¹ I am most grateful to Christopher Pelling and to the anonymous CQ referee for their general guidance and specific suggestions which have greatly improved this article; and to Evert van Emde Boas for introducing me to Text World Theory and for sharing with me his forthcoming article (see n. 12 below). Poems of Sappho and Alcaeus are referred to in Voigt's numeration.

² For example, focussing on Alcaeus, the works of W. Rösler, for instance, Dichter und Gruppe (Munich, 1980); and, with focus on Sappho, the works of C. Calame, for instance, Les choeurs de jeunes filles en Grèce archaïque (Rome, 1977).

³ See Rösler (n. 2) with the reservations of Parker, R., ‘The audience of the lyric poets’, CR 31 (1981), 159–62Google Scholar. From a different perspective, West, M.L., Studies in Greek Elegy and Iambus (Berlin and New York, 1974), 11–12CrossRefGoogle Scholar suggests eight occasions at which Greek elegy may have been performed, but see the comments of Bowie, E., ‘Early Greek elegy, symposium and public festival’, JHS 106 (1986), 13–35CrossRefGoogle Scholar, at 17.

⁴ Rösler, W., ‘Über Deixis und einige Aspekte mündlichen und schriftlichen Stils in antiker Lyrik’, WJA 9 (1983), 7–28Google Scholar. The terms ‘deixis ad oculos’ and ‘deixis am Phantasma’ are from Buehle, K., Sprachtheorie (Jena, 1934)Google Scholar.

⁵ This is because early lyric, tied as it is to specific occasions, exhibits ‘ein sehr enger, direkter Bezug des Textes auf das Hier und Jetzt’ (Rösler [n. 4], 10). The lack of such specific deixis reference-points in a later poet such as Theognis is, for Rösler, a sign of the move away from ‘oral’ occasional poetry towards generic poems capable of working on a number of occasions. For Rösler (n. 4), specific references, for example to named people, many of which he sees as incomprehensible outside the group, were evidence that the songs are composed not just for a specific group (the hetairia or Mädchengruppe in the case of Sappho) but for a specific occasion.

⁶ See Latacz, J., ‘Realität und Imagination, eine neue Lyrik-Theorie und Sappho's φαίνεταί μοι κῆνος-Lied’, MH 42 (1985), 67–94Google Scholar, responding to Rösler (n. 4).

⁷ D'Alessio, G., ‘Language and pragmatics’, in Budelmann, F. (ed.), The Cambridge Companion to Greek Lyric (Cambridge, 2009), 114–29CrossRefGoogle Scholar, at 117–18, arguing that modern readers ‘have too often seen the texts as a simple record of the performance situation’. On deixis, see also n. 12 below.

⁸ D'Alessio (n. 7), 120, with examples from other oral poetry. Cf. also L. Edmunds, ‘Deixis and everyday expressions in Alcaeus frs. 129 V and 130b V’, in V. Bers (ed.), Donum natalicium digitaliter confectum Gregorio Nagy, available at https://chs.harvard.edu/CHS/article/display/4353. Examining deictic expressions in Alcaeus, Edmunds accepts that there is ‘deixis ad oculos’ (the singer-poet's deictics do point to real objects in original performance) but notes that deixis may occur to objects outside original performance circumstances (which he calls ‘fictive’ deixis), and that we need to consider the effect that reperformance will have on an original ‘deixis ad oculos’.

⁹ Budelmann, F. and Phillips, P. (edd.), Textual Events. Performance and the Lyric in Early Greece (Oxford, 2018)Google Scholar.

¹⁰ See Budelmann and Phillips in their introduction to the volume (n. 9), 7: ‘Poems map more or less closely onto the occasion of their performance, they veer towards the fictional or the real.’

¹¹ Thus A. Uhlig, ‘Sailing and singing: Alcaeus at sea’, in F. Budelmann and P. Phillips (edd.), Textual Events. Performance and the Lyric in Early Greece (Oxford, 2018), 63–92 talks of a distinction between, and interplay of, ‘the fictional “mimetic” performance scenario created within the frame of a lyric poem’ and the ‘“real-life” circumstances in which such a song historically found voice’.

