec. 898; Supp. 348; Supp. 1182; HF 332; Ion 413; IT 1206; Hel. 1075; Hel. 1262; Or. 1598; Ar. Nub. 431. 59 Eur. Med. 731; El. 650; Hel. 744; Hel. 1294; Or. 1041; IA 149. 60 Eur. Alc. 1152; Ion 425; Tro. 87; Phoen. 748; Ar. Av. 446. 61 Eur. Med. 731; Hec. 898; Supp. 1182 (Theseus-Adrastus); Hel. 1262; Ar. Nub. 431. 62 Eur. Alc. 328; Alc. 1152; Heracl. 567; Supp. 348; HF 332; Ion 413; El. 650; IT 1206; Hel. 744; Hel. 1075; Hel. 1294; Phoen. 748; Or. 1041; Or. 1598; IA 149; Ar. Av. 446. 63 Eur. Ion 425; Tro. 87. 64 Eur. Alc. 328; Alc. 1152; Heracl. 567; Hec. 898; Supp. 348; Supp. 1182; HF 332; Ion 413; Tro. 87; El. 650; IT 1206; Hel. 744; Hel. 1075; Hel. 1262; Phoen. 748; Or. 1598; IA 149; Ar. Av. 446; Nub. 431. 65 Eur. Med. 731; Ion 425; Hel. 1294; Or. 1041. Text type Dialogue (73.91%) 66 Rhesis (26.07%) 67 It seems that the D factor is slightly more frequently negative than in the cases of jakx &|ke *ceim formulae. At the same time, instances in Aristophanes are rare. The expression with e >rsai may have a stronger but still formal tone than the jakx &|ke *- ceim formulae, and require more closeness between participants in the interaction. As for the P factor, exceptions are telling 68: the characters of low status who make use of it are old tutors (Eur. Hel. 744; El. 650; IA 149) 69. Thus, servants resort to other expressions when heeding their masters, see 8) and 9): 8) {Jk.} g #khe| ei \| >Aqco| leh\ g <lx &mja \lo +|g #rh\ a \ei *pose {Pq.} x ’d\ e >vei" jai +roi +le +met >mot| ei \li *,rx &id\g ’rrom po *rei CLYTEMNESTRA: You came to Argos with us and have been mine ever since. OLD SERVANT: That’s it; and I am well-disposed towards you, though not so much so towards your husband. (Eur. IA 870-871) 9) {Jk.}ei \|a >q\ \Iuice *meiam <Eke *mg| mo *rso| g #m pepqxle *mo|; {Pq.}pa *ms\ e >nei|" \Aqse *lidi ht *reim pai &da rg +mle *kkei pasg *q CLYTEMNESTRA: What? Is then Helen’s return fated against Iphigenia? OLD SERVANT: There you have it; the father is up to sacrifice your daughter to Artemis. (Eur. IA 882-883) Servants can also replicate a verb in the previous turn in the imperative but in the first-person singular, see 10) and 11): 10) {\Ac.}# X pqe *rbt, do *lxm sx &mde pa *qoihem rsei &ve. {Pq.}rsei *vx. si *de +jaimotqcei &|, \Aca *lelmom a >man;{\Ac.} rpet &de.{Pq.} rpet *dx. Athenaeum 109/2 (2021) 395 66 Eur. Alc. 1152; HF 332; Ion 413; Ion 425; Tro. 87; El. 650; IT 1206; Hel. 744; Hel. 1075; Hel. 1262; Hel. 1294; Phoen. 748; Or. 1041; Or. 1598; IA 149; Ar. Av. 446; Nub. 431. 67 Eur. Alc. 328; Med. 731; Heracl. 567; Hec. 898; Supp. 348; Supp. 1182. 68 Eur. Ion 422-429, uttered by Creusa, is not an exception. It is clear that she is reacting to Xouthos’ previous words but she might be already out of contact with the king, who enters the palace through the rjgmg *after his command; she is left alone and only the young Ion is present at that moment. Thus, her words can be considered «lines cast at a departing back» (Taplin 1977, pp. 221-222). Only unorthodox women utter the expression when interacting with male characters: Medea, Helen, and Electra in Eur. Med. 731, Hel. 1294 and Or. 1041, respectively. 69 See Venegas Lagu ¨e ´ns 1991, p. 134, when considering the distribution of interruptions in Gaskell per character type: «there is a clash between class and superiority and the superiority of age and experience, which is resolved in favour of the latter». AGAMEMNON: Old man, come here, in front of the house. OLD MAN: I am coming. What is new, sir? AGAMEMNON: Hurry up! OLD MAN: I am. (Eur. IA 1-3) 11) lai &a, pa *kim lot jqt *wom jeuakg *m, ai \dot *leha ca +qsa +kekecle *ma loi jqt *pse... {Sq.