¹ Cf. esp. Hefti, V., Zur Erzählungstechnik in Heliodors Aithiopika (Vienna 1950) 1 ff.Google Scholar, 98 ff.; Keyes, C. W., Studies in Philology xix (1922) 42–51;Google ScholarReardon, B. P., Courants littéraires grecs des II^e et III^e sihles apres J.-C. (Paris 1971) 381 ffGoogle Scholar.

² This antithesis between ‘plot’ (or ‘narration’) and ‘story’ is as close as one can get in English to the distinction made by the Russian Formalists between sjužet (the sequence of events as presented in the narrative) and fabula (the sequence of events as they ‘really’ happened), and taken up by French structuralists in the terms récit and histoire.

³ These rather nasty terms ‘proairetic’ and ‘hermeneutic’ are taken from the ‘codes’ of reading analysed by Barthes, R. in S/Z (Paris 1970;Google Scholar English trans, by R. Miller [London 1975]). The ‘proairetic’ code directs the way the reader follows and integrates the plot, step by step according to the logic of the action; the ‘hermeneutic’ the solution of enigmas. In redeploying the terms to denote types of narrative I am following the precedent of Brooks, P., Reading for the plot (Oxford 1984) 18 ffGoogle Scholar.

⁴ Some MSS of the Aithiopika (BPZ) include the names of heroine and hero in the title given at the beginning of the first book; and the Byzantines in general seem to have referred to the novel simply as Χαρίκλεια (cf. the citations given as Testimonia nos. x, xii, xv, xvii, xix, xx in the edition of the novel by A. Colonna [Rome 1938]). A reader of such a copy would find little difficulty in locating the main characters. But other MSS (CVM) are headed simply Ἡλιοδώρου Αἰθιοπικῶν βιβλίον πρῶτον (vel sim.), and our earliest reference to the work (Sokr. Hist. ekkl. v 22 [ = Colonna Test, i]) seems to confirm that this was the novel's original title. Anyone reading a MS of this kind would have to wait until the course of the narrative itself cast Theagenes and Charikleia as hero and heroine. It is tempting to read this as a deliberate manoeuvre to prolong the reader's uncertainty and sense of disorientation, an effect typical of the author, but one spoiled by the alternative title.

⁵ Well treated by Sandy, G. N., Heliodorus (Boston 1982) 33 ffGoogle Scholar.

⁶ ii 2.1, ii 3.4; cf. Char, i 5.2, i 6.1, v 10.10, vi 2.8 ff.

⁷ And so, of course, of being a transparent documentary record of ‘real’ events, rather than a self-referential text that constantly alludes to its own fictional status. The method of presentation is part of the realism of the Aithiopika which I analysed in my article ‘History, romance and realism in the Aithiopika of Heliodoros’, ClAnt i (1982) 221–65;Google Scholar see esp. 260 ff.

⁸ Although a certain ‘realistic residue’ of non essential material is only to be expected in the cause of the mimesis just referred to; cf. Morgan (n. 7) 250 ff.

⁹ Cf. Brooks (n. 3) 37 ff.

¹⁰ There is no surviving example of an ancient novel that does not end happily, but compare the remarks of B. E. Perry on the lost Kypriaka of Xenophon of Cyprus (The ancient romances [Berkeley 1967] 120 f.Google Scholar).

¹¹ i 1.7: οὐδὲ συνιἑναι τὴν σκηνήν ἐδύναντο; i 2.6: ἡ τῶν γινομἑνων ἄγνοια … τὰ ὄντα δὲ οὔπω ἐγἱνωσκον.

¹² i 3.5: σἱ δὲ λῃσταί [the new group] … ὑπὸ τῆς τῶν ὁρωμένων ἀγνοίας ἃμα καὶ ἐκπλήξεως τέως ἀνετἑλλοντο; i 3.6: τοὺς μὲν γὰρ πολλοὺς φόνους ὑπὸ τῶν προτέρων γεγνῆσθαι λῃστῶν εἴκαʒον.

¹³ I would not wish to argue that it was Heliodoros' intention to highlight the conventional aspects of fiction and hence the fictionality of his own discourse. It is more that, as a virtuoso, he plays the game at the limit of its rules, thus running the risk of incidentally making conscious conventional rules of which every reader was already subconsciously aware anyway.