¹² Stockwell, P., Cognitive Poetics. An Introduction (New York, 2002), ch. 4Google Scholar has a good summary of deictic analysis. For the application of deixis theory to poetry, see Green, K., ‘Deixis and the poetic persona’, Language and Literature 1 (1992), 121–34CrossRefGoogle Scholar. On deixis and archaic lyric, see the essays in Felson, N. (ed.), The Poetics of Deixis in Alcman, Pindar, and Other Lyric, Arethusa Special Issue 37 (2004)Google Scholar with a full bibliography, especially G. D'Alessio, ‘Past future and present past: temporal deixis in Greek archaic lyric’ (at 267–94); also Danielewicz, J., ‘“Deixis” in Greek choral lyric’, QUCC 34 (1990), 7–17Google Scholar; and now E. van Emde Boas, ‘Deixis and world building’, in L. Swift (ed.), A Companion to Greek Lyric (forthcoming).

¹³ Deixis theory recognizes other coordinates. Stockwell (n. 12), 45–6 lists in addition: perceptual (which includes what I have called ‘personal’), spatial, temporal, relational, textual and compositional.

¹⁴ For ‘world’-based approaches to discourse, see e.g. the essays collected in Gavins, J. and Lahey, E. (edd.), World Building: Discourse in the Mind (London, 2016)Google Scholar.

¹⁵ Gavins, J., Text World Theory: An Introduction (Edinburgh, 2007)CrossRefGoogle Scholar has a helpful account, on which the summary here is based. The methodology originates in the work of Paul Werth: e.g. ‘How to build a world’, in K. Green (ed.), New Essays in Deixis: Discourse, Narrative, Literature (Amsterdam, 1995), 49–80. On Text World Theory and Greek lyric, see now van Emde Boas (n. 12).

¹⁶ Deictic shift theory: Stockwell (n. 12), 46–9; and world shift: Gavins (n. 15), 45–50.

¹⁷ Gavins (n. 15), 27.

¹⁸ As CQ's referee suggests to me, study of the effect of music, metre and body movement would expand our appreciation of world-creation in archaic sung lyric.

¹⁹ Cf. D'Alessio (n. 7), 114.

²⁰ Cf. Green (n. 12), 130: ‘The deictic elements and terms are constantly helping us to sort from possible contexts, helping us to move from symbolic meaning to some kind of reconstructed indexical meaning.’

²¹ Stockwell (n. 12), 46.

²² As N. Felson puts it (‘Introduction’, in ead. [ed.], The Poetics of Deixis in Alcman, Pindar, and Other Lyric, Arethusa Special Issue 37 [2004], 253–66, at 260), a text challenges audiences ‘to establish, by inference, the pragmatic/contextual anchoring of the discourse in order to apprehend what is not self-evident … The resultant participation in the process of making meaning intensifies their response to what they hear, making them work harder and therefore become all the more engaged’.

²³ Culler, J., Theory of the Lyric (Cambridge, MA, 2015)CrossRefGoogle Scholar: epideictic (115), a liturgy (120), ritual (123). For Culler, ‘lyrics are written for readers to repeat’ (120). This is even more the case for performed lyric, where the singer in some sense becomes the lyric ego in performance.

²⁴ ‘Lyric can be tentatively (transhistorically) defined as a first-person utterance whose performative conditions are reconstructed by a reperforming reader, who typically positions himself somewhere in a continuum whose extremes are a generic voice and some individual idea of the author’ (A. Barchiesi, ‘Carmina: Odes and Carmen Saeculare’, in S. Harrison [ed.], The Cambridge Companion to Horace [Cambridge, 2007], 144–61, at 150).

²⁵ Cf. Budelmann, F. and Phillips, P., ‘Introduction’, in id. (edd.), Textual Events. Performance and the Lyric in Early Greece (Oxford, 2018), 1–28Google Scholar, at 22, noting ‘lyric's signature combination of “performative immediacy” and generalising claims’. Cf. Whitmarsh, T., ‘Sappho and cyborg Helen’, in Budelmann, F. and Phillips, P. (edd.), Textual Events. Performance and the Lyric in Early Greece (Oxford, 2018), 133–49, at 144–9Google Scholar on the relationship between the intense ‘here and now’ of the poem and its generic significance.

