¹ For analysis of Umbricius' speech as a rhetorical syntacticon, Cairns, F., Generic Composition in Greek and Roman Poetry (Edinburgh, 1972), pp. 47–48.Google Scholar

² Praise of Italian towns: 171 tT., 190ff., 223ff. On the relationship between preamble and Umbricius' speech, see Fredericks, S. C., ‘The function of the prologue (1–20) in the organisation of Juvenal's Third Satire‘, Phoenix 27 (1973), 62–67. Contrasting laudatio and vituperatio are also prominent in the fourth satire: A. Hardie, ‘Juvenal, Domitian and the accession of Hadrian’ (forthcoming in BICS).CrossRefGoogle Scholar

³ Fredericks, S. C., ‘Irony of overstatement in the Satires of Juvenal’, ICS 4 (1979), 178–191, at 184.Google Scholar

⁴ For ampliflcatio in Juvenal, Scott, I. G., The Grand Style in the Satires of Juvenal (Northampton, MA, 1927), pp. 37–43.Google Scholar

⁵ Plat. Symp. 198d–e; Arist. Rhet. 1368a; Isocr. BusirisA; Cic. Brut. 47. A. W. Nightingale, ‘The folly of praise: Plato's critique of encomiastic discourse in the Lysias and Symposium’, CQ 43 (1993), 112–130. In general, T. P. Wiseman, ‘Lying historians: seven types of mendacity’, in Gill, C. and Wiseman, T. P. (edd.), Lies and Fiction in the Ancient World (Exeter, 1993), pp. 126–128.CrossRefGoogle Scholar

⁶ On these issues, seeMeijering, R., Literary and Rhetorical Theories in the Greek Scholia (Groningen, 1987)Google Scholar; also Feeney, D. C., The Gods in Epic (Oxford, 1991), pp. 12–33, and Gill and Wiseman (n. 5), in particular the chapter by J. R. Morgan, ‘Make–believe and make believe: the fictionality of the Greek Novels’, pp. 175–229, at 176–193.Google Scholar

⁷ Zanker, G., Realism in Alexandrian Poetry: A Literature and its Audience (London, 1987), pp.42–50. Meijering (n. 6), pp. 37–39. For an important analogue of artifiction in Longus' Daphnis and Chloe, Morgan (n. 6), p. 218. For artificial caves and the antithesis between natura and ars, H. Lavagne, Operosa Antra, BEFAR 262 (Rome, 1988), esp. 19–22. Fredericks (n. 2), p. 63 comments on the programmatic significance of some physical features of the vallis.Google Scholar

¹ Hor. Serm. 1.1.24–1.1.25: quamquam ridentem dicere verumJquid vetaf! Bramble, J. C., Persius and the Programmatic Satire(Cambridge, 1974), pp. 153–154. In satires 2 and 4: cf. 2.3,15–2.3.16,64–65, 153; on 4.35 (res vera agitur),CrossRefGoogle ScholarSweet, D., ‘Juvenal's Satire 4: poetic uses of indirection’, CSCA 12 (1979), 283–303, at 298.Google Scholar

⁹ Anderson, W S., ‘Anger in Juvenal and Seneca’, CSCP 19 (1964), 127–196 (= Essays on Roman Satire [Princeton, NJ, 1982], pp. 293–361, at pp. 301–305), noting Juvenal's juxtaposition of claims to truth and distortions of that truth; and the need to ‘distinguish between the satirist’ “truth” and truth itself (305).Google Scholar

¹⁰ LaFleur, R. A., ‘Umbricius and Juvenal Three’, Ziva Antika 26 (1976), 383–431 remains essential reading, together with id., ‘Amicitia and the unity of Juvenal's first book’, ICS 4 (1979), 158–177, at 161–164. From a large bibliography on Umbricius,Google Scholar see Winkler, M. M., The Persona in Three Satires of Juvenal (Hildesheim, 1983), pp. 220–223;Google ScholarJensen, B. Frueland, ‘Martyred and beleaguered virtue: Juvenal's portrait of Umbricius’, CM 37 (1986), 185–197;Google Scholar and Braund, S. Morton, Juvenal Satires Book I (Cambridge, 1996), pp. 233–235.Google Scholar

¹¹ The narrative time perspective, perhaps self–evident, is reflected in hie tune (21, introducing Umbricius' speech).

