1. I would like to thank Professors E. J. Kenney and Geoffrey Lloyd for reading a protozoan draft, and the editors and their anonymous referee for their scrupulous final readings. Professor Niall Rudd told me not to abandon ship and pointed the way: so this essay is for him. For the record, S. D. Goldhill and S. H. Prescott were both in the first class [I ever gave on Juvenal 1].

2. Anderson, W. S., Essays on Roman satire (Princeton 1982) 297–8, 427–8CrossRefGoogle Scholar, points up the initial projection of the reader into the recitation-scene and the eruption into fury of the speaker before he tells us of it.

3. On Juvenal's first audience, cf. Coffey, M., Roman satire (London 1976) 124Google Scholar, ‘(Juvenal) wrote … not so much for the pleasure of friends as for the applause of a public audience that expected the extrovert declamatory manner’ (cf. 244 n. 38). Such a picture should not undermine the importance to that audience of their status as ‘insiders’ expert in the mode and the heritage of their performers, nor should it displace the importance to all concerned of the circulation of texts, books, æuvres in the wake of the recitatio; the phases were interfused from start to finish (as Juvenal is presuming). On ‘Juvénal et la vie littéraire de son temps’, cf. Gérard, J., Juvénal et la réalité contemporaine (Paris 1976) 55–115Google Scholar.

4. Did anyone go to recitations to sneer – until later? Juvenal will teach everyone to! To go, and sneer at going.

5. Rudd, N., Themes in Roman satire (London 1986) 123–5Google Scholar, points to the ‘ambiguities involved’ in attendance at recitationes.

6. The institutional organization of group or collective or somehow public reading sessions, with or without associated fora for discussion, reaction, valuation, has always presupposed unmonitored individual more or less private reading which it sought to dictate, determine and control; the ‘private’ reader has always read under and against ‘public’ orders. I would like to assert that a recurrent error from which much critical dogmatism has stemmed is to concede arenas such as the recitatio the privileged status within ancient reading culture that it certainly meant to claim for itself; interplay and contestation between these poles are necessary if we are ever to think of (any) literature dynamically, and as itself dynamic, both textually and in history.

7. Cf. Anderson (above n. 2) 199–203, 206.

8. Cf. Hirst, G., ‘Notes on Juvenal, I, III, VI, X’, AJPh 45 (1924) 276–83, at 276–7Google Scholar, vindicating the structural symmetry of the poem.

9. Cf. Hylan nautae quo fonte relictum | clamassent, ut litus ‘Hyla, Hyla’ omne sonaret, Virg. Ecl. 6.44, Wijsman, H. J. W., ‘Female power in Georgics 3.269/270’, HSPh 94 (1992) 259–61Google Scholar, on sonantem | Ascanium. With Juvenal's stress on the wild-goose chase's potential for diversionary amplification, cf. esp. Val. Fl. 3.596–7, rursus Hylan et rursus Hylan per longa reclamat | auia …

10. Ready and set for the off in his borrowed Lucilius' chariot of satire (1.20), Juvenal displays the colours of his poetics through his preliminaries: his employment of the Georg. 3's programme for his Book 1, Satire 1 shows at least he could read Virgil's ‘Proems in the middle’; cf. Conte, G. B., YCS 29 (1992) 147–59Google Scholar, on Georg. 3 and Ecl. 6.

11. The work of Thomas, R. F., ‘Callimachus, the Victoria Berenices, and Roman poetry’, CQ 33 (1983) 92–113CrossRefGoogle Scholar, later applied in his Commentary on the Georgics (Cambridge 1988)Google Scholar, is too decisive by half, arguing for ‘Callimachean (rather than Pindaric) reference …’ ‘In fact Callimachus was influenced by Olympian 1 …, and reference to Pindar may be only apparent…’ ‘The metaphor of a temple … has been thought to be Pindaric … This may be the ultimate impulse … but the image may have been developed at some length by Callimachus …’ ‘The treatment of envy is a standard epinician motif … But again “epinician” does not mean “Pindaric” … Envy is also a motif of Callimachean epinician’ Vol. 2, 37, 39, 41, 46–7). But if, as Thomas demonstrates, to ‘know’ Virgil is to know Virgil's reading of Callimachus, then what kind of reading of Callimachus is imputed to Virgil once ‘knowledge’ of Callimachus' reading of Pindar is denied him? It is ironically the case that much of our ‘knowledge’ of Pindar and Callimachus, as well as their textual relations, is ‘known’ to us through intermediary comment on their relations with Virgil, through readers of Virgil such as Statius, and through commentators on their relations with Virgil. (Thus, the mummycartonnage which preserves our fragments of the Victoria Berenices is from an edition dated to ‘within one generation of Kallimakhos' death’, presents ‘a commentary within the text’ which ‘intersperses the lines of text at irregular intervals, usually with glosses on previous lines, and … is differentiated from the poem itself only by the indentation of three letters. … [V]erse and scholia can look exactly alike’, Rosenmeyer, P. A., ‘The tradition of Kallimakhos as innovator: a re-examination in the light of Victoria Berenices’, unpublished B.A. dissertation, Cambridge 1982, 1)Google Scholar. When Thomas has Callimachus' ‘poetry of rejection … itself rejected’ by Virgil (art. cit. 101), we need not reject the view, but may constructively read him as half-wrong, since imitation of the ‘rejection of triteness’ topos is, as he has shown, a tribute to Callimachus' example, even as it gives Virgil the model for arrogating power to his own poetry in Callimachus' despite (‘How Callimachean!’): ‘rejection’ is only a fraction of the point. To reject the cliché of ‘the rejection of the cliché’ is too swashbuckling an approach to poetic traditionality for Juvenal, indeed it can only be a non-approach. After his proem, Virgil's first topic as stock-breeder will be ‘The importance of careful selection’ (Thomas, , Virgil, Georgics, Vol. 2, 49, on 49–94Google Scholar).