²⁶ D'Alessio (n. 12), 268: ‘Archaic Greek lyric is different from most modern lyric poetry in that it is more strongly related to an actual performance context.’ Latacz (n. 6), 71 notes that the ‘Situationsbezogenheit’ of ancient lyric drives a more concrete or pragmatic use of deixis.

²⁷ Cf. Rutherford, I., Pindar's Paeans (Oxford, 2001), 177Google Scholar.

²⁸ See Barchiesi (n. 24), 150–3 on the lyric strategy in Horace, and Culler (n. 23), 270–5 for an analysis of the lyric Locations in Horace, seeing them as ‘figural devices’.

²⁹ Hdt. 5.113.2, though Plutarch, who quotes these lines (Sol. 26.2–4), does not say he was present at Soloi.

³⁰ See e.g. Solon, fr. 1 W. (‘I have come as a herald from lovely Salamis’) with Bowie (n. 3), 19.

³¹ Cf. orations written in imitation of a speech actually given (Plato's Apology).

³² D'Alessio, G., ‘Fiction and pragmatics in ancient Greek lyric: the case of Sappho’, in Budelmann, F. and Phillips, P. (edd.), Textual Events. Performance and the Lyric in Early Greece (Oxford, 2018), 31–62Google Scholar, at 57 argues that this is a historical question, to which in most cases ‘it is intrinsically impossible to find an answer based on the texts alone’.

³³ See D'Alessio (n. 7), 116–20 on such references in archaic lyric and the extent to which they may correspond to circumstances of performance.

³⁴ Cf. Gavins (n. 15), 40–1; and also 85–6 on devices in fiction which invite the reader to project themselves into the deictic centre of the text world. Such devices ‘transcend the ontological boundaries of the text–world in order to enter the reader's half of the split discourse world’. In performed lyric, the identification of the ‘I’ as singer creates the illusion that there is no split discourse world at all.

³⁵ On the ‘encounters’ of an audience with a lyric text on each performance, see Budelmann, F., ‘Lyric minds’, in Budelmann, F. and Phillips, P. (edd.), Textual Events. Performance and the Lyric in Early Greece (Oxford, 2018), 235–56Google Scholar. He explores (243) the ways in which Sappho and Alcaeus construct the speaking voice as ‘Sappho’ or ‘Alcaeus’. Where this occurs, a singer who is not the poet herself, to some extent ‘steps into the shoes’ of the respective lyric personality (253). On the ‘I’ in archaic lyric in general, see Slings, S. (ed.), The Poet's ‘I’ in Archaic Greek Lyric (Amsterdam, 1990)Google Scholar.

³⁶ Addressees in Sappho and Alcaeus encourage the audience to identify themselves as the addressee, or a member of the same group as the addressee, breaking down the split discourse, and are a poetic device no less than the addressees in Horace. On addressees in Western lyric, see Culler (n. 23), ch. 5 and 201–5.

³⁷ The term means, broadly, verbal face-to-face communication: Lyons, J., Semantics (Cambridge, 1977), 633Google Scholar. Gavins (n. 15), 103 discusses the device of the recreation of the face-to-face discourse world at the text world level.

³⁸ https://shanty.rendance.org/lyrics/showlyric.php/strike. A British sailors’ song from about 1870.

³⁹ Cf. Hutchinson, G., ‘What is a setting?’, in Budelmann, F. and Phillips, P. (edd.), Textual Events. Performance and the Lyric in Early Greece (Oxford, 2018), 115–34, at 127Google Scholar. Hutchinson suggests that we can establish patterns or ‘deep structures’ in the settings of archaic lyric.

⁴⁰ Some named figures referred to in the songs certainly are real, for example Pittacus or Myrsilus, though these are not participants in the lyric Location.

⁴¹ Inscriptional evidence suggests that Archilochus’ addressee Glaucus did in fact exist (see D'Alessio [n. 7], 115).

⁴² Cf. D'Alessio (n. 32), 61.

⁴³ There is still a distinction between an object ‘conjured up’ by deixis to an addressee of the poem (like the Mytilenean assembly in Alcaeus, fr. 130, line 8) and an object presented as being at the imagined lyric Location (which we could call ‘deixis velut ad oculos’, as Chris Pelling suggests to me.)