¹² Jacoby, F., ‘Zwei Doppelfassungen im Juvenaltext’, Hermes 87 (1959), 449–462, at 453.Google Scholar

¹³ Cf. Morton Braund (n. 10) on 1–3; for laudo quod, Kiihner–Stegmann 2.276–2.277, including examples of the attraction of the subordinate verb into the subjunctive; the latter represent the terms in which the praise was articulated.

¹⁴ Thus Cairns (n. 1), p. 165; cf. Morton Braund (n. 10) on 1–3.

¹⁵ Cf. Men. Rhet. III.395.5ff.Sp.:

¹⁶ Jacoby (n. 12), p. 453; anticipated in his analysis of the movements by Munro (ap. Mayor on 10–20).

¹⁷ Nisbet, R. G. M, ‘Notes on the text and interpretation of Juvenal’, in Vir bonus discendi peritus (BICS Suppl 51 [1988]), 86–110, at 92–93; supported, with modifications,Google Scholar by Pearce, T. E. V., ‘Juvenal 3.10–3.20’, Mnemosyne 45 (1992), 380–383, and by Morton Braund (n. 10) on 10–20.CrossRefGoogle Scholar

¹⁸ Cic. Ver. 1.94; Ov. Fast. 4.543–4.544; Met. 8.636. Nineteenth–century commentators generally took the word in this sense. Pace Courtney (on 10), line 286 suggests only that Umbricius was not accompanied by slaves in the street, not that he had no slaves at all. For compono with personal or animate objects, TLL s.v., I Cla. Pearce (n. 17), pp. 381–382, sets out some cogent objections to taking raeda as subject of substitit.

¹⁹ For examples of the combination of propempticon and syntacticon, see Quesnay, I. M., ‘Vergil's First Eclogue’, PLLS 3 (1981), 29–182, at 63 (the first eclogue itself directly influenced the third satire: Morton Braund [n. 10], 235–236).Google Scholar

²⁰ Dramatic satire/dialogue: LaFleur (n. 10 [1976]), p. 393; cf. Sweet (n. 8), pp. 283–284 on satire 4 as ‘dramatic monologue’ with Juvenal speaking as a character. Peripatetic dialogue: cf. Plato Laws 625b; Phaedr. 228e; Cic. de Leg. 1.15. Peripatetic dialogues in prose and satire, S. [Morton] Braund, Beyond Anger A Study of Juvenal's Third Book of Satires (Cambridge, 1988), p. 148, with bibliography.Google Scholar

²¹ Ac 2 (Luc.) 147. With veteris...amid (1), cf., perhaps, Cic. Ac1.1(Cicero of Varro): hominem...vetustate amicitiae coniunctum; compare Ibid. 148 Catulus remansit (i.e. after the other dialogue participants had departed) with Sat. 3.29–3.30: vivant Artorius isticl et Catulus, maneant qui

²² R. Maltby, A Lexicon of ancient latin etymologies (Leeds, 1991), s.v. amoenus, citing Isid. Orig. 14.8.33: amoena loca Varro dicta ait eo quod solwn amorem praestant et ad se amanda adliciant; and s.v. amicitia, citing Cic. Lael. 26: amor... ex quo amicitia nominata est; also Ibid.s.v. amicus.

²³ The erotic element in the Numa/Egeria relationship (amicae, 12) is emphasized also at Plutarch, Numa 4.2 and 8.6.

²⁴ Men. Rhet. III.396 Sp.

