1 Most recently by Burl (2010) and Dunn (2016). For more scholarly assessment of what can (and what cannot) legitimately be inferred about the poet’s life and background, see Wiseman 1979b; Skinner 2003: xix–xxii; Gaisser 2009: 2–6.

2 E.g. Plin. HN 36.48: Mamurra Catulli Veroniensis carminibus proscissus; Macrob. Sat. 2.1.8: Veronensis poeta. See also Ov. Am. 3.15.7: Mantua Virgilio gaudet, Verona Catullo; Plin. NH praef. 1: conterraneum meum; Martial 14.195.1: tantum magna suo debet Verona Catullo.

3 Jerome, Chronica, p. 154 Helm² (58 bc).

4 On the Valerii Catulli and the Transpadane background, see Wiseman 1985: 107–11, 1987b, 2007, 2023: 87–102. Wiseman conjectures, on the basis of an amphora found in Rome in 1878 and bearing the name C. Valerius Catullus, that the family made its money through commerce.

5 On the Bithynia poems, see further pp. 15–16 below. Poems 4 and 25 both refer to the province (the phaselus traces its origins to Mount Cytorus in Pontus [4.11–16] and ‘Bithynian ?writing-tablets’ – the meaning of the word catagraphos is disputed – are among the items stolen by Thallus [25.7]); but in neither case is it a necessary inference that the items in question were personally obtained by the poet during his stay in the area.

6 Skutsch 1969 advances the theory that the polymetrics may be dated in relative terms by means of the poet’s handling of the first two syllables (or ‘base’) of the hendecasyllabic line: hendecasyllabic poems between 2 and 26 almost exclusively employ a spondaic base, in accordance with the more usual Roman practice, whereas in roughly 20% of the hendecasyllabic lines in the second part of the collection, poems 27–60, the base is either iambic or trochaic. Skutsch argued that this pattern is evidence for a development in Catullus’ metrical practice, and that poems 27–60, along with the dedication poem, 1 (which contains four iambic or trochaic bases), are, broadly, later than 2–26 (cf. Skinner 1981: 21–3). But this metrical criterion cuts across the chronology suggested by the datable references discussed above; and it is certainly possible to think of other ways of interpreting the configuration discovered by Skutsch (so Hubbard 1983: 234–5; Beck 1996: 235–41).

7 For a good brief discussion of the evidence, see Wiseman 1985: 189–91.

8 This is particularly clear in the case of the Lesbia poems, discussed in Chapter 2 below.

9 See especially Gutzwiller 1998 on Hellenistic epigram; Kerkhecker 1999: 282–90 and Acosta-Hughes 2002: 3–9 on Callimachus’ Iambi; and Gutzwiller 2005a on the new Posidippus papyrus.

10 For the manuscript tradition and the history of the text, see Chapter 6, pp. 126–8, below.

11 The figure of 113 is necessarily approximate: on the one hand, the traditional numeration includes the three Priapea, 18–20, which do not appear in the manuscripts but were inserted by the Renaissance editor Muret in his edition of 1554, and which are universally excluded in modern editions (see Zarker 1962, who argues for Catullan authorship); on the other, a number of what are on the traditional numeration single poems are now regularly divided into two. In some instances (e.g. 58 and 58b, which clearly do not belong together), this latter issue is clear cut; but many cases (esp. poems 2 and 68) remain controversial. The manuscript tradition, which is erratic and intermittent in indicating poem divisions, gives us little assistance here.

12 Dettmer 1988, 1997. Cf. (e.g.) Wiseman 1969: 1–31; Quinn 1972: 9–20; E. A. Schmidt 1973; Skinner 1981; Claes 2002; Holzberg 2002: esp. 72–87; Schafer 2020. A slightly different approach, adopted notably by Skinner 2003 and combined by Holzberg with a more traditional ‘architectural’ schema, is to emphasize the process of sequential reading arguably necessitated by the book-roll format, which – it is claimed – imposes a unidirectional reading of the collection, or of its constituent parts: thus, Holzberg (2002: 67) tropes his reading as a ‘voyage of discovery’ (Entdeckungsfahrt), while Skinner (2003: xxiv–xxvi) sees the ‘serial progression of events’ from poem to poem as a crucial factor in the construction of meaning. Claes’s principle of ‘concatenation’ – lexical and/or thematic repetition from poem to poem (Claes 2002: esp. 27–56) – is somewhat similar.