⁴⁴ Green (n. 12), 127–8.

⁴⁵ Cf. Kurke, L.V., ‘Archaic Greek poetry’, in Shapiro, H.A. (ed.), The Cambridge Companion to Archaic Greece (Cambridge, 2007), 141–68CrossRefGoogle Scholar, at 143–4: archaic lyric ‘accomplished some real social work in performance … by affirming—indeed, constructing anew on each occasion—the values and roles felt to be proper for the group’.

⁴⁶ Sappho's ‘factional’ songs: frr. 55, 71, 131.

⁴⁷ Alcaeus, frr. 6, 208a, 249.

⁴⁸ On the ship as a metaphor or proxy for the symposium space, see W. Slater, ‘Symposium at sea’, HSPh 80 (1976), 161–70; Bowie (n. 3), 17. For the symposium compared to rowing a ship, cf. Dionysius Chalcus, frr. 4–5 W.

⁴⁹ Alcaeus, fr. 6.7.

⁵⁰ See Uhlig (n. 11), 89–90, arguing that we should take the imaginative world of these sea songs seriously, creating a world that was ‘emphatically unlike’ that of the symposium, and ‘transporting’ the audience to the high seas.

⁵¹ Gavins (n. 15), 46–8.

⁵² For a summary of the physical locations suggested for the song, see Spelman, H., ‘Alcaeus 140’, CPh 110 (2015), 353–60Google Scholar and Budelmann, F., Greek Lyric: A Selection (Cambridge, 2018), 109CrossRefGoogle Scholar.

⁵³ Clay, D., ‘Lesbian armour: Alcaeus fr. 140 Voigt’, Prometheus 39 (2013), 18–24Google Scholar.

⁵⁴ Cf. Fearn, D., ‘Materialities of political commitment’, in Budelmann, F. and Phillips, P. (edd.), Textual Events. Performance and the Lyric in Early Greece (Oxford, 2018), 93–115, at 102–6Google Scholar, on how Alcaeus creates the ‘hall of arms’ as an idealized imaginative space, with the potential to have effect in various potential performance scenarios.

⁵⁵ Culler (n. 23), 131.

⁵⁶ See Gavins (n. 15), 3–4, 22–3 and 38: scripts are ‘knowledge stores containing information about familiar types of events and situations’.

⁵⁷ See Budelmann (n. 52), 94 with bibliography.

⁵⁸ Hutchinson, G., Greek Lyric Poetry. A Commentary on Selected Larger Pieces (Oxford, 2001), 195Google Scholar.

⁵⁹ See Hutchinson (n. 58), 196: ‘no ordinary pattern of prayer’.

⁶⁰ On the significance of the place (implying not just a Mytilenean but a Lesbian or even Aeolian national unity), the aggrandizing adjectives and the rewriting of myth, see Hutchinson (n. 58), 196–7.

⁶¹ Cf. Sappho, frr. 1.5, 2.1; Alcaeus, fr. 34A.8–9.

⁶² Hutchinson (n. 58), 198 suggests deixis to a cult statue.

⁶³ Rösler (n. 2), 195–6. Budelmann (n. 52), 94 suggests that the song could have been performed at a sympotic space at the precinct, but notes also (95) that the deictics ‘situate the speaker right before’ Hera's altar. D'Alessio (n. 32), 44 suggests that we are to imagine the poet present at a festival of the gods, noting the deictic ‘these citizens’ in line 6.

⁶⁴ Cf. Fearn (n. 54), 105, seeing the deictic ‘this task’ in Alcaeus’ fr. 140 as ‘open-ended’, reaching out to the reperformance of the text.

⁶⁵ On a rough count, of forty-seven songs for which the Location can be reconstructed, only three (frr. 16, 44, 44A) are not prayers or addresses to named individuals (on fr. 31 see below). Culler (n. 23), 198 calculates that 87 per cent of the songs of Horace (who imitates the Locations of Greek originals) are addressed to another person, as are 70 per cent of those of Catullus.

⁶⁶ On the Brothers Poem, see the essays in A. Bierl and A. Lardinois (edd.), The Newest Sappho (Leiden, 2016). On the importance of ‘enactment’ in lyric, see Culler (n. 23), e.g. 201–11, including the introduction of addressees to create ‘an event in the lyric present’.