²⁵ Schonbeck, G., Der locus amoenus von Homer bis Horaz (diss. Heidelberg, 1962), pp. 102–111; A. Motte, ‘Le pr6 sacr6 de Pan et des Nymphes dans le Phedre de Platon’, AC 32 (1963), 460–476;Google ScholarIsebaert, L., ‘La fascination du monde et des Muses selon Platon’, LEC 53 (1985), 205–219;Google ScholarFerrari, G. R. F., Listening to the Cicadas (Cambridge, 1987), pp. 1–34.CrossRefGoogle Scholar

²⁶ Plut. Amat. 749a; Lucian, de Oeco 4; Philostr. Vit. Apoll. 7.11; Cic. de Or. 1.28; cf. also Lib. Decl. 2.1.26; Them. 32c, 246a. Murley, C., ‘Plato's Phaedrus and Theocritean pastoral’, TAPA 71 (1940), 281–295;Google ScholarTrapp, M. B., ‘Plato's Phaedrus in second–century Greek literature’, in Russell, D. A. (ed.), Antonine Literature (Oxford, 1990), pp. 141–173, at 141–148 and 171.Google Scholar

²⁷ Trapp (n. 26), 171; Hardie, A., ‘Philitas and the plane tree’, ZPE 119 (1997), 21–36, at 28–29.Google Scholar

²⁸ Fraenkel, E., Horace (Oxford, 1957), pp. 136–137.Google Scholar

²⁹ There are no caves in the Phaedrus, but Plato may have meant the space underneath the plane tree to be an arboreal ‘cave’, perhaps of the Muses: this is suggested by Rothstein on Prop. 3.3.27, and developed by W Berg, Early Virgil (London, 1974), p. 202, at n. 17; see now Lavagne (n. 7), pp. 446–448. For plane trees as ‘caves’, cf. Plin. Nat. 12.5.9; Sidon. Apoll. 24.65–24.68.

³⁰ Livy 1.19.5,1.21.3; Plut. Numa 8.6; cf. 13.1. For the constitutio, Cic. de Leg. 2.23.

³¹ Wissowa, G., Religion und Kultus der Romer ² (1912, repr. Munich, 1971), pp. 219–220Google Scholar; Camena, s.v.; Waszink, J. H., ‘Camena’, CM 17 (1956), 139–148.Google Scholar

³² Serv. Aen. 1.8. The area retained its association with the Camenae, as is shown by the vicus Camenarum mentioned in CIL 6.975 (A.D. 136).

³³ Serv. Aen. 1.8. Richardson, L. Jr, A New Topographical Dictionary of Ancient Rome (Baltimore, MD and London, 1992), pp. 63–64 is probably wrong to suggest that a temple replaced the aedicula: the aedes Camenarum in which Accius set up his statue (Plin. Nat. 34.19) is the aedes Herculis Musarum.Google Scholar

³⁴ Livy 1.21.3: lucus erat quern medium ex opaco specufons perenni rigabat aqua....Camenis eum lucum sacravit, quod earum ibi concilia cum coniuge sua Egeria essent; see also Ogilvie ad ioc. F. Williams, ‘Vox clamantis in theatre’, PLLS4 (1983), 121–127 at n. 15, rightly notes 12–20 as ‘the ironic reversal of the locus amoenus motif.’

³⁵ Desecration: violarent (20) on aesthetic vandalism, is sacral in tone (OLD s.v. violo la; CIL 1.366.1.2: honce loucom nequis violatod; Ov. Fast. 4.649)

³⁶ For numen in streams, Wissowa (n. 31), p. 222. For praesens numen, cf. Virg. Eel. 1.41, with Clausen's note; Georg. 1.10; Ov. Met. 15.622.

³⁷ Ov. Am. 3.1.1–3.1.2 (cited below, n. 48) is a locus classicus; cf. Fast. 4.649–4.652; Lucan 9.519–9.521.

³⁸ For Juvenal's sensitivity to these issues, cf. Sat. 6.553–6.564, where the silence of Delphi is contrasted with the credence given to astrologers.

³⁹ Neuerburg, N., L'architettura dellefontane e dei Ninfei nell'Italia antica (Naples, 1965), p. 161 (non vidi: cited by Courtney on 11). For nymphaea and representations of ‘caves’, see Lavagne (n. 7), Index nominum et rerum, s.v. Nymphaeum; on the nymphaeum of EgeriaGoogle Scholar, Ibid., pp. 620–621.