13 For a useful history of the debate, see Skinner 2007a (esp. 36–8 for nineteenth-century theories).

14 See e.g. Wheeler 1934: 21–32; Clausen 1976; Hubbard 1983, 2005; Stroh 1990; Beck 1996; Hutchinson 2008: 109–30, 2012; Du Quesnay 2021; Wiseman 2023: 13–47. Proponents of this view have tended to emphasize in particular the apparent lack of meaningful arrangement in the closing run of polymetrics, 52–60, described by Skinner (1981: 74, 76) as a ‘jumble of unrelated and curiously unfinished verses’, probably constituting a ‘posthumous supplement to the Passer’, and by Hubbard (1983: 220), more colourfully, as a ‘rather embarrassing grab-bag of doggerel and fragments’. Cf. Schafer 2020: 46–62.

15 References to Catullus by other ancient authors are conveniently collected and classified by Wiseman 1985: 246–62.

16 in hendecasyllabis: Sen. Controv. 7.4.7, Charisius, Grammatici Latini I.97 Keil; Catulli nobile epigramma: Quint. 1.5.20; in epithalamio: Quint. 9.3.16. Butrica 2007: 19–21, pointing to similar patterns in references to the work of contemporaries such as Calvus and Cinna, argues that separate publication of works in different metres was the rule among poets of Catullus’ generation. Such collections likely went by the titles Hendecasyllabi, Epigrammata, etc., though Butrica thinks it more probable that these were bestowed on the works by booksellers or editors than by the poets themselves.

17 A further argument in this direction is provided by the important and influential study of Ross 1969, who points to significant stylistic differences between polymetrics and epigrams. See also E. A. Schmidt 1973: 234–8; Solodow 1989; Fain 2008: 17–64; Hutchinson 2008: 121–30, 2012.

18 So e.g. Clausen 1976; Butrica 2007: 19–22. Some evidence of a further self-standing (?) publication is offered by Pliny the Elder (HN 28.19, Mynors’ fr. 4 of Catullus), who attributes to Catullus an ‘imitation of erotic spells’ along similar lines to Theocritus 2 (the Pharmaceutria) and Virgil, Eclogue 8. For this and a possible collection of Priapea (frs. 1–2 Mynors), see e.g. Butrica 2007: 17–18.

19 Though this continues to be a matter of debate: see e.g. Tränkle 1967: 100–1; E. A. Schmidt 1979; Van Sickle 1980: 7–12; Minyard 1988; Scherf 1996:12–29; Hubbard 2005: 255–6.

20 Juventius is named in poems 24, 48, 81, and 99; verbal echoes between 15, 21, 23, and 24 invite us to read these poems as a group, with the implication that the unnamed object of rivalry in 15 and 21 is to be identified with Juventius, and the rival of 24 with Furius. Furius also appears in 26, and the pair are jointly addressed in poem 16 (discussed in Chapter 4 below, pp. 76–9). Veranius and Fabullus appear, separately or together, in poems 9, 12, 13, 28, and 47.

21 In poem 29, Mamurra is both attacked under his real name and referred to as ‘that fucked-out prick’ (29.13 ista…diffututa mentula): this is usually taken to identify him definitively as the target of the ‘Mentula’ poems.

22 Mamurra/Mentula: 29, 57, 94, 105, 114, 115 (also 41 and 43, where he is almost certainly to be identified with the decoctor Formianus or ‘bankrupt of Formiae’; for Formiae as Mamurra’s birthplace, see 57.4); Gellius: 74, 80, 88–91, 116. On both cycles, see further in Chapter 4 below.

23 See esp. Barwick 1958: 312–14; Segal 1968; Skinner 1981: 39–43; Miller 1994: 63–72; Du Quesnay 2021: 179–80.

24 See esp. Beck 1996. Cf. Stroh 1990; Hubbard 2005; Du Quesnay 2021: 179–85.

25 Martial (4.14.13–14) writes – anachronistically – of Catullus perhaps sending his ‘Sparrow’ to Virgil: it is clear from the context that passer refers to a poetry book. Cf. Mart. 1.7, 11.6.14–16 (discussed further in Chapter 6).

26 So e.g. Wiseman 1969: 7–16 (now modified by Wiseman 2023: 13–45); Skinner 1981: 39–68; Stroh 1990; Schafer 2020: 81–133.