}Jqt *psx PHAEDRA: Nurse, cover my head again, for I am ashamed of what I have said, do cover it! NURSE: I am covering your head. (Eur. Hipp. 243-250) Examples like these show a possible function of resonance in Classical Greek 70. It would be an ultimate act of obedience: expressing subordination not only to the master’s will but also to his/her discourse by means of a lexical resonance, far from creativity and one’s own lexical choices. Once the distribution of jakx &|e >kena| and other variants has been established, and once this distribution has been contrasted with other a priori similar expressions (the verb [e \p]aimx &,o \qhx &|[e >kena|/ke *cei|], la *kirsa and e >rsai), the procedural meaning of jakx &|ke *ceim formulae becomes easier to pin down. These formulae can be considered taxemic markers, since they are uttered downwards in non-symmetric interactions (Dyads I and II), that is, they are uttered by characters with a positive P factor and addressed to characters with a negative P factor. Disregarding this, the literal rendering «you spoke beautifully», «you spoke well» is insufficient to reveal the procedural meaning, as jakx &|e >kena| could apparently convey bare agreement or even a compliment. Nevertheless, it is a conventionalised idiom, a polite marker for the expression not just of agreement but of approval 71, since the participant of a higher status is not in need of approval but can grant it. Leech 1983 speaks of the «Approbation Maxim» under the framework of Politeness Theory. However, this does not provide an accurate account of the difference between approval and agreement. This is perhaps because the difference in status between participants does not play such a role in modern Western languages anymore unlike Classical languages. S. Rodrı ´guez Piedrabuena, The Expression of Approval in Euripides and Aristophanes396 70 Resonance, also known as «echoic reply» (Berger 2017, p. 25), means the total or partial repetition of the previous speaker’s words (ten Have 2007, pp. 162-163). See Drummen 2015, p. 142, for the concept in Attic drama regarding the particles. 71 Regardless of whether there is a corresponding idiom in modern Western languages that is conventionalised for the expression of approval in non-symmetric interactions. This is different in the case of i <jeset *x, whose procedural meaning can be easily rendered into English as «please» (Willi 2003, p. 25). Thus, the Approbation Maxim has a broader scope than approval as a different speech act from agreement in line with which it is here advanced 72. In the following example (12), jakx &|ke *ceim formulae are uttered again and again. Moreover, the example would be an apparent exception to the analysis proposed here (since a slave is using a taxemic marker). All this may call for an explanation and is possibly why scholars have found this text difficult: 12) {Oq.}dot &ko| x /m uobz &so +m % Aidgm, o %|r\a \pakka *nei jajx &m; {Uq.}pa &|a \mg *q, ja /m dot &ko| z #si|, g %desai so +ux &|o <qx &m. {Oq.}et #ke *cei|" r{ *fei re rt *meri|. a \kka +bai &m\ e >rx do *lxm. {Uq.}ot \ja >qa jsemei &|le +.{Oq.}a \uei &rai. {Uq.}jako +me >po| ke *cei| so *de. {Oq.}a \kka +lesabotketro *lerha. {Uq.}sot &so d\ ot \jakx &|ke *cei|. {Oq.}lx &qo|, ei \dojei &| le skg &mai rg +m jahaila *nai de *qgm. ORESTES: Despite being a slave, do you fear Death, who will set you free from toils? PHRYGIAN: Every man, even a slave, enjoys seeing the light. ORESTES: Well said, your intelligence saves you, but get out of here! PHRYGIAN: You are not going to kill me, then? ORESTES: You are released. PHRYGIAN: You said this well; ORESTES: I will change my mind. PHRYGIAN: You did not say that well. ORESTES: You fool! If you think I could stand smearing your gullet with blood. (Eur. Or. 1522-1527) On his commentary ad loc., C.W. Willink remarks: Or[estes] is not threatening to change his mind and kill the Phrygian, though that of course is how the Phrygian ‘misunderstands’ him (see vv. 414-416) 73. Or[estes] has finished with that issue and is thinking of something quite different, as his continuation makes clear (Willink 1986, p. 333, followed also by West 1987, p. 284). This interpretation raises problems. Willink does not specify about what else Orestes could possibly be thinking, nor the lines that refer to the ‘something else’ about which Orestes is supposed to be thinking. Moreover, he does not change his mind regarding Menelaus 74. In any case, if Orestes is not addressing the Phrygian, Athenaeum 109/2 (2021) 397 72 See Watts 2003, pp. 67-68. Sacks 1995, pp. 428-435, also disregards the difference between approval and agreement. Risselada 1993, pp. 48-49, considers (the granting of )permission, which is a comparable speech act in the sense that it also involves non-symmetric interactions (it is «a type of directive that can be characterized as a speaker’s positive reaction to an initiative by the addressee [...] who [...] needs the consent of the speaker») – the problem is that this speech act cannot be integrated in her own typology of speech acts (Risselada 1993, pp. 37-48). 73 This reference has no point here: it is a moment of tragic irony concerning Menelaus and Orestes and it is not comparable to this situation. 74 The plan is already plotted (cf. Eur. Or. 1155) and does not change during the interaction with the Phrygian. then we would be obliged to accept a break in ‘contact’ in the middle of an antilabe, which would be truly abrupt and exceptional in Euripidean tragedy 75. V. di Benedetto points out what he considers to be prosaic words 76 and interprets that Orestes does interact with the Phrygian and that he does threaten him in l. 1526 77, but he fails to explain the reason for his abrupt change of mind. F.A. Paley tries to overcome the difficulty by conjecturing that the text is corrupt 78. N. Wedd renders lesabotketro *lerha into «but we will reconsider it», and offers the following interpretation: «Orestes says this the better to ensure the man’s silence by still keeping the possibility of death before his eyes» (Wedd 1895, p. 154). The situation becomes desperate, as W. Biehl considers Eur. Or. 1525 (a \uei &rai. ... a \kka +lesabotketro *le- rha) an «unmotiviert Widerspruch» and compares it with another supposed contradiction between Eur. Or. 251 and 380, which is not really apposite here (Biehl 1965, p. 164). It seems less problematic to admit that Orestes is actually addressing the Phrygian. Bearing in mind the procedural meaning of jakx &|ke *ceim formulae, it seems that Orestes has a sharp reaction because he considers that the Phrygian is performing impolite behaviour: the D factor is positive and the Phrygian has no power in the interaction, it does not befit him to express his approval or disapproval. Still, the ‘barbarian’ persists (jako +me >po| ke *cei| so *de), even in the negative form (sot &so d\ ot \ jakx &|ke *cei|), which is as unexpected in this context as rarely attested elsewhere. It is not a question of whether the expressions are polite or impolite by themselves but that they are the expected (politic) behaviour of a participant depending on the addressee and the context. Watts 2003, p. 201, points out that «failure to abide by members’ expectations of what constitutes politic behaviour frequently leads to the attribution of impoliteness». In this way, Euripides builds up the linguistic characterisation of the Phrygian as a barbarian: he is not linguistically aware of the politic behaviour that is expected on his part as a minion 79. Note that the Phrygian utters varied formulae in an odd way (especially jako +me >po| ke *cei| so *de), as if he were unable to get the conventionalised idiom right. In conclusion, example no. 12 is the exception that proves the rule 80. S. Rodrı ´guez Piedrabuena, The Expression of Approval in Euripides and Aristophanes398 75 See Bonaria 1991, p. 177. For the concept of «contact» in Greek tragedy, see Mastronarde 1979. 