12. Who can like hating what they are reading? You?

13. Greatness can be a curse as well as a blessing on any statement. Can't it?

14. Otto, A., Die Sprichwörter der Römer (Hildesheim 1971 ²) 120Google Scholar, s.v. domus §4 = Cic. Ad. Q. fratr. 1.1.45.

15. Gérard (above n. 3) 67 n. 4, cf. 74, 458, accepts Niebuhr's notion from Friedländer (Vol. 1, 132) that this is T. Catius Fronto, cos. 96, defence counsel for Marius Priscus (cf. Juv. 1.49), Arval in 101 and 105 next in order to Q. Fulvius Gillo Bittius Proculus (cf. Juv. 1.40?). Or is this already the satirist's tease – aggressively putting up a brave … front? The singers of Calpurnius Siculus are bizarrely fastidious (postmodern) about the circumambience for their idyllic songs, e.g. antra petamus … (1.8), sub hac platano … (4.2), as are Juvenal and Umbricius for Satire 3, speluncus | dissimiles ueris … nec … uiolarent marmora tofum (13–20).

16. Venti do, of course, agunt transitively, e.g. Virg. Georg. 1.461–2, unde serenas | uentus a g at nubes.

17. As Sen. Apocol. 14.1 puts it, of its tribunal Aeaci.

18. Friedländer ad loc. (Vol. 1, 130–1) follows Janus Parrhasius (1470–1534) in seeing Valerius as Juvenal's prime target, ‘nicht ohne Wahrscheinlichkeit’; but, he adds, Valerius is not exclusively the target, because the triteness of trite examples is the point; Gérard (op. cit., 82) seconds Friedländer: ‘avec raison’. Mayor (Vol. 1, 92), on 8, has, ‘Iuv. still alludes to the Argonauts …; perhaps to the poem of Val. Fl.’, Duff (116), on 7–13, ‘It is possible that the Argonautica of Valerius Flaccus … is the main object of this attack’, and Ferguson (112), on 7, ‘This may be a hit at [Valerius].’ Rudd and Courtney (35) on 7–17 opine, ‘In mentioning their trite epic themes, Juvenal was not thinking solely of the Argonautica of Valerius Flaccus (written before A.D. 93); several of the references do not fit that poem, and Juvenal's whole point is that there are many works of that kind’, and Courtney (85) on 7ff. rules, ‘It has often been thought that the allusion to the grove of Mars … and the contemptuous reference to Jason's ‘theft’ of the Golden Fleece are a hit at the Argonautica of Valerius Flaccus; this is neither particularly likely (especially as Valerius does not give an ἔϰφϱασις of the grove) nor particularly unlikely.’ Further views in Ramage, E. S., ‘Juvenal and the Establishment. Denigration of predecessor in the “Satires”’, ANRW II.33.1 (1989) 592–639 at 666 n.52Google Scholar.

19. Recent studies of 1.1–21 include Braun, L., ‘Juvenal und die Überredungskunst’. ANRW II.33.1 (1989) 770–810 at 771–2Google Scholar, Wehrle, W. T., The satiric voice. Program, form and meaning in Persius and Juvenal (Heidelberg 1992) 10–14Google Scholar.

20. The finest is at Sen., Thy. 641–82Google Scholar. The topos is proclaimed a locus, e.g. Sen., Oed. 530–48Google Scholar, Est … lucus … | huc …, Luc. 3.399–425, lucus erat … | hanc …, Stat. Theb. 4.419–42, Silua … stat … | hic …

21. The Virgilian ‘Cave’ of Aeolus is post-Homeric (Od. 10.13, A.d, δώματα ϰαλά, etc.), inspired by Lucr. 6.195, speluncas, etc.

22. In Georg. 4.173, Aetna is the locale, whence antrum | for Aetna | in the self-quotation at Aen. 8.451.

23. Cf. Aetn. 445, Siculi uicinia montis, Solin. 6.1. ‘Neighbours’ are ‘alike’, cf. Williams, G. D., Banished voices. Readings in Ovid's exile poetry (Cambridge 1994) 150Google Scholar and n. 110 on Ovid and Cotys.

24. Cf. Hardie, P., Virgil's Aeneid: cosmos and imperium (Oxford 1986) 105–7, esp. 106 n. 54Google Scholar.

25. Feeney, D., The gods in epic. Poets and critics of the classical tradition (Oxford 1991) 336–7Google Scholar, ‘This direct attempt to outweigh the earthly suffering is designed to have a solace which cannot be achieved by Vergil's more diffuse images … [A] tepid and unmoving moment, perhaps, but –’.

26. Vessey, D., Statins and the Thebaid (Cambridge 1973) 248Google Scholar, ‘Valerius the classicizer has spurned mannerism and affirmed his faith in Homer and Virgil.’ (Virgil?)