27 Thus, related poems are directly juxtaposed at 2–3 (unless ‘2b’ is regarded as a fragment of a separate poem), 23–4, 28–9, 61–2, 75–6, 88–91, 97–8, 110–11, and 114–15, to cite only the more obvious instances.

28 The name may be corrupt: it appears as a me an a in the manuscripts, and is otherwise unattested. For the identification of ‘the bankrupt of Formiae’ (41.4 = 43.5) with Mamurra/Mentula, see n. 22 above.

29 Frequently invoked in support of this tripartite division is the argument (apparently first advanced by Süss 1876: 23–4) that 61 and 65 can be identified, alongside 1, as dedication poems: the epithalamium for Manlius Torquatus may be read as an implicit dedication for the second ‘book’, with its recurrent theme of marriage, while the carmina Battiadae offered to Hortalus at 65.16 could, in a broad sense, subsume all the following elegiacs and epigrams, in addition to the more obviously Callimachean 66. It has been pointed out, too, that all three poems contain addresses or reference to the Muses, assuming that the ‘virgin patroness’ of 1.9 is so identified (cf. 61.2, where the marriage god Hymen is addressed as Vraniae genus, ‘offspring of [the Muse] Urania’, and 65.2 a doctis…uirginibus, ‘from the learned maidens’ (i.e. the Muses); so Wiseman 1979a: 176–9, 1985: 266). Internal evidence for a three-book edition is perhaps corroborated by some peculiarities of MS O (a direct descendant of the now lost archetype, V), which begins a new page – contrary to the scribe’s usual practice – at 61.1, and from poem 65 onwards introduces each poem (where divisions are indicated at all) with an illuminated initial and capitalized second letter (see Thomson 1997: 7–8).

30 On ‘publication’ and the book trade in the Roman world, see especially Kenney 1982; Starr 1987; White 2009; Winsbury 2009. Also relevant here is evidence derived principally from Greek papyri, which suggests that poetry collections might have circulated in different forms simultaneously (see e.g. Gutzwiller 2012: 86–7, with further bibliography at n. 14). See also Heerink 2009, who makes a concise case for considering the term ‘publication’ anachronistic when applied to Catullus and to ancient books in general.

31 As Catullus plans to do in poem 14, where he threatens to send Calvus a collection of bad poetry in retaliation for his friend’s similar ‘gift’. See Barchiesi 2005, esp. 337–9; Hutchinson 2008: 110–11; Gaisser 2009: 30–1; Gutzwiller 2012: 86–7.

32 Book 4 of our Greek Anthology consists of what were originally the introductory poems of the separate collections amalgamated by the Byzantine compiler: 4.1 is identified in the manuscripts as ‘Meleager’s Garland’.

33 So Barchiesi 2005: 322–3.

34 On this aspect of Poem 50, see further Chapter 3, pp. 53–5, below.

35 Some critics, e.g. Gibson 1995, have seen irony in the rather hyperbolic praise of Cornelius’ three-volume historical work, the Chronica, particularly the grandiose omne aeuum, ‘all of time’ (6), coupled with the interjection Iuppiter (7). The adjective laboriosis, ‘hard work’ (7), too, might be taken to imply that the three volumes were tough going for the reader, as well as complimenting the writer’s scholarly credentials. Such ironies might be interpreted, however, as friendly teasing rather than serious criticism. On (Catullus and) Nepos, see also Cairns 1969; Wiseman 1979a: 154–66; W. J. Tatum 1997: 485–8.

36 For programmatic readings of poem 1, see esp. Elder 1967; Cairns 1969; Latta 1972; Singleton 1972; Syndikus 1984–90: i.71–7; Batstone 1998, 2007: 236. For a different approach, see McMaster 2014 (esp. 201–2), for whom the emphasis on the book’s refinement and novelty serves to enhance its value within a system of gift exchange.

37 Callim. Aetia fr. 1.21–8 Pfeiffer.

38 E.g. Ovid, Am. 3.9.62; [Tibullus] 3.6.41; Mart. 7.99.7, 8.73.8, 14.100.1.

39 On literary dedication as social practice, see Stroup 2010 and – on poem 1 specifically – McMaster 2014; Gale 2016. See also Johnson 2009, 2010, which explore the ‘sociology of reading’ among the Roman elite of the High Empire, but contain much of relevance also to Catullus and the Late Republic.