⁴⁰ Cf. Virg. Georg. 3.13–3.14 (a figurative temple of the Muses at Mantua): viridi in campo templum de marmore ponaml propter aquam

⁴¹ That this word play is intentional is further suggested by a verbal echo at 138: numinis... Numa.

⁴² Trapp (n. 26), pp. 143–144 (Dio Chrys. 1.56–1.58), 151–152 (Dio Chrys. 36.33–36.35), 157 (ps.–Lucian Amores 31).

⁴³ For a further such rejection, cf. Plut. Amat. 749a; Trapp (n. 26), pp. 158–159.

⁴⁴ Cf. Trapp (n. 26), pp. 165–170.

⁴⁵ Speech is further related to nymph–topography by the speaker (262d), recalling the characteristic activity of nymphs, namely ‘play’ (i.e. ritual dance): cf. 229c on the Illisus' waters,

⁴⁶ Ferrari (n. 25), Index, s.v. ‘rhetoric: and truth’; invective and encomium contrasted at 241e, 265c, 266a–b; Nightingale (n. 5) notes these passages (112 with n. 1), but does not treat them in depth. Compare also Phaedrus' praise of Lysias at 228a with his detraction at 257b.

⁴⁷ Theocritus, Idyll 7, a dramatic, peripatetic dialogue followed by rest in a sacred locus amoenus, reflects similar topographical interests (6–9, 10–11, 131–146), as well as complex interplay of nymphs (92,148), Muses (12,37,47,82,95,129), inspiration (37,92,148ff.) truth (44, cf. 38; 99–100, 148–153), the sensory impact of the place (138ff.), and the impact on dialogue of friendship and Eros. It may well have been influenced by the Phaedrus: Murley (n. 26).

⁴⁸ Cf. Ov. Am. 3.1.1–6: Stat vetus el multos incaedua silva per annosj credibile est Mi numen inesse loco.I fans sacer in medio speluncaque pumice pendensj el latere ex omni duke queruntur aves.1 hie ego dum spatior. I (quodmea, quaerebam, Musa moveret, opus)J.... For experience of inspiration in a sacred grove, cf. esp. Calp. Sic. 1.89–1.90; Luc. 9.564–9.565 (Cato at the oracle of Jupiter Ammon, where it is left ambiguous whether local inspiration is at work, or Cato is ‘inspired’ by his Stoic concept of the divine, or both).

⁴⁹ Parke, H. W., Sibyls and Sibylline Prophecy in Classical Antiquity (London, 1988), p. 93. Val. Max. 1.8.10; Ov. Met. 3.14; Plin. Nat. 28.147; Dion. Hal. 2.61.2. Cf. Phaedr. 278b–c.Google Scholar

⁵⁰ Morton, S., ‘Umbricius and the Frogs (Juvenal Sat. 3.44–3.45)’, CQ 40 (1990), 502–506, at 505.Google Scholar

⁵¹ F G H 813 f.B; Parke (n. 49), p. 72.

⁵² Suggested by J. Adamietz, Untersuchungen zu Juvenal (Wiesbaden, 1972), p. 41; rejected by Courtney on 13.Google Scholar

⁵³ [Morton] Braund (n. 50), following Nisbet (n. 17), p. 92, n. 9. Contrast J. Ferguson, A Prosopography to the Poems of Juvenal (Brussels, 1987), s.v. Umbricius. Like Tanaquil (cf. Sat. 6.565–71 and contrast Livy 1.34 and Sil. 13.818–13.820), Umbricius is not professionally involved in divination (cf. 44), but his name recalls a historical figure who was.

⁵⁴ dis acceptus: Cic. de Rep. 6.13; Varro R.R. 3.16.5; Virg. Georg. 2.101; Ciris 219; CIL 1.1012.8; cf. Ov. Her. 21.50; Tac. Ann. 4.64; Cat. 90.5; Plaut. Rud. 25. Maltby (n. 22) s.v. dives; Varro L.L. 5.92: dives a divo qui ut deus nihil indigere videtur.

⁵⁵ Morton Braund (n. 7), p. 232. For a speaker being ‘carried away’, see below, n. 65.