¹⁴ ‘The mendacity of Kalasiris and the narrative strategy of Heliodoros' Aithiopika’ YCS xxvii (1982) 93–158Google Scholar.

¹⁵ Winkler (n. 14) 107. I hope I am not distorting his sense by taking ‘story’ here to mean sjužet rather than fabula; cf. n. 2.

¹⁶ i 9.1: Ἠν μοι πατὴρ Ἀρίστιππος, τὸ γένος Ἀθηναῖος κτλ.

¹⁷ Heliodoros generally makes no attempt to characterise through style. Even the eunuch Bagoas, whose Greek is, we are specifically informed (viii 15.3), fractured, is presented speaking in fluent periods, with even a Euripidean allusion.

¹⁸ Winkler (n. 14) 107.

¹⁹ A perfectly valid judgment, of course, but patently not the author's own.

²⁰ Winkler (n. 14) 96 f.

²¹ Winkler (n. 14) 142.

²² Winkler (n. 14) 147.

²³ Winkler (n. 14) 108 f.

²⁴ Sandy (n. 5) 31.

²⁵ A state of affairs recalled as recently as v 4.3: ἐπειδὴ γὰρ ὁ μὲν Θύαμις ἁλοὺς ἐʒώγρητο καὶ εἴχετο αἰχμάλωτος …

²⁶ vi 3.4: ὁ τούτων ἔναγχος ἀποδειχθεὶς ἔξαρχος Θύαμις.

²⁷ i 8.4, ii 5.2, v 2.10, viii 6.4, x 20.2.

²⁸ Demainete does twice refer to Knemon as her ψυχή, but the first instance (i 9.4) is in a passage of obvious insincerity and is no more than a stratagem to wheedle him into her bed; the second (i 14.6) occurs after she has destroyed him, and is lamenting her loss, ostensibly as a mother but in reality as a lover: the irony is clear.

²⁹ i 2.4, i 4.1, ii 1.2 f., ii 4.4, v 33.1, vi 8.6, vi 9.3, viii 8.4, viii 9.8, viii 13.3, x 19.2.

³⁰ i 15.8: εἰ δὲ τύχοις ὦν βούλει, μάλιστα μὲν εἰκὸς σχολάσαι τὸν, ἔρωτα, πολλαῖς γὰρ κατά τὴν πρώτην πεῖραν ἐναπεσβέσθη τὰ τῆς ἐπιθυμίας· κόρος γὰρ ἔρωτος τῶν ἔργων τὸ τέλος.

³¹ i 15.7: τὴν Ἀρσινόην, οὗσαν μοι πάλαι γνωρίμην ἀπὸ τῆς τέχνης …

³² E.g. v 5.2: οὐδένα γὰρ χρόνον εἷναι ὅσος ἀμαυρῶσαι αὐτοῖς τῶν ψυχῶν τὰ ἐρωτικὰ γνωρίσματα; v 2.7: βιώσεσθαι … τὸ λεπόμενον ἅμα τῷ φιλτάτῳ.

³³ i 8.3, i 26.1, ii 4.2, iv 13.4, v 29.4, vii 21.5, vii 25.5, vii 26.3, x 33.2. At iv 11.3, after Charikles has expressed a wish that Charikleia should marry her cousin Alkamenes, she exclaims: Ἀλκαμένει μὲν … τάφον πρότερον ἢ γάμον τὸν ἐμὸν εὐτρεπιʒέτω, ἐμὲ γὰρ ἢ Θεαγένης ἄξεται ἢ τὸ τῆς εἱμαρμένης διαδέξεται. This has been misunderstood by translators: she is not wishing Alkamenes dead but praying for her own death, as the second clause makes clear; τὸν ἐμόν agrees with τάφον as well as γάμον (for the word order cf. iv 14.1: σὺν τέχνῃ πολλῇ καὶ σοφίᾳ τῇ ἐμῇ, and iv 18.5: γένος τε καὶ οῖκον τὸν ἡμέτερον).

³⁴ Possibly, on the basis of the parallels between Demainete and Arsake, we should read this punishment as an attempt to coerce him into complying with her desires; if so, it is implied rather than stated; cf, however, i 15.5: ὣσπερ οὐκ ἐρῶσά τινος ἀλλ' ἄρχουσα δεινον ὃτι μὴ ἐξ ἐπιτάγματος ὐπήκουσεν ἐποιησάμην.