40 On urbanitas and related concepts, see Ramage 1973; Seager 1974; Wiltshire 1977; Fitzgerald 1995: 87–113; and especially Krostenko 2001.

41 Ideologically inflected readings of Catullus include Skinner 1979, 1989, 2003; Fitzgerald 1995; Krostenko 2001; Nappa 2001; Wray 2001.

42 For a nuanced discussion of Catullus’ ‘conflicted cultural identity’, see Fitzgerald 1995: 185–211, esp. 201–7 (quoted phrase at 185).

43 See p. 1 above. The free population of Cisalpine Gaul was fully enfranchised by Caesar in 49 bc, and the province was formally incorporated into Italy only in 42, after the probable date of the poet’s death.

44 Pliny, HN 36.48.

45 Attacks on associates of Caesar: poems 28 and 47 (Piso, probably to be identified with Caesar’s father-in-law [see n. 48 below]); 29, 41, 43, 57, 94, 105, 114 and 115 (Mamurra/Mentula); 52 (Vatinius, for whom see also 14.3 and 53.2); and 54 (where the obscure Otho, Libo, and Fufidius [?] appear, on the basis of lines 6–7, to be singled out as Caesarians).

46 See e.g. Quinn 1972: 267–77; Nappa 2001: 85–105; Skinner 2003: esp. 23–4, 137–42; Konstan 2007; Roman 2014: 59–66, all of whom tend to view Catullus as ‘profoundly estranged’ from Roman politics (Konstan 2007: 78), though varying in the extent to which they perceive this estrangement as part of a broader critique of elite Roman mores and public life (Nappa, e.g., writes of Catullus’ ‘sustained scrutiny and criticism of Roman society’ [Nappa 2001: 23; cf. Konstan 2007: 75]). W. J. Tatum 1997: esp. 482–5 and 2007 argues cogently against the imposition of rigid distinctions between the ‘political’ and the ‘personal’ in the interpretation of Catullus’ poetry. For the contrary view – that Catullus’ poetry is strongly partisan and aligned with aristocratic opposition to Caesarism – see esp. Syndikus 1986 (with a useful survey of earlier scholarship at 34–5). See also Bellandi 2012, who reads poems 29, 52, 54, and 57 as ‘bitterly anti-Caesarian’, but seeks to restrict this alignment on the poet’s part to the years 56–54 bc. Benferhat 2005: esp. 139–40 writes more judiciously of Catullus’ studiously maintained independence from Caesar.

47 See e.g. Cic. QFr. 1.1.12–14; Att. 5.16.2–3, 6.1.2, 7.1.6.

48 Probably L. Calpurnius Piso Caesoninus, the target of Cicero’s invective In Pisonem, though the identification is not certain (Syme 1956; contrast Wiseman 1969: 38–40). Wiseman (1985: 99–101; 1987b: 336–40) argues that Catullus’ expedition to Bithynia was as much a business trip as it was an administrative posting; but (even if Wiseman’s deductions about the commercial interests of the Valerii Catulli are accepted) the pursuit of private business affairs was evidently not regarded as in any way incompatible with political ambitions (thus, Cicero tells us at Pro Caelio 73 that Marcus Caelius Rufus travelled to Africa as a contubernalis of Quintus Pompeius both to gain administrative experience and because his father owned land in that province).

49 ‘Socration’ should probably be taken as a mocking soubriquet for the Epicurean philosopher Philodemus, a protégé of Calpurnius Piso and similarly depicted by Cicero (Pis. 68–72) as a kind of partner in debauchery, albeit a rather reluctant one. Porcius (‘Piggy’) seems likely to be a pseudonym too: no conclusive identification has been established, though Cairns 2003: 184–7 makes a good case for Plotius Tucca, a known associate of Philodemus’, and later the literary executor of Virgil.

50 See esp. Cic. Att. 5.21.7, 6.1.2, 6.2.4–5. For Catullus as self-discrediting speaker, see Braund 1996.

51 For the broad semantic range of Latin amicitia, see esp. Brunt 1988; C. A. Williams 2012: 44–54. On invective in literature and political life, see Chapter 4 below.

52 See Richlin 1992: 26–7, 69; C. A. Williams 1999: 198; Krenkel 2006a [1980]: 213–15. On Roman sexual ideology and the classification of sex acts, see further pp. 74–6 below. Social kissing between males appears to have been the norm in Roman upper-class society at this date: see e.g. Cic. Att. 16.5.2; Cic. Fam. 16.27.2; Suet. Tib. 34.2.