⁵⁶ The general relevance of Epode 16 to the theme of ‘flight from Rome’ is noted by Adamietz (n. 52), pp. 13–14.

⁵⁷ Quirites recurs at 163, in reference to secessiones plebis (a form of flight from Rome).

⁵⁸ Motto, A. L. and Clark, J. R., ‘Per Her tenebricosum: the mythos o f Juvenal 3’, TAPA 96 (1976), 267–276, see in Umbricius' departure a reflection of withdrawal myths such as that of Astraea/Dike: the details of the argument are unconvincing, but the concept may well be correct. sarcula (311), an iron agricultural implement (i.e. aferrea sarcula) for the manufacture of which no ferrum will be left over from convicts chains, puns on ferrea saecula; similarly, the iron highwayman's sword alludes to the bronze highwayman's sword which characterizes the bronze age at Aratus, Phaen. 131–132.Google Scholar

⁵⁹ See above, n. 53.

⁶⁰ Thus his opening statement: res hodie minor est here quant fuit atque eadem crasldeteret exiguis aliquid, a gloomy prognostication of personal financial decline which is structurally reminiscent of the progressive collective decline suggested at Hor. Odes 3.6.46–3.6.48 and Ep. 16.64–16.66, but relates purely to his own fortunes.

⁶¹ Motto and Clark (n. 58), pp. 273–274, who take ‘Juvenal's’ attack on the vallis at face value, unsurprisingly reach the opposite conclusion about the inspirational role of the topography: ‘Egeria and Diana are irredeemable; the Muses silent’; ‘inspiration is lost’.

⁶² Murray, P. A., ‘Poetic inspiration in early Greece’, JHS 101 (1981), 87–100, at 87–89. The fullest treatment is E. Barmeyer, Die Musen: ein Beitrag zur Inspirationstheorie (Munich, 1968).CrossRefGoogle Scholar

⁶³ Socrates' improvising: cf. 236d: Inspiration and fluency in performance: Murray (n. 62), pp. 94–96.

⁶⁴ Emotion and improvision: Quint. Inst. 10.3.18; A. Hardie, Slatius and the Silvae Poets, Patrons and Epideixis in the Graeco–Roman World (Liverpool, 1983), p. 145.

⁶⁵ For a speaker being ‘carried away’, cf. Plato's description of the rhapsode Ion's psychological state when performing, described partially in ‘inspirational’ terms (Ion 535b–e); Meijering (n. 6), p. 9. For the inspired poet physically ‘carried away’, cf. Hor. Odes 3.4.21–3.4.22; AP 7.42.5.

⁶⁶ Barmeyer(n. 62), pp. 188–191.

⁶⁷ Quint. Inst. 6.2.25–6.2.33, esp. 29; ‘Longinus’, de Subl. 15.1–15.2. Scott (n. 4), pp. 20–24; Verdenius, W. J., ‘The principles of Greek literary criticism’, Mnemosyne 36 (1983), 1–59, at 46. For inspiration and prophetic vision, Murray (n. 62), p. 94.CrossRefGoogle Scholar

⁶⁸ Cf. 8.4: Tranq. An. 17.11. See also Meijering (n. 6), pp. 25–26.

⁶⁹ Socrates thinks that he has heard something better than Lysias' speech in an earlier poet or prose writer; this is because This cannot be his own idea, but something he has heard from elsewhere, poured in, as it were (combining inspirational sentience with a topographical metaphor). For inspiration and memory, Murray (n. 62), pp. 92–94, citing Notopoulos, J. A., ‘Mnemosyne in oral literature’, TAPA 69 (1938), 465–493.Google Scholar

⁷⁰ Cf. e.g. Plut. Mor. 14e. Prop. 3.1.1; Stat. Silv. 4.4.54–4.4.55; Theb. 12.816–12.817.

⁷¹ Umbricius is familiar with the Aeneid(25: Daedalus; 199: Ucalegon); the Iliad (279–280: Achilles' restlessness); also Pliny's Naturalis Historia (238: vitulisque marinis, cf. Nat. 9.19.41–9.19.42). He shows knowledge of plastic arts (89, 217); medicine (232–244); popular philosophy (229); etymology (71); Roman history (53, 114–115, 313–314); Roman religion (137–139); legal processes, both public (33,137–142) and private (81–82,161,273).