27. For Judge Aeacus in prose, cf. Cic. De off. 1.97, etc.

28. Ullman, B. L., ‘Psychological foreshadowing in the Satires of Horace and Juvenal’, AJPh 71 (1950) 408–16 at 415Google Scholar supposes that in 1.7–11 Juvenal first sets puzzles (7–9), then confirms the guess of Argonautics (10–11) – but ‘the revelation comes too soon’. This sets the ‘knowledge’ shared by Juvenal's audience rather low.

29. Cf. dirum Phrixei uelleris aurum |, Stat. Theb. 2.281, non Athamanteo potius me mirer in auro, | Aeolium dones si mihi, Phrixe, pecus, Mart. 8.28.19–20, Bömer on Met. 7.155.

30. Apollonius repeats the χϱύσειον … ϰῶας of his simple proem (1.4) a further seven times, with χϱύσειον … δέϱος in seven places; he builds no phrase for the Fleece on the base ‘gold-'. Valerius extrudes the Fleece from his elaborate prologue (first at 1.56, pecoris Nephelaei uellera). He produces many variations on the bases aurum, uellus and pellis (u. Helles, 1.167, Phrixi u., 1.272–3, cf. 1.328, 4.556, Phrixea u., 8.267–8, u. Graio, 1.519, aurea u., 5.200–1, 7.167–8, Aeolio u., 7.517, umida … uellera … leui … auro, 1.288ff., uellera … ardenti … metallo, 5.228ff., rutilant … u., 5.250, u. sacra, 5.629, uellera … et exuuias pecudis … sacrae, 6.18–19, uellera …, fulgentia dona, 8.101; optatam … pellem, 4.620 (from Catull. 64.5), dives p., 5.1203, aurea … p., 7.30–1, rutilam p., 8.114, uillis … comantemi | sidereis p., 8.122. Cf. uectorem … Helles. 1.425, aureus … uector, 1.281–2, auratae pecudis … exuuias, 5.490–1, aurea … terga, 5.553–4, 8.42–3, 8.131–2, nemus auriferum, 5.637, Phrixeae pecudis, 8.75, aurigerae … arboris, 8.110, aureus (Jason in fleece), 8.128. Avoidance of set phrasing is Valerius' Quest. (To review the periphrases built on proper names, Nephele is not in Ap. Rhod.; Nephelaeus is first in Val. Fl.: cf. Nepheleis, Ov. Met. 11.195, Nepheleias, Luc. 9.956; Αἰολίος is not in Ap. Rhod.; Ἕλλη is named once in Ap. (1.256: not of the Fleece; Phrixeus has no equivalent in extant Greek; Phrixus is found in no circumlocution in Ap.).)

31. For the poem-and-poet's self-investment in this many-splendoured ship, cf. Davis, M. A., ‘Ratis audax: Valerius Flaccus' bold ship’, in The imperial muse. Flavian epicist to Claudian, ed. Boyle, A. J. (Victoria 1990Google Scholar = Ramus 18 (1989)) 46–73, and for what follows, cf. esp. 65–8 on Argus' ‘decorations’ of Argo, 66–7 on the Centaurs. She successfully vindicates Valerius' kaleidoscopic intertextuality through the flagship of his figuration.

32. The disposition is controversial, cf. Strand, J., Notes on Valerius Flaccus' Argonautica, Studia Graeca et Latina Gothoburgensia 31 (1972) 55–6Google Scholar. For parte alia in ecphrases, Virg. Aen. 1.474 (with agnoscit, 470, cf. Val. 144, agnoscitur), 8.682 (with contra, 711, cf. Val. 137, Sil. 2.426. Strand (56–7) also discusses in mediis, Val. 148 (cf. Aen. 8.675, in medio, 700, medio in certamine). The scenes portend, through Peleus and Thetis, the ‘marriage’ of Jason and Medea; through Polyphemus and Galatea, the jilting of Styrus by Medea; and through Lapithocentauromachy, the carnage wrought by Medea's marriage to Jason, cf. Adamietz, J., Zur Komposition der Argonautica des Valerius Flaccus (Munich 1976Google Scholar = Zetemata 67) 10–11. They cleave a portentous trajectory of representation through the oceanic surge of desire.

33. Frank, E., ‘Works of art in the epics of Valerius Flaccus and Silius Italicus’, RIL 108 (1974) 837–44 at 837–8Google Scholar explains that Valerius alludes here to a scene cut from his narrative, Thetis' salvation of the Argonauts on the return journey (Ap. Rhod. 4.757–968); she sees the Lapiths' win for civilization over Centaur barbarism as paradigmatic for the Argonautica. But Chiron plays the music at vexed Thetis' wedding.

34. Cf. RE 20¹.513ff., Pholoe.

35. Masters, J., Poetry and civil war in Lucan's Bellum Civile (Cambridge 1992) 173–4Google Scholar, argues that Lucan is not out to fool us, but rather ‘not [allowing] the realities of geography to interfere with the plan of contracting all possible myths into the bounds of a single land’.

36. ‘Popularity’ may need qualification, in whichever of its senses.

37. Val. 140–1, promotes his Rhœtus report into the slot of Eurytus in Ovid's preliminaries, quam uino tam uirgine uisa | ardet, 220–34; Rhoetus had had the first aristeia in Ovid, 271–301. For Atracia, 141, cf. Ovid 209, Atracidas.

Val. 142–3 is short-hand for Ovid's opening paragraphs of rioting, 235–70, with signis exstantibus asper | antiquus crater, 235–6, mensœ, 222, 254, arœ, 260, cf. 258, pocula … uolant … que … que, 242–3.