76 For example, Eur. Or. 1510: jqatcg *m; 1515: diektlg *maso... 1516: jasx *lora and et \oqjoi &li; 1526: lesabotketro *leha, etc. 77 Di Benedetto 1965, p. 279. As here, threats are usually expressed in the future (e.g. Eur. IA 311- 313; Eur. Andr. 257). 78 Paley 1879, p. 102: «something is suppressed, ‘if you dare to stir’, or ‘if you call for aid!’». 79 See Maas 1962, p. 53, x76, and Kaimio 1988, p. 59, for other features that characterise the Phrygian from the standpoint of the metre and the hikesia, respectively. Pace Dover 1976, p. 361. 80 One can also add another aspect of the Phrygian’s impolite (even if unintended) behaviour: the barbarian is not only persistent in repeating the jakx &|ke *ceim formulae, but also seems to be echoing (and overdoing) the formula used by Orestes (et #ke *cei| // jako +me >po| ke *cei| so *de) as if he was usurping his 5. Discussion: jakx &|ke *ceim Formulae as Dispreferred SPPs As earlier pointed out, together with the traditional idea that the jakx &|ke *ceim formulae were employed as an expression of thanks or acceptance, Quincey 1966 struggled to determine a different origin for those instances in which polite refusal was conveyed. The theoretical framework here implemented allows for one and the same formula to express approval and polite refusal, as long as a rejection or refusal is understood as a Face Threatening Act in need of some kind of facework or mitigation 81. Thus, disagreements or criticism are usually prefaced by an agreement formula, or any criticism with a compliment or positive evaluation (13): 13) Well, in a way, I suppose you’re sort of right. But look at it like this. !(FTA) Why don’t you...? (Watts 2003, p. 89) Similarly, A. Pomerantz describes the seeming puzzle from the standpoint of CA and names the agreement components in actual disagreements as «agreement tokens», providing the following example (14): 14) D: We’ve got sm pretty (good schools). C: Well, yeah but where in the hell em I gonna live. (Pomerantz 1984, p. 72; and 72 no. 42) Accordingly, jakx &|ke *ceim formulae would potentially work as agreement tokens in dispreferred SPPs. The results here presented (see Table 4) show not only that it is possible for a single expression to convey both agreement and polite refusal, but also that the expressions labelled as ‘varied formulae’ 82, which are generally more complicated and less formulaic, present proportionally more examples of dispreferred SPPs 83. Consider the following example (15): 15) jakx &|a %pamsa sat &s\ a /me \nei *qgso *roi, ei \lg +\jt *qei fx &r\ g <sejot &ra" mt &md\e \pei + fz &,pa &r\ a \ma *cjg, jei \jakx &|ke *cei|, o \jmei &m. All these words of yours would have been well said, Athenaeum 109/2 (2021) 399 way of speech (learning manners in promptu from a high-born interlocutor?). Hence the strong reaction of Orestes who might have even felt he was being mocked. 81 Watts 2003, p. 197. The underlying principle for this has been variously addressed. Sacks 1987, p. 65, speaks of «seeking agreement» in conversation, by virtue of which Brown-Levinson 1987, pp. 112, 266, describe as the positive politeness strategy «Avoid disagreement». Likewise, Edmondson 1977 speaks of «gambits», of which the «appealers» seek the agreement of the addressee. Similarly, Leech 1983, pp. 135-136, coined the «Approbation Maxim». 82 Namely, Eur. Med. 576-578 (there is no interaction stricto sensu, since it is a choral distich); Hec. 382; Heracl. 464; Hel. 772-774; Phoen. 1625 and Eur. IA 1407, 1115-1116. 83 See Risselada 1993, p. 59 and 59 nt. 60, for remarks about this with regard to Latin.