⁷² Cf. Meijering (n. 6), pp. 8–9, who well notes the linkage of ai, madness and inspiration, its importance for the Phaedrus, and its force in suggesting that the poet's avraalai might be ‘an important source of profound knowledge, only to be rendered comprehensible through allegorical interpretation’.

⁷³ Author/reader collusion and the reader's privileged position: Feeney (n. 6), pp. 183–184; Farrell, J., Virgil's Georgics and the Traditions of Ancient Epic: The Art of Allusion in Literary History (New York and Oxford, 1991), pp. 23–24.Google Scholar

⁷⁴ LaFleur (n. 10 [1979]), pp. 162–165, suggesting some affinity with Catius in Hor. Serm. 2.4, ‘meant more to provoke than to persuade’. Anderson cited from ‘The Roman Socrates: Horace and his Satires’, in J. P. Sullivan (ed.), Satire: Critical Essays on Roman Literature(Bloomington, IN, 1968), p. 34.

⁷⁵ Frueland Jensen (n. 10), p. 191 rightly points to some moral inconsistencies in Umbricius, particularly as regards criticism of Roman legacy–hunting at 126–130; the argument could have been strengthened by reference to a series of intertextual allusions (based on Hor. Serm. 2.5) to legacy hunting in the immediately preceding lament about personal displacement from the household of a rich amicus: ironically, it is not the amicus who perishes, but the poor Roman's investment of time (124–125).

⁷⁶ The priest of Diana made the sacrifice, after tricking the cow's owner (Val. Max. 7.3.1): vaccam ipse immolavit et urbem nostram tot civitatum tot gentium dominant pio sacrificii furto reddidit; Livy 1.45; Plut. Q.R. 264c– baca nutrita Sabina (85) puns on vacca Sabina, underpinned by the ancient assimilation of ‘b’ and V; cf. the dictum vacca dicta quasi boacca (Isid. Orig. 12.1.31).

⁷⁷ See Ogilvie on Livy 1.45.

⁷⁸ Maltby (n. 22), s.v. Viminalis; further etymologizing suggested by collo (68), collem (71), collum (88): cf. Isid. Oiff. 1.376: colles prominentiora iuga montium, quasi colla.

⁷⁹ Sherwin–White, A. N., The Roman Citizenship (Oxford, 1939), Index s.v. communis patria ftpma.Google Scholar

⁸⁰ Postliminium: AEXXH.863–873; and limen: Ibid., 864–866.

⁸¹ The aqueduct (arcus)lcave parallelism is clearer when taken with Juvenal's models/artistic analogues: Ov. Met. 3.155–3.162, where arcum = ‘cave’; and Cic. de Leg. 2.2, where aqueducts are contrasted with the natural surroundings of Arpinum.

⁸² Fron. Aq. 1.4–5: [for 441 years the Romans were content with the use of waters...] quas aut ex Tiberi aut ex puteis aut exfontibus hauriebant. fontium memoria cum sanctitate adhuc exstat et colitur, salubritatem aegris corporibus afferre creduntur, sicut Camenarum.....[in 312 B.C.] aqua Appia in urbem inducta.

⁸³ But at 31–32, the activities of those who should stay at Rome all involve synonyms of egerere, from which Egeria's name was etymologized; Maltby (n. 22), s.v. Egeria. With aedem (i.e. construction, digging and removal of soil) cf. OLD s.v. egero 2(b); with flumina and siccandam eluviem, cf. Ibid.2(d); with cadaver, Ibid. 1(b).

⁸⁴ Amicuslarnica: 1,12, 57, 87, 101, 107, 112, 116, 121,279. LaFleur (n. 10 [1979]).

⁸⁵ non est... locus, echoes Umbricius' opening quando artibus... honestisl nullus in urbe locus, and the two together play on the topographical content of the satire; similarly, in the Phaedrus, Plato deploys a series of ‘topographical’ word plays, on and 229b, 229d, 230c.