53 For a range of views on the personal and/or political nature of the attack on Lesbius, see esp. Skinner 1982a; W. J. Tatum 1993; Butrica 2002: 507–13; Watson 2006: 44–8.

54 On the ambiguities of Catullus’ status as a member of the municipal elite, see esp. W. J. Tatum 1997.

55 The centrality of competitive social performance to Catullus’ poetry is emphasized, in slightly different ways, by Fitzgerald 1995: 87–113; Krostenko 2001: 233–90; and Wray 2001: esp. 56–63 and 124–9.

56 Cic. Cat. 2.22–3; Cael. 66–7.

57 Nothing is known of Volusius beyond what can be inferred from these two poems, and Catullus’ prediction in poem 95 that his work would not survive has proved accurate: no identifiable fragments have been preserved.

58 On urbanitas and social performance in Cicero, see esp. Krostenko 2001:154–232.

59 The self-conscious obscurity of the wind name is presumably part of the stylistic parody, alongside the fifth-foot spondee: compare the recherché Greek name Apheliotes for the east wind at Catullus 26.3.

60 Cic. Att. 7.2.1.

61 On the ‘new poetry’ and the usefulness (or otherwise) of the label ‘neoteric’, see esp. Clausen 1964; Crowther 1970; Wiseman 1974: 51–3; Tuplin 1976; Lyne 1978; Courtney 1993: 189–91; Lightfoot 1999: 54–8; Hollis 2007: 1–2.

62 For what is known of these contemporary poets, and the fairly minimal surviving fragments, see the relevant entries in Courtney 1993 and Hollis 2007: none of Valerius Cato’s work has survived, and Caecilius is known to us only from Catullus poem 35.

63 Note also the references to Battus and Cyrene (Callimachus’ place of birth) at 7.4–6, and the apple simile at the end of poem 65, which is (probably) evocative of the Acontius and Cydippe story related in Aetia 3 (frs. 67–75 Pfeiffer; see further Chapter 5, p. 99, below). Bibliography on Catullus and Callimachus is too extensive to list in full here: helpful starting points are Clausen 1964; Wray 2001: 186–203; Fantuzzi and Hunter 2004: 464–85; Hunter 2006: 88–108, 144–6; Knox 2007; Acosta-Hughes and Stephens 2012: 214–33.

64 Callim. Hymn 2.108–12; Clausen 1964: 188–91.

65 The earliest manuscripts read canas, which is both unmetrical and clearly anticipated in error from the following line. Most editors accept the early correction cauas; citas is the suggestion of McKie 2009: 243–4, n. 113; see also Nisbet 1978: 111.

66 Skutsch 1985: 147–50.

67 So Clausen 1964: 187–93; Wiseman 1974: 51–8. For a more sceptical review of the evidence, see Lightfoot 1999: 50–76.

68 Quinn 1969.

69 Though see now Knox 2011, who argues against labelling these works as ‘Callimachean’.

70 Ov. Tr. 2.427–41, Plin. Ep. 5.3.5; Tac. Ann. 4.34; cf. also Gell. NA 19.9.7.

71 It is noteworthy that Catullus and Calvus are regularly paired by later writers: see e.g. Hor. Sat. 1.10.19; Prop. 2.25.4; Ov. Am. 3.9.62.

72 The term ‘intertextuality’ is preferable to the more traditional ‘allusion’ as (i) broader in application, denoting as it does any relationship between texts, not only those involving close verbal resemblance; and (ii) not implying anything about the (unknowable) intentions of the historical author. Intertextual relationships (as opposed to incidental ‘allusions’) are typically perceived too as constitutive of meaning, rather than as a kind of ‘optional extra’: the reading process always, arguably, involves conscious or unconscious comparison with previously encountered texts, though some works may be said to invite intertextual awareness more strongly than others. For helpful discussion of the terminology, see esp. Fowler 1997.

73 See further Chapter 5 below. On the narrative and thematic complexities introduced via the echoes of Euripides, Apollonius, and Ennius, see especially Zetzel 1983; Gaisser 1995; Clare 1996; De Brohun 2007; Fernandelli 2012: 160–210.