Ovid's Nestor shared in his Peleus' aristeia (366–92), then helps himself to a string of scalps (429–45, 449–58), before his tale returns to its task with Caeneus (459ff.). Val. 143–4 pairs Peleus with his intrusive (non-Ovidian) father of Jason, Aeson, who is thus in the shoes of Ovid's narrator Nestor. Peleus' spear is from Ovid (fraxineam … misit … hastam|); Aeson has Nestor's sword (| ense … nostro, 42, the only ensis in Ovid's riot, before Caeneus' stand at 484–5, 492).

Then Nestor does appear in Val. 145, but on a Centaur's back, which is Theseus' stunt in Ovid (345ff., tergo … insilit haud solito quemquam portare nisi ipsum; Bianor for Valerius' Monychus; Val. has therefore ‘syncopated’ here Theseus' deeds, 342–60, with Nestor's).

Clanis in Val. 146 is from Ovid 379, where he hides in a clannish list of Peleus' victims. Actor faces real death by (fake-)etymology (peragit/Actora), with a blazing tree, as in the assaults in Ovid 327–60 (quercu, 328, cf. 342; arserunt, of Rhoetus’ brand, 275, cf. 247, 261, 274, 280, 295–6). This Actor is unemployed elsewhere in extant Latin, but his name is epic stock, Il. 2.513, Virg. Aen. 9.500, 12.94 [whence Juv. 2.100], Stat. Theb. 10.250, 257; one Actor was Patroclus' maternal grandfather, Val. Fl. 1.406–7, cf. Il. 11.785, etc.

Nessus (147) keeps his position in the rout as in Ovid 301–15. But his ‘blackness’ tallies with Ovid's Atracian black beauties Cyllarus and Hylonome (393–428, cf. pice nigrior atra …, 402ff. Valerius' conceit with Nessus and his equus-half synthesises Ovid's series of grotesqueries, uir … equi, 399, hominemque simul protectus equum que, 431, uir equo, 478, each of a single Centaur.

Finally, Hippasus in Val. 148 is hidden in the middle of Ovid's list of Theseus' victims (352: 342–60); but handed the role of comatose Aphidas (316–26).

On this analysis, Valerius has ‘accounted for’ all Ovid's scenes up to the fall of Caeneus. He has put Aeson in his centre-stage (144), in place of Ovid's Nestor. And he has shifted his ‘Aphidas’ from Ovid's start to his own conclusion and labelled it in mediis (where Ovid's classic burlesque belongs — and abides in the heart of Valerius' text).

38. The style of presentation traces to a particular brand of ‘knowledge’; the question is how closely this model, token of a certain Oxbridge, approximates to, or palms itself off as, a replication of a, or any, dominant literary mentalité in Antiquity. The question of Classics.

39. To set out these questions is (not) to insist on some multiple-choice moment, but to propose that any reading is structured in terms of the function within it of the moment of krisis. Decision is no less/more absurd, and ultimately self-confuting, than epochē.

40. E.g. Sidon. Apoll. Carm. 9.

41. Cf. Pers. 5.10–11, neque anhelanti, coquitur dum massa camino, | folle premis uentos, nee clauso murmure raucus …

42. Aeneid 8's sung hymnal narrative of Cacus' Cave cumulates into its narrated transposition, Vulcan's Forge, in the build-up toward the epic incised on Aeneas' shield: Hercules' Ara Maxima and Augustus' post-Actian Palatine empower the narrative's formative narration of Roman political destiny through Aeneas. Antra Aetnaea tonant, ualidique incudibus ictus | auditi referunt gemitus, striduntque cauernis | stricturae Chalybum et fornacibus ignis anhelat, … (419–21, cf. 449–53, esp. uentosis follibus auras | accipiunt redduntque … gemit impositis incudibus antrum. |). These textual noises grate sparks of … compositional greatness, as the narrator's lungs rouse his characters to kindle new flames in their heart(h)s – Evander, Aeneas, and every reader.

43. … uersantque tenaci forcipe massam |, Aen. 8.453.

44. For ‘combinatorial imitation’ as the compositional axis of post-Virgilian Epic, cf. esp. Hardie, P., ‘Flavian epicists on Virgil's epic technique’, in The imperial muse (above n. 31) 3–20Google Scholar, esp. 6 on Valerius' Aeolus.

45. A crafted genre-piece, and we know it. By telling us we must know it, too, the writer forces the issue on us: what brand of ‘knowledge’, what cultural formation or textual apparatus, does he, and do we, presume?

46. For Ovid, cf. Bömer on Met. 1.272, 4.663; Aeolus is missing from Lucan, and Silius has just | Aeoliis candens Austris, 1.193, after Aen. 1.51–2, Austris | Aeoliam.

47. Ovid, however, introduced the adjective, with Aeoliis … in antris | (Met. 1.262) at the start of his Flood, as in the incidental reference at Val. Fl. 1.417 to Aeolus antris |, and in Statius’ simile (Theb. 10.246), Aeolus antro |.

48. In Aeoliam … ad antra, 1.576, the island is Virgilian, the sedes and plural are from Ovid; | strata Ceres, 578 < | sternuntur segetes, Met. 1.272; Hippotades for Aeolus, 416, is an Ovidian favourite, after Od. 10.2, 36, cf. Bömer on Met. 4.663, Hippotades … carcere, = Val. 1.610 (same sedes).