⁶⁷ Page's energetic rejection of this theory (Page, D., Sappho and Alcaeus [Oxford, 1995], 32Google Scholar) did not signal its death-knell: see Nagy, G., ‘Lyric and Greek myth’, in Woodward, R.D. (ed.), The Cambridge Companion to Greek Mythology (Cambridge, 2007), 19–51CrossRefGoogle Scholar, at 28 (‘the celebrated wedding song of Sappho’).

⁶⁸ Burnett, A.P., Three Archaic Poets: Archilochus, Alcaeus, Sappho (Cambridge, Mass., 1983), 241–2Google Scholar.

⁶⁹ W. Rösler, ‘Realitätsbezug und Imagination in Sapphos Gedicht ΦΑΙΝΕΤΑΙ ΜΟΙ ΚΗΝΟΣ’, in W. Kullman and M. Reichel (edd.), Der Übergang von der Mündlichkeit zur Literatur bei den Griechen (Tübingen, 1990), 271–87.

⁷⁰ Stehle, E., Performance and Gender in Ancient Greece (Princeton, 1997), 292–3Google Scholar.

⁷¹ D'Alessio (n. 32), 59—on the difficulty of constructing circumstances where both the man can speak to the girl and the singer can address the girl.

⁷² Rösler (n. 69), 283–5, also pointing to the difficulty of an actual address in lyric Location given the ‘gnomic’ ending.

⁷³ Hutchinson (n. 58), 168–70. Budelmann (n. 52), 133 agrees. Contrast Winkler, J., Constraints of Desire (London and New York, 1990), 179Google Scholar, objecting to seeing the song as ‘a modern lyric of totally internal speech’ rather than an utterance that ‘imitates other well-known occasions for public speaking’.

⁷⁴ (n. 12). Noting the deictic ambiguities between ‘now’ and ‘whenever’, he sees this as a kind of ‘blended’ text world, and concludes that the ambiguity is deliberate: it ‘frames the speaker's concrete individual experience as part of a wider phenomenon’. Cf. also Culler (n. 23), 63 on Sappho, fr. 31: ‘cast in the present tense, an account of what happens repeatedly, it none the less impresses us as something happening now, in the performative temporality of the lyric’.

⁷⁵ Cf. Budelmann (n. 52), 133.

⁷⁶ Cf. Sappho, fr. 111—the bridegroom is ‘equal to Ares’; and fr. 105(b)—Sappho ‘likened the bridegroom to Achilles’. Latacz (n. 6), 77–9 compares Od. 6.158–9 (Odysseus addressing Nausicaa): ‘That man (κεῖνος) is in his heart the most blessed exceeding all others | Who would (ὅς κε) prevail with bridal presents and take you home.’ Both Latacz (n. 6) and Winkler (n. 73), 178–9 see the man in Sappho's fr. 31 as purely imaginary, a ‘figure of speech’ like the one in Odyssey Book 6, but this does not do justice to the urgent, closely imagined, present tenses, suggesting a real scene.

⁷⁷ On the makarismos here, and on Sappho's fr. 31 as a ‘praise song’, see Lardinois, A., ‘Who sang Sappho's songs?’, in Greene, E. (ed.), Reading Sappho (Berkeley, 1999), 150–72Google Scholar, at 167–9.

⁷⁸ The verb describes a sudden and intense psychological loss of control akin to panic and caused either by fear or by erōs (see Latacz [n. 6], 89). It does not necessarily suggest jealous panic but rather the intense erotic reaction caused by a sensory stimulus. The parallel from fr. 22 may suggest that the subject of ἐπτόασεν here is the laugh, rather than the whole scene.

⁷⁹ Cf. D'Alessio (n. 32), 59–60.

⁸⁰ ‘an unusual mix of a specific individual and a generic class of men’ (Van Emde Boas [n. 12]).

⁸¹ On the characteristics of Sapphic erōs, see Boehringer, S. and Calame, C., ‘Sappho and Kypris: the vertigo of love’, in Bierl, A. and Lardinois, A. (edd.), The Newest Sappho (Leiden, 2016), 353–67Google Scholar.