74 Conte 1986: 31.

75 On the relationship between the two poems, see esp. Hiltbrünner 1972; Carilli 1975: 942–5; Marcovich 1982; Dettmer 1989; Gale 2022: 89–94.

76 The significance of the unguentum has been much debated: see especially Vessey 1971; Littman 1977; Hallett 1978; Marcovich 1982; Bernstein 1984; Gowers 1993: 229–44.

77 Catullus may have taken his impetus here from another poem of Meleager, Anth. Pal. 7.207, on the death of Phanion’s pet hare (so Thomas 1993: 134–5), where the language is perhaps mildly suggestive. In particular, Phanion, like Catullus’ girl, is imagined as cradling her pet in her bosom/lap (ἐν κόλποις στέργοʋσα; 3).

78 The ‘obscene’ reading of poems 2 and 3 goes back to the Renaissance scholar-poet Poliziano (Gaisser 1993: 75–8); modern adherents include Genovese (1974), Giangrande (1975), Nadeau (1980), and Miller (1994: 66–70). For the contrary view, see especially Jocelyn 1980; Adams 1982: 32–3; and Jones 1998. The key witness is the second-century ad lexicographer Festus, who reports (p. 410 Lindsay) that the male member was referred to in mime as strutheum, a salacitate uidelicet passeris, qui Graece strouthos dicitur (‘evidently from the wantonness of the sparrow, which is called strouthos in Greek’); the sparrow’s salacitas is similarly noted by Cicero (Fin. 2.75) and Pliny (HN 10.107). Also relevant is a double entendre on the word passer in the epigrams of Martial, on which see Chapter 6, p. 124, below.

79 There are several similarities with the opening lines of poem 64, where the Argo, too, is said (in imitation of the opening lines of Euripides’ Medea and Ennius’ Latin version) to have begun life as a tree (4.10–17, 64.1): both ships have ‘palms’ (palm[ul]ae, i.e oar blades; 4.4 and 17, 64.7) and ‘swim’ (4.3 natantis, 64.2 nasse) through the waves; the verb imbuo, meaning both to ‘wet (for the first time)’ and to ‘initiate’, is used of both vessels (4.17, 64.11); and both are swift (4.2 celerrimus, 64.6 cita…puppi) – significantly so, since one of the several ancient etymologies for the name Argo derived it from the Greek argos, ‘swift-footed’. The route of the phaselus approximately reverses that of the Argo (to Colchis at the eastern end of the Black Sea). See Hornsby 1963.

80 See M. Schmidt 1955; Khan 1967; Coleman 1981; Fitzgerald 1995: 104–10; Davis 2002; Roman 2014: 76–7.

81 Superannuated prostitutes: e.g. Meleager, Anth. Pal. 5.204 (catalogue of body parts); [Plato] Anth. Pal. 6.1; Philetas of Samos, Anth. Pal. 6.210.3–4 (ageing courtesans dedicate the mirrors they no longer need or wish to use, a topos perhaps recalled and inverted by Ameana’s failure to look in the mirror). Catullus’ anatomization of his victim in poem 43 may also have in view a contemporary epigram, Anth. Pal. 5.132, in which Philodemus lists Flora’s physical charms (‘o foot, o calf, o thighs…’) and dismisses her lack of culture and sophistication (‘so what if she’s a country girl and…can’t sing the poems of Sappho?’). Ameana, in contrast, has neither physical attractions nor stylishness to recommend her (the last item on Catullus’ list is nec…nimis elegante lingua, ‘none too stylish a tongue’).

82 So Skinner 1979.

83 Cf. Skinner 1989.

84 See e.g. Skinner 1971; Dyson 1973; Thomas 1984; Nappa 2001: 71–5; Uden 2006; Polt 2021: 107–25.

85 This technique, we should note, is not as wholly unprecedented as is sometimes claimed: contemporary epigrammatists, such as Meleager and Philodemus, also address or write about the same mistress (or boy) in several poems. But what seems new in Catullus is the sense of continuity – of a developing narrative – created largely through the use of internal echoes, which invite us to read the Lesbia cycle in particular as a sequence.

86 The most thoroughgoing recent proponent of this view is Niklas Holzberg (2000, 2002), who sees the poem’s speaker or persona as a wholly artificial creation, not necessarily bearing any relation to the collection’s ‘hidden’ author. Similarly, Lesbia, for Holzberg, should be seen as entirely fictive: see further Chapter 2, pp. 40–2, below.