49. Many Virgiiian details, terms and ideas are reassigned by Valerius in making his storm. In particular, he concentrates the assaults on Aeneas’ fleet onto Jason's ship. Valerius' Neptune tells us Juno and Pallas coaxed him into saving Argo, but he looks forward to long centuries of deserved shipwrecks in the maritime age of greed ushered in by Argo (642–50). Silius rewrites Valerius and Virgil, as his Neptune starts his storm (17.236–9).

50. In Silius' descriptio Siciliae, 14.1–78, he mentions Lipare and Aetna (56, 58), before paying homage to Virgil (and Valerius?), Quid referam Aeolio regnatas nomine terras | uentorumque domos atque addita claustra procellis? (70–1) Valerius wittily makes his (feminine) Tempestas a tame domiseda, cf. CE 52 Bücheler (Claudia) domum seruauit, lanam fecit, Virg. Aen. 7.52 (Lavinia) sola domum et tantas seruabat filia sedes).

51. Callimachus had dislocated the Hesiodic trio of (Theog. 140) with before | Βϱόντεω (Hymn. Dian. 68, 75: he withholds names in his introduction to the forge, 46ff., releasing them in agendo). The Iliad has … (cf. Od. 8.274) (18.476–7; cf. , 410–12, ϰάμε, 615; , Od. 3.434; πυϱάγϱης; Callim. loc. cit. 144; , Nonn. Dion. 28.6). LSJ s.v. πυϱάϰμων have ‘=forfex, Gloss. (πυϱαϱμων cod.)’ (= Stephanus s.v.). Virgil's Pyracmon lives on in Stat. Theb. 2.599, lasso mutata Pyracmone …fulmina (lasso punning on ἄϰ;μων, as in Val. Fl., no Brontes or Scrapes); Claud. De rapt. Pros. 1.240–1, has Pyragmon … Steropes, De cons. Honor. 3.193–5, Brontes … Steropes … Pyracmon.

52. Cf. Acmon | feruidus ingenio (Met. 14.484–97: patientia, 486 is a ‘witticism'. At Aen. 10.128, Acmon hurls a literary rock formerly wielded by the Cyclops in Hom. Od. 9.481).

53. Acamas is a respectable epic name, a Trojan's in Il. 11.60, , a Greek's in Troy, Aen. 2.262. And in Il. 18, just after the passage with all the Hephaestean fire-tongs (476–7), there appears (484: the epithet in Homer only here, Il. 18.239, , Hom. Hymn. 31.7, ). An ancestor of Serv. ad Aen. 1.742, solis labores, in the song of Iopas, may also have referred, as Servius does, to (sic) in Homer, and discussed the sense of the epithet. Valerius is most obviously forging an etymologizing jingle in Acamans … Pyragmon: if ἄϰμων was seen as < α- privative + ϰάμνω, as is implied in Nonn. Dion. 28.6, cf. Il. 18.614, ϰάμε, then Acamans makes explicit the ‘indefatigable’ energy of the Cyclopes implicit in Virgil's Pyracmon. ἄϰμων had long been used of ‘iron men’, LSJ s.v., e.g. Herakles, in Callim. Hymn. Dian. 146Google Scholar: more specifically, ἀϰάματος occurs in Homer ten times, nine in the formula ἀϰάματος πῦϱ. Valerius' grip may draw out the sense in Virgil's ‘Fire-Indefatigable’.

54. Valerius' ‘twin gates’ are not (are not) the Virgilian exit, but one the always-open entrance for mobs and kings alike, the other springing open for the rare paragon (geminae … portae, 1.832 = Aen. 6.893–5). Paradise survives unaltered (siluas et amoena piorum | deueniant camposque …, 843–4 < | deuenere locos laetos et amœnauirecta | fortunatorum nemorum sedesque beatas … campos, 638–40).

55. In Valerius' last few verses, he directs readers to the Virgilian threshold to Hell (quot limine monstra < the huge catalogue of bogeys in 6.273–89, cf. in limine, 279, multa … uariarum monstra ferarum, 285, plus the team lined up at the doorway to Tartarus, 555–77, cf. limen, 563, limina, 575); and a phrase takes in the unburied crowd on our side of Charon's Styx (turbam … ruentem, 1.850 < turba … ruebat, 305, turba, 325).

56. At Silius' threshold, the evoked shade of a long-deceased Sibyl – the Sibyl – tells us how large a regiment of monsters haunt the hall of Hell, then names them all (579–600).

57. Aetna reads ‘through’ Virgilian loci (Georg. 1.472, 4.263, Aen. 3.579–80, cf. Ov. Met. 15.340) to the matrix in Lucretius (esp. 6.680–702, cf. 1.722–5). In the programme's enigmatic phrase quid fremat imperium, (3) Aetna fastens its project upon the political cosmology worked in epic signification between bearers of force and control, where Virgil's pent winds fremunt, his ruler king enforces imperio (Aen. 1.56, cf. Lucr. 6.199, fremitus, Val. Fl. 1.594, 608, Stat. Theb. 1.348, discussion in Serv. ad 1.52, Sen. Nat. quaest. 6.18.2ff.; cf. Stat. Theb. 10.247, imperiosus, Sil. 9.492. The demythologizing deconstructors of myth, the satirical-rationalist Lucretius and Aetna, work in and out of the bards' remythologizing symbolics. As in Juvenal, oppositional differentiation between the two positions is pointed – but blunted. (See Hardie, P. (above n. 24), passim, esp. 90–1, 237–40Google Scholar; Gale, M. R., Myth and poetry in Lucretius (Cambridge 1994) 156–207Google Scholar, ‘Latent myth in the De Rerum Natura’, esp. 187.)