³⁵ This, I take it, is the point of the diminutive, γύναιον, with which she is characterised on her first appearance (i 9.1); cf. i 9.2: τῇ τε ὣρᾳ τὁν πρεσβύτην ἐπαγομένη.

³⁶ Charikleia is seventeen (x 14.4); Theagenes' age is not specified, but he is introduced as an ἔφηβος (i 2.2).

³⁷ So, for example, Thisbe says to her (i 17.1): κόσμει ασυτἡν · ἁβρότερον ἔχουσαν ἣκειν προσήκει.

³⁸ i 25.4: εἰς δεῦρο διετέλεσα καθαρὰν ἐμαυτὴν καὶ ἀπὸ σῆς ὁμιλίας φυλάττουσα, πολλάκις μὲν ἐπιχειροῦντα διωσαμένη, τὸν δὲ ἐξ ἀρχῆς ἡμῖν συγκείμενόν τε καὶ ἐνώμοτον ἐπὶ πᾶσι γάμον ἔνθεσμον εἴ πῃ γένοιτο περισκοποῦσα.

³⁹ iv 18.5: καὶ ἔτι μᾶλλον τῶν μελλόντων ὅρκῳ πρὸς θεαγένην τὸ ἀσφαλές ἐμπεδωθείη ὡς οὔτε ὁμιλήσει τὰ Ἀφροδίτης πρότερον ἢ γένος τε καὶ οῖκον τὸν ἡμέτερον ἀπολαβεῖν ἤ, εἴπερ τοῦτο κωλύει δαίμων, ἀλλ' οὖν γε πάντως βουλομένην γυναῖκα ποιεῖσθαι ἢ μηδαμῶς. cf. also iv 10.6, vi 9.4; even in her dreams he must respect Trappier chastity (vi 8.6).

⁴⁰ i 3.1, i 8.3, ii 33.4 f., iv 18.4 ff., v 4.5, vi 8.6, viii 13.2, etc.

⁴¹ Sokr, . Hist. ekkl. v 22 (Colonna [n. 4]Google Scholar Test, i); cf. Rattenbury, R. M., Proceedings of the Leeds Philosophical and Literary Society, Literary and Historical Section i (1927) 168 ff.Google Scholar, Lacombrade, C., REG lxxxiii (1970) 70 ffCrossRefGoogle Scholar.

⁴² ii 33.4–5: ἀπηγόρευται … αὐτῇ γάμος καὶ παρθενεύειν τὸν πάντα βίον διατείνεται … ἐκθειάʒουσα παρθενίαν καὶ ἐγγὺς ἀθανάτων ἀποφαίνουσα.

⁴³ x 9.3 ff., a passage verbally and thematically connected to the scene at Arsake's stake (viii 9.13 f.), when after a judicial process expected death makes way for a salvation saturated with light.

⁴⁴ The prominence of the sun in the novel is part of this image-system, not (as argued by Altheim, F., Literatur und Gesellschaft im ausgehenden Altertum i [Halle 1948] 93–124)Google Scholar a declaration of faith by a devotee of the Emesan sun-cult.

⁴⁵ Cf. Achaimenes' reaction to the arrival of Theagenes in the palace (vii 16.1): σύνηθές τι καὶ ἀφροδίσιον διακόνημα τῇ Ἀρσάκῃ τὸν θεαγένην ὑποτοπήσας.

⁴⁶ vii 2.1, vii 9.2.

⁴⁷ i 10.2. A reference to Theseus follows in the MSS, but is impenetrable as the text stands; it should perhaps be excised, as argued by Neimke, P., Quaestiones Heliodoreae (Halle 1889) 15Google Scholar n. 1, anticipating Rattenbury's note in the apparatus of his Budé text. For discussion of the novella as variant of the Phaidra story, see Donnini, M., MCSN iii (1981) 145–60Google Scholar.

⁴⁸ viii 15.2: τέθνηκεν Ἀρσάκη βρόχον ἀγχόνης ἀψαμένη, echoing E. Hipp. 802: βρόχον κρεμαστὸν ἀγχόνης ἀνήψατο.

⁴⁹ This use of the double plot is outlined by Empson, W., Some versions of pastoral (London 1950) 53Google Scholar.