58. Hylonome's romance, egalitarian marriage, and suicide over Cyllarus, her stallion destroyed in fine fettle, is a gem in the ensemble. Ovid has semifer again only at Met. 2.633 (singular as in both Virgilian uses, Aen. 8.267, 10.212; but he compounds the half in | nubigenas feros, at his start, 211, with the other half in semihomines Centauros |, at his close, 536).

59. Ἰξιονίδης is not extant in Greek; for Latin refs., cf. Bömer on Met. 8.457.

60. Also a solitary in Statius (Theb. 2.598, soon after his Lapitho-Centauric simile at 2.559–64; cf. Priapea 68.15). Serv. ad loc. tells of horse-taming invented here – Lucan's next topic (396–9).

61. E.g. Stat. Theb. 4.140. For the brute/atavistic feta … effudit, cf. feta tellus impio partu | effudit arma, Sen., Oed. 731Google Scholar (with TLL. 5.2.222.18ff., effundo).

62. See Stephanus, s.v., RE XVI.232Google Scholar, §2. The Schol. on Juv. 1.7 take Monychus to be a Giant, but here we catch the bluffer's guide bluffing.

63. Cf. Mart. 8.6.7–8, hoc cratere ferox commisit proelia Rhoetus | cum Lapithis. For ‘Rhoet/cus’, cf. Nisbet and Hubbard on Hor. Carm. 2.19.23. Lucan twins Pholoes (the mount where Hercules slew Centaurs after Pholus' treacherous hospitality) with suh Oetaeo … uertice, 388–9 (where Hercules self-immolated), as in 7.449, petit Pholoen, petet … Oeten, cf. 3.177–8, 197–8, liquit | Oeten, linquitur … | … Pholoe (With 7.872, Pholoe x 4Google Scholar in Lucan: not in Seneca; Oeta common in Sen. and nephew). Lucan's aspera Pholoes is ‘probably’ learnèd, since Pholoe appears to connote ‘roughness’, see Nisbet and Hubbard on Hor. Carm. 1.33.7 (e.g. Tib. 1.8.67, non frangitur, Stat. Silv. 2.3.44, durae … nymphae).

64. TLL 6.1.1242.60 picks only Luc. 6.388 and Sen., Phoen. 571Google Scholar for saxa frangere.

65. ornus x 8 in Virgil, who introduced the ash into the ‘felling’ topos, cf. Aen. 11.138, where … gementibus ornos | loads the scene with sonorous after-echoes, as in 6.180, where the paragraph-end aduoluunt ingentis montibus ornos | reverberates on (seized by Silius, raptas collibus ornos |, 3.639). Stat. Theb. 6.98–106 displays | procumbunt piceae … | ornique (100, blending in Luc. 3.440, | procumbunt orni, which gathers the first and last words of Virgil's list, | procumbunt piceae … ornos |, 6.180–2). Sil. 10.529–34 has its ornus at 530.

66. Cf., e.g., Val. Fl. 1.406, Peliacas … ornos |, 2.6, Pelion ornos |.

67. Did he die, then? Was he pile-driven so far down he wound up in Hades? Or did Mopsus really see him metamorphosed into a bird? We are in the hands of Nestor, of Ovid's Nestor … We have seen Lucan's Monychus ‘smash a peak’ for ammunition; Ovid has, besides Rhoetus' altar-torch, a mixingbowl, altar, threshold, and saxum … e monte reuulsum (236–7, 260–1, 281–2, 341–2); but the tree-spears are the point, cf. Horace's Gigantomachy, Carm. 3.4.55–6, Rhoetus euulsisque truncis | Enceladus iaculator audax.

68. [Hes.] Scut. 188–90 already has Centaurs attack Lapiths, .

69. In Lucan, Pholus, from his eponymous Pholoe (in Apollod. 2.5.4, Pholus inhabits Pholoe), phollows Rhoetus, as in Virg. Georg. 2.456f., Centauros … Rhoetumque Pholumque | et magno Hylaeum Lapithis cratere minantem (cf. Aen. 8.293–4, nubigenas … | Hylaeumque Pholumque, and the Rhoetus massacred by Euryalus in his drunken sprawl, 9.344).

Pholus, barely named in Ovid's catalogue of those who turned tail with Rhoetus, 306, pairs with the (unnamed monster) Nessus (mentioned at 6.365; for Pholus, cf. Stat. Theb. 2.559–64, saxum ingens quod uix plena ceruice gementes | uertere … ualeant … iuuenci, | … qualis in aduersos Lapithas erexit inanem | magnanimus cratera Pholus. Val. Fl. 1.337–8 (Aeson), signiferum cratera minantem | non leuiore Pholum manus haec compescuit auro, reworks Virg. Georg. loc. cit. For Pholus' reception of Hercules, cf. Serv. ad Aen. 8.294, Mayor on Juv. 12.45.)

Alcidœ magni (cf. Aen. 5.214) makes explicit the Herculean connection, Lernœas you must be learnèd enough to ‘know’ (cf. Virg. Aen. 8.300; × 5 in Ovid; often in Seneca). In Ovid, Nestor takes revenge on Hercules by repressing his role; cf. Zumwalt, N., ‘Fama subversa: theme and structure in Ovid Metamorphoses 12’, CSCA 10 (1977) 209–22, at 216Google Scholar.

Nessus appears in Ovid just after the rout including Pholus, told by the augur not to run as he is to be saved for Hercules' arrows (307–9): whence Lucan's conceit, passure sagittas (passur- not in Virgil; in Ovid × 3, including 386; in Lucan × 5, including 392).

Lucan's senex Chiron (cf Anth. Lat. 617.5) starred in Seneca's climax to his chorus’ great vision of cosmic collapse, with Scorpion … Hœmonio … senex … Chiron … neruo … gelidus … Aegoceros (Thy. 858–62; using Manil. 1.269–72 (on Sagittarius; his Chiron is Centaurus, 5.348ff.), fulgentem (cf. Lucan 393, fulgens), Scorpion … arcu | … missurus … sagittam, cf. Ov. Met. 2.81–3, Hœmoniosque arcus … Scorpion. For gelido … sidere, cf. TLL. 6.2.1727.22–3; inpetere is first extant in Lucan, cf. 6.223, | inpetit. In some versions, Chiron did figure in Pholus' reception of Hercules, e.g. Theocr. 7.149).

70. Cf. TLL. 7.1.70–1: Hor. Carm. 3.4.56, of Enceladus iaculator audax (prompting Ov. Am. 3.12.27, Enceladum iaculantem), is the only other poetic use extant before Juvenal in the sense of ἀϰοντιοτής.

71. Mason, H. A., ‘Is Juvenal a classic? An introductory essay’, in Critical essays on Roman literature. Satire, ed. Sullivan, J. P. (London 1963) 93–176, at 157Google Scholar.

72. Satire turns on nothing more than this imperative to scrutinize the cultural terms of literariness.

73. This ‘cosmos’ is the product of my rewriting of the topoi of ‘what Juvenal knows’; to read Satire is always to be implicated in such displacement of ‘what was written’ in the quest for ‘what you would mean’ (the performance of ‘who you are’ that is activated by the presumptions encoded in the writing).

74. Cf. Anderson (above n. 2) 390–1, Edwards, C., The politics of immorality in ancient Rome (Cambridge 1993) 1–33CrossRefGoogle Scholar.

75. Tac. Ann. 4.34–5. Cordus was a bona fide Roman name, ‘Late-Born’, cf. Quintil. Inst. or. 1.4.25.

76. For what follows, see the provocative essay of Beard, M., ‘Looking (harder) for Roman myth: Dumézil, declamation and the problems of definition’, in Mythos in mythenloser Gesellschaft, ed. Graf, F. (Stuttgart 1993Google Scholar = Colloquium Rauricum 3) 44–64, positioning the Roman controuersia as a paradigm theatre of mythhistory discourse misrecognized by ‘our myths about myth – and about Rome’.

77. Cf. the naughty schoolboy of Pers. 3.44–7.

78. Cf. Quintil. Inst. or. 3.8.53 for the exercise (but in prosopopoeia form), uerba … Sullae dictaturam deponentis. Silius' Sibyl glimpses the future for Scipio in what must represent in nuce precisely the terms of this declamation: (twinned with his polar opposite, Marius) imperium hic primus rapiet, sed gloria culpae, | quod reddet solus, nec tanto in nomine quisquam | existet, Sullae qui se uelit esse secundum (13.858–60). Does your Juvenal mean his emphatically bracketed | et nos … et nos | verse and his caesural rhyme … Sullae …, … ferulae … to get puerility or assurance into his take-off?

79. Catullus' dark sarcasm enjoys exchanging death-threats in the forms of his alter ego Calvus' Christmas gift to him of an anthology of bad poems, and of a poem from Catullus' poisoned pen to get even with him for it (14), as he returns the lethal package to sender: why should his friend kill him with so many awful poets (tot male perderes poetis)? Just as Catullus hands on to long-suffering Nepos his collection of his own ‘lean, avant-garde gathering’ (1.1), Calvus must be playing pass-it-on with this ‘avant-garde, exquisite gift’ from some dreadful schoolmaster – namely, Sulla litterator (14.8). The perfect title for the ultimate Orbilius, limit to the school-child's world: Master Sulla. The Sullan Lex Cornelia de sicariis kept him synonymous with murder and poisoning, so that anyone bearing his name spells curtains: e.g. Suet. Gai. 57.2, Sulla mathematicus certissimam necem appropinquare affirmauit … Statius' Epicedion in Patrem Suum traipses forlornly through Virgil's Elysium, finally praying that pater carry on his tuition through dreamwork, as Egeria taught Numa, I. O. M. coached Scipio, they say, and his Apollo locket went wherever went Sulla (sic non sine Apolline Sulla. | (Silv. 5.2.293)): so Statius senior fathered a Sulla – is that why so many of the other addressees receive epicedia, from Statius, the kiss of …?

80. As Juvenal will point out at 2.28; cf. Sen. Suas. 6.3. Does detestable Sulla Felix earn his place in Lucan's Elysium because he did resign his despotism, the unprecedented precedent of a dictatura of unlimited duration (81–79 B.C.E.), before slipping away to sleep for ever just months later (in 78)? (Cf. Ahl, F. M., Lucan. An introduction (Cornell 1976), 139Google Scholar.) Sulla's Memoirs, written up to the eve of his death, claimed the astrologers had predicted a good life and an end at the height of good fortune, and his dead son had come in a dream to bid him cease his cares, come and join Metella and live with her in peace and without bother; on the other hand, he died a suitably lousy death, of phtheiriasis (Plut., Sull. 36–7Google Scholar).

81. Cf. Marache, R., ‘Juvénal – peintrede la société de son temps’, ANRW II.33.1 (1989) 592–639, at 603–6Google Scholar, ‘Passé ou présent: Juvénal et l'histoire’.

82. Cf. Kenney, E. J., ‘The First Satire of Juvenal’, PCPS 8 (1962) 29–40Google Scholar, Griffith, J. G., ‘The ending of Juvenal's First Satire and Lucilius, Book XXX’, Hermes 98 (1970) 56–72Google Scholar, Anderson (above n. 2), 198–9.

83. The rest of the poem will begin to outline the kinds of differences in terms of which we can measure the passing of two hundred years. They will include political revolution as well as literary consolidation (and vv.).

84. Kenney (above n. 67) best expounds the first poem's a-rhythmic alternation of huff-‘n’-puff streams of accusatory exposés with driblets of programmatic declaration (at 30, 45, 51–4, 63–4, 79–80, 81–6, 147–end).

85. Seen by Romano, A. C., Irony in Juvenal (Hildesheim 1979) 76–7Google Scholar.

86. Cf. Lynn-George, M., Epos: word, narrative and the Iliad (Basingstoke 1988) 209–29CrossRefGoogle Scholar, ‘The death of Achilles, the fall of Troy’, esp. 220, ‘the epic present is one which anticipates retrospection, conceives of itself as already past in a future which is the indefinite certainty of what “someone some day will say”’.

87. Cf. Gratwick, A. S., ‘The Satires of Ennius and Lucilius’, The Cambridge history of classical literature, Vol. 2 Latin literature, edd. Kenney, E. J. and Clausen, W. V. (Cambridge 1982) 156–71, at 168Google Scholar, cf. 163.

88. This Flavian courtier's two surviving verses ‘refer to the politics of the previous dynasty’, Coffey (above n. 3) 119.

89. Deferral, that is to say, is intrinsic to any proem, implied by the very recognition of a preliminary function.

90. See above all Hardie, P., The epic successors of Virgil (Cambridge 1993) 60–5Google Scholar, esp. the mega-thrust paragraph at 65, on the opening to Juvenal 1, ‘… behind the sneer Juvenal captures something essential in the Flavian epic’. Juvenal's prologue voice catches the post-Virgilian Juno's impetuous mix of frustrated menace with belligerent power, cf. Feeney (n. 25), 125–8, 131–6 (Ennius and Virgil), Steinkühler, M., Macht und Ohnmacht der Götter im Spiegel ihrer Reden (Hamburg 1989), esp. 63–4, 201–2, 352–3Google Scholar, ‘Die Ohnmachtsrede’ in Virgil, Ovid, Valerius, Lawall, G., ‘Virtus and pietas in Seneca's Hercules Furens’, in Seneca Tragicus, ed. Boyle, A. J. (Victoria 1983) 6–26Google Scholar (Virgil, Ovid, Hercules Furens Prologue), Eigler, U., Monologische Redeformen bei Valerius Flaccus (Frankfurt-am-Main 1988, Beiträge zur klassischen Philogie 187) 32ff., 39ffGoogle Scholar.

91. Cf. Anderson, (above n. 2), ‘Anger in Juvenal and Seneca’, 293–361Google Scholar.

92. See Cloud, J. D. and Braund, S. H., ‘Juvenal's libellus – a farrago’, G&R 29 (1982) 77–85, 78–9Google Scholar, ‘not … a random mess of pottage but … a mixture of thematic material within the Book as a whole’.

93. Gowers, E., The loaded table. Representation of food in Roman literature (Oxford 1993) 193, with 188–219Google Scholar.

94. Cf. Braund, S. H., ‘Juvenal – misogynist or misogamist?’, JRS 82 (1992) 71–86, at 85Google Scholar.

95. Cf. Fredericks, S. C., ‘The function of the prologue (1–20) in the organization of Juvenal's Third Satire’, Phoenix 27 (1973) 62–7, at 62–4CrossRefGoogle Scholar.

96. Cf. Motto, A. L. and Clark, J. R., ‘Per iter tenebricosum: the mythos of Juvenal 3’, TAPhA 96 (1965) 267–76Google Scholar, with the journey Rome–Cumae as a back-to-the-womb reversal of Aeneas' descent to the Underworld on entry to his Italian destiny. LaFleur, R. A., ‘Amicitia and the unity of Juvenal's First Book’, ICS 4 (1979) 158–77 at 164Google Scholar, prefers the sense of ‘umbra in the sense of leisure and retirement’, so querulous escape to ‘pastoral’. Put them together.

97. Cf. Bertman, S. S., ‘Fire symbolism in Juvenal's First Satire’, CJ 63 (1968) 265–6Google Scholar.

98. If the proem proposes that a safe era allows the satirist to speak against recourse to the safe topics needed in unsafe times, then this is revoked at the conclusion? Cf. Ramage (above n. 18) 666.

99. Cf. Sharrock, A. R., Seduction and repetition in Ovid's Ars Amatoria II (Oxford 1994) 36–9Google Scholar.

100. Cf. ‘We should understand [the last line of Satire 1] as a metaphor for Juvenal's art. The “ghosts” which are assailed in his poems are more than the dead of history; the list must include haunting nostalgic memories of virtues and ideals which had really not had authentic life for well over a century’ (Fredericks, S. C., ‘Irony of overstatement in the Satires of Juvenal’, ICS 4 (1979) 178–91 at 190Google Scholar). Satire, rather, requires ‘ghosts’ to raise against the mere presence of ‘here-now-us’.